This action might not be possible to undo. Are you sure you want to continue?
99
* O
X
o
* > MO
we o
•a
wff.
*S
o
introduction to
oo H»
MOLECULAR
SPECTROSCOPY
4
kctwiHi
G. M.
BARROW
Ccu U.C.
/">
//<»
Introduction to
MOLECULAR SPECTROSCOPY
ufY^/^Mr^f*"
n n np
(P*
f
Introduction to
M O
L
u
INTERNATIONAL STUDENT EDITION
McGRAWHILL Boofc Company, Inc. TORONTO NEW YORK SAN FRANCISCO
LONDON
KOGAKUSHA COMPANY,
TOKYO
LTD.
n n n nnnnnnrn
R
S C
O
P Y
GORDON
M.
BARROW
Professor of Chemistry Case Institute of Technology
I96L
INTRODUCTION TO MOLECULAR SPECTROSCOPY
INTERNATIONAL STUDENT EDITION
Exclusive rights by Kogakusha Co., Ltd., for
manufacture and export from Japan. This book cannot be reexported from the country to which it is consigned by Kogakusha Co.,
Ltd. or by McGrawHill Book
Company,
Inc.
or any of
its
subsidiaries.
I
Copyright
©
1962
Inc.
by
the
McGrawHill
reserved.
Book Company,
All
rights
This book, or parts thereof, may not be reproduced in any form without permission
of the publishers.
Library of Congress Catalog Card
Number 6212478
Tosho Printing
Co., Ltd.,
Tokyo, Japan
PREFACE
Molecular spectroscopy cannot be regarded as the private domain of the
physicist or physical chemist
who
describes himself as a spectroscopist.
of infrared
Both organic and inorganic chemists make use spectrometers as if they were standard tools
chemists base
spectroscopy.
and
ultraviolet
of the trade.
Analytical
of
many
of their analytical techniques
all
on the methods
Furthermore, almost
chemists
make
frequent use of
molecularstructure data, and
spectroscopic measurements.
much
of this
has been obtained from
One
with
can,
it is
true,
make very
considerable use of spectroscopy as a
tool for the characterization, identification,
little
is
or no understanding of the
way
in
and analysis of materials which the observed specif
trum
related to the properties of the absorbing, or emitting, molecules.
Likewise, data on molecular properties can be used even
the
which they were obtained
is
not understood.
The tendency
of
way in many
chemists, who are not molecular spectroscopists, to proceed with this lack of appreciation for the elementary theory of molecular spectroscopy
persists, in part,
because of the lack of suitable introductory material.
This book
gap between the very cursory treatment of spectroscopy generally given in undergraduate textbooks and the detailed treatments written for the specialist and research
is,
therefore, intended to bridge the
worker that are given in books in the various areas of molecular spectroscopy. The reader will here be provided with the basic theory which makes understandable the relationship of the amount and wavelength of
by a sample and the properties of the moleThis introduction should not only allow the organic, inorganic, or analytical chemist to make surer use of spectroscopy as a
cules of the sample.
radiation absorbed or emitted
It is hoped that the student will be able to under stand and appreciate advanced and current spectroscopic material if. are used. on spectroscopic results. greatly influenced by I have attempted. . it is recognized. This introduction to molecular spectroscopy will show how molecular spectra can be interpreted in terms of molecular behavior and how. or results obtained by these techniques. one's particular interests and background. The problems of selection of material and depth of treatment are. is but should also provide that comfortable feeling of knowing "what happening" when spectroscopic techniques. is previous knowledge of this subject mechanics and group theory is carefully confined to areas of specific value In a similar way. No special background is necessary for the study of this book. relative of course. In fact the goal of much of the material presented could be considered to be that of providing readier access for the student to the two important molecular spectroscopy books. to use a specific example. frequently referred to throughout the book. by G. some of the theory in this introduction to spectroscopy. molecular properties can be obtained. no prior knowledge of quantum mechanics is assumed. The level of treatment is main tained by avoiding the many additional avenues that these subjects open up. The organic chemist. considerable in a book which attempts to serve as an introduc tion to a field as extensive as spectroscopy. should be able to make increased use of the information on the nature and energies of excited states that are being discovered and analyzed by the spectroscopist. to a large extent.vi PREFACE tool. and some analyses are based on the quantum The use of necessary. however. known as "group theory" is given Again no methods of group theory. from such interpretations. he pays some additional attention to the details and nomenclature surrounding a particular spectroscopic subject. for example. after studying this book. Although the Schrodinger wave equation is introduced and simple problems to illustrate its relation to quantities that are important in spectroscopy are solved. be able to appreciate the very important guides to the nature of bonding in coordination compounds that are now being based. Judgement as to the importance of the topics to be treated and the depths to which these treatments should be taken are. The level of understanding of the theory of molecular spectroscopy that this book seeks to achieve is that from which the student can enter into the special ized reference books and the research literature of molecular spectroscopy. With this background the inorganic chemist should. Herzberg.
and electronic spectroscopy that are generally of interest to chemists. W. can be omitted without destroying the continuity of the book.PREFACE vii to maintain a uniform introduction to the areas of rotational. Miss Kim Vo. Thompson. Mrs. for the of the presenta Gordon M. Abrahamson and Dr. I would like to express my appreciation to my colleagues. vibrational. for the assistance they have given me by carefully working through E. Barrow . Carol Haberman. Julia Dasch for typing the manuscript. and to the graduate students at Case Institute. and Mr. Dr. Harriet. I am also indebted to my wife. Wm. Chu. improvements that she made in the style tion and to Mrs. These sections. I. Several sections are concerned in some detail with subjects that are perhaps of special rather than general interest. many sections of the manuscript. marked with an asterisk.
.
CONTENTS Preface v INTRODUCTION Classifications of Spectroscopy 1 Wavelength. and Energy of Radiation 3 Chapter 1 INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS 11 7 7 The Recognition of Quantum Restrictions 12 13 The Bohr Theory The Wave Nature of the Hydrogen Atom 11 8 of Particles 14 The Timeindependent Schrodinger Equation 12 The Particleinabox 15 Problem 14 17 16 17 Normalization of Orthogonality Wave Functions 18 18 19 Symmetry Properties of Wave Functions 21 20 The Quantummechanical Average 110 The Boltzmann Distribution 23 25 Chapter 2 THE VIBRATIONAL ENERGIES 21 OF A DIATOMIC MOLECULE The Vibrations of a Single Particle (Classical) 26 of 22 The Vibrations Two Particles Connected by a 30 for a Spring (Classical) 23 The Potentialenergy Function Bond 34 Chemical . Frequency.
x CONTENTS 24 The Quantummechanical monic Oscillator 37 Solutions of the Har 25 Vibrational Absorption Spectra of Diatomic Molecules 26 40 of Molecular The Anharmonicity Vibrations 43 47 Chapter 3 THE ROTATIONAL ENERGIES OF LINEAR MOLECULES 31 The Rotation (Classical) of a Linear System System (Quantum 53 48 of a Linear 32 The Rotation Mechanical) 50 33 34 Rotational Energylevel Populations Rotational Spectra of Rigid Linear Molecules 35 54 57 The Nonrigid Rotor Chapter 4 THE ABSORPTION 41 AND EMISSION OF RADIATION 61 61 The Timedependent Schrodinger Equation Induced Quantum Transitions 63 42 43 The Interaction of Electromagnetic Radiation with a Molecular System 44 45 46 66 Comparison with Experimental Quantities 70 The Basis of Selection Rules 72 The Integrated Absorption Coefficient for a Transition of a Particle in a 47 48 Box 74 76 Induced Emission and Induced Absorption The Integrated Absorption Vibrational Transition Coefficient for a 76 49 The Intensities of Absorption Bands due to Electronic Transitions 80 83 Chapter 5 ROTATIONAL SPECTRA 51 Rotational Selection Rules for Linear Molecules 84 52 The Stark Spectra Effect in Molecular Rotation 89 53 Molecular Rotation Coupling 92 —Nuclearspin 54 The Positive Wave and Negative Character of the Functions of Linear Molecules 96 .
Spectra. Chapter 8 MOLECULAR SYMMETRY AND GROUP THEORY 81 Symmetry Elements.CONTENTS 55 xi SymmetricAntisymmetric Character and Statistical Weight 98 of Homonuclear Linear of Molecules 56 The Rotational Energy Molecules Symmetrictop Symmetrictop 104 57 Selection Rules and Spectra of Molecules 58 106 Results from Analyses of Rotation Spectra of Symmetrictop Molecules 59 109 The Energy Levels of an Asymmetrictop Molecule 109 Selection Rules. for 510 and Structure Results Asymmetrictop Molecules 112 115 Chapter 6 THE VIBRATIONS OF POLYATOMIC MOLECULES 61 The Number The Nature Coordinates of Independent Vibrations of a Polyatomic Molecule 62 of 115 Normal Vibrations and Normal 116 63 Quantummechanical Treatment of the Vibrations of Polyatomic Molecules 122 64 The Symmetry Coordinates Properties of Normal 125 131 Chapter 7 ROTATIONVIBRATION SPECTRA 71 and Transitions for the Rigid RotorHarmonic Oscillator Model 134 Selection Rules 72 The Relative Intensities of the a RotationVibration Absorption 73 74 Components of Band 136 138 Coupling of Rotation and Vibration Parallel Bands of Linear Molecules 142 146 75 76 77 Perpendicular Bands of Linear Molecules Parallel Bands of Symmetrictop Molecules 149 Perpendicular Bands of Symmetrictop Molecules 152 159 Operations. Symmetry and Point Groups 160 82 Symmetry Operations on Molecular Motions 163 Species and Character Tables of a 83 Symmetry 168 84 The Nature Group 172 .
xii CONTENTS 85 86 87 The Symmetry Operations as a Group 174 Representations of a Group 175 179 Reducible and Irreducible Representations 88 The Irreducible Representations as Orthogonal Vectors 89 181 Characters of Representations Classes 183 810 811 186 188 Analysis of a Reducible Representation 812 The Characters Reducible Representation of Molecular Motion for the 190 813 Number of Normal Modes Symmetry Types 194 The The Symmetry of of Various 814 815 Infrared Active Fundamentals 195 198 Group Vibrations Chapter 9 CALCULATION OF VIBRATIONAL FREQUENCIES AND NORMAL COORDINATES OF POLYATOMIC MOLECULES 91 202 The Kineticenergy Expressions Polyatomic Molecule 203 for a 92 The Potentialenergy Expression Polyatomic Molecule 204 for a 93 Use of Lagrange's Equation in Molecular207 of a Linear vibration Problems 94 The Stretching Vibrations Triatomic Molecule 208 95 96 Use of Symmetry in Vibrational Problems 218 213 Solution of Vibrational Problems in Internal Coordinates by the Method of Wilson 97 Formation of the Method 223 G Matrix by a Vectorial in Wilson's 98 Use of Symmetry Coordinates Method 226 Chapter 10 THE ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 101 230 The Vibrational Structure Bands 231 of Electronic 102 Rotational Structure of Electronic Bands Electronic States of 238 103 104 Atoms 244 251 Electron Orbitals in Diatomic Molecules Electronic States of Diatomic Molecules 105 260 .
CONTENTS 106 xiii Potentialenergy Curves for Electronic States of Diatomic Molecules 261 264 Chapter 1 1 ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 111 Electronic States of Localized Groups Electronic Transitions and Absorption 266 112 Bands 113 271 Conjugated Systems by the Freeelectron Model 276 Electronic States and Transitions in Aromatic 114 Systems 115 279 Electron Orbitals of Coordination Compounds 116 289 Electronic States of Coordination Compounds 117 293 305 Nonradiative Processes Fluorescence 118 307 Index 315 .
.
spectra can be obtained and used without a detailed understanding of the behavior of the components of the instrument. Only brief mention will be made of the experimental techniques used to obtain the spectra that are discussed. With such equipment. furthermore. shape. listed at the end of this section. as size. It will and electronic arrangement of a molecule can be become apparent that spectroscopy offers one of the most powerful tools for a great variety of molecularstructure studies. furthermore. Classifications of Spectroscopy to When the theory of molecular spectra is treated. deal in part or wholly with these techniques. primarily from studies of infrared and ultraviolet spectra. obtained. be seen. There will. it is convenient to is classify spectra according to the type of molecular energy that being . References some treatments and collections of these data are given below. In many commercial spectrometers operating in the various spectral regions are available. practice. will In this way. be no attempt to deal in detail with the tremendous body of very useful empirical data that has been accumulated. One can often understand the nature of the molecular changes that are responsible for the emission or absorption of the radiation. In such the experimental spectroscopic data can be used to determine quantitative values for various molecular properties.INTRODUCTION Molecular spectroscopy is the study of the absorption or emission of electromagnetic radiation by molecules. A number of books. of the radiation and the amount of radiation emitted or absorbed by the sample. The experimental data that such studies provide are the frequencies. or wavelengths. cases. remarkably detailed and exact measurements of the flexibility.
photomultiplier detector It —electronic spectra of should also be mentioned that vibrational spectra can be obtained by means Raman spectroscopy and that this technique uses a visible or ultraviolet spectrometer. might be given as: Microwave spectrometer: klystron source. Thus an absorption of energy due primarily to a change in the vibrational energy may show the effects of accompanying rotational energy changes. and components as has been implied by this classification. that spectra One can used. from the action of a magnetic field on the molecules of the sample. and electronic spectroscopy. In nuclear field is to magnetic resonance spectroscopy the effect of the magnetic orient certain nuclei in certain directions with respect to the direction of . however. categories The energy result. and ultraviolet spectrometer: tungsten lamp or hydrogen dis charge tube source. levels that are studied in these spectroscopic in contrast to those normally studied in rotational. vibrational. In addition to the types of molecular spectroscopy listed above. wave guide. electronic see. and crystal —molecular rotation spectra thermocouple detector— molecular vibration spectra detector Visible Infrared spectrometer: hot ceramic source.: 2 INTRODUCTION TO MOLECULAR SPECTROSCOPY altered in the emission or absorption process. These are nuclear magnetic resonance (nmr) and electron spin (esr) resonance spectroscopy. We will may result from transitions in which more than one type of molecular energy changes. there are two closely related types that play a comparable role in chemistry. It also divide spectroscopy according to the instrumentation happens that the categories obtained in this way are similar to The instrumentation classification those based on molecular energies. energy of a free molecule as made up of separate rotational. rocksalt prism or grating. vibrational. glass or quartz prism or grating. to treat the to different electron arrangements It is allowable. These are Rotational spectra molecule —due to changes the rotational energy of the Vibrational spectra— due to changes in the vibrational energy of the in molecule Electronic spectra —due to changes in the energy of the molecule due generally to a good approximation. In this way the principal headings for the material that is to be presented can be arrived at.
1959.INTRODUCTION 3 These field. is oriented one way or the and again the resulting energylevel separathese two orientations is studied by radiation of suitable frequency. Inc. and Energy of Radiation Before the emission or absorption of radiation by molecular systems can be treated. X E = E cos2irv(tf) .. "Nuclear Magnetic Resonance. Ordinary radiation traveling in the z direction. suitable frequency can then be used to study the energySimilarly.. York. 1959.. New tions of Nuclear sion of nuclear magnetic resonance and electronic spin resonance will be included. and radiation of a the level separations. 1 Planepolarized electromagnetic radiation. can be obtained. no discus Benjamin. New in HighResolution Inc. A and are Inc. which more con venient for discussion here. to nuclear magnetic resonance Of particular interest electronic spin resonance spectroscopy. York. 1961. and electronic spectroscopy given here. some of the terms used to describe electromagnetic radiation must be summarized. by M. New York. fields perpendicular to each other and to is Polarized radiation." W. Figure 1 indicates the FIG. and "An Introduction to SpinSpin Splitting Nuclear Magnetic Resonance Spectra. for example. number of recent books have appeared which provide excellent introductions. Roberts." McGrawHill Book Company. both by J. Wavelength. Jackman. and "ApplicaMagnetic Resonance Spectroscopy in Organic ChemisIn try. which are at the same level as the treatments of rotational. can be treated in terms of electric and magnetic the direction of propagation. The wave nature oscillating electric description of electromagnetic radiation associates fields and magnetic with the radiation. suitable for the chemist. vibrational. Frequency. A. L. in electronic spin resonance spectroscopy the spin of an electron of the sample molecule other with respect to the tion of field. view of these treatments." Pergamon Press. different orientations correspond to different energies. D.
4 INTRODUCTION TO MOLECULAR SPECTROSCOPY and magnetic fields that are associated with polarized electromagnetic radiation which has only the xz plane component of the electric electric field. traveling with the velocity of light. The value of the electric field along the z axis at a given time and at a given point on the z axis as a function of time can be expressed by the formula E =E cos 2wv (t  J (1) This formula. In spectroscopy almost all 1 attention is centered on the electric but. of the radiation. the direction of propagation. The corpuscular description of electromagnetic radiation views this radiation as a stream of energy packets. and also of the associated mag netic field. Basic to an understanding of spectra brings together the is the relation of Planck that theories of radiation. travels out in the z direction. As can be readily verified by considering a position with some fixed value of z and asking how many cycles pass this position in the time c seconds. often customary to deal as well with the wavelength. while at a given value of z the is E with wave respect to time given. the relation between the wavelength and the frequency of oscillation of the wave is "I Although. t. for completeness. shows that this plane polarized radia tion has the magnetic field of the electric field. called photons in the visible region. With sists of this picture of radiation. component in the plane perpendicular to that The oscillation of the electric field. W the frequencies of the radiation that are absorbed or emitted in a spectral study are more directly related to it is the molecular energy changes that cause the absorption or emission. gives the behavior shown in Fig. 1 for a given value of t = for example. electromagnetic radiation con waves of varying electric and magnetic field strength traveling with the velocity of light and having a given wavelength and frequency. wave and corpuscular His equation Ae = hv (3) . as will be seen later. Fig. oscillation of it should be recognized. field. with a velocity c.
it is a historical fact that each area of spectroscopy developed pretty venient set of units. with the value 6.624 X of 10 27 ergsec. such as the cgs wavelength unit of centimeters. are shown in Table 1. TABLE 1 The Spectral Regions of Electromagnetic Radiation Microwave X (cm) Infrared Visible Ultraviolet 300.000 megacycles/sec V 3004. tion.6l X 10. called reciprocal centimeters or v is essentially wave numbers. where the "particles" may be molecules. and each acquired own con frequently necessary to convert from one description of radiaX. or by the value of Ae = hv.0003. differs from by the It is designated as v and is defined as (4) *\ The units of v are invariably those of It cm1 . being different from light. In the various types of spectroscopy one makes use of Ae.0030. interconversions that are likely to be encountered are It should shown in Table be pointed out.000 1000300. or quantum.00025 X(A) V X 10" 53 X 10" 6 7. with the wave nature concept v. with regard to the entries in Table 2.000 7 1 3 X 10. by the constant factor of the velocity of The ranges of electromagnetic radiation. atoms. the energy of a radiation energy packet.0001. such as given by a value of or v. independits ent of developments in other areas.000 cm" . is not used throughout. such as given by v. expressed in the units usu ally used for each region. This unit. the frequency of the radiation. Conversion factors for all the 2. for example.5 3. characterize the particular radiation emitted or absorbed.1 0. tional unit or X to One addi which v is proportional to v has convenient numerical values and is often used. quite commonly called a frequency unit. While it may seem unfortunate that a consistent set of units. to another.INTRODUCTION 5 where h. a constant known as Planck's constant. should be kept in mind that v a measure of frequency. that a term such as cal/mole is intended to mean calories per Avogadro's number of particles. factor c. ties together the corpuscular quantity Ac. or quanta. It is much on its own.
2. R. A. 1954. Inc.067 10" 10" " 1..242 10" 2 cal/mole erg/molecule cv 1 X 1 10 11 5.. Verify the conversion factors along a row of Table Calculate typical energies of an Avogadro's number (6.2396 X 10" 4.. Inc. Harrison. of Complex Molecules. N. Cliffs. John Wiley 3. . and E..439 X 1 10 16 1..." Edward Arnold & Co.. K. Sons. and J. Englewood Cliffs. Sawyer. Englewood Applications 1.. 1. Lord. and R. S.986 6. V.949 cm 1 cal/mole.J. 2.338 X lO" 6 23. of Inorganic Substances. E.602 1.3499 Exercise Exercise Exercise 10 23 ) 1.: "The Infrared Spectra York. 2." Reinhold New York. Gordy. .. Complete Table 1 by filling in all the blank spaces. W. Lawson. Trambarulo: "Microwave Spectroscopy.023 X of quanta in the four regions of electromagnetic radiation shown in 3. these values with the classical thermal kinetic energy 600 cal/mole per degree of freedom at room temperature and with chemical bond energies of 50 to 100 kcal/mole. R." John Wiley & New 2. 1961.036 X 1 10 16 1. & New York." Sons. 1954." PrenticeHall. G. Gillam. Loofbourow: "Practical Spectroscopy. Inc. R..INTRODUCTION TO MOLECULAR SPECTROSCOPY TABLE 2 Energy Conversion Factors erg/molecule 6. E. Stern: "An Introduction to Electronic Absorption Spectroscopy in Organic Chemistry. . Smith. A. W. 1953. N. X X X 8.858 0. 3. F. London. 1944.: "Infrared Absorption Publishing Corporation. L. C. J.060 2. Bellamy. Table of Compare ?RT = PRINCIPAL REFERENCES Experimental Methods 1. .: "Experimental Spectroscopy. 1948.J. Inc." PrenticeHall. R.
the general nature of theoretical treatments of atomic and molecular systems will be introduced to the extent needed for the treatments of molecular systems given in later chapters. as was later stated. driven to assume. The Recognition of Quantum Restrictions The earliest recognition of discrete energy jumps or quantum restric tions stems from Planck's studies of the radiation emitted by hot bodies. An appreciation of the restrictions that are placed on the energies of atomic and molecular systems is basic to an understanding of the spectra of these systems. In this chapter. been closely tied to the study of their spectra. Planck's constant. Furthermore. where v is a frequency of He was oscillation a proportionality constant. since about 1900. that the oscillating atoms of a hot body cannot have any energy of oscillation but can have only the energies that are integral multiples of hv. radiation from such an oscillating system consists of the and h is 7 . a better under usually obtained if some of the general features of these theories are known.1 INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS The development of theories of atomic and molecular systems has. 11. The principal aspect of these theories that must be introduced here is their statements about the energy that an atom or molecule can have. While standing is many spectra can be understood to some extent without any further appreciation of the theories of molecular systems.
or quantum. therefore. however. however. Planck's blackbody theory was. His reluctantly proposed ideas of quantum restrictions were accepted and expanded into more elaborate theories of molecular behavior. each allowed orbit having a certain energy. By 1913 it was recognized. In the next few years after 1900 the ideas of Planck were applied and extended. by Niels Bohr in 1913. in particular to the hydrogen atom. of energy given out in such is. by the constant h were basic features of Planck's blackbody radiation deriva They seemed. in some manner out side the nucleus. = hv idea of discrete energies and the prominent role played The tion. or resided. they could not be justified. since. Bohr's interpretation of the behavior of the electron was based on the rather arbitrary assumption that the nucleus in such a multiple of h/2w. as heat capacity of in Einstein's theories of the photoelectric effect solids. that an atom. in particular the hydrogen atom. in classical to be awkward and troublesome features treatments. a jump from one energy state to the next lowest one Ae therefore. worthy of study. the Bohr theory provides a clear and concrete illustration of one of the most important quantum rules and is.8 INTRODUCTION TO MOLECULAR SPECTROSCOPY energies emitted levels to a lower one. consisted of a small heavy nucleus carrying the positive charge and that the electron moved. and his interpretation of the continuous spectrum provided a hot body was to be followed by analyses of spectra of great detail by and complexity. is as follows: the angular momentum . as a result of the then recent work of Thomson and of Rutherford. in brief. and the The next major step of spectroscopic interest was the application of these ideas of quantum restrictions to atomic systems. 12. The derivation. a preview of the developments in molecular theory and spectroscopy that were to take place in the next half century. it way that its angular moved in a circular orbit about momentum was an integral This statement and the ordinary rules of dynamics and electrostatics lead to allowed orbits for the electron. when an oscillator jumps from one of the allowed energy The unit. The Bohr Theory of the Hydrogen Atom Although detailed theories of atomic structure are not pertinent to the discussion here.
the electron r is allowed to travel as (4) — n' 47r 2 me 2 Substitution of numerical values for the constants gives the radii of the Bohr orbits as r = 0. (1) m the mass of the electron. . v its velocity. (3) (7) Use can now be made and the quantum restric tion of Eq.INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS 9 postulate requires that mvr where orbit. for the total energy of the electron. according to Bohr. 3. defined so that A = 10 8 cm. (1) to obtain « = fcivsy w = 1> 2 3 ' — < 8> . . = n jjis n = 1. 2. If the potential energy at infinite separation of the electron and proton is taken as zero. the result t = KE + PE = g mu2  7 of the relation of Eq. . has Of more interest in spectroscopy are the energies of the allowed orbits. by means of the quantum Eq. unit of angstroms. gives the radii of the orbits in which. 2. restriction of Elimination of 2 i> . the potential energy at separation r is given by Coulomb's law as PE = . . 3. 2 angstrom n = 1.r (6) Addition of the kinetic energy gives. (1).529n. (5) 1 The convenient been introduced. and r the radius of its The requirement that the electron should have the coulombic force of attraction to the nucleus balanced by the centrifugal force gives t = r2 ^ r (2) or mv 2 where e is (3) the charge of the electron. . .
the total angular momenby tum of the electron is 2(h/2r). With the Planck relation Ae = hv. be treated here. the orbit radii and energies for an electron of the He + ion. the Bohr theory predicts that an amount of energy Ae = e„.\ down to that with n = ni. more. must is also be quantized in units of n = 2 orbit. if atom is now assumed to jump from one orbit to another.677 (11) expression. which is not quite at the nucleus. for example. Derive. and there the possibility of showing the application of a directed field that this orbit can be inclined in five . cm (11) This corresponds almost exactly to the empirical expression known Rydberg atom spectrum. according to the Bohr theory. on substitution of numerical values. a very generally applicable one in atomic and molecular systems.10 INTRODUCTION TO MOLECULAR SPECTROSCOPY of e where the fact that each value of n implies a value nized has been recog by writing «„. A more complete statement of the principle applied to the simple Bohr hydrogen atom would recognize that not only must the total angular momentum momentum Thus for the be quantized but that. angular It is interesting to see that the condition that the momentum Further be quantized in units of h/2K leads to a set of allowed energies. Elaborations of this result have been orbits. as the which summarizes the observed hydrogen Exercise 11. to take into made to allow for elliptical account that the electron and nucleus should be treated as rotating about the center of gravity. as be seen. the possible energies that the electron of the hydrogen atom can assume. according to the Bohr theory. — e must be emitted. say from that with n = n. This result gives. and to include the relativistic dependence of electron mass on These details need not. W2 ^ has the prediction (10) In terms of the more often used units of cm 1 one 2w*me /l 1\ "c'ehrltf*) > cm which. in that direction if a direction field. gives = 109. the Bohr theory leads to the electron of the hydrogen (l! the prediction of emission of radiation with frequencies ' = At T v = 2ir 2 f^\n\ ~ </ l me*/l 1 \ Wlth ni > . is imposed on the atom. velocity. however. as by an applied electric or magnetic the component of angular h/2w. The will principle of quantized angular momentum used by Bohr is.
as imposed by Bohr. protons. postulated that the For wave associated with an electron in a Bohr orbit should form a standing wave around the nucleus. became . proposed that electrons. It seems best here to present and in 1926 in illustrate this approach by means of It is of great an equation given appreciate the described by Erwin Schrodinger.e. i. value to way which atomicscale problems are handled and by the wave mechanics . l(h/2w). or state of the electron.INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS different directions corresponding to angular 1 momentum components The of 2(V2ir). and so forth. The Wave Nature of Particles In 1925. the condition that 2irr = n\ where n is an integer and 2irr is the circumference of the is imposed. has a multiplicity of 5 or fivefold degenerate. Louis de Broglie. interestingly. all with the same energy. Instead. while ing the hydrogenatom spectrum. example. One says that the n is = 2 orbit. momentum mv.e. as well as radiation quanta. 13. 0. to the same requirement. l(h/2w). did not recognize in a sufficiently basic way this wave nature of the electron. i. of Schrodinger. splitting of the energy of the original n that the original state is = 2 state by the application of a field reveals to be thought of as having five states. will The details of atomic theory and spectra not be developed further here. The importance of the wave nature led ultimately to a formulation of a general approach to the mechanics of atomicsized systems. and 2(h/2w).clear that the Bohr condition. some further general theoretical developments which led finally to the more powerful and less ad hoc formulation of quantum restrictions than that of Bohr will be outlined. would be given by X = — mv if it is (12) This interesting relation can be used in a number of situations. particle with He further suggested that the wavelength associated with a mass m and velocity v. Even more shown It indicative of the validity of the de Broglie relation were the experiments of Davisson and Germer in to give diffraction effects corresponding to a which an electron beam was wave with wavesuccessful in explain length given by the de Broglie relation. electron orbit mvr = n(h/2w). reasoning from the generally symmetric nature of the physical world. should have wave properties associated with with them. This stipulation together with de Broglie's wavelength relation leads...
The position of the however. The Schrodinger equation accomplishes this extension. like/ = ma. It will. of a derivation showing its validity. or eigenfunction. say the x coordinate. to results in agreement with observation. represented by \//. is not really derivable and should be looked applicable to atomic. or quantummechanical. equation. to consider a simple problem in which the particle can move only along one dimension. is used and accepted. be used to solve again the hydrogenatom problem and predict the position and the energy of the electron exposed to the coulombic potential of the nucleus. According to M. will be a function of x. is (If \p is a complex function. be extended to more complicated systems. the solution function. For emphasis one sometimes writes 2 ^/{x). In a study of spectroscopy there molecular systems. The Schrodinger equation upon as the counterpart. The Schrodinger equation. however. The Schrodinger particle is. The Timeindependent Schrodinger Equation Although the equation originally presented by Schrodinger allowed for the calculation of timedependent behavior. the probability given by ^V. for simplicity. given only in terms of a probability function. be very helpful to have performed a simple illustrative calculation.12 INTRODUCTION TO MOLECULAR SPECTROSCOPY 14.) . The probability of the particle being at a given value of x along the x axis 1^ is at the given value of x. Bohr's method could not. Born. for example. as will be seen. calculations of the properties of however. (The Bohr theory of hydrogenlike atoms obtained just such information. If we choose. not because leads. as then equal to the value of is sometimes <j/* the case. One can then better appreciate the nature of more general quantummechanical methods and solutions. the value at a given point of the square of the trigonometric or algebraic function that solves the Schrodinger equation gives the probability of the particle being found at that point. Use of the Schrodinger equation implies that is we are interested in learning about the energy of a particle. is little need to perform wave mechanical.) The Schrodinger equation might. it is convenient first to demonstrate the simpler timeindependent features of the complete equation. of such classical formulations as Newton's / = ma it expression. but rather because wherever properly applied.and molecularscale problems. which subject to some potential energy. and the position adopted by the particle. and the information was in agreement with experiment. yields directly values for the allowed energies of the particle under study. where is the complex con jugate of ^.
however. but not identical with. the Bohr orbits. and the probability that is found is related to. the probability of the particle being at various positions.) Solution of a particular problem requires values of m and an expresfunction ip sion for the potential function to be substituted. + 2 dy U(x)+ = e* (13) (The fore. that solves the resulting differential equation must then be found. Briefly. The ^ 2 function would give the probability of the electron being at various positions. lem is. The e. as mentioned. which turn out to be identical with those obtained by Bohr. in fact. In the first case one would deduce from the infinite value 2 an infinite probability of the particle being at a given position.) Solution of the problem would give the allowed energies. Before the equation is illustrated by a simple. but important. One understands these restrictions from the fact that the equation involves the second derivative of and that this would go to infinity at a discontinuity in the slope. perhaps better at first merely to use the equation to calculate desired quantities much as one does with / = ma. of while in the second case one would have two different probabilities at the same position. first term can. ' . three dimensional. the second with the potential energy. there with the total energy. must be "well behaved. The hydrogenatom problem would and the expression (The hydrogenatom prob/r to be put in as the potential function. go to infinity or be double valued. the function ^.INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS 13 The Schrodinger equation for a particle of mass m constrained to one dimension and subject to some potential function U(x) is fe 2 8ir 2 m da. The equation and its application are more understandable when require the value of for the electron to be used m —e 2 \f/ \[/ applied to a specific problem. and the sum. Neither of these would be physically reasonable." It must not. The square of the solution function gives. to be an acceptable solution. for example. Further restrictions are that the function must be continuous and that discontinuities in the slope can occur only at points where the potential energy goes to infinity. example. it is necessary to state some of the limitations that are imposed on the function f. Such a function will generally exist only for certain values of and these values are the allowed energies of the system that are being sought. be associated with the kinetic energy of the It is particle.
for example. infinitely high outside this region. expression rmx/a has. In the region < x < a the potentialenergy function in this region reduces to is U(x) = 0. as can be checked. (14) to give T . and the Schrodinger equation h2 87r 2 dV dx i m 4 (14) find wellbehaved solutions for this equation. . as will be shown in Chap. 3. is that of determining the allowed energies and the position probability function of a particle that can in move This problem implies a potentialenergy function that has some value. . is subject to a potential which for some purposes can be so represented. Of more spectroscopic interest. been arranged so that the function goes to zero at x = and at the = a for any integral value of That the function satisfies Schrodinger equation can be tested by substitution in Eq. or t. which can conveniently be taken as zero. . (15) A is some constant The n. nirx V o 2 ) sin / o n 2h 2 8ma 2 Right side = c I A sin J (17) . . the function yj/ To be in the region between i = A and x sin where n factor: = 1. between x = and x = a. now necessary to The function must be It is \f/ zero outside the potential well since there the potential is infinitely high. electrons of a conjugated system of double bonds in a molecule behave approximately as though the potential which they experience is such a simple squarewell function.U 15. in 4/. being in such a region. which in fact has some counterparts of interest in real molecular problems. is the fact that the doublebonding. Functions which solve the differential equation and also satisfy these boundary conditions can be seen by inspection to be well behaved. 2. . 11. INTRODUCTION TO MOLECULAR SPECTROSCOPY The ParticleinaBox Problem A very simple problem... and is is only one dimension and confined to a region of length a. An electron in a piece of wire. and there is no probability of the particle and prevent a discontinuity and a must be such that it equals zero at x = and at x = a. Lett side = — h2 „ »7r OT „ 2 ( I nV \ jA 2 .
FIG. This leads to a zero probability of a particle being anywhere in the box and is therefore unacceptable. A and molecular problems. (15) provides a solution to Eq. (14) are equal. The quantum phenomena. 11. 3. and no energies other than these will result. (18) No really different solutions can be found. which were so arbitrarily introduced in the Bohr theory.2. are seen to be quantized as a result of the quite natural intro similar situation occurs generally in atomic duction of the integers in the solutions of the Schrodinger equation. (14) if m3A J 8ma s n = 1. and probability functions for the particle. (14) but gives a wave function that is everywhere zero. e4 = 16(A /8 2 ma 2 ) t3 = 9(r/8ma) £2 = 4<ft J /8ma*) . wove functions. UTTX and the expression ^ = A sin gives solutions of Eq. [The value n = in Eq.s INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS 15 The left and right sides of Eq.box problem. are seen to result much more naturally in Schrodinger' approach. )•] Energies.in a. which are represented in Fig.] The allowed energies t.
or eigenfunctions. features of Schrodinger equation solutions are appreciated. In these cases. One can therefore approximate such a system by representing the molecule as a conjugated system. As an illustration of this Schrodinger equation problem the one dimensional squarewell. Since the spin of an quantum number of +^ or — two electrons can be accommodated in a state represented by a single value of n. i. energies of the ir electrons. sufficient to it will be have the solution functions and allowed energies stated with out derivation. remains to recall the Pauli exclusion principle which requires that no two electrons of a mole quantum numbers the same. 11.e. by the previous approxima « = n 2 2/8ma 2 /i To apply this squarewell it tion to the w electrons of a conjugated system.— 16 INTRODUCTION TO MOLECULAR SPECTROSCOPY The ii solution functions. The success of this simplification of the factors affecting t electrons in conjugated systems is judged by cule have all their electron can be oriented to give a spin J. onedimensional region of uniform potential bounded by regions of infinitely high potential. are then those given derivation. One finds that solutions exist for the Schrodinger equation e only for certain values of and that the solution functions and the cor responding energies are characterized by an integer. or "particleinabox.. the comparison of the energy calculated for the promotion of one of the highest energy electron pair to the next higher energy state with the . 12. called a quantum number. — A sin o and the probability functions ^ are = A* sin 2 — level in Fig. A number of situations will arise in later chapters where soluthe general tion of the Schrodinger equation will present plexities that some mathematical comif cannot be dealt with here. The allowed . ignor ing electronelectron repulsions. The occupancy of the energy levels by the six k electrons of hexatriene is represented in Fig. (19) shown opposite the corresponding energy This simple problem illustrates many of the characteristics of more difficult quantummechanical calculations on atomic and molecular systems." solution can be applied to the question of the energies of the doublebonding or r electrons of a These electrons are apparently delocalized and are relatively free to move throughout the length of the molecule.
is remarkably good. occurs The dashed arrow shows the transition that radiation is when absorbed. Compare the observed value of 2.INTRODUCTION TO THE THEORETICAL TREATMENT OF MOIECUIAR SYSTEMS 1 7 energy of the radiation quanta absorbed in an electronic spectral study. In terms of a onedimensional problem this implies that particle. absorbed in the lowest energy irelectron transition of hexatriene according The length of the molecule can be taken as to the square. when a wave function f is obtained for a shows that there is a value of unity for the total probability of the particle being found somewhere in space. Normalization of It is Wave Functions frequently convenient. about 7. 12 The squarewell model for the w electrons of hexatriene. to arrange it so this function /: if/* dx is (20) For the particleinabox problem. gives a total probability of unity is said to be FIG.580 A. since ^ region " zero everywhere outside the to a. as Exercise 12 illustrates. with A.well model. the limits can be reduced to give / r dx = 1 (21) A wave function that normalized. Calculate the wavelength of the radiation that will be Exercise 12. . The agreement.3 16.
wave functions for other molecular problems can.18 INTRODUCTION TO MOLECULAR SPECTROSCOPY For the particleinabox wave functions. if dr represents the differential element of volume. = A*sm*—dx a 1 (22) The an integration can be performed (or read from tables which give." /_ i*dr or. functions that solve the is Schrodinger equation for a particular problem. therefore. where I and m imply different quantum numbers. a property of considerable importance in spectroscopic problems. and usually tion are. or A = ^ "i" (24) The normalized wave ^ functions for a particle in a box are. normalization consists of evaluating A such that o /. if = 1 (26) the wave functions are complex. normalized.e. Two wave functions \f/i and ^/ m . that of orthogonality. 1 + = /_ „" ^** dT 17. the normalization condi on a wave function \[/ is + . Z (23) For normalization. = Va Sm (25) In a similar way. < 2? ) Orthogonality A general property of wave functions. In the general threedimensional that case.. i. therefore. are such that /+« Mm dr = . / if m is integer. sin 2 my dy = t/2) to give o /o rin*^«fa = £ o.
I and form can be shown to be zero by. be readily illustrated by the particleinabox eigenfunctions. (e imy <r im ») (30) and performing the integration on the exponential form of the integral.INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS or. It is necessary to show that sin o JC — a sin a ax = = rx/a converts the (^9) for of I ?± m. m are integers. These properties. It can. Substitution of the variable y left side Eq. 19 more generally. replacing the sine terms by sin ly = jp (e"» —  e"") and sin my = ^.dT /_ if I = (28) and m describe states with different energies.. corresponding to a single energy of the system.. however. I = for I a* m /o n it will be seen. The integral in for example. (29). (29) to / " sin ly sin my dy where. wave functions cor responding to nondegenerate states of a system /+». tional features must be considered when there are several states. A summary of this 4> t statement that if and the previous section can be given by the and 4*m are normalized. . this it should be kept in mind. real.. an example of the general orthogonality result that is applicaSome addible to the wave functions for states with different energies. lead to considerable simplification in a number of spectroscopic problems. + . i. This is wave functions." +*+. is This property stated without proof.e. In this way one obtains zero for the integral and verifies the orthogonality statement of Eq.
antisymmetric i =_„**££ Even function. symmetric #2 = sin *M 2 Odd function. one sees mathematically that of the replaced by —y 3 The symmetry properties of the partideinabox wove functions. the potentialertergy function is a symmetric one.) Wave function 4' a Symmetry nature of ^ Odd function. When this is the functions shown in Fig. Generally. it is more satisfactory to use the center of symmetry. 13.20 INTRODUCTION TO MOLECULAR SPECTROSCOPY 18. It is then possible to investigate the sym metry properties when y is done for FIG. as the origin of the coordinate system. The potential of the particleinabox problem. 1 wave functions by investigating what happens in a. 13. antisymmetric *— ? 1 I Even function. the solution functions are sines and shown in Fig. If this is cosines as done in the squarewell case. for example. is symmetric about the point x = a/ in Fig. 1T. symmetric a/2 6 +«'2 —u +y • . particular wave function. such as the midpoint of the square well. 11 that those with even values of n are functions that are antisym metric about the midpoint of the well while those with odd values of are symmetric. (The wave function! shown are not normalized. n Such symmetry properties are characteristic of wave functions that arise from a problem with a symmetric potential function. as is often the case. It is apparent from the plots of the wave functions shown in spectroscopic problems Fig. Symmetry Properties of Wave Functions is A general property of is wave functions that their important in many form when.
The method be introduced. particle for We will. function is of an even.INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS 21 For n odd Hv) = Hy) For n even (32) My) = Mv) Again one sees that for odd values of n metric. we have occasion to want to know the average value of some other properties. in some cases. the value for the variable g times ^. such as position. If what is called an denoted by g and the operator for this function by G. to In quantum mechanics every variable. which will also be encountered later. Tx = ~ h d &r^ te 2 2 2 (34) The calculation of the average value of the kinetic energy of a particle in a box can be used to illustrate this operator. function. quantummechanical average. function. furthermore. ' The kinetic energy operator. by Tx gives * 87r 2 m (^Xt)'* sin 5 ? (35) wVa ir = w* . i. allows the concept of operators. G+ = g+ Only a few functions and these are listed in Table 11. the. *l9. (33) and The operator is. want to know the average position of a which we have obtained an eigenfunction of finding a \f/ from solution of the appropriate Schrodinger equation. or The Quantummechanical Average of the Although the values quantummechanical will results of allowed energies of a system are the prime importance for spectroscopy.. for example. has associated with is it momentum. kinetic energy. and operator.e. for kinetic energy can be G\f> used to illustrate the basic operator relation = gf. their operators will be encountered. operation on this function Since ^ = A sin nrrx/a. then operation on the wave function of the system by one of these variables G will give. denoted by T for onedimensional motion along the x axis. while for even values antisymmetric. or symn the function is an odd.
2fridZ 2vidy h a Angular momentum: M x M M v " — VriVaz — — — 2xiV — h / a \y I 2 a\ — or ay/ } h / sin 6 x a a ^ cot S cos d> a\ — I 2iri V ae 34. out. agrees with that obtained previously./ h / a \z h / — a\ J or ax at) d> I cos a cot fl sin d> a\ — I 2irt\ ae a<t>J h ' / a\ T X 2^A 7y.z) .z) ^Kw+W+6*) V(. z and polar coordinates Variable Position 6.x.V 7x)° ±/±) 2irt \a<t>/ M« = M I + Ml + Ml — j^_ 2 r_i_ a_ / ". <f>) Operator x v z x y z Linear momentum: h 2jr» a dx h_d_ Pv P. titx\ T) = h 2Vi [2 /rar\ nwx C° S VH \T) IT r. Since TYis the kineticenergy operator. since the potential energy was taken as zero. 4t Lsin e ae h* v ( m i a\ ae) a' l_ sin* e aa 1 a* 2 J Kinetic energy: a2 \ i^vl+vl+vl) Potential energy: V{x. such as re 2 /i 2 /8mo 2 in the example.„„. are the allowed This result. (36) TABLE 11 Some Quantummechanical Operators (In terms of cartesian coordinates x.y. One further feature of the nature of these operators must be pointed Sometimes an operator acts on ^ and does not lead to a number. y. the values of kinetic energies of the particle. momentum operator acting on the particleinabox Thus the wave function is .: a 22 INTRODUCTION TO MOLECULAR SPECTROSCOPY Here operation on ^ by Tx gives a quantity nW/Sma* n 2 /i 2 /8ma 2 times ^. P° _ / /2 W« Sm .y. times the wave function.
occupy the n = 1 level. ) nh = i. (37).INTRODUCTION TO THE THEORETICAL TREATMENT OF MOLECULAR SYSTEMS 23 is In such cases one is to understand that the variable being investigated not a constant for one of the allowed states. will be spectroscopic problems. + "lKtydr (38) this relation as One obtains the average momentum with „ = P " f" ( J2 jo ia 2 W« fa / . therefore. Sm . real functions. sin ax . the average value obtained according to the general rule / +" f*G+dT (37) + y*idr f_ For normalized functions the denominator is of course equal to unity.e. Similar applications of Eq. mcx\ h rmx a /2 (nir\ IT) 2S V« cos j U T C°S nvx dX . variable is Rather it has various of the For such functions. Jo _ nh ia? / a \ \2mr) titxI" _ a Jo (39) = The result of classical picture of the an average momentum of zero reflects the fact that the motion of the particle would ascribe both positive and negative values to the momentum since the particle would move in either direction. Exercise 13.. made in a number of The Boltzmann Distribution Spectra can be interpreted in terms of the energies of the allowed states if one can decide which of the allowed states will in fact be occupied by the molecules of the sample. In the particleinabox example the question that would have to be asked is: If some large number (frequently it is convenient to consider an Avogadro's number 6 X 10 23 ) of particles are placed in such a potential well. i. nwx a [" . the average value of a variable is. how many will behave according to the n = 1 wave function. n = 1 state using the operator Obtain the average position of a particle in a box in the method of Eq. values at various positions. or (38). For normalized. (38). 110. how .
mr ( 40 ) where k is Boltzmann's constant. 12.giNoe^^ ltT [Note that the number in each state at energy e. compared to the number with energy Ni=. There are. 1. however. the number of molecules with is energy «. by the relation . occupying some state with energy ts which is lower than a. (43) is still N e ( " . Most often one is interested in the number of molecules in a state with an energy t compared to the number in the lowest energy state to. equal to Nk.) state with The answer to such questions is given by the Boltzmann distribution which states that the number of particles N. The appropriate form of the Boltzmann distribution is then t Ni = JVoe<«*»* r (42) with the same energy. the number of states with energy and as e is usually the case. If it is more convenient to measure the energies of the states in calories per mole. i. Sometimes.)IRT t (41) where R. the appropriate expression is ^! = e (E E.. there are several states The population of the energy level is then cor Thus is if gt is the multiplicity of the ith energy «*.24 INTRODUCTION TO MOLECULAR SPECTROSCOPY many the n = 2 level. the lowest energy level a single state. 15 the tacit assumption that the lowest available levels would be occupied was made. respondingly greater.e. as was mentioned in Sec. occupying a energy n will be related to the number N. ^f = e(..380 X 1016 erg/deg molecule. and so forth? (In Sec. is 1.] .«.987 cal/mole deg. g { states at energy a. ° )/ * T . level.
Then the corresponding problem with molecularsized units and quantum mechanics will be studied. but chemically more interesting. It passing will be dealt with in greater detail. and Chap. which' large molecules show The rather "mechanical" treatment of simple vibrating systems in this chapter constitutes an introduction to these more complex. 21 b. First the classical behavior of a single particle and a pair of particles will be treated according to classical mechanics. like that of Fig. well be already familiar with the complicated absorption pattern. 4 will deal with the transition process that takes a molecule from one energy level to another. 25 . to absorption of radiation in the infrared spectral region. seems desirable.THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE The general introduction to the theory of atomic and molecular systems of the preceding chapter can now be applied to the particular case of simple vibrating systems. Chap. We will see that changes in the vibrational energy of the simple systems dealt with here lead. The chemist may in this region. In later chapters a number of points that are here mentioned in Thus. systems. in spectroscopic studies. The energy levels deduced in these calculations will then be levels indicated compared with the separation between these energy by spectroscopic transitions. 6 will recognize that gasphase molecules engage in simultaneous rotation and vibration and that it must be shown to what extent these motions can be treated separately.
The simplest classical vibrational problem is is that illustrated in Fig.e.0001 3.400 . to present some introductory material before these questions are raised. 21. ordinarysized. the greater the number of atoms the a molecule. the more ways the molecule can vibrate and the more complex infrared absorption spectrum.. The question that the particle of mass is asked is: What type of vibrational motion does the nature of the particle is m undergo? among other things.200 1.100 1.000 5. on many springs are such that if it The answer the spring. general.26 INTRODUCTION TO MOLECULAR SPECTROSCOPY however.500 l. 21 The absorption of radiation (b) in by (a) a diatomic is molecule and in a polyatomic molecule. Wave numbers 4. found that its removed a distance from ing force that is equilibrium position.000 1 1 1 1 800 1 700 1 62 100 I 80 60 x V • < 40 20 COin CCI 4 n 100 (a) ft 7 8 in 9 microns 10 16 Wavelength (b) .1.3O0  1. 22. particles which can be described by classical mechanics is very helpful when the quantummechanical behavior of molecules is encountered. It is clearly depends. The effect of gravity here ignored.000 1. be apparent from the next few sections that an understanding i. experiences a restor proportional to its displacement from the equilibrium the infrared spectral region In FIG.000 in cm 1 900 1 2. The Vibrations of a Single Particle (Classical) It will of the behavior of macroscopic.
. written explicitly in i. it gives the restoring from the equilibrium (2) so The minus Eq. (2) where is the measure of the displacement from the equilibrium position. and A. the displacement coordinate. which will also appear position. force for unit displacement sign. It should be appreciated that k. is a proportionality constant called the force constant.THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE is 27 position. that k will be positive. The displacement is from the in equilibrium position (a) measured from the equilibrium position and in (b) from the point of attachment of the spring. measures the stiffness of the spring. 22 The restoring force and potentialenergy function for a boll andspring system obeying Hooke's law. .e. A spring which behaves in this manner For such behavior one can write <* said to obey Hooke's / or x (1) f = — kx x. law. enters because FIG. in molecular problems. f is the force which the spring imposes on the particle.
where / = — kx. place the particle a distance dx potential energy U. is /.d dx to deal with the force / that the spring exerts more convenient particle and. Sometimes a problem is more conveniently set up in terms of the point than position of the particle measured from some other reference . In the particular case of Hooke's law. The statement of Eq. dU = kxdx If the equilibrium position is taken as that of zero potential energy.pP It is ii. (5) gives U = Px* This potential function is illustrated in Fig. the potential energy derivative is dU j. Before proceeding to the problem of the nature of the vibrations. it is important to point out that Hooke's law implies that the potential energy of the particle increases parabolically as the particle tion from the equilibrium position. is equal to the curvature of the potentialenergy function.= dx or kx (5) . measuring the stiffness of a spring. that. both corresponding to Hooke's law. on the one has / = /. (6) 2 shows. (6) is therefore equivalent to that of Eq. integration of Eq.28 INTRODUCTION TO MOLECULAR SPECTROSCOPY is as x increases in one direction the force increases but directed in the opposite direction. since it is this force that the applied force acts against. 22. k is equal to d* U/dx The force constant. (6) (2). since it will be encountered again . It is well to recognize here. PI)iied and this work is stored as Thus (3) dU = /.ppii.d and dU = or (/) dx (4) dU dx This is _ J an important and general relation between force and potential energy. later. moves in either direcThe work that must be done to disdx. as differentiation of Eq.
226 indicates. When this is done one obtains an —4f7r 2 v 2 m = —k (14) or 1 Ik 2t \m . Eq.+£One can note that. Newton's / = ma equation to — kx = mji d %x or — kx = mx to start with statements particle. since dT ax .VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE If. the length of the spring is I. (12) second derivative with respect to time cos (2tv< £ = 4tVA if + *>) (13) back into the equality differential equation. 29 the equilibrium position. % (8) It is often more convenient and potential energies of the about the kinetic One then writes .THE . is obtained. = mx and d (dT\ = mx dt\Tx) the same expression mx as set + kx = or — kx = mx of the (11) up by Newton's f = ma relation. (11).. Kinetic energy Potential = T = %mx energy = U = %kx 2 An equation equivalent to the / = ma equation that can then be used to is solve for the nature of the motion of the particle that of Lagrange i(S). has a solution x form (12) = A cos (2wvt + <p) The and correctness of this solution can be verified its by substituting Eq. The differential equation. is I as Fig. and the equilibrium length / the previous results are expressed as $k(l = k(l  k) and U=  l.)* (7) The equation One way of doing obtain describing the motion of the particle can this is to substitute into now be set up..
for simplicity.e. but will be seen that the quantum(14). 22. It shows that a held by particle with mass according to Eq.30 INTRODUCTION TO MOLECULAR SPECTROSCOPY This equation is the important classical result.. (12) with the frequency v given by Eq. is The displacement coordinate q defined as q = r  r. (14). the amplitude A of the vibration. Again it is asked What undergo? : is the nature of the motion that these particles FIG. move only along the line of the system. Only this frequency is allowed.. The quantummechanical result will differ it from this (not all vibra tional energies will be allowed). . be allowed to The particles will. which provides the counterpart of the diatomicmolecule problem. 23 The shape of the potential energybond length curve for a diatomic molecule. Consider the macroscopic system of a spring and two particles. i. mechanical solution retains a considerable similarity to Eq. The energy with which the particle vibrates can m a spring with force constant k will vibrate be shown to depend on the maximum displacement. The Vibrations of Two Particles Connected by a Spring (Classical) It is worth while treating one additional problem in a completely classical manner before we proceed to molecular systems.
problem X\ exist. can be used.. which will be important in later work.. the and potential energies can be written as %{m\x\ \k(x. = — 4x 2 i' 2 Ai cos + + 2 and Xz (19) = — 4jr 2 v 2 A. the use of Lagrange's equation.kA = kAi + ( 45rVw 2 + k)A 2 = .r 2 — — if x\) = (17) and m X2 + 2 k(xi X\) =0 Let us again see solutions of the type found for the singleparticle try. . will equations in two unknowns will be obtained.THE VIBRATIONAL ENERGIES OF If Xi A DIATOMIC MOLECULE 31 and Xi represent displacements of the particles of m 2 from initial positions in if equilibrium distance. Substitution of to be found. T = and + wi2X?) " (15) U =  Xl y Although one could solve the problem by writing/ = ma for each particle. Since the simultaneous motion of two particles is being considered. therefore. (17) gives. For each particle i. one obtains miXi — fc(. Two methods which are presented here because they introduce methods that encountered in molecular vibrational problems. With either Lagrange's equations or / = ma applied to each particle. be Direct Solution.dU . and kinetic mass wii and which the particles were separated by their Hooke's law is assumed for the spring. two of solution. one writes the Lagrange equation as d/dT\.. and (18) into Eqs.2 COS {2wvt <p) and Eqs. the frequency will be that of the for particles (2irvt 1 if a vibrasystem and must be the same £1 2.. on rearranging. We — A\ cos COS (2irvt + <p) ' and Xi (18) = A2 (2irvt + <p) where the amplitudes ^i and A% tional solution is may and </>) be different but where. a. will be illustrated here. (20) (4ir 2 v 2 m + l k)A t .
yields the equation (47rV) 2 wiim 2  4rVfc(mi +m + t) k1  fc 2 = (22) which has the roots v = (23) and 2~ \/~ 2t \h where n = ^ wi2 mi + ^ _^_ (24) The quantity p which is introduced here will be frequently encountered and is given the name the reduced mass. to a translational motion of the entire system. Xi sists of = x2 . The v = root corresponds. This is A 2 from these two equations to get a relation most conveniently done by recalling that for such linear nontrivial solutions for the A's will exist only if the is zero. and xi = — z2 . after Ay = A. (20) gives. (18). homogeneous equations determinant of the coefficients of the A's That is. Thus the motion corresponding to v — con displacement of the particles by the same amount and in the same direction. therefore. on expansion. one gets Ay v = — At. therefore. The motions that correspond if to these frequencies can be recognized the relation of the two amplitudes Substitution of v A\ and A 2 is found. (20) gives.s/k/y. Substitution of substitution of y. The motion corresponding to = 1/2*. (47rVmi + k) ( k —k — 4jr 2 p 2 m + 2 k) (21) This determinant. — in either of the equations of Eqs. v = l/2x \/k/n in either of Eqs. in view of Eqs. two natural frequencies for the system. can be generated. one of the frequencies being zero.32 INTRODUCTION TO MOLECULAR SPECTROSCOPY One can now eliminate Ay or for v. and Xi m 2 my (26) TO 2 = Xi If my the two masses are equal. therefore. There are. Ai = A 2 (25) and. after rearranging. = mjm 2/m\ + rn 2 and rearranging. by displacing the particles from their equilibrium position by .
Obtain the relation between A and A 2 for the two roots by comparing the appropriate minors of the determinant. and the problem can be set Q. (26) shows that the amplitude of particle 1 will be correspondingly greater than that of one of vibration and .J 6fc + ^l. Exercise 21. Xi = A = A j q (29) and Xi + mi + m 2 * These allow the kinetic energy to be written as r. or by forming the inverse transformation. + m z) 2 This expression must be converted to one involving q and X. x Solution in Terms of an Internal and a Center ofmass Coordinate. .*. and X is a measure of the displacement of the center of mass of the system. rather than displacements xi and x 2 one could choose to use the coordinates . will The motion that will occur will clearly be have the frequency v = 1/2t s/k/ii. (27) one obtains by rearrangement. Thus if Wi is less than w 2 Eq. «^.THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE 33 equal and opposite amounts. + J(. existence of a translational The mode and a vibrational one can be recog up in terms of coordinates which are especially suited to describe these motions. In terms of these coordinates one can immediately write down the potential energy as q is a Now particles U= The 4*(zi  xt)* is = ikq* (28) kinetic energy always easily written in terms of simple cartesian coordinates. (21) b. (26) that if the masses of the two particles are not equal. Thus. It should be recognized from Eq. — Xi — Xl (27) X = m&i + mzXz mi + m 2 measure of the displacement of the distance between the from the original equilibrium value. nized at the outset. From Eqs. the lighter one will move with a greater amplitude than the heavier one. of Eq.30) . Thus T = i(m. particle 2.
as that for Eq. two simultaneous equations were obtained. For problems of much complexity it is of great value to eliminate the translational and rotational motions before the vibrational problem is developed. t Alternatively one can write the solution velocity as X — const representing a uniform velocity of the center of mass of the of this procedure of recognizing the existence of system. and these had to be solved to extract the desired information on the vibrational motion. This can be recognized as identical to Eq. x's in By finding the inverse of the transformation matrix for the terms of q and X. therefore. as with a is a consideration of the nature of . the same we can write down the solution as q and = A cos (2wvt + <p) (32) and ' *4 for (33) The Lagrange equation and leads to (mi or X retains only the kineticenergy term + mi)X = X= A solution which is (34) formally like that obtained previously is X= A with v cos (2rvt + <p) = 0. The Potentialenergy Function for a Chemical Bond A preliminary to macroscopic mass and spring system. The advantage translational and vibrational modes initially is that one ends up with one equation which determines the vibrational motion and another which determines the translational motion. 23. the study of the vibrations of a molecule. (31) is. (11). Previously. (29). then gives i + + kq = (31) or nq kq = particle. obtain the expressions of Eq. The solution to Eq. applied q to the coordinate q. (11) for the vibration of a single Only a change in notation from m to n and x to q has occurred. Exercise 22.34 INTRODUCTION TO MOLECULAR SPECTROSCOPY The Lagrange equation.
Except for the case of H 2 . however. that will Fig. it has not been possible to calculate the energy of the molecule as a function of internuclear distance from the The interactions between the bonding electrons and the atomic nuclei. or The Morse curve with the values of curvature. that if a bond is stretched enough it will break. (The quantity we curves for H 2 as has been done for HC1 in Fig. of the potential well.e.and D e = 38. for example. by the relative and more exact ones be developed. and that bonds strongly resist compression. D e and j3 that are appropriate to HC1 is shown in Fig. W^ 1 .e. More particularly. and Morse Plot the harmonic oscillator. M. ideas. from the minimum of the curve. one asks about the change in potential energy of the molecule as a function of the distortion of the of the molecule bond from its equilibrium distance. The value of /? in the Morse equation can be conveniently calculated from the relation 0_. 24.. Exercise 23. as incompressibility of solids.. lead to a potential curve of the form shown in is revealed. to look for functions that solve the resulting differential equaThe pattern of allowed energies. 23 has been given by P. One knows. D e is the dissociation energy measured from the equilibrium position. for i. known in terms of e and /3. 23. is related to the vibrational level spacing as will be shown in Sec. for a diatomic molecule. 26. a molecule can be dissociated. 1 for H 2 for which <b e = 4. for example.395 cm. a simple and often convenient one that yields curves of the shape of that in Fig.2 k X 10 6 dyne/cm. D . U(q) He has suggested the relation D„(l =  e<") 2 (35) • where. q measures the distortion of the bond from its equilibrium length. i. Such qualitative Although no exact mathematical expression for the potential energy all molecules is known. and /3 is a constant any given molecule and can be said to determine the narrowness. general shape of the potentialenergy versus bondlength curve can.) A straightforward procedure for deducing the allowed vibrational would be to substitute the Morse function into the Schrodinger equation and. as before. tion. 23.THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE 35 the variation of the potential energy with the position of the atoms. in much the same way as was done for the squarewell energies problem. far be sketched. curves of Morse.310 cmDetermine = using 5.
+ (dU\ \dq).000 i 30. and therefore (dU/dq) g =o must be zero. a Maclaurin series expansion about q = q* „ gives U(q) = !7_. re Furthermore. 24 The harmonic oscillator (dashed line) and Morse (solid line) potentialenergy functions for HO. therefore.36 INTRODUCTION TO MOLECULAR SPECTROSCOPY could then be compared with the observed spectrum. If q is the displacement of the bond from equilibrium length. sets t/ g _o equal to zero.000(p  re y\ . and a simpler and more fruitful procedure is followed. particularly an expression for the potential energy near the minimum This suggests that a suitable expression for the potential energy can be obtained by a series expansion about the its minimum. at q — the potential energy is a minimum.000 10. 40. + (36) or q = 0.000 E u c i « e 3 20.<> 9 + 1 /dHJ\ 2\dq*) Q= . It will be seen shortly that the vibrations of a molecule result in only its small distortions of the bond from equilibrium length. In a study of the vibrational motion of a molecule interested only in in the potentialenergy curve.000 V= 134. If only the next higher term in the expansion is The usual choice of U= at r = FIG. Some mathe matical difficulties arise in this procedure. we are.
i. the parabolic potential function is used to approximate the potential of a chemical bond. the distortion of the distance between the two particles from the equilibrium bond length. (For a true Hooke's law spring. The force constant is then the value of d 2 U/dq 2 near the minimum. is When oscillator. This result should be recognized. (The solution functions for the two problems will are in fact similar.) . by comparison with Eq. correspondingly. The Quantummechanical Solutions of the Harmonic Oscillator The solution of the twoparticle problem by method b in Sec. the displacement of the single particle from its equilibrium position. The procedure should be recognized to be essentially that followed in the simpler squarewell problem. as identical to that which would have been obtained if it had been assumed that a chemical bond behaves as a spring following Hooke's law. ?& + (I *A+* = ^ ^ 8* 2n dq 2 fcQ. The Schrodinger equation for a single atom of mass m subject is identical in to a potential of U= %kx* is while that for the vibrations of a diatomic molecule with reduced mass ju is. U is a For a chemical bond the curve can be said to approximate a parabola only near the minimum. the system said to be treated as a harmonic 24. is equal to k at all values of q. valid near the equilibrium position. and those which be given as solutions to the har monicoscillator problem should be compared with those previously obtained for the squarewell problem. we have the approximation. (6).e.) Figure 24 shows the parabola that fits the Morse curve for HC1 near the minimum. It is only necessary to replace m by /i and x. The same situation is true in the quantummechanical problem. 22 showed that the vibrational equation that arises for the vibration of the twoparticle system form to that for a single vibrating particle. by q = Xi — X\. d 2 U/dq 2 parabola..2 V2 (39) ) The deduction of solution functions will not be followed through here but can be found in any textbook on quantum mechanics.THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE 37 retained.
38
INTRODUCTION TO MOLECULAR SPECTROSCOPY
It is found, as
would be expected from the similar squarewell probe and that these functions, and the values of t, are characterized by a quantum number. Thus, there are again certain allowed states, and each of these states has
lem, that solution functions exist for certain values of
a specified energy.
represented by
v.
The
It
integer that describes these states
0, 1, 2,
. . . .
is
usually
can take on the values
The
energies of the allowed states are given
by the expression
(' +
5)s>
where
v
=
0, 1, 2,
(40)
These allowed energies are shown, along with the potentialenergy function, in Fig. 25. The appearance in Eq. (40) of the same quantities as
in the classical calculation of
a vibrating system should be noted.
This
p
=3
i>
=2
v
=l
FIG. 25
The harmonicoscillator
wave
v
functions
and the funda
=
mental vibrational transition.
THE VIBRATIONAL ENERGIES OF
A DIATOMIC MOLECULE
39
relation to the classical
problem
is
sometimes made use of by writing
(
«.(»
where w
+
i)fo*
is
41 )
=
1/2t y/k/p
written for the frequency with which, according
(In to Eq. (14), the system would vibrate if it behaved classically. what follows, to will be used for a molecular vibrational property and v
will
be reserved for the frequency of radiation.)
The form
in Fig. 25.
of the solution functions, or eigenfunctions,
is
also
shown
Although little will be done with the solution wave functions themselves, it can be mentioned that each involves a set of series terms known as a Hermite polynomial. If one introduces a as
a
=
27r
^
h
v v
(42)
the solution functions can be written as
4> v
=
N el°<>'H
(V*q)
where
v
=
0, 1, 2,
v.
.
.
.
(43)
and
H
v is
the Hermite polynomial of degree
The normalizing
factor
2V„is
^ = (^^\ V2V.VV
"
J
(44)
For most purposes
functions
it
will
be enough to recognize the form of the solution
shown
it
in Fig. 25.
polynomials
first
may
For those unfamiliar with the Hermite be informative, however, to write down a few of the
solution functions.
*o
=
=
(i)'
/
1
~
e iaq'
ft
qeM V 2a("V K;)'
(2a?2
4>i
= y= (Y =
~
1)e_i "'
!
(45)
** fa
Vi (»)*
(f\'
{2aiq3
~ 
Zaiq)e
~ iaq '
= ^=
(4a 2 g 4
12ag 2
+
3)e*°"
!
The
spectral absorptions or emissions attributable to changes in the
vibrational energy of a molecule will
now be
discussed in terms of the
energy levels of Fig.
25.
40
INTRODUCTION TO MOLECULAR SPECTROSCOPY
25. Vibrational
Absorption Spectra of Diatomic Molecules
section has
The previous
tion that
is
shown that the allowed energy
by the
relation
levels of
a
vibrating molecule form a pattern of equally spaced levels with a separarelated to the molecular properties
A«
= ^,«. =
A^
(46)
The
energylevel pattern can be predicted
be expected to arise from such an if one also knows what transitions can occur with the absorption or emission of radiation and what energy
spectral transitions that are to
levels are appreciably populated at the
temperature of the experiment. Although the interaction of molecules with radiation will be dealt with in Chap. 4, some of the results that will be obtained can be mentioned here.
First of all, it will be shown that a vibrating molecule cannot interact with electromagnetic radiation unless a vibrating, or
oscillating, dipole
litatively,
moment accompanies the molecular vibration. Quaone can picture an oscillating dipole as coupling with the
energy can be exchanged between the
electric field of the radiation so that
This stipulation that the vibration be accompanied by a dipolemoment change implies that all homonuclear diatomic molecules, which necessarily have zero dipole moment for all
molecule and the radiation.
bond
lengths, will fail to interact with radiation
and
will exhibit
no
vibrational spectral transitions.
On
the other hand, a heteronuclear
diatomic molecule will generally have a dipole
the dipole
moment
and, generally,
moment
will
be dependent on the internuclear distance.
will, therefore,
The
vibration of a heteronuclear diatomic molecule
be
generally accompanied
can interact can thereby change their vibrational state to one of higher energy. They can, also, emit radiation energy and thereby change their vibrational state to one of lower energy.
by an oscillating dipole moment. Such molecules with radiation and can absorb energy of the radiation and
4,
One further transition restriction, which will be discussed in Chap. must be mentioned. Even for those molecules that have an oscillating
dipole
ble
moment
there
is
the further restriction, which
is
rigorously applica
only to harmonic oscillator type systems, that the vibrational
unit.
quantum number can change only by one
therefore, only
if
Transitions are allowed,
Av
= ±1
(47)
THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE
41
This restriction is an example of a selection rule. For absorption experiments, which are those most commonly performed, the applicable part of the selection rule is, of course, An = +1.
Now that the allowed transitions have been stated, it is only necessary to decide in which state, or states, the molecules are likely to be found at the beginning of an experiment. It will be shown that the
energy spacing between vibrational energy levels is typically of the order 13 erg/molecule or 3000 cal/mole. With Boltzmann's disof 2 X 10 tribution one can calculate the number of molecules in the.v = 1 state
compared with the number
"»»
in the v
=
state at
25°C as
_
/)
2X10 1S /298(l.S8X10 1 •)
=
Thus,
less
0.008
1
(48)
than
per cent of the molecules are in the v
still
=
1
state,
and
negligibly small numbers will be in
that, in experiments not
higher energy states.
It follows
much above room
temperature, the transitions
that begin with the
v
=
state will be of major importance.
The
selection rule
arrow of Fig. 25 to be
transition for
and population statements allow the vertical drawn to indicate the expected spectral vibrational
an absorption experiment with a heteronuclear diatomic
molecule.
found experimentally that such molecules do lead to the absorption of radiation quanta in the infrared spectral region and that this absorption can be attributed to the vibrational transition. Thus, HC1
It is
absorbs radiation with
v
v
=
2,890
8.67
cm"
1
or
cycles/sec
is
=
3
X
hv
10
10
(2,890)
=
X
10
13
The energy
Ae
of the
quanta of this radiation
=
= =
v
6.62 5.74
X X
10"(8.67 10 13 erg
X
10 13 )
.
(49)
It is this
energy that must correspond to the energy difference between
the
v
=
Ae =
and
h
=
1 levels, i.e.,
to
lie
(50)
equating these two results for Ae and using
Mhci
=
We
+
1.627
Wei
X
10 24 g
42
INTRODUCTION TO MOLECULAR SPECTROSCOPY
one obtains
k
=
4.84
X
10 6 dynes/cm
(51)
In this way it is seen that the spectral absorption is understood terms of a molecularenergylevel pattern and that new quantitative information about the molecule is gained.
in
The frequencies of radiation absorbed by some diatomic molecules and the deduced force constants are shown in Table 21. One should
recall that the force constant
measures the
stiffness of the
bond and
should see that the force constants of Table 21 are to some extent understandable in terms of qualitative ideas about these chemical bonds.
It should be mentioned finally that it is not customary to convert from radiation frequency to quantum energies in comparing spectral
absorptions with energylevel patterns.
It is
more convenient
i.e.,
to use
the conversion factor h between energy and frequency,
Ae
=
hv,
and to convert the energies
of allowed energy levels to frequency units.
TABLE
21
Frequencies of Fundamental Vibrational
Transitions v
=
to v
=
1
and the Bond Force Con
stants Calculated
from These Data
(Values for homonuclear molecules are from
spectral studies)
Raman
Molecule
p(cm _ ')
4,159.2 2,990.3 3,958.4 2,885.6 2,559.3 2,230.0 2,143.3 1,876.0 892 556.9 321 213.4 1,556.3 2,330.7 246.3 157.8 378 278
k (dynes/cm)
H D HF
2 2
HC1
HBr
HI
CO
NO
F,
CI*
Br*
I*
o.
N
a
lis
Nas
NaCl KC1
5.2 5.3 8.8 4.8 3.8 2.9 18.7 15.5 4.5 3.2 2.4 1.7 11.4 22.6 1.3 1.7 1.2
.8
X
10 5
THE VIBRATIONAL ENERGIES
OF A DIATOMIC MOLECULE
43
Thus the energy
levels of the
harmonic
oscillator
can be written as
(• +
<v
s)s^
2) 2^
ergs
(52)
=
(
»
+
J
cycles/sec
(53)
The symbol
are written as v
I
will
be used for energies expressed in units of
cm1
to be consistent with frequencies or
if
expressed in
cm1
.
quantum energies With the last of
of radiation that
these expressions
one can compare the energy separation between two levels directly with
the
wave number
P
of the radiation absorbed or emitted in a transition
i.e.,
between these
Exercise 24.
levels;
At will be the same as
v for
the radiation.
compared
to the
Calculate the numbers of molecules in the v = 1 state number in the v = state at 25 C C for Br 2 which has the
,
rather small vibrational energylevel spacing of 323
cm 1 and
,
for H2,
which has a vibrational spacing of about 4,159
molecule.
cm 1
,
the largest for any
26.
The Anharmonicity of Molecular Vibrations
That the harmonic
oscillator expression
U=
\kq*
is
only an approx
imate representation of the correct potentialenergy function shows up in the presence of weak spectral absorption corresponding to Av = +2,
+3,
.
.
.
,
in violation of the Av
= +1
= +1,
selection rule,
is
and
in the fact
2, 3,
.
.
that the frequencies of these overtone absorptions
not exactly
.
times that of the fundamental, Ay
absorption.
The
One
retention of one additional term in the Maclaurin series expan
sion of Eq. (37) provides a better approximation to the potential energy.
has, then,
This expression for the potential can be used in the Schrodinger equation
to deduce the energy levels of the allowed states of the
oscillator.
anharmonic
tion,
The solution is obtained by an approximation, or perturbamethod and leads to an energylevel expression that can be written
as
«.
=
«.(»
+ *) 
^J.
(v
+ i)
J
(56)
14. therefore. .000 £ 6. It should be recognized that the broadening out of the potential curve. confines the FIG.000 10. as a bond stretches. Pattern (a) is based on the transition. levels. harmonicoscillator approximation and the frequency of the fundamental Pattern (b) is based on the anharmonic values of Table 22. or equilibrium.000 < B 8.000 Harmonic (a) Anharmonic (b) .000  2. the spacing of the if the potential curve were a parabola with the curvature that the actual curve has at the minimum. It is With the minus sign than the principal term £> e written explicitly in Eq. that would occur quantumnumber term always is known as the anharmonicity constant. This quantity w e is.000 12. 23 and 24./\i where = (d 2 U/dq 2 ) q „o. the separation between successive vibrational levels is not constant but much less rather decreases slightly with increasing values of v. 26. (56) it is found that w e x e is always a positive The energylevel expression of Eq. The coefficient oieXe of the squared energy expressed in cm1 . shown in Figs.000 4.44 INTRODUCTION TO MOLECULAR SPECTROSCOPY o> e Here fc 8 has been written for convenience in place of 1/2tc y/k. 26 The vibrational energy levels calculated for HCI. as shown in Fig. (56) shows. position. therefore. that quantity.
885.THE VIBRATIONAL ENERGIES OF A DIATOMIC MOLECULE 45 vibrating particles less closely than would a parabolic curve. levels. that Eq.6 14. of the higher energy shown one observes some of the overtone bands. and so forth.e. (57) provides a very satisfactory to the observed frequencies of HC1. i.70 5...6 13. therefore.923.60 cm1 .347.90 cm1 and fit u^x e = 51. more intro The anharmonicity term duces.988. i(0) The distinction is that &>« is a measure of the curvature of the potential v curve at the very bottom of the curve. as If v an effect which decreases the spacing in Fig.5 (2.9 5.8 8.7 11. One notices that «„ is considerably larger than the quantity «(1) — which would have been identified with the coefficient of the (v f. four overtones of HC1.668.20 8. It follows that the force constants calculated from these two quantities will be different. 26.988.5 2. (56) for the energies of the transitions from v = to v = v.90 cm K = 5.885. and Table 22 shows the data obtained for the fundamental and first These data can be compared with those derived from Eq.396.543.396. v = 3. transitions from one can check the success of the and determine the constants o>.50 10.e. for a>„ = 2.9) 5.429.347.1 13.5 .1574 X 10 6 dynes/cm Comparison (v TABLE 22 lator Frequencies of the Vibrational Transitions of HCI.0 8. (56) cteX.771.^) term in the expression based on a harmonic potential. Such loosening of the restrictions on the motion of particles always leads to closely spaced allowed energy levels. i(v) — e(0) = ii e (v)  WeX^iv + 1) (57) One finds.885.60v(v + 1) Soto cm * Av Description fobs cm ' Harmonic oscillator Anharmonic oscillator 0> 1 Fundamental First overtone 0>2 0»3 0—4 0»5 Second overtone Third overtone Fourth overtone 2. The v harmonicoscillator approximation takes the difference in energy of the = potential curve u. of the Observed Frequencies with Those Calculated from the Harmonic Oscil Approximation and with Those from the Anharmonic Expression — «o= 2.988.668. 1 levels as a measure of the curvature and therefore gets a lower value.923.90(v)  51. where a hypothetical = — ^ level of the would be. Thus for HC1 v and = = 2. to v — = 2. i. energylevel expression of Eq..657.0 10.
" chap.900. stretching vibration of The fundamental and overtone frequencies of the CH CHC1 3 are reported by Herzberg to be at 3. What value of u e can be deduced? PRINCIPAL REFERENCE 1.90 cm 1 fc = 4. 11.300. 13.J. Inc.806 X 10 6 dynes/cm is. 5..: "Spectra of Diatomic Molecules.860. Fit these frequencies with a suitable equation that allows for anharmonicity.. Exercise 25. 8. G.46 INTRODUCTION TO MOLECULAR SPECTROSCOPY while «(1) of i(0) Jj e = 2. 1950.700.315. For polyatomic molecules. instead of the observed fundamental energy in a number not more satisfactory.885.019. cm'. but its use requires information on the overis tone frequencies. Princeton. . The use of ways. N. Herzberg. Van Nostrand Company. D. this information always available. and 16. 3.
is A treatment that assumes that the molecular dimensions are independent of molecular vibrations and undisturbed by molecular rotation known 47 as the rigidrotor approximation. if what molecular property can be deduced from a study of the frequencies of radiation emitted or absorbed. also quantized. or emission. of radiation in the infrared spectral region and that information on the frequency stiff of the radiation absorbed. A final section will consider the slight . or emission. can be used to determine the ness of the bond holding the atoms together. No attempt be made in this introductory treatment to allow molecules undergo simultaneous rotation and for the fact that the vibration. classical It again convenient to consider the behavior of a linear mass system and then to see the difference is in behavior that imposed when the system is of molecular dimensions and the quantum restrictions will become important. Although diatomic molecules provide most molecules. is It is now natural to ask whether the rotational energy of a molecule of radiation. and. the theory that is is of the examples of linear first developed is equally applicable to linear polyatomic molecules. whether changes in the rotational energy can lead to the absorption.THE ROTATIONAL ENERGIES OF LINEAR MOLECULES In the previous chapter it has been shown that changes in the vibrational energy of simple molecules can lead to absorption. or emitted. so. The nature of the rotational energies of molecules can be introduced by a consideration of the particularly simple systems of linear molecules. The rotational effects will be considered by themselves.
is v/2vr or <a/2v. mi . Such a treatment is said to deal with the nonrigid 31. usually a second.G. For a particle moving with a linear velocity v a distance r from the center of gravity. the value of / can be worked and then applying Eq. of the angular velocity It will be recalled that a> is defined as the number of radians of angle swept out in unit time by the rotating system. of inertia of a The moment I system is defined as = n Y. for instance.. C. rotor. For a diatomic molecule. mr ' i (2) where out by is the distance of the ith particle from the center of gravity of locating the center of gravity the system. (i) It follows also that the the dis tance traveled per unit time divided by the circumference of the particle path. 31 Location of the center of gravity of a diatomic molecule. are given by v/2vr and the angular The rotation of <*> velocity by 2 = '(^H number of revolutions per unit time. The Rotation of a Linear System (Classical) any system is most conveniently treated in terms and the moment of inertia /. the masses and distances of Fig. first For any particular system.48 INTRODUCTION TO MOLECULAR SPECTROSCOPY modifications that occur as a result of the centrifugal distortion the rotation produces. (2). 31 lead to the center of gravity being located so that FIG.e. i. the revolutions per unit time.
_ = mfm 2 (nil +m 2 2 2) . (m t m +m main\ y m\ — 2) 2 r2 mim 2 mi iur 2 +m j r2 2 .y?. Eq.126 and 1. . respectively. more conveniently treated in terms of the angular One then can write the kinetic energy T as =  £ m. . which which have an equilibrium bond length of 1. No restriction. the particle velocities are velocity w of the system.m7 +^r 2 kinetic energy of a system of wtim 2 («) The given by moving macroscopic particles is £ i ^m. for Application of the denning equation. = where m n.275 A. since r x OF LINEAR MOLECULES 49 +r = m n = mi + m r 2 2 : r. and is (b) and has N—N and N— bond lengths of 1. If the system performs a rotational motion.THE ROTATIONAL ENERGIES or. the moment of inertia now gives . DC1 36 2 . (4) is the reduced mass. defined as in the previous chapter as . It can now be expected that when the same problem is treated for a molecular will appear. system some restrictions on the allowed rotational energy Exercise 31. furthermore. to the relations 2 and r2 (3) = mi mi + r r wi2 (2). Calculate the linear moment of inertia of (a) HC1 85 HC1 87 and . is placed on the angular velocity with which the system can rotate. all of N 0.191 A.coV? = t WX i "W? (6) = i/" 2 Equation (6) shows that a system of particles that obeys classical mechanics has a kinetic energy that is dependent on the moment of inertia of the system and on the angular velocity.
50 INTRODUCTION TO MOLECULAR SPECTROSCOPY 32. is (9) by solution of the Schro dinger equation. can be obtained by solving the Schrodinger equation for a rotating linear system. the quantized angular momentum 0. (8) where the quantum number. the wave functions and energies of the allowed states. The Rotation of a Linear System (Quantum Mechanical) Although a complete quantummechanical description.e. The angular momentum J) (rnivdu i of a rotating system is which. i. 2.. has been designated by J. \/J(J + 1) ^> . following customary notation. obtained. term. This stipulation leads to the allowed rotational energies of 'Si* The correct result. if J an integral quantum number. angular momenta given by rather than in early by the simpler form J(h/2w) introduced by Bohr and used If this quantummechanical treatments. . for example. in view of the definition of I as ^niirf and w as Vi/r i} is Thus. first. being (10) ej = ^jJ(J + D The is correct quantummechanical treatment turns out to allow. condition requires that Iw = j£ with/ = . it is. only slightly different. 1. at more informative to see what result is obtained by applying Bohr's defined as condition that the angular momentum be quantized in units of h/2w. .
are shown schematically in Fig. Solution of the molecularrotation problem. orbitals. Any textbook on quantum mechanics can be consulted for such a derivation. 33. (10). potentialenergy function can be set equal to zero. These allowed energies. that they are identical to the angular part of the solutions for the behavior of the electron of the concern us. for the resulting differential equation only if Solutions are found the energies have the values indicated in Eq. It hydrogen atom.= sAc The group ^ +i > cm1 is (11) of terms h/2nr i Ic sufficiently frequently encountered in FIG. and hence derivation by application of the Schrodinger equation will not be carried out here. (10). 32 Polar coordinates. using the convenient polar coordinates For a free gaseous molecule. in these units we have the expression . For J = 1. Again it is more convenient to express these energies in wavenumber units. tional states allowed For spectroscopic work. the rotational wave function ^j_ has the same form as the angular part of an s orbital wave function. The wave functions themselves need not now can be mentioned. however. as given by Eq. obtained. (10). rotation is unhindered. Thus for J = 0. the correct quanis tummechanical result of Eq. . of Fig.THE ROTATIONAL ENERGIES is OF LINEAR MOLECULES 51 introduced for Iu in the classical energy expression. The problem consists of setting up the Schrodinger and the equation for three dimensions. the wave function is the same as that for the px py and pz . . the expression for the energies of the rotaby the quantum restrictions is of prime importance. 32. and so forth.
= J&/t7 + 1J energies 30E U 20B 3 12JJ 6B 2B OB .1 molecule can now be written as allowed rotational energies of a linear Again the barred symbol lj = BJ(J + 1) cm 1 (13) FIG. each rotational level is The degeneracy. 33 The energylevel pattern for a rotating linear molecule. and one uses the symbol defined as B B = h Sx^/c em~ is . Allowed rotational 2/ Energy pattern if I degeneracy + were removed (schematic) Multiplicity «. or multiplicity.52 INTRODUCTION TO MOLECULAR SPECTROSCOPY molecular spectroscopy to merit a symbol. of if indicated by the number of levels that would appear an electric field were applied to the rotating molecule. (12) The used to emphasize the units of cm.
kT. to be 66. that ponent of the angular momentum in a direction imposed on a rotating molecule he quantized in units of h/2ir.THE ROTATIONAL ENERGIES OF LINEAR MOLECULES 53 33. J . /! * w (14) be seen later that the allowed rotational energies typically have small energy spacings compared to the roomtemperature value of For example. 0. Now eral. 34 as a function of 7 in Fig. (11). 12. This can most easily be seen from the condition that the comIt follows. —7 times h/2r that would appear if an external field were applied. (10). as with the hydrogen atom treated in Sec. 7 — 2y. 2 as suggested in Exercise 31. or (13).— 7 + 1.. there are. In view of this 27 + 1 degeneracy of the 7th energy level. The moment of inertia can be calculated.8 X lO40 g cm 2 and . The relative populations of rotational energy levels of 25 r N2 at 25°C. The number 27 + 1 from the states with angular momentum components of 7. 33 indicates. N has an N— N bond length of 1.J < J+1 » 000202 > is shown (15) The population of the levels of N 2 FIG.191 A. —1. there are arises 27 + 1 states with energy BJ{J + 1). Rotational Energylevel Populations The next tional states step in laying the foundation for an understanding of rotational spectra involves a discussion of the populations of the rota whose energies are given by Eq. 34. AV^o 10 15 20 25 30 35 40 45 Rotational quantum number. as Fig. .. 7 — 1. — 2. at 25°C the energylevel population expression is Nj = (27 + l)JV e. the Boltz mann distribution expression for the populations of the rotational energy levels is Nj = It will (27 + i)JV e"'+ 1 >* . in gen a number of rotational states corresponding to a given rotational energy.126 A and an N— O bond length of 1.
35 can be obtained from the The allowed energies given in Eq. (13). we must expect transitions starting from energy levels corresponding not only to J J values to be important. 34 34. the experiment that is invari when rotational spectra are obtained. absorbed radiation. previous section. that quite a is few of the allowed rotational generally valid. Make J value diagram like Fig. The requirement is seen to be equivalent to that stated in the previous chapter where the vibration of a molecule had to be accompanied by an oscillating dipole for such interaction to cule occur. the energy difference in cm + . Rotational Spectra of Rigid Linear Molecules be obtained It is again necessary to anticipate the results that will in the following chapter when the transitions that electromagnetic radia tion can induce will be studied. The selection rule for rotational transitions of linear molecules that occur with the absorption or emission of electromagnetic radiation is AJ = ±1 For ably done (16) studies of the absorption of radiation. and the population results obtained in the With this selection rule. The first requirement for the absorption of radiation by a rotational energy transition is that the molecule must have a permanent dipole. the expected transitions can be indicated by the vertical arrows of Fig. The rotating dipole can be thought of as interacting with the electric field of the radiation and allowing energy to be exchanged between the moleand the radiation. in result of this example. Exercise 32. general expression for the energies that are absorbed in the transitions indicated by arrows in Fig.BJ(J + l) ( 17 > . 35. If the quantum number for the lower level involved in a transition is J and that for the upper level is 1 and therefore the value of f for the J 1.54 INTRODUCTION TO MOLECULAR SPECTROSCOPY The that. v is = B(J + i)(7 + = 2B(J + 1) 2) . There is a further restriction on rotational transitions even for those molecules which have dipole moments. the appropriate part is of the selection rule AJ = +1. It follows energy levels are appreciably populated. for = but also to higher HC1 a population versus at 25 and 1500°C. an absorption experiment.
. . Table 31 shows the frequencies of absorption. since the levels J — 0. and should therefore observe a set of lines spaced by a constant amount. . 2. The symbol B used to represent the quantity h/87T 5Jc. absorptions of radiation at frequencies of 2B. observed for the Unear molecule OCS. will generally 4JB. &B. This amount can then be identified with 25. . Energy — 30B Ai=105 Spectrum . 35 Rotational is energy levels and transitions for a rig id rotor linear molecule. One observes in the microwave spectral region such series of absorption lines when linear molecules with dipole moments are studied. populated. 1. be .THE ROTATIONAL ENERGIES OF LINEAR MOLECULES 55 We thus expect. FIG. attributable to rotational transitions. .
l>2 2* 3 .92 36. one deduces that in the OCS molecule r(CO) r(CS) = = 1.162.001 A A (21) one will discover. and the assumption that the bond lengths are independent of the isotopic species.08 4>5 * From Wentink.. 75: 270 (1949).4 lO" 40 g/sq these momentofinertia data. Phys.559 ± + 0. C 12 S 32 Thus.0 X X 10. 1 Rev. since B = h/STr 2 Ic.202864 em" 5 From i 6c i2 S 34 and and / = = 138.64 3>4 60. Wentink.7 megacycles/sec and S = 12 16 7 2( 2 9 9 790 X Z) = ° 20285 ^ of inertia of the (18) Therefore.40 g/sq cm cm (20) = 0. and Strandberg.. Rev.197910 cm 1 7 0l6c i 2s 34 141.651. and Kyhl [Phys. more detailed analyses and 16 C 12 S 34 reported by Strand berg. as from the moments of inertia. what is done whenever possible is to obtain the rotational spectrum of different isotopic of the rotational spectra of 16 species.82 48.56 INTRODUCTION TO MOLECULAR SPECTROSCOPY From line these data one can obtain an average value of the absorption spacing of 12.325.40 g cm 2 (19) should be apparent that from this single experimental result it is not possible to deduce both the CO and CS distances of the molecule.488.001 0.0 X 10.161 1. the moment OCS mole cule can be calculated to be /ocs It = 138. In such cases. to verify that these distances are conmomentofinertia data than to derive the bond lengths the sistent with It is easier. Measurements of the spacing of rotational spectral frequencies have Absorptions of TABLE 31 the Microwave Region Due Transitions* 16 C 12 S 32 in to Rotational Transition Frequency v J — J + > 1 (megacycles/ sec) 0» 1 24.814. Kyhl. 75: 270 (1949)] give i6 cl 2 8 32 5 0l6cl2s n = 0.
315.40 megacycles. The constant spacing expected from the treatment given here is that obtained from the rigidrotor approximation. and can be expected to stretch under the influence of the centrifugal forces that aris&when they rotate. in fact. and R. Some Bond Lengths Obtained from wave Spectroscopy of Linear Molecules* TABLE 32 Micro Molecule Bond Bond distance (A) 1 1282 2. Close examination of data such as those of Table 31 reveals that the spacing between adjacent spectral lines is not.629 1. but are flexible.97 megacycles and for N Deduce moments of inertia for these molecules.064 1. F. in many bonds of molecules. Gordy.057 1. The Nonrigid Rotor The stretching effect of the centrifugal forces on the bond lengths. Now we can see (although this is a minor matter that can be passed over) what alterations to the theory arise from the recognition that molecules are not rigid. . and Trambarulo [2] for HC 12 14 as 44. Inc.632 1.157 .207. W." John Wiley & Sons.382 1. calculate the H C and C N bond lengths. and therefore on the moment of inertia of a rotating system.203 1.163 1. Smith.052 1.211 1. Tram"Microwave Spectroscopy. Smith. for the lengths of the moments of inertia and. barulo. CO NaCl C=0 Na—CI HCN C1CN HC=CC1 HC^C—C=N C— C=N C—CI C=N C— C=C c— C— c=c c— C=N * From W. The rotational constant term B = cB is given by Gordy. can first be calculated on the basis of classical behavior. Some of these results are shown in Table 32. exactly constant. DC N 12 14 — — *35. New York. Exercise 33. 1953.3606 1. as 36. V.156 1..H c a THE ROTATIONAL ENERGIES OF LINEAR MOLECULES 57 led to very accurate values for the cases. for simplicity. Consider. Assuming that the bond lengths are independent of isotopic substitution.
with Eq.58 INTRODUCTION TO MOLECULAR SPECTROSCOPY a single particle of mass velocity of u. to r by means of Eq. the particle a distance r from the fixed point and that this length increases to a value of r when the particle rotates. the total energy can be written as e = £Jw 2 + ik(r  r„) 2 (24) which. The k(r centrifugal force of — r ) stretching www 2 is balanced by a restoring force of that accompanies bond stretching. In this way. given by the expression k(r — r ) = www 2 (22) On rearrangement. minor term of Eq.J^'V + 3ar*mVJIb l)' (27) . becomes t = 2 2 Iu> + + 2~¥~ = I 7o)2 I(W 2 fcr 2 (25) The quantum tized according to restriction that the angular momentum I<a be quan y/j(J + 1) (V2t) will convert this classical result to a quantummechanical result. (23). therefore. (26). and approximating r by r in the second. deduced to be The correct allowed energies are. is m rotating that. (22). = which ^^ of + 1) + 32^^ + 1)2 (26) It is finally necessary to relate the distorted distance r in the first term. (26). one obtains is tj = '(' + ^D gSSFT^ "2 ' . this gives the distorted bond length as r = _Jr° k — rrua* is (23) Since the energy of the rotating system made up of kinetic and potentialenergy contributions. The extent of bond that occurs for an angular velocity of w is. Assume when about a fixed point with an angular there is no rotation. major importance in Eq. therefore.
36 distortion The effect of centrifugal shown by the comparison of the allowed rotational energy levels of a rigid (solid lines) and nonrigid (dashed lines) rotor. as Fig. B = h/WIc. Rigid rotor Nonrigid rotor . and the and nonrigid rotor treatments can. this result becomes = BJ(J + where o> 1)  ^ J*(J + 1)* (28) is the vibrationalenergy term l/2xc s/k/m and. as before. It is customary to write this expression for the rotational energy levels as tj. THE ROTATIONAL ENERGIES OF LINEAR MOLECULES 59 In terms of wavenumber units. according to the above by the derivation. is a quantity that can be evaluated from spectral results and. known as the centrifugal distortion constant. = BJ{J + 1)  t)J\J + 1)2 (29) where D. 36 less much J 8 — ft 5 FIG. can be related to other spectroscopic molecular parameters relation D = difference 4B>/tf of (30) The value D is always very between the rigid than that of B.
: "Spectra of Diatomic Molecules. 1950. N. 1950. Van Nostrand Company.0004 em') 83.30 145. Van Nostrand Spectroscopy. according to Eq. The expression for the energies of the rotational transitions AJ = +1 for a nonrigid rotor are. Inc.75 124.' Inc.86 (with = 2B(J + 1) B = 10.08 144.J. Herzberg. New York. W.40 124.30 226. Table 33 illustrate the improved to experimental for centrifugal absorption frequencies that can be obtained by allowing PRINCIPAL REFERENCES 1.. 3. F.03 165. D." D. suggests. .1 124. John Wiley W. and R. 1953.94 206." Company. Princeton.55  11 206. Gordy.34 cm"') 82.395. V. G.. be expected to show up only when high J value levels are involved.06 103. "Spectra Diatomic Molecules.J.. Trambarulo: "Microwave & Sons. Herzberg.03 104.38 226. Smith..80 227..98 165.51 185." chap. D = 0. Ai(J » J + of 1) = 2B(J + 1)  4D(J + l) 3 fit (31) The data distortion. (29).50 of From G.4D(J + l) s = 10. N.72 103..76 165.„ J+J + 3>4 4>5 5>6 6>7 7»8 8* 9 9> 10 10 * 1 (cm" 1 ) 83.60 INTRODUCTION TO MOLECULAR SPECTROSCOPY Absorptions Due to Rotational Transitions of TABLE 33 Transition HQ* •\>b» »„.48 Pcalc (B = 2B(J + 1) .50 185. Princeton.39 144. 2.. Inc.12 206.44 186.
system and. that transitions between allowed energylevels can occur with the absorption or emission of electromagnetic radiation. is and illus known as the timeinde A more complete expression for Schrodinger's equation allows the calculation of atomicscale phenomena as a function of time as well as of space. an expression that can often be used in a qualitative. or diagrammatic. selection rules governing the transitions that an thus occur have been quoted. Furthermore.) The Timedependent Schrodinger Equation trated The form previously given by the particleinabox for the Schrodinger equation. it is approprie to investigate the theoretical basis >e understood. the intensity of the absorption or emission will be observed.4 THE ABSORPTION AND EMISSION OF RADIATION In the two previous chapters it has been necessary to state. (The reader will notice that in the previous chapters the emphasis has been on the frequencies of absorption bands little attention has been paid to their intensities. calculation. moreover. For one dimension the equation pendent. Now that some introduction to lowed energy levels and to simple spectra has been given.t) U(xn(xt) _ _ j_g*M m 61 . written for x and t is _ v a*9(z. tion is. A considerable amount on which spectral transitions can of manipulation with quantum mechanical expressions will be necessary. The net result of the deriva however. without adequate explanation or proof. way to understand what transitions will occur in a particular band that and 41. or amplitude. equation.
^(x.e. gives &*&*> + "»«**«>&*(*!& or '(x) [ <Kx) L 2 (3) 8ir m dx 2 J dx* + ^ 17(35) v ' =  h 1 <*>({) d*(<) dt x. 2ti The x and since t.. the solution <t>(t) is = e< 2 "'»«>' (6) € where «„ is written because the values of that are required n. with the supposition that U a function only of x.« The will INTRODUCTION TO MOLECULAR SPECTROSCOPY explicit indication that It will * is a function of x and t. (1). and . (5) The solution . The constant is called e with the energy of the system. gave information on the position of the particle. is a function only of and the right side is a function only of only if each side is it can be identified The equation can be valid for aH values of equal to a constant. * is a function of x and and that ^ is a function only of Previously it was shown that 4/*^ described the system. it is customary to try to separate them two equations each involving only one of the variables. t be understood that x. Substitution is of this relation in Eq. in the onedimensional squarewell potential. not be continued. In a similar way. involve a quantum number which can be denoted by by Eq. + U(x)f (* A 2 dV &w 1mdx i = e+ (5) The second of these equations can be recognized as the timeindeThe timedependpendent Schrodinger equation introduced in Chap. The two (3) equations that are obtained by setting Eq. i. *** describes the system in space and time. 1 ent part of the complete Schrodinger equation can be readily solved. respectively..e.t). i. With this tries the general substitution (2) approach one ¥ = where *(*) «(f) 4>(x) and <f>(t) are functions of x and t. left side of this equation t. To reduce tion to into equations such as the timedependent Schrodinger equa more manageable forms. equal to the constant * are 4jja and «*«.
THE ABSORPTION
AND EMISSION OF RADIATION
63
to the complete equation for a general onedimensional problem can,
therefore, be written as
*„(*,*)
If
=
*„(:r)e< 2 "««»'
(7)
solutions that
one thinks of the particleinabox problem as an example, the would have been obtained from the timedependent Schro
dinger equation would have been
*,(x,0
=
J^in^U'
nn
8ma 2
2 "'")'"'
where
(8)
The implications of the solution in x and t can be seen if the quantummechanical description of the system, as given by **'4 is obtained.
r
,
This
is
given in general by
\j/*\£
= =
^*^e (2Ti "
,)
'» <
e
_(2 '' M) '"'
*V
of the solution to the space
is
(9)
The timedependent part
tion
is
and time equaSystems, like
seen, therefore, to
drop out, and the system
again described
simply in terms of the timeindependent wave functions.
the particleinabox system studied in Chap.
1,
which lead to solutions
that are time independent, are said to have allowed stationary states. These are the allowed states that are obtained as solutions to the timeindependent Schrodinger equation.
For a spectroscopic process to occur, a procedure must
time dependence
is
exist
which
causes a state to change from one stationary state to another.
How
a
introduced into a quantummechanical problem will
now be
42.
investigated.
Induced Quantum Transitions
Much of spectroscopy is concerned with the absorption of radiant energy as a system goes from one stationary state to another under the
influence of incident electromagnetic radiation.
tigate, therefore,
It is necessary to inves
how
radiation can disturb, or as one says perturb, a
system so that a transition is induced. In general any atomic or molecular system will have many allowed energy levels corresponding to the stationary states of the system. It
64
is sufficient,
INTRODUCTION TO MOLECULAR SPECTROSCOPY
and the notation is simplified, if a system is treated that is assumed to have only a lower energy state, with a wave function and an upper energy state, with a wave function m where I and m are two different values of the quantum number for the system. The complete time and space Schrodinger equation for one dimen\[/i,
\f/
,
sion
is,
as before,
8w 2 m dx*
+
'
= U(x)* K '
 ^.^2wi dt
symbol H,
(10)
It is convenient to introduce the
called the Hamiltonian, to
represent the terms
h*
d2
8w 2m dx 2
(Although
that
it is
+
,
TT , U(x)
,
it is sufficient
to treat
H
as a symbol, one should recognize
the quantummechanical operator for the total energy of the
system.)
With
this notation,
Eq. (10) becomes
The
$1
supposition that only the two stationary states described
by
and
yp
m exist for the system implies that the only two solutions to
Eq. (11) are
and
(12)
Since such solutions to the Schrodinger equation, as pointed out in Sec. 17, form an orthogonal set of functions, a general solution to the equation
can be written as a
series of these functions.
In this case, where only
two solutions
exist,
the general solution
is
simply
(13)
* =
t
a&i
+
am ^r m
where a and am are weighting coefficients which are not functions of x The two allowed stationary but, as we will see, can be functions of time. states of the system are described by the general expression with Oj = 1,
am
=
or a
t
=
0,
am
=
1.
The potential energy of the system before it is disturbed, or pert turbed, by electromagnetic radiation is now represented, more specifically,
THE ABSORPTION
AND EMISSION OF RADIATION
i.e.,
65
by
Uo,
and the corresponding Hamiltonian by Ho;
H '£a£' +
as
ff
™
case
is
(14)
The Schrodinger equation for this initial, unperturbed
then written
"*aSStwo solutions fy and
ty m
.
(15)
It is this equation that has the
Now
let
us assume that
I t
it is
the lower energy state
initial state,
0(
with the
known that initially the system is in wave function ¥j and energy ei. The
by Eq.
(13),
at time
0.
=
0,
is
therefore described
with
=
1
and am =
As was seen from the previous
discussion, the solu
tion *; will lead to a description of the system as being time independent
with the wave function
^
and energy
is
«;.
This state of
i.e.,
affairs will con
tinue as long as the system
undisturbed,
as long as
U =
J7o.
Now
tem,
let
us suppose that electromagnetic radiation
falls
on the
sys
As
will
radiation
may
be shown later in more detail, the electric field of the act on the system and change the potential energy. The
U' and that in
H'.
change in
U is denoted by
is
H is denoted by H'.
and ¥i
The new
will
Hamiltonian
satisfy the
then
Hq
+
The
is
differential
equation corresponding not
to the initial state of the system
therefore changed,
new
equation.
The more
general expression
* =
ai{t)* t
+
am {t)*m
(16)
where O; and am are timedependent parameters, can be used to provide a solution to the new Schrodinger equation that is applicable to the perturbed system.
This wave function, with undetermined a and o m can be substit
,
tuted into this
new Schrodinger equation
to give
(#o
+
H'Haflr,
+
am * m )
= 
AA
(„,*,
+
o„*m )
Expansion of
a
this equation gives
t
H .*, +
am
H *m +
+ amH'*m =  Zrm *, *5 at __fe_dflU_fcd¥, h
a,ff '* (
A
2«
"
dt
2wi
l
dt
2ri
am
d^m
~dT
.
.
{
'
+
66
INTRODUCTION TO MOLECULAR SPECTROSCOPY
first
The
fied
two and
last
two terms cancel because ¥j and
Wm
are solutions
is
of the
unperturbed wave equation.
multiplied through
The remaining
expression
simpli
if it is
from
x=— «>tox—
+00.
by V* and integrated over all space, i.e., The orthogonality of ^m and fa eliminates
one of the integrals, and the timedependent part drops out of the ***„, integral as in Sec. 41. On rearranging the remaining terms, and
substituting Eqs. (12), one has
da"»
dt
?« fljer<
h
, H/A)
(
„_« m)
,
f

^H h dx
,
~
It is
2m
h
/"+«
known
that initially a
t
—
not
initially contribute.
It
1 and a m = so that the can be shown also that the
final final
term does term does
not contribute appreciably at longer times, and one has
dam
dt
2«
e
h
_ (2]riM)(ej _ 0i f + °^* Hlx f, idx
,
(19)
This important equation gives the rate at which a system can be changed from one stationary state to another under the influence of a
perturbing
rate at
which a m increases corresponds to the system changes from fa to m To proceed it is necessary to be more specific about the perturbation H'. Some features of the interaction between the electric field of the electromagnetic radiation and the molecular system that absorbs
effect.
The
rate with
which the description
of the
\[/
.
the system are
43.
now
treated.
The Interaction of Electromagnetic Radiation with a Molecular System
falls on a molecule, the oscillating can in some cases disturb the potential energy of the molecule and allow it to escape from its initial stationary
When
electromagnetic radiation
electric field of the radiation
state, here
will
assumed to be characterized by the quantum number I. It be recalled from the discussion presented in the introduction that
field of
the electric
the radiation oscillates at the point occupied by the
v.
,
molecule with a frequency
radiation can be described at the position occupied
For example, the x component of the by the molecule by an
equation which
is
usually written as
(20)
E.
= 2E° X
cos 2wvt
It is here
more convenient to use the exponential form
for the time
THE ABSORPTION
AND EMISSION OF RADIATION
dependency and write
Ex =
The
£°(e 2 ""
+
e*™')
(21)
derivation of the influence of the electric field on the molecule
can, for simplicity, be followed through in terms of the x
of the interaction.
component
At
the end of the derivation, the effects of the inter
actions in the y
can be added on. can act on a dipolemoment component /** This term adds to the potential to produce a change in energy of E x n x that occurs energy of the system and is responsible for the change in
and
z directions
The
electric field
Ex
.
H
when
radiation falls on the system.
One can
write, therefore,
(22)
H' =
If
ExP x
,
the frequency dependence of
Ex is
shown
explicitly, this
becomes
(23)
H'
=
£°(e 2 ""
+
ei*
M )nx
Substitution of this perturbation in Eq. (19) gives
^2 =
(£t
~E°
th
J
+" f— tltx^i
00
da;(e< 2
* i "'>«»«'+A'><
+
e (*r«/« «„.,«<)
(24)
The
integral over x that appears here is customarily represented
by
\n x im\; i.e.,
Im.1
=
/_*"
Kn*h dx
and integration
to
(25)
With
this notation,
of the timedependent functions of
t,
Eq. (24) over the time interval
[1
one obtains
1
_
p(.2Tilh)(e m ei+hv)t
«m
—
ti
xr + hv
is
+.
_
e
/
>(2irtM)(««,«fc>')!l
m
—
e{
—
hv
r
(26)
J
The
process of interest here
that in which the system goes from a
lower energy level u to a higher one e m For such an arrangement of energies, the denominator of the second term in the square brackets
will
go to zero when the radiation frequency
hv
is
such that
(27)
=
t
m

•
e,
For such conditions this second term can take on large values and be of major importance in determining am (t). On the other hand, if tm were lower than t t the first term in the square brackets would be of major
,
importance in determining am (t), for then hv equal to the energy
differ
*8
INTRODUCTION TO MOLECULAR SPECTROSCOPY
It follows that the first
I
ence would lead to a zero denominator.
in the square brackets is important
if
the transition from
to
m
is
term one
of emission, whereas the second term is important if the transition is one of absorption. Since we are here investigating an absorption process, the energy of the to state is greater than that of the I state, and it is only
necessary to retain the final term of Eq. (26). the assumed energylevel pattern
[1
We
have, therefore, for
_
t>iirilh)ii m —T\,)t~]
described
As the system changes from the description ty to *m it can be by the product ¥*&, where * = a&i + am ^rm and the a's are
,
time dependent.
tion,
as a function of time
This quantummechanical description can be given if it is integrated over all values of x. This integraall
moreover, removes
the spacedependent
wave functions because
only with
of their orthogonality
and normalization.
l
One
is left
/_«"
*** dX =
a
ai
+
am a ™
(
29 )
is is
The importance
given, therefore,
of the
m
state in the description of the system to Eq. (28),
by the product a*am which, according given as a function of time by
al{t)a m {t)
=
mm HEIY
L

,
(e m
—
ei
— 1—hvy 
(3o >
rrtwi
The above
Integration over
all
expression shows the effect of a given frequency of radiation.
frequencies, treating
E° x as a constant
since the
absorption will usually occur over a narrow frequency range, gives, with the definite integral
+"
sin 2 x
/:
the result
—5— ax
,
a* m {t)a m (t)
=
~
\ii* m
\KK)H
(31)
This
radiation
is
is
the desired result except for the characterization of the
its electric field
by
strength instead of
its
energy density.
It
the latter measure of the
are
when comparisons
amount of radiation that is more convenient made with experimental determinations of the
(32) where p is the energy per unit volume that is irradiated by the electromagnetic Replacement of E° x in Eq. . i. p (33) For isotropic radiation the three components interaction are equal. moment. becomes (35) J (OrnOm) t = B lmP with 8ir B tm = gp Wm 2   3 (36) This expression obtained for the time dependence of the coefficient of ¥„ for a system originally described by ¥j shows the factors that are responsible for the change from one stationary state to another. To complete the general study of the process by which radiation energy is absorbed. intro(34) duced so that Eq. and on the energy density of the radiation. wV (34) This rate of change of the system as a result of absorption of radiation under the perturbing written with effect of the electric field of the radiation is usually B tm called Einstein's coefficient of induced absorption.e. radiation. and one writes of the radiationdipole J (o£Om) t jj£j \(Hxlm) 2 + {HylmY + G*»lm)*p % . is The necessary relation. it remains only to compare this derived expression seen to depend on the term called the transition with the quantities usually encountered in the experimental determinations of the absorption of radiation. P = ^ (KY the energy density.THE ABSORPTION AND EMISSION OF RADIATION 69 amount of absorption. (31) by this relation gives the rate d(aZam )/dt with which the importance of state increases as m in the description — Jt w = = *ft. rate of transition is The \fn m \.. which is here given with out derivation.
70 INTRODUCTION TO MOLECULAR SPECTROSCOPY Exercise 41. is. C moles/liter . The dependence of a on the frequency is here emphasized cell a(v). versus frequency near hv = e m — n. and to the path length Introducing a(v). intensity. incident intensity. and see that the frequencies that contribute most to the integral are those close to the Bohr condition em — e = hv. hv) which must be integrated over all frequencies to obtain Eq. and the intensity of the radiation after passing through the the solution. The as in decrease in intensity of the radiation as Fig. as the proportionality constant dl = by writing a(v)IC dl (37) to be written. — «i — — — hvY «. or the intensity with no absorbing material. according to Beer's law. as cell containing The integrated form of Beer's law is obtained in this way aOOijln^ FIG. 41 Source (38) The absorption of radiation by a Sample solution. Spectrometer ir dl Solution of concentrate. the molar concentration. the absorption allows the equation coefficient. to C. Integration of this equation over the length . dl. Comparison with Experimental Quantities The quantities used in reporting the experimental results for the absorption of radiation can be introduced by considering the derivation of Beer's law for an absorbing solute in a it nonabsorbing solvent. t 44. 41. Plot the function sin 2 (2ir/h) (e m («». (31). proportional to the radiation dl. I allows the absorption coefficient to be measured in terms of 7 the I. penetrates a distance I.
since / of 1 sq is the energy flowing through a crosssection area related to the energy density cm in 1 sec.)p If = iW P one introduces the symbol N' to represent the number of molecules in a cubic centimeter of the sample and recognizes that an amount of energy hv tm is removed from the radiation by each transfer. gadro's number is N. or energy. The theoretical result of the previous discussion is that the rate of transfer of molecules from state ! I to state m is (40) (B (. it is by (42) I = cp where c is the velocity of propagation of the radiation. this relation is N' = NC 1. /. of the beam passing through a length dl of the sample is given by the theory of the previous section as dl dl = p Wm V^JV'  (41)  Furthermore. the decrease in intensity.e. absorption band for a given transition usually extends over a The total intensity of the band is obtained by measuring a(v) in the region of the absorption and determining. usually by graphical integration. (37). over band or a(v) dv (39) A = J over band [ a(v) dv = A c It is now necessary to relate the experimental quantities a(v). i. the integrated absorption coefficient A.. or A. Eq. can therefore be written as Equation (41) dl = §£ Wm 2 ()   hv lmN' dl (43) For comparison with the experimental expression.000 (44) .THE ABSORPTION AND EMISSION OF RADIATION 71 An range of frequencies. v . to the theory of the previous section. it is If Avofinally necessary to relate N' to the molar concentration C.
allows the terms corresponding to a(v) to be identified.d7 Sw@^S a differential (45) Comparison with the form of Beer's law. In this way we 2 obtain 3/ic(l.000) yjmlWml (46) where v tm is the center of the absorption band and.72 INTRODUCTION TO MOLECULAR SPECTROSCOPY Substitution for N' in Eq. both procedures 45. and the identification should be made with A = ja(u) dv. Eq. and for a given system they can be deduced by deciding for which transitions the integral \m m must necessarily have a zero value. and therefore fail to absorb radiation. can be calculated the wave functions for the initial and final states are Conversely. The dipole moment . 47 ) iMtml = /_ „T **"*. be used. (46) is that the integrated absorption coefficient. will In subsequent work. The Basis of Selection Rules The results of the previous section provide a quantitative relation between the transition moment of the absorbing species and the absorp by a given sample. (47) can be performed. furthermore. The theory than predicts. the total absorption for the transition rather that for a given frequency. and thus lead to the absorption of radiation. something about the  known and wave functions of the states involved in the transition. if the integration of Eq. as it will be recalled.. by means of Eq. some very important qualitative statements can be made about which transitions can be induced by radiation. and which transitions cannot be so induced. however.dx result given + ( The important for a transition if by Eq. m 2 and of + /xim and thereby learn. The deduction of selection rules for the particleinabox system will tion of radiation often \ conveniently demonstrate the deduction of selection rules. (43) gives . or simply the band intensity. From this relation. one can use the measured value of A to obtain a value of ju. Such general statements are called selection rules. Let us consider the particle to be an electron and the balancing positive charge to be located at the midpoint of the well or. to be symmetrically distributed about the midpoint. at least. (37). (47).
function since it changes sign at the origin. .' when a One can here illustrate that for the deduction of selection rules it is often sufficient that appear to consider the symmetry of the three functions if/*. is clearly an odd.+ ++* +. y of the potential well. with y at the midpoint is convenient for the symmetry discussions that are necessary here. the distance of the electron from the center with n = m For a transition from a state with » = ItO one to be induced by electromagnetic radiation. or antisymmetric. as Fig. and ^ in Eq. the integral W= The /_°^ 2 MejMm <*z (49) must he nonzero. The coordinate system. 42 The symmetry proper ties of the porticleinobox wave tranrule function and some of the sitions allowed by the + **—. (49). 42 shows. and therefore the function ey. —44* — . integration will be carried out in the next section is quantitative result for the absorption intensities desired. 42. The coordinate y. ey. of Fig. All the wave — (odd function) FIG.THE ABSORPTION will AND EMISSION OF RADIATION 73 then have the form ft = ey is (48) where.
\ from the integration over the right Thus \m m will necessarily be zero for a transition from n — I to n = m if I and m are both odd numbers or both even numbers.e. The allowed transitions. one writes. in of the have some simple direct relation to the symwave functions corresponding to the energy levels. 42. for For the dipole moment term p. symmetric or antisymmetric they are either even or odd functions. allowed by these selection rules are indicated in Fig. 15. as was shown i. connect a state the transitions for which \m m is not necessarily with an even value of the quantum number to a \ an odd value. The selection rule can be conveniently expressed as + where <+> +. 18. /» = •1{' ~ I) = ex ~ ea 2 (50) The wave functions are .e. are either about this origin. in Sec.. «(» wave wave function and — The transitions + indicates a state with a symmetric indicates one with an antisymmetric function. *„ = 1 sin \ja nwx a (51) . As for this particleinabox example. The symmetry of the system will be important in this regard. Particle The Integrated Absorption Coefficient for a Transition of a a Box To evaluate the theoretical integrated absorption coefficient for a transition. zero.74 INTRODUCTION TO MOLECULAR SPECTROSCOPY functions. one must work out the integral involved in /*j„]. and the selection rules will usually metry A6. using the coordinate system which the wave functions of the square well were given in Sec. state with i. recognize that the selection rule stating that transitions One should can occur between energy levels only if one level has an odd quantum number and the other an even quantum number is equivalent to the statement that only levels of different symmetry can combine. If the product of the three functions of the integral of Eq. The transi tion for a particle in a box can be easily treated.. it will often be possible to deduce selection rules without a detailed working out of the integrals. (49) is an odd function. the result of the integration over the left half of the potential well will exactly cancel the contribution half.
for which 12.000) "" 4e 2 a 2 ** (58) _ 32NeWvlm SwhclflOO Using X (lB = 2.THE ABSORPTION AND EMISSION OF RADIATION 75 and the . one gets A = 14 X 10 18 sec 1 cm1 mole1 liter .l" a Jo \ x cos (I  I m)— a x cos (l + m) — \dx y (55) Integration. (56) of the various even and odd combinations for and m. therefore. transition . moment \tu m \ is. (56) _ 96 ea ^ 2ea is ~~ ( The integrated absorption coefficient then given. I = 3 and m= Substitution of these. gives ea Ml* I" I cos (I (i i* — m)ir — . (46). (56). one is left with the expression (53) I"' = T /„" x ( sin 1?) ( sin "?) dx of the relation </> Making use sin B sin = Mcos (0  *)  cos (0 + <£)] (54) one gets ImiI = . by Eq.3 A. the frequency was calculated in Exercise application of Eq. as Fig. corre sponds to gives . 12 indicates. can be used as a The 4. making use of the integral (y cos "J = cos dy y + y sin y. transition. as 8**N 3Ac(l.my I 1 _ cos (I (J + w)7r— 3 +m ] 2 ) (56) where the fact that I — m and + m are integers has been used to elim inate the sine terms. 2e I Mi* Since the orthogonality condition eliminates the integral arising from the a/2 term. The selection rules deduced from symmetry considerations in the I previous section can be obtained analytically by the insertion into Eq.600 A and a = 7. specific The intensity of the ireleetron transition of hexatriene. . values in Eq.
. there are the same number of molecules in the state to as in the state the effect of subjecting the system to electromagnetic radiation will be to cause just as many molecules to go from from I to to. using the squarewell potential model. The probability of such emission is usually here. when one studies absorption spectra. tional The application to a vibrawill.3 X 10 18 sec" 1 cm. The interaction of the quadrupole with the electric field of the radiation is smaller than the dipole interaction by a factor of about 10 8 and one can therefore almost always ignore these quadrupole effects. 47. 2 used the selectionrule result that for a vibrational transition to be induced by electromagnetic radiation Av = +1. much less than for the induced process. that the absorption. problem of the important relations of Eqs. and it need not be treated Finally it can be mentioned that the treatment of induced absorption has been based on the interaction of the electromagnetic radiation with the dipole moment of the atom or molecule. The Integrated Absorption Coefficient for a Vibrational Transition The treatment in of vibrational energies and vibrational spectra given Chap.76 INTRODUCTION TO MOLECULAR SPECTROSCOPY is The observed value A = is Ja(v) dv = 4. one generally deals with a system where each state of lower energy has a larger population than each upper state. I m to as bility of In addition to induced absorption and emission. Exercise 42. 43. Calculate the frequency and the integrated absorption coefficient for the lowest irelectron transition of octatetraene. It can \ii now be \ recognized that this implies that only for Av = ±1 can tm for a vibrational transition be nonzero. there is the possispontaneous emission. the order of magnitude agreement quite satisfactory. (46) and (47) how . as was mentioned term for induced emission is equal to that for induced if I. This result can be readily obtained by those familiar with the general properties of Hermite polynomials. in Sec. This means that. where v is the vibrational quantum number. A net absorption of energy is then observed. Emission or absorption can also occur as a result of the quadrupole of the atom or molecule.mole1 1 liter In view of the simplicity of the model. Induced Emission and induced Absorption In some experiments it is important to recognize. In practice. 48. No net absorption of energy will be observed.
and m= 1. necessary for the evaluation of moi. .+ (£)_ <'*)+•: Mr*. In Sec. the quantity usually referred to as the perma nent dipole moment. if it is required that" the dipole moment must be expressed only for small displacements of the molecule from its equilibrium configuration.e displacement coordinate (62) = rr. to represent the dipole moment by the series expansion m *. with units of sec . In terms of v. . (61) where essentially. of a heteronuclear diatomic molecule must be expected to be some not easily determined. is. In terms of q th. the expression for A is = 3h!(lfi00) i> ° lM ° l12 (60) The dipole moment.JHE ABSORPTION AND EMISSION OF RADIATION 77 will ever. function of the internuclear distance. 44 the result was obtained that the integrated absorption coefficient (y) l dv Jo ° A dv see 1 cm1 mole1 liter is related to the transition moment * according to the relation A  31^0) vM SeC_1 Cm_1 m°le_1 Uter In the experimental study of vibrational transitions one usually deals 1 with frequencies v. An important result be obtained by a derivation of the relation between molecular properties and the amount of to v — 1 vibrational absorption caused by the fundamental v = transition. It is possible. with units of cm 1 instead of v. or even expressed. the integrated absorption coefficient A is 1 3Rlfj0) v ^W are I «« nudcr^ liter (59) For the fundamental transition for which the values of the vibrational quantum number Ao1 = 0. be demonstrated. however.
jic and /i„ d . written to distinguish the reduced mass from the dipole expressions for moment With these can be set up. This leads to the result of ^o The orthogonality constant term /j g _ M= (l) o/. 24 to be expressed in terms of the displacement coordinate as *° = w e ~ aqm and where 2t \/iwfc 4x 2iu„ d . as . (63) are retained. however.78 INTRODUCTION TO MOLECULAR SPECTROSCOPY is the dipolemoment expression "='" + (!) 9_„ 9+  («) The wave functions for the v = and the v = 1 states that are involved in the fundamental transition are shown in Sec.0#ldg = (iL v^@7r^. (64). ^ two terms of Eq." + * 9 aa! ^ (65) The necessary integration can be performed with the tabulated integral i o Xi ^" dx = i^ta < 66 > One obtains  Wl =  _i=(£) (67) . IMoi if only the first and ^ x the transition moment p.  = + /_ >[^o + (g) g]^ 9=o (64) and ^i leads to a zero contribution from the and shows that a constant dipole moment is not sufficient to lead to a vibrational transition being induced by electromagnetic radiation. result from the second term of Eq. A nonzero value of moi can. . is it.
one calculates (71) \dg/«=o X 10.000)Mred. The result obtained for BrCl It is consistent with this expectation. 23: 363 (1955)] to be given in Eq. in fact.2 mole 1 liter (70) this value and the reduced mass 76 of 4. recognize that at r = the dipole moment /i is necessarily is. V. (72). the order of magnitude that is to be expected for values of (dn/dq) q= o can be obtained by assuming a is constant charge of atom. The measured value of A for the fundamental transition of BrCl 1 at 439 cm for example. ~ \dq) qa This result suggests that values of (dn/dq) q= o should be of the order of magnitude of the electron charge. rather unsatisfactory to . (69) desired to obtain the . however. Crawford [J. Chem.1 X 102 * g. Phys.10 esu Although distortion of a chemical bond from its equilibrium distance undoubtedly generally accompanied by a complicated redistribution of the electrons of the bonded atoms.THE ABSORPTION AND EMISSION OF RADIATION 79 and with this value for the transition moment the integrated absorption coefficient A is found to be related to molecular properties according to in£S^(%l V «A=» 3c 2 (l.o _ ~ * \ / 3c 2 (l. (68) is the how the absorption intensity of a fundamental vibrational transition related to the proper and (dii/dq) q=a of the molecule under study. or a diatomic molecule. d m (69) or /djA \dq) q. 4.000)Mred. M With this + be on one atom of the bond and model the bond dipole moment is r„ — be on the other = (Se)r (72) r and. attempt to express the dipole moment of a bond.A *N important result that shows is Equation ties M™a. since g = — \dr) r=r ~ . One can. by means of Eq. has been reported by W. Brooks and B. F. it is The transposition is that which is used when dipolemoment derivative from the measured absorption intensity. A = From 105 + 14 cm..8 X 1010 esu.
The calcur. The value of n at the equilibrium distance can be determined. The Intensities of Absorption Bands Due to Electronic Transitions Detailed discussion of the changes in the electronic structure of molecules that lead to the absorption of radiation. It is at present a difficult matter + {dn/dq) q=a from the measured value to decide whether the curve with a positive value of (dn/dq) q= o or a negative value is correct for a given molecule. 49. but even this determination does not lead directly to a determination of the sign of this equilibrium dipole moment. It is. lation is based on a simple model for the behavior of an electron in electron that molecule or ion. Nevertheless. The information available. with a Hooke's law type of force. here assumed to be spherical. 10 and 11. (71). 43 Types of ft versus r curves that can be drawn from measured values of fi and dn/dq. since in inert solvents or gaseous media essen bonds are expected to break to give neutral particles. With this model. however. Mobs. will not be given until Chaps. g — 0— +9 . the measof values of A and the deduction of values for ±(dn/dq) = a q according to Eq. the FIG. including a deduced value of of A. siderable potential value in consideration of the electronic structure of molecules. n is also zero. lead to the accumulation of data that are of con urement . usually in the visible or ultraviolet regions. convenient to obtain here an oftenused guide to the intensities of absorption bands that result from electronic transitions. can be illustrated by the curves for possible dipolemoment versus bonddistance dependence shown in Fig 43.80 INTRODUCTION TO MOLECULAR SPECTROSCOPY zero and that at r tially all = °o. One assumes that the is responsible for the absorption of the radiation is attracted to the center of the molecule.
the result A = or A iVe 2 tt^f." These ideas be enlarged upon in Chap. 11. Likewise. that many /. — ez. (67). l. y. mole 1 liter Insertion of numerical values gives the prediction as A = Ja(v) dv is = 2.THE ABSORPTION AND EMISSION OF RADIATION is 81 behavior of the electron given by the harmonic oscillator wave func tions used in the previous section. [iVe 2 /l. (67) and the expression for a.x = ex. one replaces /tw. With these variations the transition moments for this simple model of a bound electron are. for the integrated absorption coefficient predicted by this model. the mass of the electron. 1. (46) gives. To do one introduces the term oscillator strength. that of Eq. by e. (76). 3e 2 ft 8ir 2 TOvC Moi' (75) Substitution of this expression in Eq.y = ey. This is denoted by / and is denned as the ratio of the observed integrated absorption coefficient Thus one has to the value predicted by Eq.000c 2 wi * cm 2 . IMoii! = Im«J  *S^ = « (74) Addition of the equal cartesian components then gives . since the mass of the electron is much less than that of the remainder of the molecule. and p. however.) dv (78) One very finds.000c 2 m] 4.31 X 10 8 cm2 mole1 liter (77) This result this often used as a reference value against which the actual intensities of electronic absorption bands are compared. . since p. 3 = = Ua(v) di>U. by m. one can replace (d/u/dx) x _o according to Eq. Each cartesian coordinate contributes a dipolemoment integral like Now. in fact.33 X 10 9 /a(i. One also finds electronic absorption bands have / bands with will much smaller values of and in these cases one expects to assign the absorption band to some "forbidden transition. and so forth. electronic absorption values near unity..000cm — sec 1 cm1 mole1 liter (76) = Ne 2 n„ n . .
1935.J. Inc. PrenticeHall. and E. S. Cliffs. L. Wilson. K.. N. chap. New York.. C: "Quantum Theory of Atomic Structure. B. 1960. 1. Inc. Slater. Pitzer.82 INTRODUCTION TO MOLECULAR SPECTROSCOPY PRINCIPAL REFERENCES 1. . chap.ll. 3. New York. 6.." 1953. Mc GrawHill Book Company. Inc. Jr.: "Introduction to chap. J. Pauling.: "Quantum Chemistry." 6. Quantum Mechanics. Engle wood 2... McGrawHill Book Company." vol.
83 . Now the heading will be used principally to introduce a number of finer features of rotational spectra that were not mentioned in the earlier introductory chapter. on the rotational states that exist. 3. interacts with the molecular rotation and for commenting on the effect that any nuclear spin has.ROTATIONAL SPECTRA Some states of the principal features of the energy levels of the allowed rotaof the transitions 3. LINEAR MOLECULES The selection rules for transitions between rotational energy levels and the multiplicity of these levels have been stated in Chap. tional states of diatomic molecules and have been discussed in Chap. simple situation of linear molecules. it is perhaps advisable to look in somewhat more detail at these two features for the relatively These molecules also provide a is convenient basis for mention of the fine structure that rotational spectra observed in when nuclear spin. as a result of the statistics that are followed. symmetrictop. . Before proceeding to molecules of general geometry. The rotational energies between these and rotational spectra of general polyatomic molecules are conveniently treated under headings that specify the relative value of the three principal is moments of inertia of the molecule. On this basis. and asymmetrictop molecules. 3. if present. of linear molecules A number of aspects of the rotational spectra have already been dealt with in Chap. the chapter divided into the three principal topics: linear.
to do this it is only necessary to investigate the value of the transition moment for the transition under consideration. 3. One part involves only the and of the electron considered to move about a fixed nucleus. Each characterized by two quantum numbers which are here denoted by J and can be assigned integral values. set The former part is identical to the equation that up for a freely rotating rigid molecule which. 32. p. Those familiar with the hydrogenatom wave functions that the angular part of these solutions associated is will recall related to functions solution is known as Legendre functions. as was mentioned in Chap. M and which than. generally solved when the hydrogen atom is be given here since it is studied.84 INTRODUCTION TO MOLECULAR SPECTROSCOPY 51. As was shown in Chap. 51. the other involves only the distance between the electron and the would be variables. The rotational wave functions can be obtained by Schrodinger equation expressed in polar coordinates for a molecule freely rotating in threedimensional space. M is less of the lowest The Schrodinger equation solution to the rotating linear molecule problem leads to allowed energies that are identical to those given in Sec. Rotational Selection Rules for Linear Molecules The requirement that a molecule must have a permanent dipole moment and that the rotational quantum number can change only by one unit for a rotational transition to be induced by electromagnetic radiation must first be investigated. /. of course. . 32 to describe the wave functions and solution of the the dipole moment that appear in the transitionmoment integral. The problem is. would of these equations involve only the angles describing the orientation of the molecule as The correspondence means that the rota tional wave functions that are of importance here are identical to the hydrogenatom wave functions. nucleus. Solution of this problem will not somewhat intricate and can be found in any of the standard texts on quantum mechanics. The threedimensional nature of the problem suggests the use of the polar coordinate system of Fig. . 4. 51 and Table 51 . The wave functions corresponding to some energy states are given in Table 51. d. appropriate to atomic problems are given. These are where both the J quantumnumber notation appropriate to molecular rotation problems and the s. notation familiar angular part of the illustrated in Fig. . The solution functions illustrated in Fig. with the provision that or equal to. It turns out that the Schrodinger equation for the hydrogen into <t> atom can be separated angular coordinates 6 two parts.
a S E o i5 o r "l".CM C o i V > "5 = o o B e a. HI 6 85 .
average angular it is momenta from the wave found that Total angular momentum = \/J(J + 1) h ~ Component TABLE 51 Molecule of angular momentum along an axis = Wave ¥r (2) Some of the Functions for a Rigid Rotating Linear J =0 Af =0 = = +1 1 <P 2Vi 7= / = 1 M Af .=i^COS9 * = 8mee 2\l%r /i 5 \ 1/3. . 2 (3 cos 91) or 4 \ir . if quantummechanical operators for angular momentum given in Table 11 are used to compute functions given in Table 51. /3 /sin 9 • sm * J  2 M M = ^ = «. momentum Furthermore... .. 2. .. (1) the same as that obtained previously from the quantized restriction. = .±1 1 (16 =A — sin 2 \2x / 1 \  9 cos 96*'* „ /l5 .*/sin 2 '"" .86 INTRODUCTION TO MOLECULAR SPECTROSCOPY if « satisfy the Schrodinger equation only in that equation has the values 8wU This result angular is J(J + 1) / = 0. / \ 2 / \ — sm • 9 cos 9 sin „ „ • . / — sin 9 cos 9cos ./ 2 9e ±2i+ or 4 . <t> 1 1 1 /1 5 . * IT Af = ±2 &..» 4 \2ir / 1 I — sin /l5 . ) / I \ — sin IT 2 9 cos 2* 1 (15 «/— sin 2 \ 4 \ 9 sin 2c> jr . I OT ./i. . 9 cos 4> r i = S 1 1 1 r  . 1.
of The diagrams that result for quantization in terms + 1) and are illustrated in Fig. 52 The total and component angular momentum vectors for components. sometimes convenient to illustrate the angular momentum components implied by the rotational wave functions by means of a vector diagram. +1. the remaining term that necessary for the evaluation of the transitionIf the moment integral the dipole moment. /* . After the the average angular wave functions is is of the states involved are specified.ROTATIONAL SPECTRA 87 where M= 0. Angular momentum in direction component . of appl ied field Angular momentum of rotating molecule KA/27T)  2(ft/2T) . and confirm that momentum is y/J(J + l)(h/2ir) in these cases. M Apply the operator for total angular momentum to the J = and +1 wave functions. The number of angular momentum components along an axis that can be obtained from a given value of J can be seen from such a diagram to be 2 J + 1. permanent dipole moment of the molecule is represented by the components along the three axes FIG. momen The first of these the while the second corresponds to the stipulation that the angular tum component It is in a direction in space be quantized. is ±J now familiar total angular momentum stipulation. J = 2. 52. ±2. \/J(J Exercise 51. For 1 = 1 an applied 2) field reveals five More generally for the Jth rotational level + 1 states would be revealed.
requires a more detailed investigation of the integrals to see that they also vanish unless J' J"±l (6) that is. 32 are given. by Mx nv y. . (4) that the transition moment for It rotational transitions will be zero unless mo has a nonzero value. if a transition between a state characterized by J' and With the rotational wave functions (3). and will not be given here." of this result is outlined in "Spectra of Diatomic by G. 51 Deduce from the symmetry of the wave functions of Fig.) transition These components allow the calculation as of the total moment l„J"Af'J"ikf"2 _ „/'JV/'J'"M"2 __ l„/'Jlf'J"Jlf"2 __ „J'Jtf'/"Jf"2 (5) It is immediately clear from Eq. how ever. be carried out in a straightforward manner. Herzberg. (3). t = = = mo sin 6 cos /xo 4> sin 6 sin <f> (3) mo cos 8 of Table 51 and the dipoleone can investigate the contributions of the transitionmoment components to the total transition moment. one obtains the  M = MQ Mo 2 f ' ^J'M' [J 2 CQS ^J'M" gm (20 df d« (4) ^j'Mjjf'i J'*<j"<jf" = = ' o Jj 2x o j" ' o fw sin 8 sin W"*" sin de dtf) dfl  M Mo j" ^/'Af cos 0^j"K" sin (It will be recalled that the element of solid angle is sin 6 dd d<t> in polar coordinates. M" can.' 88 INTRODUCTION TO MOIECUUR SPECTROSCOPY of Fig. such verifications are suggested as exercises. M' and J". The verification of this selection rule for particular values of J'. moment expressions of Eq. for any orientation of the molecule. In this way. and the antisymmetric nature of the dipole moment. as can be verified to J = 1 will be from Eq. Exercise 52. that the transition moment for J = nonzero while that for the J = to J = 2 will be necessarily zero. A/ = +1 The deduction JMolecules. M and one characterized by J" and integrals J'Jf'J"M" M" gj n fl is investigated.
due to the various values of for a given value of J. can be made apparent if an external field is applied M to the molecules while their rotational spectrum field is is observed. therefore.ROTATIONAL SPECTRA 89 Since the energy of a freely rotating molecule is. or reference to deriva. An electric used. indicated here for linear molecules. and the shifting and splitting of rotational spectral lines that results is known as a Stark effect. The calculation of these energy . It is found that the integrals of Eq. that this selection rule operates for transitions between states described by the wave functions Exercise 53. only on J. it will not be given. When. as will be discussed in the in rotational spectroscopy. in a classical picture. The quantum number M gives the component field. The Stark Effect in Molecular Rotation Spectra Although the rotational states of linear molecules are characterized two quantum numbers. of the rotational spectra of generally The treatments other selection rules. can be found in ref 3. Verify the AJ = ± 1 and the AM = selection rules for J = to J = 1 transition. next section. the selection rule for J is of prime importance dependent (1). of these selection rules 52. according to Eq. (4) vanish unless AM = (7) Again. without derivation. since the general proof of this requires a knowledge of the behavior One can. on the same considerations as those Derivations. the of Table 51. an external electric field imposes a direction on the rotating have different energies molecule. it can be expected that these different orientations relative to the applied electric field will cor respond to the different energies. to different orientations of the rotating molecule relative to the applied field. however. degeneracy of these rotational levels. the energies of these states depend only by The 2J + 1 the on total angular momentum quantum number J. however. Since the molecule has a dipole moment. of the angular momentum in an applied is direction. the states with different values of and the selection rule for M M becomes significant. and understood that they are based tions. shaped moleit is cules that will be presented in the remainder of the chapter will introduce These will be stated. and in a Stark effect experiment states with this direction that of the applied electric The different values of M correspond. verify of spherical harmonics.
and this is done in Fig. 53. is. ref The result of the calculation is that the energy of a state characteris shifted in an electric field of £ volts/ cm by by values of J and M where for / ?* 01 _ 4. 3 was assumed that a simple if series of equally spaced lines could be expected is obtained. J levels of the molecule are responsible for the In Chap. for AM = AM = (9) absorption experiments. 53. The derivation of the expression beyond the scope 2. With these rules as mentioned in the previous section. and the calculation of the energy requires a secondorder perturbation calculation to be made. of this treatment and is outlined. with references. While this is a rotational spectrum of a linear molecule true for linear molecules. The effect on the energy shifts levels given by these equations is shown for a particular case in Fig. Observation of the number of components that rotational line splits into when an it electric field is applied is often a great help in deciding which transition.'IgjS* fe J(J + 1) 2 (300) 2 J(J + 1)(2J . seen. (10) AJ = +1. These energylevel rules show up when spectral transitions are obtained for molecules subject to an electric field. are important. however. The effect of the applied field as will be very small. in ized Se. for the energy shifts is .3M» . not simply done. Now. the unit usually used in microwave spectroscopy. both selection AJ = ±1 or.1)(2J + 3) e g while for / = (8) 5£ = " «00P er & can be converted to frequency shifts of These energy shifts in ergs cm1 or megacycles per second. sufficient complexi ties are introduced into the rotational spectrum of generally shaped . the allowed transitions can be added to the energylevel diagram.90 INTRODUCTION TO MOLECULAR SPECTROSCOPY shifts that the states with given J and various values of M experience is.
Dakin. W. In addition to illustrating the space quantization of Stark effect splittings allow. measurement with the aid of Eq. Good. effect. of often rather precise values of the molecular dipole first such determination was made by T.00Q volts/cm on the rotational energy levels of HCI. and FIG. M = ±l 50 M= No applied field 1. 200 6X y 10 6 =4 150 M 0 4X10 6 / = 3 ±1  ±3 +2 100 E u e § M 2X 10° E ^^^^ so ±2 J = ^ . (8).000 volts/cm 2X 10 6 .ROTATIONAL SPECTRA 91 molecules that a simple matching of the spectral lines with the expected pattern is seldom possible. E. \ moment /x The W. the determination . 53 The effect of a Stark field of 1.
. oscilloscope pattern of the The J = 1 to J = 2 line is shown for different electric field strengths in Fig.. must take into account molecules. One finds." John Wiley & Sons. Some dipolemoment determinations Exercise 54. 54 the dipole moment of the OCS molecule. Inc.00 0. experimental evi dence in the hyperfine structure of rotational transitions of some moleThe basis and nature of this fine cules for the effect of nuclear spin. .710 3. such of rotational states both the molecular and nuclearspin angular to the total momentum contributions angular momentum.23 1. 54. Trambarulo.87 1. CH 8F CH 3 C1 CH.29 0:57 0. Smith. * 0. V. as a result of an applied electric field of 5. "Microwave Spectroscopy. as we will see. 70: 560 (1946)] for the molecule OCS.80 1. multiple^ of tum. made using the Stark effect in rotational spectra are given in Table N 2 = 1 energy level (a) Calculate the shift in the J = 1.92 INTRODUCTION TO MOLECULAR SPECTROSCOPY D. of that have been 52. Gordy. CHF. detail of rotational spectra will here is be only briefly outlined. — Nuclearspin Coupling + has been recognized that the allowed rotational st ates of a mole1) *yP e cule are those for which the angular momentum is a \/J(J momenangular rotation molecular this to addition In h/2*. It is quantized in units of TABLE 52 Some Molecular Dipole Moments from Stark Splittings of Rotational Transitions* Molecule Dipole moment (Debye) Molecule Dipole moment (Debye) FC1 FBr BrCl OCS HCN NO s NH.47 0. 1953. (b) Estimate from Fig.64 1. some molecules have angular momentum because of the nuclear A complete characterization of the spin of one or more of their nuclei. W. and R.13 From W.000 volts/cm. K.Br CH.166 1. Its introduction to appreciate the many rotational student intended to allow the spectra studies reported in the literature where this effect occurs. The angular momentum of a nucleus results from the spinning of the nucleus and is a characteristic of the nucleus.79 1.55 NF.I CH CCH 3 B5H9 0. Coles [Phys. in fact.75 2. Rev.65 0. PH. Molecular Rotation It M 53.88 1. F.
. and the two markers on each curve are spaced 6 megacycles/sec apart. is The Stark splitting of the J = 1 W.] The line for no field centered at about 24. K. interaction between the orientation of the nucleus and that of the molecule. Phys. If. No electric field 750 volts/cm 1. f If there is no coupling.. and D. In such a case the energy of a given molecular rotation state. on the other hand. %. the spin angular momentum 1) of a particular nucleus has one of the values Vl(I + where I = 0. and the total angular momentum of the system is then y/F(F +1) (h/2ir).070 volte/cm Frequency  . This is usually denoted by F. 560 (1946). 1. Rev. would be unaffected by the nuclear spin /. i. Cotes. W. A molecuFIG. 70: * 2 absorption line of OCS. if the nuclearspin quantum number is designated by /.. [From T. the molecule will rotate and leave the spinning nuclei unchanged in orientation. there is an energy of inter action the energy of the system will depend on the orientation of the nuclear spin relative to that of the molecular rotation. ROTATIONAL SPECTRA 93 h/2w.320 megacycles/sec.e. and. 54 Dakin. E Good. designated by J. This dependence can be expressed by introducing a quantum number for the total angular momentum of the system.
bring their electrons close to the nucleus.. they are ineffective in orienting the nucleus because of their spherical symmetry. The next most effective electron orbits in producing an electric field at the nucleus are the p orbits. however. would be imposed on the nucleus.// (11) with the provision that the total angular momentum cannot be negative and. of course. 55 with by attributing the extent of coupling. the directional asymmetry is of the electric field of the molecule " acting on the nucleus that now important. J+I . The hyperflne spin / splittings (schematic) of rotational levels resulting from coupling a nuclear = s.94 INTRODUCTION TO MOLECULAR SPECTROSCOPY lar rotation state with a given value of J can then lead to states with various values of F according to 1.. Those with / greater than ^. 55.. have a charge asymmetry that gives them quadrupole moments. Some inter esting results have been obtained FIG. F = J+ I. however. therefore. Although s orbits. For almost all molecules. a completed p shell has spherical symmetry. that the J — I type terms are not always realized for states with low J numbers. moments. and no directional effect Again. this interaction occurs between the electric field that the molecule exerts at the spinning nucleus (All nuclei lack dipole and the quadrupole moment of the nucleus. the exceptions being odd molecules such as NO. Something must be said about the source of the interaction between the orientation of the molecule and that of the nuclear spin. F J 3 — —^h^^h 9/2 7/2 5/2 3/2 7/2 5/2 3/2 1/2 . The splittings of the J levels that result from a nuclear spin of f are illustrated schematically in Fig.) It is.
of the atom is more or less complete in a particular The spectral effect produced by the coupling of J and I to produce like that a resultant of Fig. Phys.1 ROTATIONAL SPECTRA for a given nucleus with a given quadrupole 95 moment. is [The absorption spectrum from W. and W. No detailed treatment of this quadrupole coupling be attempted in the following treatments of fine structure will.. Frequency  F= FIG. 56 The J = > y= i 1 ab F=2 F= l sorption and transitions for DCN showing the effect of the nuclear spin / = 1 of the nitrogen nucleus. : interaction. 56.] /=0 F=l . symmetric. therefore. 55 F can be understood with an energylevel diagram selection rules and the AJ = + 1 AF = 0.. observed by Simmons. J. and Gordy [Phys. show such hyperfine structure due to nuclearspin simple illustration is A N 14 has a spin of unity. + (12) provided by the molecule DCN. in which The energy levels and J = —> 1 transition. In summary: any molecule containing a nucleus with a spin greater than £ will. 77: 77 (1950). Rev.and asymmetrictop molecules. Simmons. Anderson. Rev. W.Anderson. E. 77 77 (1950) are shown in Fig. if the nucleus is embedded in an asymmetric electric field of the molecule. with the extent to which the p subshell molecule [2]. Considerable complexity can be introduced into rotational spectra by this effect. Gordy..
10. = Mvfr (13) where the subscripts e.e. y. symmetry the allowed states were labeled with a plus or a minus. in Sec.! through the origin. that the selection rules for the transitions induced (+ < — operation. 57). plished by rotating the entire molecule dicular to the internuclear axis a plane perpendicular to this nuclear axis (see Fig. z — — x..) 96 INTRODUCTION TO MOLECULAR SPECTROSCOPY *54. It was found. > — + .i are reversed. important weights of of a i. be assigned to the wave functions that describe diatomic molecules. of the plus or minus character of the states involved We will now see that plus and minus signs can. the very simple system of a particle in a box were either unchanged or changed in sign the origin. but adequate separated function lAtatal for the present argument.. does not alter ^„. ior. as by the is substitu x.e. and rotational. when the wave function was subjected to an inversion through On the basis of this behavi. — z. — «+> — by electromagnetic radiation could be deduced in the on the basis transition. diatomic molecule when —y. The effect of this on ^tota i best deduced by writing the approximate. and r stand for electronic. We now when tion all consider the effect on the total wave function ^t. considered in the following section. this function is inverted \f/ the coordinates involved in at$ . and the result of this reflection depends on the particular wave function that describes the electronic state of the The nature of the many different electronic waye functions that occur in the ground and excited states of diatomic molecules will be considered in some detail in Chap. The Positive and Negative Character of the Wave Functions of Linear Molecules It has been shown. For the present it is enough to molecule. 18. that the wave functions for. v. <—I—> +. in an entirely This positive and negative character is analogous manner. in connection with selection rules and with the statistical rotational levels. . We can now investigate the effect of inversion on each factor and then combine these results to see the effect on ^ Mta Inversion of the electronic part of the wave function can be accomi. by 180 deg about an axis perpenand then performing a reflection through rotation axis and passing through the interfirst Since the step in this process leaves the it coordinates of the electrons unchanged with respect to the nuclei. when y was replaced by — y. The second step reflects the electronic wave function through the designated plane. depending on their behavior with regard to this furthermore. vibrational.
24. either basis. Electronic functions that are symmetric with respect to such planes are labeled with a plus sign. —2 is obtained by replacing r. must now be conThe polar coordinate equivalent to the inversion implied by x V> z —* ~ x —V. r — 6. <p e therefore. designed as 2+ states. 10. and one finds that most groundstate electronic wave functions fact. 57 Illustration that rotation fa. It follows that fa will also be unaffected by an inversion through the origin since this operation leaves the magnitude of the internuclear distance as unchanged. 6. This common groundstate function is positive both above and below any plane passing through the internuclear axis. \j/ The vibrational component v of the total wave function. and * + The effect of this inversion on fa can be seen either analytically by this replacement in the expressions for fa given in Table 51 or diagrammatically from the wave functions depicted in Fig. on the magnitude of the internuclear distance. are are of this type. and by r. After rotation • ~~* Initial position . . The remaining factor fa.ROTATIONAL SPECTRA 97 concern ourselves with the electronic wave functions that occur in the ground state of almost all diatomic and linear molecules that have even numbers of electrons. depends. 51. of J the function fa remains J the function fa changes inversion. and fa are unchanged by axis by 180 deg around an and reflection through a plane perpendicular to the axis correspond to inversion through the origin. will be discussed in Chap. shown in Sec. \f> e we conclude that an inversion of through the origin leaves unchanged. On > <t> <j>. one sees that for even values of unchanged by inversion. sidered. With the recognition that FIG. in The basis for the use of the letter 2 For these often encountered 2+ states. Most groundstate wave functions. 51. discussed in Sec. while for odd values sign.
This feature. show ing the behavior of ^totai on inversion. is perhaps best known to a chemist in terms of an interchange of electrons. Deduce that the interaction of a molecular dipole with electromagnetic radiation can lead only to transitions of the type *55. such as CO2 or HC=CH. 58 The positivenegative character. .98 INTRODUCTION TO MOLECULAR SPECTROSCOPY for the molecules considered here. We are concerned now with the effect of an interchange of identical nuclei on the wave function of the molecule. consistent with that of of the rotational with electromagnetic radiation in the same way as illustrated for the This result can be recognized to be A J = +1 obtained from a more functions. otill on inversion FIG. one deduces that ^ to tai behaves in a is manner dictated by fn as shown > in Fig. 58. An immediate (+ i > result of this assignment the selectionrule result — + . Mention of this operation for the two electrons of the Behavior of ^. 45. *Exercise 55. detailed analysis wave We now will consideration and will proceed to another important type of symmetry make use of the plus and minus assignments obtained above. of the rotational levels for those molecules with a S+ ground electronic state. or a linear molecule with identical nuclei. <(—> +. SymmetricAntisymmetric Character and Statistical Weight of Homonuclear Linear Molecules Another rather nized different when the molecule under type of symmetry behavior must be recogconsideration is a homonuclear diatomic molecule with identical nuclei. which must be considered whenever a system contains like particles. — (  — ) that can be deduced for interaction onedimensional example in Sec.
(14) for the helium Furthermore. the total wave function contains no vibrational and rotational factors and might be written as </w = +.ROTATIONAL SPECTRA 99 helium atom may clarify the parallel treatment that will be given for like nuclei of molecules. properly adjusted Thus one would write ls(l)ls(2) for the increased = *. from these four possibilities. to the orbital functions. **. two elec ls(l)ls(2) = Thus far our above discussion of the helium factor of the spin of the electrons.. one must make use of the fact that the behavior of if electrons in nature can be accounted for only wave functions that ^toui are antisymmetric with respect to the exchange of any pair of electrons are used. Bin will go to +^. one can write various possible spin functions for «(1)«(2) 0(1)0(2) a(l)j8(2) as +  a(2)/J(l) a(2)/S(l) all a(l)0(2) These have been written so that f. given immediately by the . describe!s only the orbital atom has lacked the important wave function given behavior of the electrons. one can write a suitable approximate expression for ^. It follows that a suitable antisymmetric combining the orbital and spin factors to give can be written by *«* = l«(l)l«(2)[a(l)0(2) this  «(2)/8(l)] (17) One deduces by means the familiar result. or any atom. clearly unchanged by the interchange ls(2)ls(l). To attach correct spin functions. and since the angular momentum associated say by the symbol a. and one should The electronic write the more complete expression tfWl = lM'. pi„ when the electrons are interchanged. For the helium atom. i.*. atom by regarding each electron as behaving according to Is hydrogenlike nuclear change. designated \f>^ n by the symbol p. designated or the other way. with this spin can be directed one way. wave functions.e. = is (15) of the This function trons.pia (16) Each electron has a spin quantum number £.
Some electronic wave functions. electrons and protons are the most often encountered examples. is the index which determines whether the wave function must be unchanged. are said to follow FermiDirac statistics. wave function for a molecule must behave. are. whereas if the spin quantum number is zero or integral. for 2+ states. symmetric. will be introduced in a way so that the complete wave function behaves in a manner suitable to the statistics followed by the nuclei. are symmetric with respect to this inversion through These symmetric electronic which constitute the most often encountered ground states.said to follow BoseEinstein statistics. NaCl.e. those of the second type.100 INTRODUCTION TO MOLECULAR SPECTROSCOPY usual form of the Pauli exclusion principle. If the spin quantum number of the particle is half integral. antisymmetric. electrons and through the origin.e.. or must be changed in sign. the wave function must remain unchanged. i. The first step of this process is nothing more than the symmetry operation treated in the previous section. It is informative to analyze the effect of an interchange of identical this nuclei of a diatomic. in a manner dictated by the nature It turns out that the spin of the nucleus. Particles of the first type. the center of symmetry of the molecule. with a g (for the German gerade). 53. and then (2) of the inversion of the electrons back through the origin. that the electrons of the helium atom. Those electronic states which are antisymmetric with respect to inversion through the center of symmetry are labeled with aw (for ungerade).) The net effect of the twostep process is. The effect of an interchange of like nuclei on ^toU1 will now be anaThen the nuclear spin. and so forth. the particles. and so forth. in In a similar its ground state. The second step of the process has an effect that depends only V>n the electronic wave already know that its effect function. with respect to an exchange of identical particles. lyzed. are labeled. . i. such as Of 6 (with zero spin) andH 2 (with unit spin). such as the lowest particle inabox state. (Note that electronic functions only have g or u properties when there is a center of symmetry as there is in H 2 CO2. as is that of the electron. molecule by imagining all exchange to occur as a result (1) of an inversion of nuclei. way the total for the interchange of like of the spin of the nucleus. or linear. We therefore on the total wave function. nuclei. must have opposite spins. or any particle. the wave function must change sign.. if the electronic state is . reflection through a plane containing the molecular axis. The property plus and minus assigned to electronic wave functions in the previous section depends on states. is plus or minus depending on whether J is even or odd. Such behavior can be considered even for molecules like HC1. dealt with in another connection in Sec.
J can exist. to leave the total of 101 J and to change its sign for wave function unchanged for even values odd values of J. such as O u C0 16 that have identical nuclei with zero spin. The rotational energylevel diagram should show. this Symmetric or antisymmetric behavior with regard to by an s or an a. exclusive of the effect of nuclear spin. 5 $ A U . be observed in absorption like rotational spectra because such symmetric molecules with identical nuclei have zero dipole moments and do not give FIG. 59 rise to rotational The rotational energylevel diagrams for linear molecules with weight factor due to the nuclear spin / nuclei. For u states this conis clusion is reversed.a ra .e.3 tu / (5 a) 5i y 5 a 85 3 ft Ortho 4 « 1 4 « 1 f Para (3 a)— 3 a 3 ft 0rtho 2 « (1 1 2 1 8 1 ft Para a) a « 3 1 ft ft "0 « 1 Ortho Para O 16 CO 16 type H2 type (a) (b) . i. The / statistical is shown. since the twostep process equivalent to the exchange of the identical nuclei. in (b) they have = if. It follows. since nuclei with zero spin behave according to BoseEinstein statistics which require sym metric behavior as a result of nuclear interchange.. €> % . In (a) the like nuclei have = 0. of course. that only the states with even values of J. only the states with even values of The absence of the odd levels cannot. no spin function is exchange is indicated . involved in the total wave function. For molecules.ROTATIONAL SPECTRA of the g type. as a result of this exchange. 59. we have found the behavior of the total wave function. as illustrated in Fig. Thus. the above treatment is complete.
102 INTRODUCTION TO MOLECULAR SPECTROSCOPY It spectra. Furthermore. Thus ortho and para states cannot be interconverted. the have an additional multiplicity factor of 3 while this factor is only 1 for even J levels. which have J odd.lMv(a/3  /3a) and for the o states. Thus one writes. for molecules like HC=CH is that will be discussed in Sec. transitions between s states and a states. confirm the absence of the odd /levels for such molecules. by any mechanism. as *. as Since three spin functions are available to provide the correct for the symmetry whereas only one spin function makes the total wave function antisymmetric when J is even. 74. the odd J levels have a statistical weight due to the spin function of 3 while the even J levels have a levels odd J weight of 1. that rotational will Raman not be discussed here. Now states with both even and odd values coupling \f/ of toU1 J can be brought to the required antisymmetry by suitably with spin functions. A — — . One should notice that the selection rule s < ated. which have opposed ][ nuclear very strong selection rule exists which prohibits. recognize that the analysis given above has led us to One should parallel spins. which are those having nuclear spins. antisymmetric functions for the s states. spectra. such as 2 and H H. Let us now consider molecules. in addition to the 2 J + 1 multiplicity that all levels have. three symmetric spin functions aa. Again this alternation in multiplicities. (The C 12 nuclei of will H —C=C — for the /3/3. however. which leads to an alternation in populations. acetylene have zero spin and can here be ignored. the absence of these levels which be noticed in rotationvibration spectra that will be studied in Chap. Furthermore. illustrated levels odd J in Fig. 7. ]\ recognize the ortho states of RYlike molecules. unless the molecules are dissocia< s. 59. that have like nuclei with spin quantum number ^. Thus. > o. which have J even. nuclei with a spin quantum number of ^ follow FermiDirac statistics that require a total wave function which is antisymmetric with + /3a respect to the interchange of the nuclei. for g electronic functions. an alternation in the intensities of the rotational components of rotationvibration absorption bands observed. a/3 two and and one antisymmetric function a/3 — /3a can be written.) As electrons of the helium atom. should be mentioned. cannot be observed in rotational spectra but. and the para states.
$. the third being different from these two. Benzene and the methyl hal Figure 510 shows the principal axes of these two examples. prolate symmetric top.ROTATIONAL SPECTRA s < * which applies to homonuclear molecules.e. those with U?* In = to. oblate symmetric lopj . have different SYMMETRICTOP MOLECULES wave functions with the Again.Most symmetrictop molecules that will be treated belong to this class because of a arrangement of some of the like atoms of the molecule.. \ . however. ides also are symmetrictop molecules. FIG. (1 A > l B . Again. 510 Examples of symmetrictop molecules. Symmetrictop molecules are defined as those having two equal principal moments of inertia. and the two axes with equal moments of inertia pass through either the carbon atoms at opposite corners of the square or through the opposite sides of the square. and <+— molecule. + — — 103 . The unique moment of inertia is generally represented by IAl while the two equal moments of inertia are I B = Ic. symmetric A typical example is cyclo butane. nuclei with spins of and so forth. one finds that all / levels ca'n be combined with spin functions to give total symmetry required by the nuclei. A similar treatment can be applied to molecules containing like 1. (I 4 < /„ . the s and a states. elimdiatomic any > applies to which — + «_)_» inates the possibility of pure rotational spectra of diatomic homonuclear «_[_> a. which has its unique axis perpendicular to the plane of the carbon atoms. i. The cules will classical first treatment of the rotation of such symmetrictop mole be given so that the quantized angular momentum restricA statetion can be applied to give the allowed rotational energy levels. with even and odd J values. statistical weights. t molecules.
2. For a symmetrictop molecule I B = Ic.104 INTRODUCTION TO MOLECULAR SPECTROSCOPY ment of the selection rules will then allow the spectral transitions drawn in and a comparison with observed spectra to be made. (19). The rotation of a system with and Ic is written classically as =  A+A+A 2IA . the allowed rotational energy levels classical can be obtained from the equations by the appropriate angular principal momentum moments e restrictions. 56. to be The Rotational Energy of Symmetrictop Molecules As with linear molecules. + ft that the total angular ("•) The energy levels of a quantummechanical symmetrictop system can be obtained from Eq. it is assumed momentum is quantized and. . . . (20) correct quantummechanical solution for symmetrictop molecules obtained by further assuming that the component of the angular momentum about the unique axis is quantized and that this quantization is is K 27 K==0 ±X ±2 > > > > ±J < 21 > where the plus and minus values correspond to the possibility of clockwise and anticlockwise rotation about the unique axis. ' ilA^\ + \Ib<»b + %Ic<*c 2IB ' 27. . where pA Pb. 1. and this general expression becomes "&+#. These quantum restrictions lead to and Pi (22) +P c > = j(j + i)(Ay_^(Ay .c as. and p c are the angular momenta about the three axes. is given by VJ(J + The 1) £ J = 0. As for linear molecules. of inertia IA Ib. with the quantum number J.
and K= ±1. (19) to the quantummechanical result and tJK (23) — &* 2 cIb . 1) . . (24) . and for a given value of J the higher the value of K. 35. An example is 2 is For this class of symmetric tops. ±2. is positive.. one would also have found that the quantum number J determines the total angular momentum of the system while the number K determines the component of this angular momentum along the unique axis of the molecule. +J This expression gives the same set of energy levels that would have been obtained if solutions to the Schrodinger equation for a rotating symmetrictop molecule had been sought. 2. . 1. It will only be mentioned here that Eq. the energies of the allowed states of a rigid rotating symmetrictop molecule are given by ijk with = BJ(J + J = 0. The significance of Eq. + (I  E)K* 0. (24) is best illustrated by a diagram of the energies that are predicted for various values of situations are recognized. _ greater than I B Molecules that have the unique moment of inertia IA — Ic are said to be oblate molecules. Two different Molecules that have the unique moment of than I B = Ic are said to be prolate molecules. . . benzene. ROTATIONAL SPECTRA 105 These relations convert Eq. and for a given J value the higher the value of K. is As for linear molecules treated in Sec. (24) For such molecules the coefficient of ple is CH 3 CN. inertia IA less J and K. (24) corresponds to the assumption of a rigid molecule and that additional terms must be intro duced to allow for centrifugal stretching. . the coefficient of negative. the shifting of the energy levels due to this effect always very small. the lower K . An exam2 in Eq. the higher K the energy. By the application of the appropriate quantum mechanical operators.^ + 1 ) + (8^8^r> and 2 cm" It is again convenient to introduce rotational constants B and A defined by cm_1 B = iJcT B A = s^n cm_1 With this notation.
NH. In such cases the selection rules for the absorption or emis sion of electromagnetic radiation are AJ = and 0. . 'Microwave Spectroscopy. R. inertia. rotation about the figure axis contributes K no rotating dipole moment. few rotational energy 57.106 INTRODUCTION TO MOLECULAR SPECTROSCOPY the energy. 511.016 From W. CF CC1 CBr CC 1.460 1.056 1. TABLE 53 This component of the rotation cannot. It should be pointed out that the occurrence of in Eq. 1953. Selection Rules and Spectra of Symmetrictop Molecules dipole Most symmetrictop molecules that have a this dipole moment have moment in the direction of the figure.938 1.767 CC s=CH CC1 ± 2° NH 1. axis of the molecule. and calculate the three principal moments of Also calculate the rotational constants and plot to scale the first levels. Thus all levels with =* are doubly degenerate. New York. or unique. CH 1. Gordy.C=CH CHC1. * HNH & Sons.460 1. Smith. and F. John Wiley V.092 CN 1 158 CH (methyl) C=C 1.207 CH 1. for the molecules considered here. the part of the selection rule that corresponds to an increase in the energy of the system and is appropriate is AJ = +1 The AK = (26) fact that electromagnetic radiation cannot induce a transition between different values can be understood from the fact that. locate the center of gravity Given the geometry of the CH 3 C1 molecule in Table 53. K K Exercise 56.." . Inc.385 1. . The energylevel diagrams for these two cases are shown schematically in Fig. *Trambarulo.109 CH 1 103 CH 1 101 CH 1.097 1. W.782 1. (24) as a squared term leads to the same energy for a state with positive or negative values of K.F CH C1 CH Br CH CN 3 8 3 HCH HCH HCH HCH HCH C1CC1 110°0' 110°20' 110°48' 109°8' 108°14' 110°24' 107° CH. +1 AK = AK = for Z^ (25) AJ = + for K = For an absorption experiment. Some Molecular Structure Results for Symmetrictop Molecules* Molecule Bond angle Bond distances CH.073 .
2 £ c 6 ~ ° e > . o O E o 3£ *» IS « IO ^ J: «> a* c T > o P> OS E o C 1^ £ \ SI 107 . to in 3 u V IT) >* o> J 4 J 4 J JJJJ : _o — * o E . Q. *• t 4 3 £ ^ — Q 8.f*.
) therefore. [From (1951). Phys. however. F. Trambarulo.. the absorption spectral frequencies are obtained as = 2B(J + for 1) is cm (27) The same pattern as expected. The details of part of the J = 8 —> 9 transition of CF CCH 3 are shown in Fig. Inclusion of centrifugal stretching effects leads. v selection rule AJ = 1 +1 and the energylevel expression of (24). for a symmetrictop molecule was found a linear molecule.108 INTRODUCTION TO MOLECULAR SPECTROSCOPY 1 I K(AK = 0) \i *i l 2 12 3 0v\ 0v\l/ l2 3 A/ 0 — — »z z — *j i — »§ * — »» I FIG. (From G. Inc. With the Eq.. 513 distortion. (28). Princeton. 1945. and W. The alterna FIG.) J. = is 8 —* 9 transition of CF 3 CCH showing the effect of centrifugal centered at about 51. N. interact with the electromagnetic radiation. the centrifugal distortion effects are resolved." D.2D KJ K\J + If 1)  4Dj(J + l) 3 (28) for the transition frequencies. Part of the J (The diagram R. 512 I I I l_ transition absorptions The effect of centrifugal distortion on the rotational of a symmetrictop molecule according to Eq. the splittings indicated in Fig. Rev.J.] W. to the expression v = 2B(J + 1) . Herzberg. "Infrared and Raman Spectra.800 megacycles/sec. 512 will provide a distinction between linear and symmetrictop spectra. §13. Sheridan. f. and the value of K does not change. therefore. Gordy. 82: 58 . Van Nosfrand Company.. Anderson.
such a procedure requires the assumption that bond lengths and angles are independent of isotopic substitution. it introduces some uncertainty when data for the v = states are used. Some ASYMMETRIC MOLECULES A molecule with three different It will moments of inertia is known as an asymmetrictop molecule. 59. siderable accuracy. be apparent that the rotational energies and. In this way one can often deduce additional bond length and angle data and. 58. of course. Another procedure is to make use of different isotopic species to obtain further momentofinertia data. i 21B Vl 21 c (29) . Results from Analyses of Rotation Spectra of Symmetrictop Molecules According^to Eq. P% . many structures have been determined with conof these structures are listed in Table 53. for example. to provide data on the internuclear distances between the heavy atoms and to supplement these data with the momehtof inertia result from the rotational spectrum. The rotational energy of an asymmetric system can be expressed by the general. insufficient for the determination of the bond lengths and bond possible to angles of a polyatomic molecule. classical formula Va 27 A —+ ^ #: + #. Again it is make use of As we will see. the rotational spectrum of such a molecule are expressed with much more difficulty than was encountered for linear and symmetricThe Energy Levels of an Asymmetrictop Molecule top molecules. either xray or electron diffraction. measurement of the frequencies of the rota tional transitions of a symmetrictop molecule leads only to information about the two equal moments of inertia about the axes of the molecule perpendicular to the figure axis. however.ROTATIONAL SPECTRA 109 tion in intensity of the different statistical K components is due to the different weights of the levels involved. (27). In spite of the inadequacy of the data provided by the rotational spec trum of a given molecule. fix the positions of the hydrogen atoms of a molecule. This one determined quantity is. therefore. while this would undoubtedly be true to a very good approximation if the atoms were at their equilibrium positions. diffraction data.
This "adiabatic" argument shows that. the 2/ + 1 states will still exist for each value of J. K is not a "good" quantum number for an asymmetrictop molecule. for example. . —J for the is lowest energy of the set to +J for any but the lowest few energy levels of an asymmetrictop molecule. the 2J + 1 states +J.) One can now schemat ically connect the energy levels of the symmetrictop patterns to show what would happen if the moment of inertia along the B axis were grad . which has. The nature of the energy pattern can. it is customary to keep This index t is of them by adding a subscript to the J value. however. and the corresponding symmetrictop diagram for the prolate symmetrictop molecule. closed general expression available for No at the right of the diagram. Solution of the Schrodinger equation for an asymmetrictop molecule has been accomplished. must be expected that J there will be 2 J + 1 states. therefore. is drawn at the left of the diagram. with A = 3 and B = C = 1. 514. is quantized according to the relation Total angular momentum = y/J(J + 1) 5 (30) Unlike the symmetrictop case. no component of this total angular momentum is is quantized. and for each value of J it is found that there are 2J + 1 solution functhere is tions.e. no available to characterize the rotational energy states. however. in characterizing these states arises in the asymmetrictop case because no quantized component of this total angular momentum. be illustrated by a diagram of the type shown in Fig. K= . it quantum number other than J As for symfor each value of metrictop molecules. The energylevel pattern for an oblate symmetrictop molecule. i. We are brought to this result by the fol lowing statement. for example. usual agreement to have the (Notice that this procedure violates the A axis unique. +1. although the problem is one of some difficulty. characterized A small change of the moments of inertia from a sym metrictop system. Since these 2 J + 1 solutions for a given not characterized by quantum numbers.110 INTRODUCTION TO MOLECULAR SPECTROSCOPY Conversion to a system which obeys quantum mechanics requires the stipulation that the total angular momentum. can be expected only to 0.. each with J are track own energy. for a given value of J. with A = B = 3 and C = 1. even for an asymmetrictop moleA difficulty cule. shift the energy levels and would not eliminate any of the allowed states. by . . its given integers going from the highest energy. ±2. ignoring nuclearspin angular momentum.
The labeling of the asymmetrictop energy levels in Fig. Inc. W. the interconnecting lines of Fig. A correlation diagram illustrating the (From W. "Microwave Spectroscopy.) Prolate symmetric top Asymmetric top symmetric top . 514 sugt. and the index t can be entered as J value as shown. V. 514 rotors. 514 are drawn.ROTATIONAL SPECTRA 111 ually changed so that B varied from the value of 3 to the value of 1. Trambarulo. In this way. Gordy. Smith. F. and R. a subscript to each The J value of any line remains unchanged. Oblate New York. 1953." John Wiley & Sons. gests an alternative to the use of One can notice from that figure that the value of t is related to the K values of the symmetrictop moleenergylevel pattern for asymmetrictop FIG..
514 necessary to retain all three possible changes in J shows. k it is convenient to have an abscissa scale to show the "degree of asymmetry. 11 : molecules of various degrees of asymmetry. k = —1. applicable also to asymmetric tops. k lies and therefore A > B between +1 and — 1.te — KobUte (31) With t this recognition one can identify an energy level with the notation oblate . M. 510. W. 512. Phys. Using the expressions of Table several correlation lines of Fig. For all asymmetric for the most asymmetric case. King. 3 K notation as Jk .112 INTRODUCTION TO MOLECULAR SPECTROSCOPY cules according to T ~ Kprol. Both types of notation may be . even in absorp tion spectroscopy. Ray [Z. For quantitative treatments of asymmetrictop molecules. energies have been tabulated for Chem. k and therefore A <B= C = = 1 . such A = B = = 0. or with the notation is 3i. 78: 74 (1932)] and by G. and P. Chem. Cross 27 (1943)] for the energy levels of some of the lower J states of asymmetric molecules are listed in Table 54.. Hainer. 1. tops. 2. Phys. (33) readily verify that for an oblate symmetric top with Ia> Ia as Ib = Ic C. the energy levels are not necessarily ordered ." A convenient K quantity is the asymmetry parameter defined as = 2[BUA + C)] A C C are the rotational constants where A. • as J. it is Now. S. while for a prolate symmetric top with < Ib = 3. . 54. B. King. 17: 826 (1949)]. plot to scale the first Exercise 57. [J. For states with higher J values. prolate' encountered. 11 : 27 (1943). Hainer. Spectra.. The expressions that have been obtained by B. and Ic. k = C. Physik. Selection Rules. R. See. and A = 5JV One can cm 1 etc. for example. and Cross [J. ±1 encountered with symmetrictop molecules. as Fig.k „ Thus a level 3_2 with the with the K notation. and Structure Results for Asymmetrictop Molecules The selection rule is AJ = 0. since. C.
B) + 9(A . to a corresponding complexity in the observed rotational spectrum.B)' + 9(A .C)* + (A .C)' — (A .C) .By + (A .C)(A .C)» + (A . The complexity of the energylevel diagram of an asymmetric top leads. quite a few molecules have been TABLE 54 metric Tops Some Closed Solutions for the Rotational Energy Levels of Asym Jku — i)K<«_+i) Energy ero t.e. Selection rules for the t values have also been given. difficulties characteristic of In spite of the encountered in analyzing the rotational spectra of asymmetrictop molecules.B)* + (A .C)(B .B)(A 4A 5A 5A 2A 10A 4« 4„ 4„ 4 2S 4.i 2„ 2o2 +2B +2C +2 V(B +B+C A + IB + C A +B +4C 2A+2B +2C 2 V(B 2A 4A C) s + (A  C)(A  B) C)» + (A  C)(A  B) C) C) C) 3 30 3 31 32.B)(A .ROTATIONAL SPECTRA in the order of J. 3 S2 3i 2 3i„ 3o.C)(B 5A +2B +5C +2 V4(A . The this work is the splittings of the the J and and the relative intensities of the components are t numbers involved in the transition. with the selection rules for induced transitions.C) + 5B + IOC + 2 V4(A . The assignment of each of the observed lines to the energy levels involved in the transition is now a problem of some splitting observed lines difficulty. 2. and their derivation..C) + 5B + 5C . are sufficiently elaborate so that it is adequate here merely to indicate that they are treated in ref. 5A + 5B + 2C + 2 V4(A .C)' .(A .2 Vi(B . Ooo lio ln loi A +B A +C B +C ' 2M 2*i 2.C) +5B +5C +2 V4 (B . These depend on the components of the dipole directions of the principal axes. The most helpful aid in when a Stark field is applied.B){B .C)(fi . 113 an energy level with a higher value of J may occur at lower energy than that corresponding to a lower value of J.B)(B .2 VA(A .B)(B 2A + 5B + 5C + 2 V4(B .C) + 2B + bC . i.B)(B .B) + 10B + 5C + 2 V4(A .2 V4 (B .C)(A .2 V 4(A .9(A .C)» + 9(A .C) 1 ) ! .C)» + 9(A .2 V4(A ^ C)i .B) + 10B + 5C .C)(B .C) + 5B + IOC .9(A .» 5A 5A 10A 4m 5A 5A + 48 + iC + 5B + 2C .2 V4(A . moment of the molecule along the The selection rules themselves.C + 5B + 5C .
G.. New York. Van Nostrand Company. Gordy. energy studied and the observed transitions identified with the participating When this is done.082 CC 1." D. Trambamlo. Smith. Trambarulo: "Microwave Spectroscopy.. John Wiley 3. constants and thus the three three data. Hainer.436 SO 1.472 CO 1. 1953. especially moments These when supplemented by isotopic studies. V.114 INTRODUCTION TO MOLECULAR SPECTROSCOPY TABLE 55 Some Molecular Structure Results for Asym metrictop Molecules* Molecule Bond angle Bond distance S0 2 0.: "Spectra of Diatomic Molecules. W.. 2." John Wiley & Sons.J. G. W. New York. Princeton.. .J. Inc.. 1953. Inc." D. V. one can use the equations of Table 54 levels. Inc. W.. "Microwave Spectroscopy. 1950. 1945.. Inc. Gordy. or the tables of King. N. CH2CI2 0S0 00 '0 HCH C1CH 119°20' 116°49' 112°0' 111°47' 116°41' SO OO' CH2 \/ CH2 HCH COC 0S0 FSF 61°24' S0 2 F 2 129°38' 92°47' 1." & Sons.772 CH 1. and R. F.278 CH 1. Herzberg. Some of the results that have been obtained from such studies are shown in Table 55.: "Infrared and Raman Spectra. F. provide a good basis for the deduction of the structure of the molecule.570 * From W. Smith. Princeton. and R. Van Nostrand Company.433 1. and Cross to deduce the three rotational of inertia of the molecule.370 SF 1.068 CC1 1. Herzberg. PRINCIPAL REFERENCES 1. N.
the motions of the system might best be described in terms of the motion of the n. One can recognize that if the bonding of the hypothetical molecule with very weak bonds were to be gradually strengthof these coordinates. absorption spectrum in the infrared region. simultaneous vibration dealt with. of changes in both vibrational 61. of solutions and pure In the follow ing chapter. 21. like that of Fig. For actual molecules the 3n cartesian coordinates are not. in fact. is 115 .THE VIBRATIONS OF POLYATOMIC MOLECULES An introduction to the classical and quantummechanical nature of the vibrations of polyatomic molecules will retical now be presented. will be studied. and rotation of molecules will be and the absorption spectra of gases. nearly independent individual atoms. could be described in terms of the 3w cartesian coordinates. that Any additional displacement or velocity. very convenient. Each of these 3n coordinates represents a degree of freedom of the system in that any arbitrary displacement and velocity can be given along each these 3w. This theodevelopment provides the background for an understanding of the liquids containing polyatomic molecules. or on the molecule. If three cartesian coordinates are used to describe the position and motion of each atom. beyond imposed on an atom. The Number of Independent Vibrations of a Polyatomic Molecule If a molecule containing n atoms is imagined to have these atoms held together by extremely weak bonds. there will be a total of 3n such coordinates required for a description of the set of the n atoms. which show the effect and rotational energies.
— 5 internal degrees of freedom can be further described. as will be shown in the following section. For such molecules. to use three coordinates to describe the motion of the center of mass of the molecule and.116 INTRODUCTION TO MOLECULAR SPECTROSCOPY it ened. 61. a particular set of coordinates. depending on what system. as in Fig. there will be left a total of 3n — 6 degrees of freedom which must be accounted Displacements and velocities for by internal coordinates of the system. displacements are given to the particles of the The general motion of such a system can be analyzed in terms of a set of internal coordinates. and. independent One likes. for example. The degrees of freedom are merely reclassified. If the molecule is linear. three coordinates to describe the rotation of the molecule about this center of mass. in accordance with these 3n — 6 coordinates constitute vibrations of the molecule. or vibrations. the number of vibrations of the molecule. as the previous section indicates. The Nature of Normal Vibrations and Normal Coordinates A collection of many mass initial units interconnected by springs can be expected to undergo an infinite variety of internal motions. . translational It follows that if three and three rotational degrees of freedom are recognized. one will see that none of the 3w degrees of freedom is destroyed. which are very important in vibrational problems. for nonlinear molecules. For any molecule there furthermore. 3n 62. — 5) independent internal coordinates. three overall translations and two overall rotations are subtracted from the total of 3n degrees of freedom. simple example. called is normal coordinates. The nature of normal coordinates. molecular rotations can occur only about the two axes that can be drawn perpendicular to the molecular axis. They are. of the molecule rather would become increasingly important to recognize the existence than to treat the system as consisting of n rather particles. which especially convenient for the description of the vibrations of a system. It is now necessary to see if these 3n — 6 or. and there remain 3m — 5 internal degrees of freedom. The number of internal degrees of freedom of a molecule corresponds to the number of ways in which the atoms of the molecule can be given independent displacements and velocities relative to one another. there are 3n — 6 (or 3n is. If one imagines a gradual strength ening of the bonds between the atoms as one goes from a set of inde pendent atoms to the rather firmly held set that constitutes a molecule. can be illustrated by a If Consider a particle of mass m held by springs. for linear molecules.
THE VIBRATIONS OF POLYATOMIC MOLECULES 117 in the plane of the spring system. can be deduced from the potential and kinetic energies which are written as (1) (2) U = i&*a: 2 + ik yy* T = $mx + \<my 2 % where the springs have been assumed to obey Hooke's law and to have and kv [It should be noticed that the spring system. One can see that the potentialenergy expression of Eq. ~ k** and /. nate. arises only from a displacement in the x direction and. oneparticle spring system. . x and y coordinates are set up as indicated. 61 A twodimensional. the equation For the x coordi 1(^ +^= dt\dx) ^ dx gives (3) mx This is + fcjX = (4) the now familiar differential form which indicates vibrational FIG. forces in the y direction only from y displacements. the motion of the particle. for small displacements. (1) leads to this result by forming J' =  (j^r \*v). Kv] can The motion of the particle now be deduced by applying La grange's equation to the coordinates of the system. is such that the restoring force in the x direction force constants k x . similarly.
for the y coordinate. therefore. will execute the displacement coordinates of more than The normal coordinate will. i. in general. if Hooke's law is applicable. and a displacement of the system according to the normal coordinate mil lead to a simple motion in which all the particles move in phase with the same frequency and. often (For some systems. lead to Displacement in any other simple sinusoidal vibrations along this axis. and such motion is said to be a normal vibration or Motion of the normMl mode. Similar normal coordinates and normal vibrations will exist for more complicated vibrating systems. however. and they are the normal of the particle to be most coordinates for the system. . In terms of this coordinate. The x and y coordinates allow the motion simply described. as we will see. the vibrational motion of the system will be simply described.e.118 INTRODUCTION TO MOLECULAR SPECTROSCOPY motion with the displacement as a function of time given by x = Ax sin (2nvJ + <p x) (5) where '• = «T \P (6) Similarly. that can be interpreted in motion direction will lead to a complicated Initial terms of sinusoidal components in the x and y directions. In such cases a normal coordinate will. and the motion is described when the appropriate values of A x and A v for the displacement are used. consist of one atom of the system. Lagrange's equation gives my + kyy = + (7) which leads to the solution y = Ay sin (2wv y t <p v ) (8) where *v = */^ 2* 0) displacements along the x or y axis will. therefore. be a complicated combination of atomic displacements. Eqs.. it will be necessary to modify this description of normal coordinates so that it is lead stated that displacement according to the normal coordinate will simple harmonic motion. (5) and (8) will simultaneously apply. system along a normal coordinate consists of simple harmonic motion.
side by side. 62. 61 has been turned through an angle as shown in Fig. did know the convenient directions in which to choose coordinates. 61 it is clear that displacement FIG. the allowed energies for the two normal modes treated quantum mechanically. k x = 1 X 10 8 and k y = 2 X 10 6 dynes/cm. 62 can be recognized by considering the restoring force that would act for displacements along the coordinate axes. In Fig. 61 with = 1 atomic mass unit equivalent to a hydrogen atom. what atomic displacements will lead to normal modes of vibration. 62 The spring system of if Fig. The essential between the axis choice of Fig. showing. M Calculate the classical natural frequencies of this system. values that are of the order of magnitude encountered with chemical bonds.THE VIBRATIONS OF POLYATOMIC MOLECULES t0 all to motion for which the cartesian coordinates of the atoms of the molecule will execute simple harmonic motion.. 61 described by an awkward coordinate system initially such as might be used the directions of the springs were not . 62. Restoring force components and resultant for a displacement along x' 9X . . The simple calculation that is necessary will.¥ known. in fact. 61 and that of Fig. The axes relative to this orientation of the system are labeled x' difference and y' as in Fig. Let us suppose that the spring system of Fig. therefore.) Exercise 61 Consider a spring system like that of Fig. therefore. 61 in way in which one would proceed if one did not know not advance the directions along which the springs lay and. What would be the wave numbers of the radiation absorbed in the fundamental transi(a) (b) Draw a diagram tions of this system? the It is informative to consider also to a description of the vibrating system of Fig. anticipate the problems that will arise with molecular systems where one generally does not initially know what coordinates are the normal coordinates and.
a selection of t = im(x')* + iMy'y x' (12) Lagrange's equation for the (12). solutions. (4) where the by trying the functions (15) = = (2rrf A' x sin + + 9) and y' sin A' y (2irvi <p) . from displacerestoring force in the a? direction can result. displacement along lead to restoring forces nate axes will not. and since the coordinates is therefore. coordinate now gives. we can again look for vibrationtype frequency of vibration x' is v. Fig. (13) and.Ktf = coordinate. convenient than those of Eq. that axis and directed toward the form of the potential written in Eq. function has the potentialenergy the that along these axis. for the my' KX  Kv'  o (14) Although these equations are clearly less and Eq. 62. (10) and ma?  k'j! y' . or the y axis. (10). It follows form u = %KWY + KWKv') + iW) 2 (10) the restoring forces This potential expression can be seen to lead to one has Thus 62.]20 INTRODUCTION TO MOLECULAR SPECTROSCOPY force acting along along the x axis. as the force vectors suggest. of orientation expected for the spring (id A ments in both in advance. terms but contains. coordinates. still are Eq. products cross addition. using the coordinates the kinetic energy orthogonal. in vibrations of this oneparticle Let us proceed now to study the The potential is written in 62. with Eqs. of system. results in a restoring the basis for the was This origin. the x' and y' directions. (7). the coordiFor the coordinate system of Fig. Thus. (1). will where the normal coordinate directions are not recognized is not just the sum of squared lead to a potentialenergy expression which between the coordinates. Fig.
(17) can be calculated. When the spring and particle system of Exercise 61 is rotated counterclockwise by 30 deg with respect to the axes. the appropriate minors of Eq.THE VIBRATIONS OF POLYATOMIC MOLECULES 121 Differentiation twice with respect to time gives functions for x' and y' and. Exercise 62. 4tVto  k' y (17) Expansion of the determinant gives (4rVm The k' x )(4w* v *m  k' ) y 2  (k' xv Y = (18) roots of this quadratic in v are found to be v = ± (*: / + K) + kk k' y y + 404)? Thus the two natural calculated. the roots given or. and values for the ratio of A' can be obtained for each vibration x to A' y frequency. same values are obtained as This rather detailed treatment of a very simple system will later be recognized as an illustration of the procedure that . and see that the in Exercise 61. the directions in which the particle moves in each vibration. the determinant of the coefficients of the amplitudes A' x — A' y must be zero. k' x xv A' k' A' y y = For these equalities to hold. = (16) and 4**.. and x y see that one can deduce. leads to the expressions 4**p*mA' m  k' x x A'  k'^A.*mA'. (b) Calculate the relative values of A' and A' for each vibration. from the given potential function. with Eqs. (13) and (14).is necessary when the relation treated. the potentialenergy function (1. Finally. for other than the trivial solution with = 0. now labeled x' and y'. (19) can be substituted into Eq.e.25 is V = (a) X 10 6 )(z') 2 + VI X lW{x'){y') + £(1. 4rrVm k' xy k' k'. between the force constants and vibrations of a molecule is The immediate purpose of the example is to illustrate that . i.75 X 10 B )(y') 2 Calculate the natural frequencies of the system. (16). alternatively. vibrational frequencies of the system can be by Eq.
with the normal coordinates.. separation of external from internal coordinates is carried out will not can be found in many texts on quantum mechanics.122 INTRODUCTION TO MOLECULAR SPECTROSCOPY is there a convenient set of coordinates. called normal coordinates. Inc. 1935). designated by Q„ of the molecule. 63. from the simple example of the preceding section the important fact that normal coordinates are such that both the kinetic and potential energies of a system are expressed by the sum of terms each involving the square of a normal coordinate. . for the kinetic treating the vibrations of a system. be written for n as particles. treat separately can. however. We know. as in the the three coordinates that describe the motion of the center of mass of the system and the three (or two for a linear molecule) that describe the rotation of the system about the The derivation that shows how this principal momentofinertia axes. B. classical case. however. Furthermore we formally. for example. or will proceed." by L. "Introduction to Quantum Mechanics. coordinates of a molecule. the position of each being described by three is cartesian coordinates. or normal. and the normal coordinates can then be deduced from the amplitudes that are obtained for the natural. It Let us now proceed to investigate the form that the solutions of the Schrodinger equation will take when it is applied to the internal. Pauling and E. such that they allow both Furthermore these coordinates are and potential energies to be written even if these convenient coordinates Finally. the 3n cartesian coordinates are not convenient for describing the molecular motions. Wilson (McGrawHill Book Company. as. as the of the problem known. Quantummechanical Treatment of the Vibrations of Previous applications of Polyatomic Molecules the Schrodinger equation have been restricted to onedimensional problems. New York. be given here. The equation can. vibrations of the system. vibrational. of can be carried through initially solution are not with any other set of coordinates. In this way the Schrodinger equation written ££*(3 + 3 + £f) +l «"* where the function ^ will involve >*=<* As One <2 °> the 3n coordinates. in the classical discussion of the previous section. In general one is in terms of the relative motions of the not initially able to describe these normal coordinates atoms of the molecule. sunt squared terms.
in terms of normal coordinates. is in the denominator of this generally absorbed in the normal coordinate The operator itself. X's are coefficients that X< & ( 22 ) depend on the force constants of the Equation (21) can now be written as 3n6 3n6 x. be a function of the 3n — 6 This manydimensional equation is greatly simplified because of the fact that. not only is the kinetic energy a sum of squared terms.(<23„6) (24) .is given by an expression of the type h2 a 2 &r 2 dQ? The mass term that normally appears kinetic*energy operator Qi. will. We write * = *l(Ql)*l(Q0 or • *3». one can write Since. The form of the potentialenergy function can be shown by writing UiQtQt Q 3n _ e ) = i^Q 2 3n6 + iXzQ + 2 • • • + iA „_. but so also is the potential energy. in general. for the potential energy is the potential energy func tion is terms of the normal coordinates.<2.?  £ X H+ X (5 • • ) * = ** (23) It is profitable to try to separate this equation into 3w — 6 equations. the kinetic energy a function of squared terms.Q§n^ 8 =i X where the molecule. in 3n6 h2 Y* d 2 J/ (21) where the solution function ^ normal coordinates. each involving only one of the normal coordinates.THE VIBRATIONS OF POLYATOMIC MOLECULES 123 The correct form of the SchrQdinger equation for treatment of the internal motion of a molecule in terms of the 3n — 6 (or 3n — 5) normal coordinates can be deduced from the fact that the kineticenergy operator for a coordinate Q..
and the observed spectrum of H It are shown schematically perhaps should be pointed out that in general we do not know the normal coordinates of a molecule. since normal coordinates allow both the kinetic and potential energy to be written as the sums of squared terms. for (We symmetric molecules some of these transitions are forbidden. an absorption of radiation to the v = 1 level of each v = will see that. (23). each pattern cor responding to one of the classical normal vibrations of the system. (25) are. therefore. 24. t such as shown for the onedimensional case in Fig. one gets 3n — 6 expressions of the type or (25) Each such expression is identical to that already encountered in onedimensional vibrational problems. the harmonic Furthermore. therefore. these coordinates. 24. however. the selection rules for transitions between the vibrational states for each normal coordinate are are the same Ai> = ±1 = +1 (26) For absorption experiments. (24) into Eq. A 3n — 6 (or 3n — 5) vibrational energylevel patterns. in fact.) 2 The energylevel diagrams in Fig. The previous derivation of the separation of the vibrational problem does not. the classical solution at the beginning of the previous section) is separable in terms of molecule containing n atoms will have. For some of the simpler molecules they have been deduced. 24. know that there are coordinates in terms of which T and U are both sums of squared terms. The important result has been obtained that. the quantummechanical problem (cf . require us to be able to describe the It is sufficient to normal coordinates. can expect. therefore.124 INTRODUCTION TO MOLECULAR SPECTROSCOPY and substitute Eq. coordinate Q will lead to a set of energy levels with a constant spacing. each normal oscillator wave functions listed in Sec. is When this is done and the entire equation divided through by $. We corresponding to the transition from the energylevel pattern. Since the wave functions written in terms of normal coordinates as those discussed in Sec. 63. With this knowledge one deduces . the pertinent rule to is ( 27 > again obtained. The solution functions for each of the 3n — 6 equations like Eq. but their deduction is often rather difficult.
. like COj.652 em it = l. d4. Here it will be enough to consider the molecule in a linear system. as many do.54Scm 600 cm . One way in which symmetry enters into treatments of molecular systems will here be illustrated by a discussion of the nature of the normal coordinates of a When the molecule has symmetric molecule. are often complicated functions in that they involve the displacement coordinates of many. of the atoms of the some symmetry. The Symmetry Properties of Normal Coordinates Normal coordinates molecule.756 cm 1 A« =3. its The energylevel patterns and schematic representations of the corresponding and a schematic infrared absorption spectrum of HsO.THE VIBRATIONS OF POLYATOMIC MOLECULES 125 is that the vibrational energy of a molecule to be treated in terms of 3n — 6 energylevel patterns. spectroscopy are. A? = 3. A complete classification of the ways in which a molecule can be will be given in symmetric Chap. 63 vibrations is a plane of symmetry. in fact. 8. i. a great The consequences of symmetry in molecular simplification occurs. or all.e. and to recognize that a plane perpendicular to the molecular axis FIG. widespread and profound.
particularly if one has a model to work with. as in In this simple example it is obvious. as shown for a general dis To deal with the s in Fig. 65fc. only the displacement vectors will undergo the symmetry operation. and this operation also shown in Fig. 64 Two symmetry elements of COi: o plane and an axis of symmetry. one introduces Thus the symmetry operation corthe idea of a symmetry operation. It should The symmetry operation for an axis of is symmetry will is rotation about the axis. 65a. 64 in the manner shown. It is customary to such a plane an element of symmetry. In vibrational spectroscopy we are concerned with displacements of the atoms of a molecule from their equilibrium positions. that the coordinates along which the system must be distorted in order for the particles to vibrate in phase and with the same frequency are either symmetric or antisymmetric with respect to the plane of symmetry. be pointed out that the numbering of the atoms be kept by symmetry operations. That symmetry and normal vibrations are related can be readily If one were illustrated by reference to the spring system of Fig. The axis of the molecule of COj is another such symmetry element call of the molecule. Plane of symmetry Axis of symmetry . FIG. unaltered asked to set this system vibrating according to its natural or normal modes of vibration. 64. as in Fig.126 INTRODUCTION TO MOLECULAR SPECTROSCOPY is equilibrium configuration identical on both sides of the plane. symmetry elements. placement of an oxygen atom of relation of such displacements to CO This operation transforms the displacement vector of Fig. responding to the plane of symmetry is a reflection through the plane. one would. 65. distort the system symmetrically. A similar conclusion can also be drawn for the less easily visualized normal coordinates of molecular systems. without any detailed analysis. 64. or antisymmetrically. These displacements can be depicted by vectors. Symmetry operations corresponding to these elements transform the solid arrows into the dashed arrows. Fig.
This is apparent where the velocities of the atoms are represented vectorially and the the magnitude of the velocities of atoms of given masses. if a its molecule has kinetic energy as a result of motion of operation cannot change this energy. 65 A twoparticle spring system with a plane of symmetry perpendicular to the Distortions are indicated that are {a} symmetric axis of the springs. since the normal coordinates represent the displacements of the atoms during a vibration. 66. does not alter the total energy. a symmetry in Fig.. therefore. If a symmetry operation is performed on a distorted molecule. symmetry It follows operation changes the directions of these in space but does not affect then that a symmetry operation. ^Q\ + iX. contain only squared terms. In a similar way. some higher potential energy than the equilibrium configuration. since kinetic energy of it does not alter the potential or a molecule. which has.QJ + iQ\ + &tQi + T ^"3B _6y Sn _e (28) Now. Plane of symmetry LwwB/SAA/1 IaA/S/B^VW (a) (h) . it is first necessary to point out that both the kinetic and potential energy of a molecule must be unaltered by a symmetry operation. and (b) antisymmetric with respect to the plane of symmetry. atoms. it In the previous section is was shown that the vibrational energy most simply expressed in terms of the normal coordinates of the In these coordinates both the kinetic and potential energies i. will is This have the same potential energy as the original illustrated in Fig. ticles that The arrows indicate the initial motions of the par would follow such displacements. they can be represented by displacement FIG. the distorted molecule that is obtained by the symmetry operation distorted one.THE VIBRATIONS OF POLYATOMIC MOLECULES 127 To do this. molecules.e. 66a.
as in Fig. immediately represent two of the normal coordinates of the 67. seen that for the energy to remain constant. change sign. CO 2 molecule. 64 or 66. however. when a symmetry operation or. in the general case all where Thus. vectors such as those of Fig.128 Distorted INTRODUCTION TO MOIECULAR SPECTROSCOPY J~ position Reflection 7* — Diane of symmetry through fnj Reflection through plane of symmetry CM FIG. X< are different. 66 kinetic Illustration that a symmetry operation does not change the potential or the energy of a molecule. One needs only to draw arrows. (28). for example. 67 Arrows indicating the relative motions of the atoms of COa in two normal vibrations with displacements along the molecular axis. . (28) is unchanged as a result of a symmetry operation (29) Q*» ±Q> This important conclusion or is that normal coordinates are either symmetric to antisymmetric with respect a symmetry operation. of Eq. each normal coordinate must remain unchanged <= most. at if is performed. lead to no FIG. With this result one can. which keep the center of gravity of the molecule fixed. it is In view of Eq.
ties of a molecule can be used to classify the normal coordinates accordoperation . The bending vibrations of two perpendicular planes. i. and in this case. when a symmetry For simple molecules with a high degree of is performed. (28). is In later chapters we will sec that such a classification a tremendous aid in studies of the vibrations of polyatomic molecules. (30) gives = HQt + QI) + 1MQ1 + «©+• + (3i) Now long as the FIG. since 2 degrees of freedom of the molecule COs in the are involved and. The previous discussion has been based on the general case where all This the normal coordinates have different values of X in Eq. 68. two is identical sets of energy levels will exist. as symmetry operation does not change the value of Q\ Q. and are such that they are result in a simple unchanged or change of sign. the symmetry properof the normal coordinates. as shown in Fig. provide an The illustration of normal coordinates that have equal values of X. the requirement that the energy remain unchanged during a symmetry operation does not lead. however. The two components of the doubly degenerate bending vibration of COs. symmetry. quantum mechanically. When such degeneracy occurs. to the restriction that Qi —> ± Qi. of direction. one has * = IQ\ + i^Ql + iQl + of i^iQl + • (30) where the equal potential energy coefficients have been designated Xi. the vibration said to be doubly degenerate. 08 the energy remains constant.e. Rearrangement < Eq. . (28). as for C0 2 one can often write down immediately the form In general. supposition corresponds to cases in which all the normal vibrations have different frequencies. frequency expected for each vibration is therefore identical.THE VIBRATIONS OF POLYATOMIC MOLECULES 12? left rotation of the molecule. accordIf the normal coordiing to Eq. Such vibrations are said to be degenerate. nates of the degenerate vibration are Qi and Q%. as far as Qi and Q$ are concerned.. ing to whether they transform symmetrically or antisymmetrically with the various symmetry operations of the molecule.
For nondegenerate vibrations. subject to a restriction imposed by the particular sym metry operation. C. the behavior of normal coordinates with symmetry operations can be stated as 1. New York.. E. 2. Princeton. arrows showing the relative displacements of the atoms of the linear molecule CS2 and the bent molecule H2S in the normal vibrations of each molecule. N. each normal coordinate of the set nates of the can be transformed.. J." Inc.130 INTRODUCTION TO MOLECULAR SPECTROSCOPY FIG. Van Nostrand Company. Decius.. 68 with only the arrows on the oxygen atoms shown. . 1945. into combinations of the coordiset.: "Infrared and Raman Spectra. B. Jr. and P. 1955. For degenerate vibrations. Exercise 63. the normal coordinates remain unchanged or change sign. by the operation. 69 The behavior of de(The figure Q\ Q'z =Q (cos9) + e2 (sinfl) 1 generate motions under a symmetry operation. D. G. 2. Wilson. Draw. C0 2 about its axis where the symmetry by an angle 8 is respect to In summary.. Herzberg. Inc. 69 operation consisting of a rotation of considered. 2.. McGrawHill Book Company.) Within this restriction. PRINCIPAL REFERENCES 1.sin9) + Q2 (cos0) corresponds to an end view of Fig.J. Cross: "Molecular Vibrations.<. C. = «. it is allowed that a symmetry operation produce the transformations Qi » aQ x + bQ t (32) and An illustration of this behavior is given in Fig." chap. to the extent that one can on the basis of symmetry and zero translation and rotation.
716 trates. are patterns energy various vibrational spectra of fairly simple molecules are shown in Figs. to v . The general features of the rota131 . In band can be associated with a fundamental transition in an energylevel pattern corresponding to a specific normal mode of vibration of the molecule. Furthermore. and rotational structure is not observed on a vibrational absorption band. The structure of such absorption bands can be attributed to changes in the rotational energy of the molecule accompanying the vibrational transition. show such illus smooth. structureless absorption bands. be of great aid when we attempt to match up the observed vibrational absorption bands with the normal and this information will vibrations of polyatomic molecules. In the liquid phase no welldefined rotational energy levels exist. welldefined regions of absorption which can. the resolution of the instrument is adequate. and we will see that molecularstructure information can be deduced from an analysis of the structure of such bands. The infrared absorption spectrum of a simple molecule in the gas if phase does not. be attributed to the v = for Typical liquidphase observed.ROTATIONVIBRATION SPECTRA When the infrared radiation absorbed by a pure liquid or by a material in solution is measured. absorption regions of considerable complexity. as Fig. the nature of the rotational structure of a rotationvibration absorption band will be seen to reveal much about the type of vibration that is occurring. We will now investigate in some detail the nature of rotationvibration bands.1 transitions in the the most part. but rather. 21 the previous chapter it was indicated that each absorption and 7la.
UOtSStlUSUBJ* tU&iJfyj .3 3 .
S 5 UOjSSjUISUej) U33 J9cj S2 u. 133 .
+ £) and [J (J + These correction terms however. the selection rule Ay = +1 is appropriate and the fundamental transitions. The complexities that arise in the rotational energylevel pattern of asymmetrictop molecules lead to corresponding complexities in their rotationvibration bands. For absorption spectra. linear. to terms involving are. harmonicoscillator energy expression is deduced. • absorption lines to be calculated. both unlike the situation in pure possibilities rotational absorption spectroscopy. axis. obtained by treating the molecule as a rigid rotor. . and now. and fairly satisfactory. will be dealt with separately. These will not be analyzed here.. i. 2 transition. and symmetric top. With this approximation. The selection rule for J was given AJ = +1. together with an expression for the rotational and allow the energies.) These rules. remain applicable when both types of energy are changed in a given transition. from v as to v — 1. the anharmonic and the nonrigid (v rotor.134 INTRODUCTION TO MOLECULAR SPECTROSCOPY tionvibration bands of the three molecular types. could 2 be introduced and would lead l)] 2 . This electron contributes angular momentum have an about the molecular and this in turn allows AJ — transitions to occur. harmonicoscillator system. of will less importance than a rotationvibration coupling term that be introduced after the rotationvibration spectrum cor responding to the simple rigidrotor. of the vibrational energy levels. one writes (2) The two improved models oscillator treated earlier. energylevel expression is The simplest.e. given in Chaps. Selection Rules and Transitions for the Rigid RotorHarmonic Oscillator Model and vibrational and The selection rules governing the allowed rotational 3. or frequencies. diatomic. Thus the Av selection rules for a rotationvibration absorption band are (1) = +1 AJ = ±1 (Exceptions to this rule are provided by molecules like NO that odd electron. is again often studied. DIATOMIC MOLECULES 71. as before. are important.
the central AJ = and it is known as the Q branch. be a gap in the band center corresponding to the absent to label AJ = which would have had a frequency <a. as we will see. that can be attributed to the v is. . are best shown on a diagram such as that of Fig. + 2B(J + 1) J > J i. and a corresponding set on the lowfrequency side with a similar constant spacing.J=0tov = in the J = transition can be deduced from the gap band and that this value gives the spacing of the vibraFurthermore. the factor the symbol o> \/2t Wm by —* v this notation. that there should be a set of absorption lines. all of There should. some molecules. l. 73. given in Fig. There it is clear. 72. (2) and l/2rc y/kh* by the symbol u>. (4) is One should note that in these expressions the value to be inserted for J the rotational quantum number for the lower vibrational state. the separation of the tional energy levels of the molecule. that predicted on the basis of Eqs. as the above expressions show. From moment of inertia of the molecule can be deduced as they were when the spacing of pure rotational lines was determined from studies in the and the bond length microwave region. = +1..e. = &+ = w  5[(J  . With and the selection rules of Eq. as was done in Sec. the lowfrequency set of lines as the P branch It is customary (For of the rotation vibration band and the highfrequency side as the R branch. 5> S. branch is allowed. 24. + B[(J + 1)(J + 2) . 2.J(J + 1)] where J = 0. . 72. The general shape of this experimental curve (3) in fact. the energylevel expression of Eq.ROTATIONVIBRATION SPECTRA It is 135 convenient to replace. These transitions.e.) The infrared absorption spectrum of HBr shows the absorption. further more. (3) For AJ i> = —1. «/ . .. spaced by the constant amount 25. —> J — l)(/) 1. . which occurs at v = u. 2. of the v and (4) or Fig. and the resulting absorption lines. transitions. . It is clear that the frequency = 0. = = + 1. on the highfrequency side of the band center. 1. = to v = 1 transition. central components of the P and R branches can be identified with the molecular this quantity. For AJ . (1) give the allowed transition frequencies v v = At for the = = 1 transitions as i. values for the quantity 25. 2B(J) where J 1.J (J + = 1)] .
k ' ' u = l . Fig. the components are not found to be exactly constant must be dealt with. 73 The tion can be made that the intensities of the components of the rotation . linear molecule based on the assumption independent of Two features illustrated yet been accounted for intensities of the by the HBr example in and must now be considered.i h t Energy levels  00 1 "5 10 1 '5 1 a i& ii '8 '5 to + i& ii + i& ii + l& II 00 1% II + la. II i=> II i& II < 20B < l« < i» < i* < l<u \1l <1 < < 12B 6B IB 1 1 1 _ * 2fi Spectrum < (P branch) (A branch) FIG. 72 The transitions that lead to rotational structure that B is in a vibrational band for a y. Fig. 72. The Relative Intensities of the Components of a RotationVibration Absorption Band relative intensities of the components of bands such as that of can be understood on the basis of the population of the various The approximarotational energy levels of the ground vibrational state. 73 have not First. second.136 INTRODUCTION TO MOLECULAR SPECTROSCOPY Energy . the relative fact that the spacings of the components remain to be investigated and.
o II 5 S uojsniusuej) jure J»d o E 137 .
on the vibrational energy. that the dependence pretty well . on this basis are not in complete accord with the observed bands. HBr at 25°C. with the rigidrotor approximation. 73. it This term is usually introduced by recognizing that B depends. As was shown in Sec.138 INTRODUCTION TO MOLECULAR SPECTROSCOPY vibration band are proportional to the populations of the rotational energy levels from which the components originate. (Much of the disagreement that will be found because of the experimental problems that arise when very sharp absorption lines are measured. according to the Jth rotational energy level is 2J + 1 Boltzmann distribution. One finds. 6.. since involves the effective treats bond length r. i*'"»<T ( 5) The momentofinertia value for the molecule being studied and the temperature of the experiment can be substituted into this expression. 33. The fact that the spectral transitions derived i. can be attributed to coupling between the rotation and the vibration of the molecule. the popula tion of the Jth (2/ level is Nj = + l)N e. 2. This coupling introduces a cross term in the energy expression for a rotatingvibrating molecule that involves a term of the type (v + $)J(J + 1). Coupling of Rotation and Vibration The energylevel expression of Eq.jr ( J +» h. the spacing of the components is not constant. on this basis.. will correspond rather closely to that observed. Exercise 71.* = (v + i)l/2rc VVm + BJ(J + is 1) (6) where S v implies that the rotational term a function of the vibrational is quantum number. the multiplicity of the and it follows that. and 10 energy levels of Calculate the relative populations of the J = 0. 4. and the value of Nj/N can be calculated for various J values. The intensities of the proportional to the values of components starting from various J values will be Nj/N The absorption band predicted . 8. Thus one usually the coupling by writing *.) 73.e. (2) is based on the assumption that the energy of a molecule can be treated in terms of separate contributions from the rotation of the molecule and the vibration of the molecule. and compare these results with the relative values of log j for the components of the P and R branches is shown in Fig. in fact.
It is these terms that are primarily responsible for the nonconstant spacing of the components of a rotationvibration band and for the asymmetry of the expressions for the B <B P and R x . difference in the frequency of the component of the R J branch and that of the P branch that start from the same value of gives information on the rotational spacing in the upper vibrational level. It should be clear from Fig.5„J(J + + B )J + 0i . J — J + > 1 vB = = i £> + [S^J + 1)(J + 2) . 74 suggests. is a small. to bunch up.1)(J) . important so that the components of the R sufficiently can become farther down to the lowfrequency side. therefore. and as a consequence this branch extends t The high J components of the frequencies than would be lower somewhat are at branch similarly R expected for B x = B = B. 74 that the fii and B . positive constant. vp J —* J — [5x(J 1 1)] S = «+ = ib  (B 1 . and the high J components of the R branch (It can be mentioned that this effect tend. = B. the difference in the frequency of the component of the R branch and that of the P branch that end at the same value of / in the v = 1 level gives . It follows that Ii > Jo and that branches are therefore negative.5„)J (9) Since the higher energy vibrational state has a greater vibrational amplitude. .5 )J + (JSi . The band is then said to have a band head.BoJ(J + 1)] + 25 + (35.) lead to values of Analysis of the rotationvibration band can. it will have a larger effective value of r. branch start moving back to lower frequencies for very high J values.5o)/ x 2 (8) P branch. Such behavior is observed only in electronicvibrationrotation bands. With this energy expression the selection rules give the allowed transition frequencies P = Ae for the v = to v = 1 transition as R branch. The high J components on P branch are at frequencies that are lower than would be expected the basis of 6 = So = B.(v + i) where B„ is the rotational term corresponding to the equilibrium bond length and a. Similarly.ROTATIONVIBRATION SPECTRA 139 represented by an expression of the type (7) B. as Fig.  a. 73. The coefficients of the J 2 terms in the band that of the is apparent in the band of Fig.
The difference frequency of the components shown as heavy frequency of the state. analytically. to values of The combinations that lead by considering Fig. 74 in 5 and B\ can be deduced by taking the appropriate differthe P and R branch components. 74 or. one can deduce values of B„ and B u Furthermore. * t I ! 4 i i _ t I ! i _ I I I i I i 1 1 1 1 ' i I i i 1 i i I 1 ! 1 i i I I 1 o = i i i 1 1 i . ences between the frequencies of FIG. and any effect of centrifugal distortion can be recognized. from various pairs of transitions as a function of J. therefore. values of B and S t can be obtained. the moment of inertia the v = and v = states can be deduced from in rotation absorption band. while the difference in the dashed lines gives the spacing in the v = J 4 =1 1 ) . the rotational structure of a vibration An 1 illustration that the rotational spacing and. From the rotational spacings in the v = and v = 1 levels. solid lines gives the spacing in the v = 1 state.HO INTRODUCTION TO MOLECULAR SPECTROSCOPY information on the rotational spacing in the lower vibrational level.
7668 0. as was mentioned Chap.7417 0. These results correspond to the equilibrium bond length. = B. spectral studies.2838 1.1282 1. by means of the formula B v = 5.429 10. corresponding to the minimum in the potentialenergy curve.3019 0. One then can obtain.655 30. a substitution usually what is studied in pure rotational isotopic substitution of. R be which B can from on level end the same the J 2 starting from P component + J level and a Thus for X x J level in the upper state.993 1.e. 5.010 2. (cm') a.9314 2.7416 1.1118 0.7495 0.7414 0. one has (10) 9b(J) ~ 9p(J) = 25 (2J + 1) starting from the component An evaluated. values of B can be obtained.3028 1. which are deduced. D for H lowers the v = energy level in the potential curve and moves the effective bond length part TABLE 71 way from the value r Dependence of § and the Bond Distance on the Vibrational State of a Molecule [Values of a.(» + j)] Molecule 5.809 45. by writing J in the expression for the R components and J +2 in the expression for the P components and performing the subtraction.102 .295 1 1359 1.7481 1.7702 0.7616 1. H D 2 HD 2 HC1 DC1 CO N 2 60.0187 0.ROTATIONVIBRATION SPECTRA Ml P and R components that start from the same J value.445 1. of I The dependence and therefore and r.275 1. The assumption of constant bond length with to obtain additional momentofinertia data for would appear to be valid to a very good approximaFor the v = state. One This should particularly notice that difference r t is significantly different from r . i. vr(J)  v P (J + 2) = of B.01748 0.094 . polyatomic molecules from which additional bond lengths or angles can be determined.7505 0. the bond length are given in Table 71.049 0. which is tion for the equilibrium bond length.993 1. 25 (2J + 3) (11) From these differences. for at least two vibrational levels. for example.282 1. (cm 1 ) (A) ro(A) 0. — a e (v + i) and the measured values of £„.1307 1. and r. r.. are for the formula 8. — o. makes in itself known when one uses isotopic substitution. on the vibrational level turns out to be appreciable.5909 5. Some results for diatomic molecules Also shown in the table are values for B.097 n(A) 0.27460 1.
27 A. for the fundamental vibrational transition. is apply and the rotationvibration band the vibration is said to be a parallel band. the situation. AJ = + 1.. (a) Draw to scale rotational levels of the v an energylevel diagram showing the = and v = 1 vibrational states. 4. LINEAR MOLECULES Two molecules. results for isotopic species. P and R branches and no Q branch. and the cules. bands can can be obtained for different and these can be safely assumed to remain constant with re compound to r e If rotationvibration isotopic substitution. as far as rotationvibration selection rules concerned. force constant of first five X 10 6 dynes/cm. a certain connected with the vibration set of selection rules. a different set of selection rules apply and the rotationvibration 74. (b) Draw vertical arrows to represent the transitions that would occur in an absorption study of the fundamental band. selection rules are Av the diatomic molecules.) (c) Pick out two transitions. If such that the oscillating dipole moment is perpendicular to the molecular axis. band is said to be a perpendicular band. principal types of dotationvibration* bands occur for linear If the vibration is such that the oscillating dipole moment to the molecular axis. 72. The treatment of the previous bands of linear molecules and leads. as discussed here. Exercise 72.'« for the Hcontaining INTRODUCTION TO MOLECULAR SPECTROSCOPY be analyzed. in no way different and resulting band shapes are from the simpler linear molecules. rotationvibration band has the same appearance. as is done in Fig.e. 75. to values of S x and S a (Values of Si and 5 would be expected to depend somewhat on the particular vibration being studied. and therefore absorption lines. Thus the = +1 and i. and a of 3. 14 X 10 24 g. Parallel Bands of Linear Molecules When dipole the vibration of a linear molecule results in an oscillating that is parallel moment is to the molecular axis.8 an equilibrium bond length The molecule DC1 has a reduced mass of about 1. . (Place the arrows so that the one corresponding to the lowest energy change is at the left and the highest one is at the right. This is as that observed for diatomic mole illustrated in Fig. is parallel which will be discussed in the following section. that could be used to calculate B and two that could be used to calculate Si. the value deduced for B e should be section can also be applied to the .
E u 6 c a _o ee o c o q z o o c o < 8 UOISSIIUSU&ll S IU» J9J d 143 .
as discussed in Sec. momentofinertia data for these nonpolar molecules can be obtained. are examples of vibrations which lead to rotationvibration absorption . linear symmetric molecules can if exhibit rotationvibration absorption bands even the molecule possesses no permanent dipole moment. It is of interest to point out that. the same as that encountered in studies of pure rotational spectra. The statistics followed by the like atoms of molecules such as CO 2 and C2H2 lead. to the decrease in statistical weight of some of the rotational levels or to the absence of some of these °— 5 — 1 »= 1 4—4. 2 J.I It' I "I" 3 !. (antisymmetric stretching vibration) ^Nmhpf 7— 5 J. and that of acetylene.. 55. of course.._. The antisymmetric stretching vibration of C0 2 0=C=0. 76a The energylevel dia gram and like transitions of a molecule C0 2 . in contrast to the situation that exists in pure rotational spectroscopy. H H. is. —C^C— bands. B or /.) The problem of deducing bond lengths from the single experimental result. 444U— FIG. . It is only necessary for the vibration to be such that there is an oscillating dipole moment.T f h '44 INTRODUCTION TO MOLECULAR SPECTROSCOPY the same for all vibrations of the molecule. T . From analysis of such bands.
76b A parallel band of the linear molecule statistics. Although the appearance of the band does not immediately indicate the effect of the missing rotational levels.ROTATIONVIBRATION SPECTRA levels. and this FIG. The absorption band that results from the is shown in Fig. do not have an effect. (The carbon nuclei have no spin and. 76a. CO2 which has a center of symmetry and like atoms which obey BoseEinstein 100 E M C s . 766. 145 The appropriate energylevel diagram for COJ 6 an example . in which the like atoms have zero spin and obey BoseEinstein statistics. The molecule acetylene illustrates the case where Fermi statistics apply since the hydrogen nuclei have spin %. one finds that the spacing of the components is twice 2B and that the central gap is half again as large as the spacing in the P and R branches. asymmetric stretching vibration is also shown in Fig.) The alternation in statistical weight of the rotational levels is like that of H 2. therefore.
+1 (12) The perpendiculartype band addition to of a linear molecule has. components shown in Figs. The appearance of such bands is illustrated by the absorption band corresponding to the bending vibration of N 2 shown in Fig. a central principal new feature. the / —* J transitions . The bands. in R Q branches that are similar to those found in parallel branch corresponding to the AJ = transitions. 77 a and b. 77a The energylevel diagram and transitions 1 i of a molecule like H —Cs=C— H. the Q branch. therefore. 78. in the case of acetylene. to the alternation in intensity of the 75. Perpendicular Bands of Linear Molecules Bending vibrations of linear molecules have oscillating dipole moments that are At. I ! alternation leads. corresponds to the overlap of the AJ = transitions.146 INTRODUCTION TO MOLECUUR SPECTROSCOPY 7 5—— A3 2 1 j1 r __ t o " lift 5 1 4 3 2 l — j 1 j FIG. are in«a direction perpendicular to the molecular axis. If the rotational spacing in the upper vibrational state were the same as that in the lower state. For these vibrations the selection rules for rotationvibration transitions = +1 P and AJ = 0.
c s 6 i X o 1 o 5 E o 2 < uoissiujsuej) )u» J9d .a I E E i o .
840 2. therefore. This The perpendicular angular momentum is tures of perpendicular bands. 78 about the figure axis leads to some additional feaThe lowest J value for the v = 1 state.760 ) 2. 1 J = because the one unit of angular momentum that A perpendicular band of the linear molecule N 20.148 INTRODUCTION TO MOLECULAR SPECTROSCOPY all would have the same frequency Po The fact that Si is smaller than B leads to a slight decrease in frequency of the somewhat Q com ponents with high slight / values. The Q branch usually.820 2.740 2. and the two perpendicular motions of the molecule can be coupled to show that in this vibrational state the molecule has angular momentum as a result of this vibrational motion. 64. FIG.780 1 2. in Sec. for example.720 .) Q branches can be attributed to 100 NjO molecules that are in a vibrational^ excited 2. vibrations of a linear molecule are. (The two satellite state.800 « — P(cm" 2. has a asymmetry. as mentioned doubly degenerate.
Eq. 2. 1. have their oscillating dipole moment either parallel to the unique axis of the molecule or perpendicular to this axis. band can then be represented on the energy diagram of Fig. . best They will occur again in our study of the rotationvibration spectra of symmetrictop molecules and will be discussed at greater length there. these two types of vibrations have different rotationvibration selection rules and therefore different rotationvibration 76.VIBRATION SPECTRA U? the vibration contributes disallows the total angular of zero. . ±1. . . momentum value rotation is split The angular momentum. and A is the rotational constant £ is the rotational constant for the axes perpenThe transitions that constitute the parallel dicular to the figure axis. can be attributed to rotation of the These finer details can. couples with the overall of the molecule in such a way that each J level of the v — 1 state levels that result and the two molecule in opposite directions. Chap. K= where 0. 710. Parallel Bands of Symmetrictop Molecules Vibrations that lead to parallel bands can be illustrated by the displacements indicated in Fig. . ±J / and K quantum numbers. for the rotational energies of a symmetrictop molecule ijK = S J(J + are l) + (I _ S)K> J = 0. (14) . the figure axis of the For such vibrations the selection rules for the rotational changes in a rotationvibration transition are AK = and AJ = 0. moment along the unique axis. 5. All these vibrations have an oscillating dipole molecule. furthermore. the spacing within each set of levels with a given value of K will be almost . 79a. however. be passed over here. which are symmetric tops because of the symmetric arrangement of the atoms.ROTATION. for the unique axis. SYMMETRICTOP MOLECULES The infrared active normal vibrations of molecules like 3 C1 and benzene. CH band shapes. . ±1 if K *0 (13) AK = (14) in AJ = ±1 if K = These rules can be combined with the energylevel expression. If the dependence of the moments of inertia on the value of K is small. As for linear molecules. ±2.
79 Vibrations of symmetrytop molecules that lead to (a) parallel absorption b ana's and (b) perpendicular absorption bands. .150 INTRODUCTION TO MOLECULAR SPECTROSCOPY V .// J FIG.
band from the transitions within each set of levels with a K can be shown as in Fig. that for the higher K values the low values not occur.ROTATIONVIBRATION SPECTRA 151 the same since they are governed by the same term. and J(J + 1)B terms of the energy expression. i. J 11 10 87 6 54. 711.. 710 J do K The energy levels of a symmetrictop molecule and the transitions that are allowed for a parallel band. 2. This follows from the requirement that. since determines a component of the angular momentum while / determines of FIG. One should notice. as both 710 and 711 show. the J(J + l)Bi The contributions to the total given value of Figs.e. 3. 4= f 10" K4 K=3 K=2 K=l K= .
or at least an aver age value of B. Ml i 1 1 1 1 I 1 1 I i L i i I 1 I I i i K=2 K=3 I I 1 1 i L Jh.1 152 INTRODUCTION TO MOLECULAR SPECTROSCOPY 11 J_L 1 I . = for the (except branches P. 712. J must be equal K K contribution which shows no Q branch). Perpendicular Bands of Symmetrictop Molecules Vibrations that have oscillating dipole moments perpendicular to the unique axis of the molecule are shown in Fig. the selection rule for rotational transitions For such vibra accompanying the vibrational transition is AK = ±1 The &J = 0. the components of the P and R branches of the parallel band can be resolved.11 L I 1 I lliii I I I i K=i l I . and R in Fig. in the same way as previously indicated for diatomic With such data the moment of inertia perpendicular to the molecules. These . to or greater than contribution shown feature each this Outside of for any allowed state.iLilUUu^ FIG. 713 shows the transitions that undercan occur in a perpendicular band. The appearance expected for the total band is seen to be borne out by the absorption band of CH 3 Br shown If in Fig. figure axis is obtained. tions. iii J 1 1 1 1 1 ——————— I I I I I i «T =4 ' K =5 ^nULUlh. 79b. 71 1 The components of a parallel band showing the contributions from each of the K levels of the v = state. 77. one can deduce values for 5i and B . the total angular momentum. I— K = . 711 consists of simple Q. The complete band can be of consist subbands stood in terms of a summation of subbands. ±1 (15) energylevel diagram of Fig.
ROTATIONVIBRATION SPECTRA 153 S FIG. while only the contour of the ft branch is obtained. . partly resolved. 712 S . 3 lim J»d 8 The P uoisnuKueji branch is A parallel absorption band of the symmetrytop molecule CHsBr.
and type subR 1. the 0. Similar subbands occur For each initial K value. for K bands result from the AJ = 0. 713 The energy levels of a symmetrytop molecule showing the / 7 transitions that are allowed for a perpendicular band. except K = 0. and so forth. AJ = +1 transitions that occur for a given change in = * one gets two subbands corresponding to the sets of transitions that occur within AK = +1 and AK = — 1. Thus. ±1 transitions. The total band. as the experimental FIG. as shown in Fig. the P. 714. Q. for K = 1 —> 0.154 all INTRODUCTION TO MOLECULAR SPECTROSCOPY K. K= .
E 8 £ » S . 8 a o o a.C M • o II e II „ II II I *7 h^ > A) O _ u . c *•» O 155 .+ II 1 + II I + II I + II II II II * * « * X ^ < < <\ < < < <\ fetj o I.
K = BJ(J + 1) + (A . : strong variation and the intensity ratio of 2 1 is the same as that dis cussed for linear molecules in Sec. Q the other diagram. (Draw the rotational spacings to scale for both the v = and v = 1 states.1. 2. 55. one has For Q branches. for CHsBr. B = C= (a) 0. weak. 715 shows. 715 shows. (A (A . easily resolved. (See ref. indicate the transitions that lead to the branches in a perpendicular type band. be identified with 2{A branches of a perpendicular band can. 3. 1.B)K* that result in the of v (16) A and the changes AK = + 1.nch„ = = &+(A + l) .K*\ w+(AB)+2(AB)K B)[(K 2 iorK = 0. 715 can be attributed to the different statistical weights of the levels with different K values. If and if one again designates the AK = +1 P 0ta .) Exercise 73.B) .\y . AJ = and B are taken as independent band origin by to. weak.B)K for K = Q 1.K* . the Q branches are well sep arated and. 79. (c) On .31 The rotational constants of cm.") but reduce the v = (b) On one diagram. to v = 1 energy difference. of the ing to this result. CH 3 Br are A = 5.08 cm1 and Make two energylevel diagrams.5) (K . For molecules like the larger where A is than B.2. indicate the transitions that occur in a vibration in which the dipolemoment oscillation is along the figure axis. = «+ = + S. in Detailed analysis of the situation will. page 406. usually reveals the Q branches of the subbands superimposed on an unresolved background.156 INTRODUCTION TO MOLECULAR SPECTROSCOPY curve of Fig. however. The basis for the strong. as Fig.1 . The spacing level expression of these Q branches can be calculated from the energy h. (18) The spacing methyl halides. accord— B). The alternation in intensity of the Q branches of the CH 3 C1 band of Fig. like that of Fig.2(1 . (17) ForAK= 1 Sflu—. which there are three rather than two identical nuclei not be given here.
UOISSIIUSUeJI »U9D 1»J 157 .
. The treatment given here. which leads to the expectation of a constant spacing of 2(A — B). Van Nostrand Company. The effect of Coriolis coupling can be calculated. Herzberg. PRINCIPAL REFERENCES 1.: "Infrared and Raman Spectra. N. D. The coupling occurs through Coriolis forces. Obtain the infrared spectrum of a gaseous methyl halide. Princeton. and in general the analysis of a perpendicular band does not yield molecular structure information. . and the basis and effect of such forces branch spacing is are discussed in ref.158 INTRODUCTION TO MOLECULAR SPECTROSCOPY If one investigates in detail a number of perpendicular bands of a given molecule. 4. Only simple molecules can therefore be handled. and see that the values obtained are not identical. Compare with the value of A — B calculated from values of the bond lengths and bond Exercise 74. is. one might find the rather disturbing fact that the Q nowhere near the same in the different bands. that lead to perpendicular bands and the rotation of the molecule. for some molecules. angles reported in the literature. Determine A — B for as many of the perpendicular bands as possible. G.J. which can be combined to show an angular momentum. but the calculation requires a knowledge of the form of the normal vibrations of the molecule. 1945. Inc.. 1." chap. in error because of neglect of a coupling that occurs between the doubly degenerate vibrations.
or vibration. in this connection. molecular symmetry for spectroscopy. and that the symmetry properties of the electronic states an important and helpful characterization of the A now be systematic study of the symmetry properties of molecules will undertaken. Furthermore. 11. To avoid a rather abstract treatment. the appearance of a rotationvibration absorption determined by the symmetry nature of the vibration.8 MOLECULAR SYMMETRY AND GROUP THEORY The previous two chapters have suggested the importance of. for a symmetrictop molecule. of the molecule either must be symmetric or antisymmetric with respect to any symmetric element. in view of our of general of the vibrations of studies of the previous chapters. the ways in which symmetry elements of a molecule are described and the ways in which the symmetry nature of molecular motions is determined and t 159 . that deductions based on the methods presented here are applicable to many areas of molecular spectroscopy and molecularstructure studies. In moreover. we will see that the wave functions that describe the behavior of the electrons of a molecule must conform to the symmetry is band Chap. This study can be divided into two major parts. the previous chapter has shown that. convenient illustrations of the applica tion of symmetry arguments. First. It has been shown that for nondegenerate vibrations each normal coordinate. of the molecule of a molecule provide states. applications methods and results will continually be made to the study symmetric molecules. It should be emphasized. Vibrational modes merely provide.
operations of a molecule form what be shown that the symmetry known in mathematics as a "group. and It is Point Groups 0. such as a reflection through a plane of symmetry.) Various moleincludes a brief statement of these operations. such as CO2 and H 2 metric. which all molecules Its inclusion as an element of symmetry is TABLE 81 The Symmetry Elements and Symmetry Operations Symmetry elements Symmetry Symbol Description operations E a i Identity Plane of symmetry Center of sym No change Reflection through the plane Inversion through the center metry C. treated in terms of the five symmetry elements along with the symbols used to denote them. 81 symmetry element E. and symmetry operations.elements that the molecule has its equilibrium configuration. SYMMETRY PROPERTIES OF MOLECULES 81. Illustration of the have. Second. Symmetry Operations. All molecular symmetries can be listed. Symmetry Elements. These symmetry elements Each symmetry are best explained in terms of symmetry operations. would be trivial. in Table 81." is it will The mathematical treatment that can be given group is for a general. symmetry elements are shown in Figs. cule in necessary to be able to describe in more detail the nature of the symmetry of a given molecule. such as a plane of cules drawn so as to exhibit and 82. element has a symmetry operation associated with it. and some of these applications are illustrated. symmetry elements. is The symmetry of a mole i discussed in terms of the symmetry. then developed to the extent that it is needed for application to problems of molecular spectroscopy." To proceed with a it is systematic study of the consequences of such symmetry. and Table 81 (The distinction between symmetry. are "sym apparent that some molecules. is necessary for an orderly treatment of the consequences of symmetry. Axis of symmetry Rotationreflection axis of symmetry Rotation about the axis by 360/p deg Rotation about the axis by 360/p deg followed by reflection through the plane . or abstract.160 INTRODUCTION TO MOIECULAR SPECTROSCOPY tabulated will be given.
Each combination of symmetry elements that can occur is known as a point group. as we will see. (The term point group is used because. 81. where some of the symmetry cell. or a unit location in the crystal. for the various It is worth while to memorize the symbols used is sym metry elements. as is This is in contrast found in a crystal.MOLECULAR SYMMETRY AND GROUP THEORY 141 become apparent later. it is customary to set up the principal axis of symmetry vertically and.) to a new RG. The onlycomment that need be made about the symmetry elements is that. If a great variety of molecules investigated. if a coordinate system is dictated by considerations that will further used. and aliens. is is a sixfold axis and also twofold axes perpendicular the sixfold axis considered to be the principal axis. . it will be found that only a few different combinations of symmetry elements occur. carbon dioxide. to have the z axis in this vertical direction. to a space group. where there to it. cbc. operations result in a translation of a molecule. 81 Some of the symmetry element* in term* of the examples nansdichloroethane. In a case such as benzene. the symmetry operations that are associated with the symmetry elements of the molecule leave a point of the molecule fixed in space. in diagrams such as those of Fig.
] and ZCi (coincident with the C 4 [Co(NH. E. D. 3S 4 axes). 4C.). and So (all coincident with the C« axis).) c Cg» Cj„ E. (mutually perp. CH 4 [PtCl. Ct.).. t E> C„. »<r« E. BC1.). axis) E. (perp. (perp. 3S 4 (coincident with the C 2 axes).) H2C ===C^CH2 CeH^ D.162 INTRODUCTION TO MOLECULAR SPECTROSCOPY A summary that can occur. on the principal symmetry element of the group and that the subscripts added to some It is of the pointgroup symools further tie in the symbol with the elements in the group. 4 (cyclobutane) C«H 6 (benzene) H. S t (coincident with one of the C 8 ). % . to the C.e.)..j (cyclohexane) Za* 3<r D*»=V» E..0. E. to the C» axis). 2a & (through the S. axis). C%. for the most part. axis). C 3 3C 2 . 3Cj (mutually perp. t E. to the C« axis).]'+ 6C 2 . to the C.—CHClBr t Cs H 0. . 4C». and the symbols used to It will denote these are given in Table 82. 2a. St (coincident with the Cs axis). tually perp. the possible point groups. E. be noticed that the boldface symbols used for the point group are based. HC=CH <*>a„ ah. (trans. «> C s (perp. E E CH. Cs. 6ff„ an. Ct. 4S. i T* O* E. C t 6C. 3C. C 4CS C H. (coincident with the C»). i E. 6a. Ei C„.. Za. 9ir.i CH. OCS C1HC=CHC1 (trans. C t and C. CsSW Cj» T>u HCN. 2 E. of the important combinations of symmetry elements i. NOQ H.). t (mu Xl2C == CH2 D sk Da D. 3C 2 (mutually perp.* (perp. cs C.).a E.» E. 3C 4 (mutually perp. not necessary to commit to memory the symbols and symmetry TABLE 82 Some of the Point Groups That Are Important in Molecular Problems (The number of times a symmetry element occurs is indicated by a number before the symbol for that symmetry element) Point group Symmetry elements Examples c. <r». 3ff„. to the C„ axis). Ct and St (both coincident with C 4 ). i. 3Cj (perp. CHjClj NH„ HCC1. CO. a\ t.—CHC1—CHC1—CH. 4*r„ an.
and the third of a by a clockwise or a of revolution. symmetry implies different symmetry operations. should be pointed out that the C3 axis of a two is a other revolution. will be used in later These molecules. which belong to these point groups. atoms a molecule. One is counterclockwise rotation 82. are illustrated in Fig.. point group (The coordinate systems are drawn in ways that will d„ and CHjCI. at this stage. 82. such as occur in a translation or rotation of the molecule. point group Ci. be convenient for later treatments. Symmetry Operations on Molecular Motions The symmetry elements etry of a molecule in its of a molecule are determined by the geomDisplacements of the equilibrium configuration. can be related to these symmetry elements by means of the symmetry operation associof ated with each element. twothirds a counterclockwise rotation by one of onethird of a revolution.) elements for all the point groups. and the symmetry elements they conIt tain. or a vibration of the atoms of the molecule.MOLECULAR SYMMETRY AND GROUP THEORY 163 C2 \C FIG. The operation symbol C% designates the first of these and C\ the second. 82 The symmetry elements of H 20. It is. symmetry operation to a molecular motion is best by means of arrows along the cartesian coordinates associated with each Atom of the molecule. illustrations. One of the rotations and one of the translations. of the H 2 molecule are so described Application of a first done by describing the motion . sufficient to notice particularly those of the C 2„ and C dv groups since molecules of the type H 2 and CH 3 C1.
(Note that the coordinate system of Fig. 82 is used.) CI y s ^7 . 84 Representation of the translational motions of CHjCI by means of displacement vectors. 84. by +1 for the symmetric result of These two no change or —1 for the antisymmetric result of a reversal in direction. Also. H2O molecule by means of displacement in Fig. since the features associated with degenerate vibra tions are then encountered.'S T. a +1 or —1 In a can be associated with the result of each to the remaining translations symmetry operation acting on the displacement vectors that correspond and rotations of the H 2 molecule. which an example of a molecule with no degenerate motions. similar manner. 83 Representation of the translation T„ and the rotation R„ of the vectors. (As the FIG. .164 INTRODUCTION TO MOLECULAR SPECTROSCOPY £Translation in the direction Y (T ) Rotation about the principal axis in the Y direction (R„) FIG. The symmetry operations corresponding symmetry elements can be performed on the displacement vectors such as those of Fig. 83. one sees that the displacement arrows either are left possibilities are represented unchanged or are reversed in direction. 83. Let us consider the effect of the symmetry operations on the is molecular motions of the to the various H 2 molecule. When this is done for the illustrated motions Ty and Ry for H 2 0. the molecule tration CH 3 C1 is also used as an illusand the three translational motions are represented by vectors first along the atomic cartesian coordiriates in Fig.
R v ." The result. the vibrations of the H 2 molecule will similarly transform. .(fl.(rr ) = (i)(T») (i) and C. for example. When molecules have a threefold or higher axis of symmetry. FIG. for is the motions of Fig. The vectors of Fig. it is found that degenerate motions occur. one recognizes that rotations about the x axis and the y axis will lead to identical sets of rotational complications'. known Draw the cartesian displacement vectors that correspond 2 molecule in the x and z directions and show that to the symmetry elements of the molecule result in either no change or a simple reversal of these vectors.MOLECULAR SYMMETRY AND GROUP THEORY 165 discussion of normal coordinates in Chap. will be recognized that the +1 and —1 trans formation factors that appear in the H 2 example are in fact onebyone as transformation matrices. 85 represent displacements that have suitable symmetry behavior for molecular vibrations. 83. that the vectors change direction as a result of the opera can then be indicated by writing c.Or** ^H H ^H n^ H H . Do the same for rotations about axes in the x and z directions and for the to translation of the H symmetry operations corresponding vibrations of Fig. 85 Displacements that correspond to possible vibrations of the H2O molecule.) = (D(B. matrices and are examples of what are Exercise 81 the . 6 indicated. C*(Tt ) to the and C 2 (ffv ) These symbols are to be interpreted as "the operation corresponding C 2 symmetry element acting on the molecular displacements Ty and tion. it When degenerate motions are considered.) The results of such diagrammatic analyses can be put in a form that more convenient for mathematical extension by writing the operation of a symmetry operation on a molecular displacement as. and these introduce a number of In CH 3 C1. for example.) one recognizes the reason for the use of (2) With this procedure +1 and —1 to represent symmetric and antisymmetric behavior. 85.
symmetry. as a result of the molecular tical energylevel patterns.=>+t{Etx . (Similar arrows can be drawn on the H atoms. as we degenerate motions must be treated as a set properties are now see. x =±T t.XTt *~TX Operation oj x =>±Tx +~&Tt « = jT. 86 molecule CH3CI.) Let us now consider the operations C 3 (T X ) and C 3 (T y ). when symmetry will considered.= OT.+or„ r. This behavior can be illustrated by considering the effect of the symmetry operations of the CH 3 C1 molecule on the degenerate set T x and T v . Operation C% 7"=— 1 x —X?T Operation C\ Operation * „ n+"^?i!r§»v r x + Y«. from the top. A vector displacement indicating the motion of the C. that such and that. or CI. The FIG. 86. H»~TX <r. iden It follows.. The diagrams here correspond to end views. and since the effect of operations on them is a little harder to imagine and draw. 82. certain vibrations of such molecules will necessarily have.166 INTRODUCTION TO MOLECULAR SPECTROSCOPY energy levels because the moments of inertia about the axes in these two directions must be identical. they cannot be treated separately. The effect of the symmetry operations on the translations Tx and Ty of the Cj. they are not included. These will behave in an identical manner to the vectors located on the axial atoms.±T„ 2 * rx =ir. Similarly. atom in the translation Tx and T„ is shown in projection in Fig.Y T r x +±T„ u =^T . of Fig.
The student with no background if follow the developments be able to he learns matrix multiplication enough to understand in this subject should equations like Eq. the result of the transformation can be described in terms of the vectors T x and Ty by W.)  ^ I (r.!£ In a similar if)® describe the effect of the way one can corresponding to the remaining symmetry elements of the symmetry operations C 3 . the vector displacements shown by dashed vectors in Fig. the result of rotating the Ty vector by 120 deg can be described _ C»(Ty ) = + ?£ (T x) (T y) (4) One sees that T x and Tv cannot be kept separate. As the geometry of Fig.MOLECULAR SYMMETRY AND GROUP THEORY 167 C3 operation consists of a rotation by 120 deg. and this is the char acteristic feature of degenerate motions. group by transformation matrices.) as =  i (T. It is therefore con venient to treat their transformation properties together and write the result of the operation corresponding to the symmetry element C3 on the pair of motions Tx and Ty by the matrix equation* *(£)(. and when the operation acts on the vectors of Tx and Ty . . 860 are obtained. 86. (5). with the aid of Fig. writes. Now. 86 shows. the vectors of Tx are not simply unchanged or reversed of in direction. all Thus one C3.) (3) Similarly. for the operations of the group * Only a few features of matrix methods are necessary for the treatments given in this chapter..
_ V3\ 2 1 will play an important role in later enough to recognize that they allow the effect of the symmetry operation to be represented in mathematical rather than diagrammatic forms. Some of the reasons for setting up the system that is actually used will. 1 1 1 1 T. with respect to symmetry operations. Rv) Tx T„ 1 Rx . The behavior of the motions of 2 0.168 INTRODUCTION TO MOLECULAR SPECTROSCOPY and. is 83. 1 1 1 1 1 1 1 R. as we will see later. TABLE 83 Summary of the Transformation Matrices for the Symmetry Operations for the Translations and Rotations of the C 2„ Molecule H2 <r» Csu E 1 1 1 1 C. arise when the motions of a molecule belonging to a given point group are investigated can be arranged in tabular form. a number of different molecular motions. v Since. The transformation matrices For the present it work. or in The transformation matrices that H fact any C 2„ molecule. however. and trans. for R„ they are +1 for the . and a' and so forth. Rot. including for completeness the C3 operation given above. Tz This table summarizes the fact that the transformation matrices are +1 for each symmetry element. . symmetry elements E and C 2 but —1 for a. Symmetry Species and Character Tables A brief introduction can now be given to the ways in which the behavior of molecular motions with respect to symmetry operations is usually tabulated. is given in Table for 83. . not be completely understandable until later material is studied.
with this added notation. Ry. script g for a operation. to the summary 86. not to be confused with the identity symmetry element. instead of the subscripts 1 and 2. point group. and a as TABLE 84 The Transformation Matrices for Cj. and also the various electron orbitals for the molecule. where the completely symmetric set of transformations written. A slightly different notation is used if the molecule contains a center Then one uses. This procedure leads. Thus the first two rows of Table 83 would be labeled A and the second two B. and trans. as we will see later. the sets of transformation C 2r symmetry are given in Table 84. T* Rx. of the symmetry behavior for C 3. Rot. molecules shown in doubly degenerate symmetry species is generally labeled Table E.MOLECULAR SYMMETRY AND GROUP THEORY 169 including vibrations as well as translations and rotations. . It turns out. Molecule with Symbols for the Different Behaviors with Regard to A Symmetry Operations c». R. as illustrated for to tabulate. matrices of a molecule with Thus.1 1 1 1 1 1 1 1 1 1 T. Ax A. to be sufficient for most purposes a group which. Bx B. the two B rows are distinguished by Bi and B 2. The rows of such a table are said to be symmetry species to which motions of a molecule or electron orbitals of a molecule belonging to the point group of the table can be assigned. as can be readily checked from the matrices of Table 85. as Ai and the second row as A%. but merely the sums of the diagonal elements of the transformation matrices. In order to distinguish the two A rows one labels the is first row. leads to degenerate sets of translations and rotations. not the complete transformation matrices. E 1 1 1 Ct 1 1 a. Tz and T„ in the previous section. Similarly. are summarized in Table 85. Ty 1 . The transformation matrices that are found for the C 3. a subsymmetry species that is symmetric with respect to inversion at the center of symmetry and u if it is antisymmetric with respect to this of symmetry. The procedure is to designate a row by A if the transformations are symmetric with respect to the principal axis of symmetry and B if antisymmetric with respect to this axis. transform with the transformation matrices given in one row of such a table. it is convenient to introduce symbols for each row.
8a  a a ^ w h 1 1 + IN ICC 1 ^b" > 1 i >r i w Q. rt — N 1 i ICO >r i 1 ^^ o— Kl .' ' 73 a ° • ^* ' * ." a* &.^^ o h j» iH i— 1 1  b o + N  r\ ' t* 1 in i 1 i 1 o* i >r + £ O *>l« <\ l IIN 1 5 o o + '". <*i win ~~ + ~b° 1 + h I IN ICO o e s i >r + 1 "5 . s 03 6i s as 6.1 rH H + V— + O 00 a CO < 170 » o .
and uses of tabulations such as Tables 84 87 will be the subject. 1 At E 1 1 1 T. Eyring. C. and G. T. Inc. Cross. Walter. J. (It will.. Thus for the C 3. Decius. K. be seen later that each entry for the nondegenerate species in tables such as Tables 84 is and 85 is. is Their is introduction here used. Inc. Van Nostrand Company.. Wilson. this of course. point group one abbreviates Table 86 and writes the information as in Table 87. (T„. (See. "Infrared and Raman Spectra. +1 or —1. of the remainder of the chapter. for example. and P. "Quantum Chemistry.Ty) E 2 1 X triply degenerate set as F or T. CI 1 1 <r» Rot. B. in fact.J. and trans. symmetry elements in the class is then indicated at the top of the column. R. therefore possible to abbreviate such tables only one column for each type of symmetry element. (Rx. C. identical with the transformation matrix. 1945. Kimball. really to be interpreted as the If sum of the diagonal elements of the is transformation matrix.. TABLE 87 The Sums of the Diagonal Elements of the in Con Transformation Matrices of a C>« Molecule densed Form Cg« E 1 2C." McGrawHill Book Company.MOLECULAR SYMMETRY AND GROUP THEORY TABLE 86 The Sums of the Diagonal Elements of the Transformation 171 Matrices of a Cj„ Molecule Cj. itself.) The and identical transformation matrices for the two rotation elements in for the three planes of It is symmetry that can be noted Table 85 are typical. "Molecular Vibrations. intended principally to establish the notation that One will notice in many advanced texts on spectroscopy and quantum mechanics that such tables are given for all possible point groups. 1955. N. Inc.Rv)t (Tx. (R„. A. E 1 c. 1 3ff« Rot.T v ) 2 1 ." D. E. New York. and trans. by showing The number of The and basis. New York. 1 1 1 1 1 A. Herzberg.Ry). J. Princeton. 1 1 1 1 R. and H.." Wiley & Sons. properties. G. the matrix a onebyone matrix.
an extensive digression into some of the general mathematical properties of the transformation matrices and character tables will be made. 84. The introductory summary of molecular symmetry has now been Before making use of the quantities introduced here. Little difficulty will be encountered in checking some of the smaller.172 INTRODUCTION TO MOLECULAR SPECTROSCOPY The first of these references also gives a detailed discussion. groups will be pointed out. symmetry operations that have been treated in the preceding section and can be used to draw a number of important deductions regarding the nature of the vibrations of symmetric molecules and the electronic properties and electronic transitions of such molecules. some of the general grouptheory The Nature of a Group To be a mathematically useful term. The character table for any point group can be deduced by finding the transformation matrices for various motions of a molecule belonging to that point group. or deal with the abstract. nondegenerate character tables by this method. will be made. and it will then be easier to understand the larger character tables containing degenerate types. these tables summarizing character tables for the symmetry behavior are known as the symmetry point groups. GROUP THEORY An apparently rather abstract mathematical subject called group theory turns out to be directly applicable to investigations of a number symmetric molecules. which . of symmetry elements and point groups. which are important for molecular spectroscopy. however. a collection of items. The theory can. and deductions of properties of such groups. usually To be a mathematical group. for example. Some of the mathematical features that can be ascribed to general. given a precise definition than the meaning. many matrix. We will then see how the character table can be used to deduce some results completed. the word group must be more associated with it. a set of elements. The previous treatment of symmetry elements and symmetry of properties of operations provides material with which results can be illustrated. be developed later. since the sum of the diagonals of a transformation matrix is known as the character of the transformation 1948. A number of general properties of character tables will. with examples. that are of great importance in molecular spectroscopy.) It can also be mentioned here that.
MOLECULAR SYMMETRY AND GROUP THEORY 173 can be objects. Q. the inverse of Q (8) QS = SQ = E . makes PQ equal to an element of the group. following conditions are satisfied: a rule for combining any two elements of the group. as we will see. moreover. One must.) 3. must conform to four requirements. There is the combination procedure multiplication although. Commonly one calls 1. and so forth. operations. combined and that the element they form then combined with (It should here be mentioned that the commutative law will not be stipulated as a requirement for a group. therefore. (7) (Note that the element E commute with 4. When the elements P and Q are combined one writes PQ. perhaps. combined accordand then P multiplies the element formed rule. S. for all elements of the group.) is also For every element of the group there must exist an inverse which an element of the group." That they make the term group a useful mathematical concept will be apparent as our study of the theory that can be built up for groups develops. It should. for is example. . R. quantities. which we will designate by E. all elements of the group. . means that Thus if P(QR) Q and R are multiplied.e. the procedure need not be the familiar multiplication. must hold for the combination of elements. . the associative The associative law P(QR) = (PQ)R The righthand side implies that is (6) P and Q are first R. is said to form a group if the A set of elements P. The first requirement for a group is that there be a rule which makes the combination PQ meaningful and. An inverse is defined so that if S.. and the combination of any two elements according to this rule leads to the formation of one of the elements of the group. 2.. be pointed out immediately that these requirements may appear "unreal. The set must contain an identity element. Thus PQ will not necessarily be equal to QP. i. An identity element E is an element with the properties such that EP = PE = P EQ = QE = Q ER = RE = R and does so forth. ing to the appropriate requires law from QR. be careful to preserve the order in which elements are written.
3. 2.e. and the symmetry operations for such a molecule are those corresponding to the symmetry elements E. EC* = C 2 <r. <t' v (9) etc. The associative law tested by investigating such combinations as <ri(<V. The identity operation. i. now see if the symmetry operations conform to the requirea group. and then with the element on the left. <r. ..174 INTRODUCTION TO MOLECULAR SPECTROSCOPY The inverse of the element Q will for the set of elements of which Q sometimes be designated as Q~ l and.)*. i. is written to operate first with the element on the right. the combination CV. is a member to be a group. in which no change It is is imposed on the so that molecule. nized as forming a group if the rule of combination is successive operation on the molecule by the symmetry operations. One can verify.. by operating on various rotational and translational motions of H 2 0.) Let us of ments . for example.1 must (Notice that an element and its be one of the elements of the group.e. has been included in the set of symmetry operations apparent that these operations will constitute a group.)?(«r:C. The Symmetry Operations as a Group The rules that must be obeyed by a set of elements if they are to constitute a group can now be illustrated by the symmetry operations belonging to a given point for a molecule of a particular symmetry. they obey the associative law. The H 2 molecule was seen to have the symmetry C 2 „.. 1.e. (11) . We find Left side = <r' v (C*r v ) (<t' v = Right side = C 2 )* v = E = (*>„ = E <r'„(«ri) (10 > In a similar manner it can be verified that various ways of associating the elements in a sequence of operations lead to the same element. .. EC = dE = C 2 * etc. C 2 <f v and c' These operations will be recogt . .e. Q. is a. i. group..<ri = C 2 = CW£ = C*r. (One generally agrees if. C 2 <r„ or <r' Thus t C2.) 85. inverse do commute. i. that successive operations are always equivalent to the single operation of one of E.
the multiplication of the For the representation to be trice. therefore. verify that the elements of the Using T. and the fourth requirement Exercise 82. It will be found that these sets of symmetry operations similarly obey the group requirements. or of a group and can . postulates. C 2. produces the identity. can be represented erally matrices. every element is that there some other element of the must be such group which. = E tb^is will Thus. by numbers. It is this therefore.e. tion. 86. if PQ — R. In a similar manner. which can be assigned to the elements A set of numbers. obey the group Exercise 84. have shown. C 3v Using T x and T y as a basis. Any deduction that can be made about the elements of a group is. 175 To satisfy the inverse requirement. the symmetry operations associated with other point groups can be tested. such as the symmetry operations of a molecule. number which represents the element R. in fact. as a basis on which to apply the group operations. that the set of symmetry operations conforms to the requirements for a group.MOLECULAR SYMMETRY AND GROUP THEORY 4. binding on the symmetry operations of a molecule. matrices. i. or more genand these can then be combined by ordinary multiplicaor faithful. written before or For the after the first element. and is its not generally be so. C 3 „ group do. Representations of a Group Of particular importance to be put is for the purposes to which group theory is the fact that the elements of a group. involving molecular problems theory to group that ties abstract result symmetry. C 2v group own inverse. = E oW. they are a group.. Verify by drawing the appropriate displacement diagrams that two successive symmetry operations on a translation or rotation of molecule are equivalent to a single symmetry operation of the the 2 molecule and that this single symmetry operation is a member of the H C iv point group. point group we see that C2C2 = E (12) w. numbers representing elements P and lead to the Q of the group must. each element of the is satisfied. verify that the elements of the group obey the group postulates. We Exercise 83.
Thus one would write E (1) (1) (1) (1) (The parentheses are placed about each number because.) Since any combination of symmetry operations leads to one of the symmetry operations of the group. to illustrate this feature. representations are designated by T and the components of a particular V are written under the group Again the symmetry operations of the elements that they represent. in fact. r 2 r 3 and r 4 can be given as . tant representation of the An constructed by assigning C 2t +1 to apparently trivial but. as we will see.. group correspond to this equality. said C 2 „ point group can be used Customarily. (1) (1) (D (1) (1) (1) (1) (D (D (1) r4 (D (D It can easily be verified that these representations all properly represent the set of the symmetry operations of the point group C2 . . they are onebyone matrices. imporgroup of symmetry operations can be each element. (1) (1) c2 (1) (1) / a.17* INTRODUCTION TO MOLECULAR SPECTROSCOPY is properly represent the multiplications of the elements of the group to constitute a representation of the group. Thus the representations Ti. E Ti Vt r. To do this one writes. this is properly represented by the multiplication (1)(1) = (1) (13) Three other onebyone representations (the way in which they were deduced will be shown shortly) can also be written and a table of representations can be constructed. and investigates the ways in which the various representations rpu Thus — of the For the Ti representation: For the r 2 representation For the r s representation: For the Ti representation (D(l)  (1) = (1) = (1)(1) (1) (1)(1) = (1) (1)(1) . for example.
of the C«r v (Tx ) = c' v (Tx ) (U) The operation C*r„(Tx ) of <r„ on Tx will give (+l)Tx and one can . that the combination of symmetry operations corresponds to simple multiplication of the corresponding transformation matrices and that this combination is faithfully represented by transformation matrices. Although the transformation matrices dealt with so far do form . is if = (+i)(i)(2V> = (i)r.. write (15) = C. it is interesting to see that they are nothing more than the transformation matrices discovered for H 2 in Sec.e. and one can write The operation C 2 c«. acts only on Tx and cannot affect the pure number (+1). (r'„ are represented by trans formation matrices. Exercise 85. since as Table 84 shows. for example. is equal to the simple product of the transformation matrices To do this. 83. the operation <r' v on T x changes Tx displacement vectors. It follows. It is necessary to show. C2 is (16) <r„ way one represented the combination C*r v represented by (—1) and by the simple multiple by (+1). A similar test of other combinations will show that multiplication of the symmetry operations is always faithfully represented. they commute. one represents (r'v (Tx) by — 1) Tx . i. This implies that (+1) and C 2 can be written in either order. in each case.(+l)r.. one might investigate the operation equation for C 2 and <r„. of the transformation matrices of C 2 and <r v and this is an illustration of a general result. that the transformation matrix for CV.) In this = (+i)c. to give the number which that representation attaches to a'. While these simple representations might have been discovered by trial and error. This example shows that. by the product therefore. (+1)(1)  (D the direction of the ( Furthermore. We must now see why such transformation matrices should form a representation group of symmetry operations. Test that the representations of the C 2 „ group faithfully represent the multiplication of several pairs of symmetry operations.{T. the product CV.(r„) sees that. = will be properly represented . if C s and a.MOLECULAR SYMMETRY AND GROUP THEORY 177 The result of the four multiplications is seen.
that xh transforms as shown into x»< and = 4 Operation ZB' SO. There are. One sees. Some of these that These genmight be built up from those of the previous table are E c2 o» GO' (lit) \0 1/ G /l ?) 0\ G /l ?) 0\ 1 G /l 1 0\ (O \0 10} 1/ (o \0 o) 1/ o) 1/ \0 G!) and so forth. 87 The xh>. the Zn X 3n transformation matrix for the particular operation. 87. in fact.178 INTRODUCTION TO MOLECULAR SPECTROSCOPY representations of the group. erally are square matrices of various orders. (J t) G ?) G 0 faithful representa Again one can verify that these form tions of the group C 2v . Other large matrix representations can be found. for symmetry operation on the 3n cartesian displacement example. tesian coordinates as a FIG. the operation <r„ on H 2 can be seen to lead to the transmolecule. xii "»^H* . an infinite number of representations. by looking at a diagram of the molecule and the 3n displacement vectors. A particularly important one for a given molecule is that based on the Sn cartesian displacement coordinates that can be associated with the atoms of the Such a representation can be set up by writing the 3n carcolumn matrix and deducing. effect of the <r„ coordinates of" that x H H 2 0. in view of Fig. For example. they do not form the only representations.
B. Thus. and those which involve matrices whose dimensions cannot be further decreased are said to be irreducible representations.0 1. y« zo The corresponding transformation matrices of the for the other operations group can similarly be found. We will find that for any given group. group. Representa tions that can be so broken down are said to be reducible representations. there will be only a . forms a representation of the group. . New matrices can be obtained from these by what are called similarity transformations. 87. Exercise 86. In this way one would obtain a representation of the C 2 „ group consisting of four 3n X 3n matrices. is broken down into representations which dimension. if . such as the 3n X 3n one based on cartesian displacements. there is a square matrix /3. matrix multiplications can be performed to give . Suppose that the set of largedimension matrices E. could be built up to give any number of representations involving matrices of larger Of much more importance is the process by which a largedimension matrix representation. Xo 1 . Verify that the transformation matrices for a given molecular motion provide faithful representations of the symmetry operations of the C 3 . although there are an infinite number of reducible few irreducible representations. representations. consist of matrices with as small a dimension as possible. Reducible and Irreducible Representations In the preceding section it was shown that representations of small dimension.MOLECULAR SYMMETRY AND GROUP THEORY 179 formation matrix ' Xh Vh zh 1 ' xH 1 1 ' Vh zh Xh' xH 1 Vh> zh' = 1 1 1 Vh' Zh' x y« «o . . such as those involving onebyone matrices. A. Do this for the onebyone matrices that are obtained when T z or R„ is considered and also for the twobytwo matrices that are obtained on the basis of (T X .T V ) or (R X) R y ). C. with an inverse /J 1 .
. . '(ai) might be possible to find a transformation matrix such that are of the form . on the by p~\ Since /3/3. B' the eleif B = did. fti. one has by /3 and AB = D Thus A'B' = D' is (22) equivalent to AB = D. which led we substitute the expressions necessary to show that (18) and D' into the equation A'B' = D' and obtain p. To do AB = D. c h If such transare square matrices of the same dimension. o 2 62.1 = p~ l P = E. C2 oi. this D'. are square matrices of the same dimension. (19).. . a (o 2) (a. connected to the old ones by a similarity transformation. B' = ( (ft. now be shown It is to be also a representation of the group with A' representing the group element that A did. . .l AWBp /? = /3 p~ l D0 each other. B'.e. one has. (21) left ptABp = P'DP Now we on the right can multiply both sides of this equation. (20) and since EB = B.) (23) /(&i) • • \ .) • \ where .180 INTRODUCTION TO MOLECULAR SPECTROSCOPY A' B' C = = = 0r l A0 p~ Bp l (17) etc. (6 3 ) • • 7 etc. and so forth. . will jS'CjS The new matrices ment that then A'B' to A'. Now all it the matrices A'. i. B'. and so forth. (19) Now since and 1 are the inverse of fl • 8 : = E = °  Q I J 1 . from Eq. and the new matrices. also form a representation of the group.
the previous multiplication result AB = D equal. The transformation matrices deduced for the C 2 „ and C 3r point groups in Sec. . Any other matrix representation can be reduced to one or more of these irreducible representations 88. 61. The symbol g will be used to represent the order of a group. . . sentation is . . Since they are composed of matrices of . respectively. They are. by a suitable similarity transformation. the only irreducible representations for these groups. ai&i 0262 03&3 or A'B' = D' if ' becomes. it by reference to the irreducible groups. C. c 2. the elements of the irreduc behave like the components of orthogonal vectors. From the treatment of the C 2„ and C 3. . C. smaller dimension than A. it should be apparent that their orders are 4 and 6. as will be shown later. since matrices are equal only their corresponding elements are = = = di ds <^3 (24) It follows that ai. . . groups in Sec. 83. the repre said to have been reduced by the" similarity transformation. The Irreducible Representations as Orthogonal Vectors We come now to some of the most important properties common to all irreducible representations. ri(fl)iyfl) = X B o o etc. As the heading ible representations of this section suggests. also form representations of the group. ment of these properties. . if I\(iJ) is the matrix which represents the . The orthogonality property requires that. Vr 1 (/j)r. and so forth. If it is not possible to find a matrix /? that will further reduce all the matrices of a given representation. Before proceeding to a general statemight be helpful to illustrate these properties representations of the C 2„ and C 3„ point First it is necessary to define the order of a group as the number of elements in the group. . or A'.MOLECULAR SYMMETRY AND GROUP THEORY 181 formations can be performed. 83 are irreducible representations. Ci. . .ffth symmetry operation of the group in the j'th irreducible representation. . B. . .(«) = (25) . B'. a2 b 2 . the representation is said to be irreducible.
which have mnth component of the t'th representation for the ation been illustrated.182 INTRODUCTION TO MOLECULAR SPECTROSCOPY Inspection of the irreducible representation table for C 2„ confirms this property which. for example.(B) = R for i *j C 2„ (26) Furthermore. . notice that if the representations only by treating Thus one can first the elements of the representation listed in Table 85. the first row. the sum any component of the matrices r 3 the . for example. which we will now label as r 1. the order of the group. for any component. first column elements of the third representation r 3 the sum of the products Likewise the sum of the squares of the elements of T is again zero. t or of T 2 gives the value of the squares of first 6. These results for the C 2 „ and C 3 „ groups are examples of an important row.(#)mv designates the m'n' component of the jth representation for the R symmetry operation. can be represented as £ r<(fl)r. can be expressed as Ir R i (i2) m „rJ (i?) m v= o i^j (28) I R r i («)„„r (R) mv i =0 m^ m'. each representation of the the group. T^RMR) = I * r X R 2 4 (27) (R)T 2 (R) = 4 etc. the order of the group divided by 2. first general property of irreducible representations. in general. n iA n' I URUURU =  ( 29 ) . all orthogonality condition for separately each component of the twobytwo matrix. the general properties. multiply. If Yi(R) mn designates the R symmetry operand r. entries in the column.. gives the value 3. group corresponds to the components of a vector normalized to a value equal to the order of i. Let us now see if 3„ these orthogonality and normalization properties also apply to the C irreducible representations group where degeneracy occurs and one of the One can test the is a twobytwo matrix. of However.e.
Do the same for the matrices of the representations of the C 3 „ group. one writes (31) Xi(R) =X r. the two the Thus in g dimensional space there can The four onebyone representations of onebyone and one twobytwo matrices. for the sixthorder C3. or their components in a g dimensional space. Exercise 87. Verify that the elements of the transformation matrix C 2 « group behave like orthogonal vectors. If I\ is not neces is the ?'th irreducible representation. therefore. One sees that we now have a a proof that there are no irreducible representations other than those that have already been reported for these point groups. One important 1. The character is of a representation of the Rth symmetry operation denoted by x(R) The definition of the character can then be written as x(R)  T R )™ X m ( (30) where r without a subscript implies a representation that sarily irreducible. It turns out that the characters rather than the complete matrices can be used for many important deductions. This fact introduces a considerable simplification into the applications of group theory. apparent that one can have only as many orthogonal vectors as one has dimensions in space.(fi) mm . One has way of determining when one has found all the irreducible representa tions of a group." given in Table 85. representations of the 89. Characters of Representations The characters of the matrices of a representation are defined as the sums of the diagonal terms of the matrices.and threedimensional it is space. only be g orthogonal vectors. relations are always It is obeyed by irreducible representation will not be given here. the latter contributing four "vectors. Likewise. presented in some detail in characteristic of the set of irreducible representations that exist for a point group can be immediately drawn from the if fact that the irreducible representations. The proof that these Appendix II of ref. group.MOLECULAR SYMMETRY AND GROUP THEORY 183 is where g is the order of the group and U the dimension of the t'th irreduc ible representation. C 2r group given in Table 84 form. behave as orthogonal vectors By analogy with two. form a complete set. they are twobytwo or threebythree matrices.. a complete set for this fourthorder group.
k. (38). need not be followed through.l Q is obtained as X i Qu = = J kl G9)*0»)« p « 22 Kfl*G8)«l p« i — 2 &tiPw = ZP» S. although the similarity transformation if changes the matrices.184 INTRODUCTION TO MOLECULAR SPECTROSCOPY We the have already seen.lP$ is (33) If the process of matrix multiplication involved in written out explicitly. the similarity transformation matrices and normal. involving more matrix manipulation than previous work. in Sec. the characters of the matrices unaltered. Equation (34) can now be written as Qu = I kl 0»)*G8)kP« of the diagonal elements of (36) Finally. (0)k /S (T% where /3' = 09% = (35) indicates the transposed matrix. Now we will show it that. [The derivation. 87. since is presumed to be an orthonormal matrix. product of two rows of the orthonormal matrix .] We wish to show that if P ity transformation and Q are matrices related by the similar Q = pWe then (32) xq = xp $. (37) [The Kroeniker defined such that Su equals 1 for k = I and for k ^ I. Eq. are orthogonal leaves. the sum i. one expresses the t'th diagonal element of Q by (34) Qu = X u /3 G8" W*i(/8)« Now. The student can proceed directly to the result. has been used to express the result of £ i (j8)«(/S)k which corresponds to the 0\. that is if a set of matrices that constitutes a representation subjected to a similarity transformation new set of matrices that is obtained is also a representation for the group.
Alternatively. Inspection of the characters of the representations of the C3. same as the character of the reducible This result will be of great value. like the representation matrices themselves. Similarly. the product of the characters of one representation by itself can be written out. one can say that the sum of the characters of the irreducible representations that can be obtained from a reducible representation will be the representation. We have. of the partially diagonalized matrix will be the same as that of the original matrix. one maintains the orthog onality result of the expression because of the orthogonality of the components that are summed when the characters are formed. ? where k Ti(R) nm T.= R The proof from the results ff (40) of these character relations is easily given. in terms of the diagonal components .. C 2„ and groups will show that the characters. If the sums of the diagonal elements are first now taken. on irreducible representations. in view of Eq. We see that if the breaking down of the large reducible representation is depicted by '( ) o ( ( ( the character. (29).) ) ) MOLECULAR SYMMETRY AND GROUP THEORY 185 We xq have now reached the desired result = xp (38) This important result will be used when we investigate the irreducible representations that can be obtained from a reducible one. their behavior can be written as 5) The general expression for Xi(R)xAR) = for i *j (39) R and X»(«)»(«). behave as orthogonal vectors.{R)„w = for i ^j (41) TiiR^ViiR)^ = f is (42) the order of the matrices of the ith representation.
This follows from the fact that the character of the repreis sentation of the identity operation tion. i. Exercise 88. a' v identity operation. it will In almost grouptheory applications. tion. X \{E) + the X \{E) + + xKE) = g (44) where c is number all of representations in the group." as discussed in Sec. belongs to a class by . given 86. symmetry operaany other operasymmetry The and the operations <r„. that the representation contributes. and those for a v . behave in the manner indicated by Eqs. belongs to a class by itself. (40) is obtained. see. in Tables 84 (40).. 88. each itself. Classes One fies further concept. 810. It is interesting to notice also that one can relate the number of irreducible representations to the characters that appear in the column under the identity element of the group. <s'„ and <r" are also identical. (44).e. as we will be possible to make use of the characters of the representations and the relations involving the characters that have been obtained here rather than the representations themselves. one can write. therefore. The C z and Cjj are said to belong to one class of . that of classes of symmetry operations. having characters unlike that of <r" and are said to belong to another class. 88. equal to the order of the representagives the The square of this number number of ''vectors. In view of the discussion at the end of Sec.186 INTRODUCTION TO MOLECULAR SPECTROSCOPY that are involved as / / Ti(R) mm Ti(R) mm I = k m = g (43) In this way the result of Eq. (39) and . one sums over the various operations of the group. Verify that the C 2t> and C 3„ group characters behave according to Eq. Exercise 89. and Verify that the characters of the C 2 „ and C 3t groups. (39) and (40) give properties of the characters of two representations as one goes across a character table. The expressions of Eqs. since all four operations of the C2 „ group have different columns of characters. Inspection of the character table for the C 3„ group shows that the characters of the representations for identical C3 and C\ are operations tions. Similarly. simpli grouptheory applications.
aE ^C\C E~ i = a 3 = C 3 = CI c\ (CIV'CICI ff^CJir. C 3 and C\ to show that they C 3l) . the previous result that a similarity transformation leaves the character of a matrix unchanged can be used of is to deduce that the representations of all members a class have the same characters. It follows that these this In way all two matrices must have members of a class can be seen to have the same characters for all irreducible representations. This general definition can be applied to constitute a class in the group investigated. = 3 C\ = = C\ (O'cv: = c\ (oi)>C*: = C (O'CX = c 3 w:y*cy: = c 3 We see that. X as <r„. . The number of elements in the class is. Thus C 3 and C\ constitute a class. When the character table is written with the characters for the different classes exhibited rather than the character for all symmetry . and one will find that rotations such as C 3 and C\ and reflections through similar planes such Cj. is a member of the set. In a similar way one could verify that E by itself and a v a'„ and a" form classes. The detailed working out of the combination X~ PX is not usually necessary in order to decide which symmetry operations belong to a class. say of the group which. It is this result that allows character tables to include only typical elements of each class rather than all the elements of the class. Geometric consideration will usually be sufficient. however. When the multiplication of the representation matrices replaces the corresponding combinations of symmetry operations. l In abstract group theory a class consists of a set of elements. The matrix of each member of a class transformed into another member of the class by at least one of the transformation matrices. <r'„ <t" constitute classes. indicated as was done in the and C 3v character tables. The following combinations are is and the operation to which they are equivalent noted E*C 3E = C 3 Cj 1 C 3 C 3 = C 3 (Cfl'CCJ P7K7*r. when C 3 and C\ are subject to a similarity type operation by all the elements of the group. P or Q.: MOLECULAR SYMMETRY AND GROUP THEORY 187 P and Q. the same character.e. when subjected to the operation X~ PX where and X~^QX which is X is in turn all the elements of the group. the result is either C 3 or C\.. gives a result i.
Eqs. one could find a similarity transbefore irreducible representations appeared explicitly. of its class. way is. 87 it was pointed out that. Since any reducible representation can. it should be one that there can be only as many representa tion vectors. (39) If R now denotes a typical symmetry element and (40) can be written as Over all 2 lVnkxi(R)][Vnixi(R)] = for i *j (45) and Over all classes X [Vr*xt(R)]lVnixi(R)] = 9 (46) Thus the \/n^Xi(R) terms behave as components of orthogonal vectors. In Sec. one can write x(R) . class. in each the characters behave as orthogonal vectors in a space with dimension equal to that of the ber of different classes. there is any convenient way to deduce what irreducible representations make up the reducible one. and so forth. a simple and important procedure for dis covering in' how many times each of the irreducible representations occurs a given reducible representation. say n. also be It should mentioned that the factors n« are usually placed at the top of the character table rather than as part of the characters themselves. in principle. Analysis of a Reducible Representation The final aspect of abstract group theory that must be dealt with we can proceed to some applications is the question as to whether. recalled. but this does not alter the argument given here. given a reducible representation. the character table a square array. allowance is made for the num of elements. 811. counting one for a onebyone representation. to the previous This conclusion is parallel. if That this must be number so can be shown from the fact that.188 INTRODUCTION TO MOLECULAR SPECTROSCOPY is operations exhibited. Thus we conclude that there can be only as many representations as there are classes. which leaves the character unchanged.l aiK(R) W . and there can be only as many such vectors as the dimensionality of the space allows. in principle. of finding the transformation matrix that accomplishes There however. formation to convert the reducible representation to a form in which the In practice there is no convenient this. as there are symmetry elements. be reduced it contains by a similarity trans to the irreducible representations that formation. four for a twobytwo.
11 when these subjects are studied. (39) ?*<«>*<«> j. Similar applications to electronic structure and electronic spectra will be made in Chap.° tit] w can be used on the right side of Eq. (48) to convert Eq. (48) to x(R)xi(R) ? or at = ckg = \ X if x{R)xi ( R) (5o) Thus. For expression requires the operations. summation over all symmetry more convenient to have the corresponding expression where the summation is over all classes of symmetry operations. We over first multiply both sides of this equation by x>(R) and then sum all symmetry operations R. of the irreducible representations are available and the characters from a character table. and the a/s are the number of times the jth irreducible representation occurs in the reducible representation. In this way we obtain x(R)xi(R) I R micmxim =lt R j=l and (40).MOLECULAR SYMMETRY AND GROUP THEORY 189 where x(R) is a reducible representation. It is The above this one writes 1 V Over all classes nnx(R)xi(R) (51) where n R operation is is the R. number of elements in the class for which a typical We now have the basic results from abstract group theory and can proceed to apply them to problems in molecular spectroscopy and molecular structure. The methods of application will be illustrated in the remainder of the chapter with problems based on molecular vibrations. that (48) Now the results of Eqs. . x3 (R) are the c irreducible representations of the group. the number of times each irreducible representation occurs in the reducible representation can be readily calculated. the characters of a reducible representation are known.
states: if A helpful result two representations only in their basis coordinates and if the coordinates of a group combinations of those of the other base. 1. the two linear are of one base representations are said to be equivalent and the characters of the two (The result is based on the fact that two such to be related by a similarity transformashown be can representations can resort to the much simpler procedure one this theorem With tion. and 812. dimensional representations. 3ra The normal coordinates are said to form a basis for the representation Since each normal coordinate must transform according of the group.190 INTRODUCTION TO MOLECULAR SPECTROSCOPY Exercise 810. representations are equal. Cz. will be the same as that for coordinates. We cannot. to an irreducible representation.) representation on the basis of the 3n reducible the of constructing Such transformation matrices were cartesian displacement coordinates. Ci. group are of.. etc. representations that particular interest. deduced for the H 2 molecule in Sec. however. 4. irreducible Deduce the representations that the reducible representation is made up for the classes of E. — 2. and The Characters for the Reducible Representation of Molecular Motion The 3re effect of the symmetry operations on the motions of a molecule could be represented by the transformation matrices that show how the normal coordinates (including translations and rotations) are changed when the various symmetry operations are performed. 0. The characters of a reducible representation of the . i. normal the 3n on representation based . transform according to the various irreducible representations. —1 for the classes of E. The character of this reprethe sentation.. which is found rather easily. 86. +2 d v and Exercise 811. Deduce the irreducible representations contained in the reducible representation of the C 3„ group which has the characters <r„. or a degenerate set must transform according to two. a v a' v . three.e. These transformation matrices would constitute a reducible representation. 7. construct the reducible representation on the basis of the normal coordinates because these are generally is initially unknown (except for the translations that overcomes this difficulty differ and that which rotations). the irreducible the Zn dimensional representation are of make up It follows that it is only necessary to see how many times the various irreducible representations occur in the reducible representation to discover how many of the 3n normal coordinates belong to the various symmetry species.
Xfj 2/'o Transformation matrix Zh' x 2/o J . The four transformation matrices for the H 2 molecule are found to be 10000000 010000000 001000000 000 100000 000010000 000001000 000000100 000000010 000000001 E 0^1 . matrices. Zo can be found by inspection of diagrams such as that shown for the symmetry operation ct„ in Fig. 4 XH Xh' = z'„. 4 1. To show how 2 0.MOLECULAR SYMMETRY AND GROUP THEORY 191 To obtain the character of the 3n dimensional representation for the it molecule being studied. on the diagonals done. this can easily be we again consider the transformation matrices. which constitute the reducible representation. for vectors are indicated as H The 3n cartesian coordinate The matrices which type XH y'n describe the transformations by equations of the Xh Vh Zh . It is is not necessary to write out the complete only necessary to deduce the elements that appear of these matrices. 87.
therefore. not atoms that change positions as a result of the symmetry operation. One can deduce the character of the representation matrix. by concerning oneself only with the cartesian displacement vectors of the do.192 INTRODUCTION TO MOLECULAR SPECTROSCOPY 1^2 ' 1 0' 1 1 1 1 1 1 1 1 1 (T 1 1 1 1 1 1  1 1 .0 < ' 1 1 1 1 1 1 1 1 o 1 One should first notice that diagonal elements occur only if the sym metry operation leaves the position of an atom unchanged. respectively. along the diagonal. one can write . With this recognition and the fact that the vectors of unchanged atoms in molecules like effects lead to H 2 are left unchanged or are reversed and that these +1 and — 1.
Slightly —3 will be made to the more difficult to visualize are the characters conso forth. general statements can be made about such characters. and . One might further notice that simple.e. Furthermore.»0> "* (Rotation fay = (1 + 2cos 0> ( 9 and reflection through plane perpendicular to axis) — 1 + * COS v) .MOLECULAR SYMMETRY AND GROUP THEORY 193 down. three cartesian coordinate vectors will be reversed. their positions are unchanged. it if it if will contribute a net —1 contribution any atom is unchanged a in position by a reflection through a plane. to the character of the 3n dimensional representation for the operations C„ and S„. as a result of a C 3 d. When an atom is unchanged in position as a result of a to the C2 C 2 character. operation. FIG. cos 6 sin 9 • sin 6 cosd 1 *<Rolatio„l. i. the result E C2 av 9113 The sums of the diagonal elements of the complete transformation matrices given above confirm these values. by inspection of the displacement vector diagram. 88 The contributions for each unshifted atom. will contribute +1 to the appropriate a character. a molecule has a center of an atom is situated at this center. and a contribution of character. rotation tributed by atoms when.. Figure 88 shows the geometry that leads to the transformation matrices and the contributions to the character in such cases. its position will be unchanged all symmetry and by inversion at the center of symmetry. Similarly. each atom on the axis.
of the Exercise 812. transforms according to one of the irreducible representations. 3n dimension representation Number of Normal Modes it of Various Symmetry Types In the previous section sentation for a molecule could easily be set up. one can readily write down the characters of the reducible representation without visualizing the effect making the cartesian displacement diagram of the various symmetry operations. shown that the number of times the various irreducible representations are contained in a reducible representation could be deduced. and in Sec.. The characters of the irreduc ible representations are given in the character table. Table 84.IM INTRODUCTION TO MOLECULAR SPECTROSCOPY With these unshifted atoms of order 3w. c\ 2 .e. The characters previous section as of the reducible representation were given in the B Ci av it. belonging to the different reducible representations. belong to the symmetry TABLE 88 Contributions to the Characters of the 3n Dimensional Reducible Representation Operation per Unshifted Atom Character contribution E a i 3 1 C2 C3. the contributions that make is to the characters of the reducible representation This done in With these and recipes. 811 was shown that the 3n dimensional repreit was These results. The H 2 example can again be used. generalizations we can summarize Table 88. C3 ci. allow the number of normal modes of the different symmetry types. Deduce the characters CH 3X.. for molecules of the type 813. i. to be determined. along with the fact that each normal mode. c\ 3 1 u 1 a. including translation and rotation as well as vibration. i. 911 Now one calculates the number of normal modes that transform first according to the representation.e.
of the vari band contours which pointed out the great and perpendicular bands. Thus M = = > i[9 n RX (R)xi(R) Over all classes  1 +1+ 1 1 3] = = = 3 Similarly. The Infrared Active Fundamentals apparent that some vibrations of symmetric molecules.MOLECULAR SYMMETRY AND GROUP THEORY 19S type A\. 63 do. CH 3 C1. (54) are.e.e.. = i[9 = #9 + = i[9 1 1  3] 3] 1 +   2 (52) and a4 + 1 1 + 3] = 3 These results can be summarized by saying that the reducible representation contains representations of the to symmetry types according (53) r If = 3Aj + A 2 + 2Bi + 3B 2 the assigned translations and rotations. indicated in the last column of Table 84. ai (51). in fact. Determine the number of normal vibrations ous symmetry types for the molecule 814. *"" 0=C=0 — >. which symmetry types the fundamental transition be active. to decide for i. two vibrations of the symmetry type A\ and one symmetry type B 2 One can verify that the displacements dis. a2 a. perhaps more apparent for difference. by application of Eq. for . therefore. molecules of the type CH X 3 because of the previous discussion of rotaof the different types of vibrations. transform according to these symmetry types. Exercise 813. such as It is the symmetric stretching oscillating dipole mode of CO2. Furthermore. the necessarily zero oscillating dipole It is possible. as moment is a result of the symmetry of the molecule and the symmetry type of the vibration. i. played for H 2 in Fig. have no moment and therefore cannot interact with electro magnetic radiation.. are removed. In a similar way one can of classify the The importance tionvibration parallel such classifications is normal modes of any molecule. i.e.. one has r vlb There of the = 2Ai + B. we will now will see.
The lAvib vibrational wave function for a molecule can be described as = v(Qi)v(QtMQ») state. a fundamental vibrational transition. from v = to v = 1. <p((?3»6) all (55) For the ground where vibrational quantum numbers are zero. as indicated intensity is an observable quantity. 45. the value of the integral must be unchanged for all symmetry operations on the molecule. indistinguishable orientations of of the three transition by Eq. It follows that the representations T/x x . It is now necessary to see the effect of symmetry operations integrals. Tii y .. FIG. It should be clear that these dipole components will transform under the various symmetry operations in exactly the same way as do the displacement arrows representing Tz Ty and T„ . and ry2 will be the same as the representations I>„ T T „ and Tt. on the quantities in the integrals and then on the complete The dipole components can be represented by the charge separations shown in Fig. 89.196 INTRODUCTION TO MOLECULAR SPECTROSCOPY oscillating dipole which symmetry type the be zero. (446). 89 The dipolemoment vector of its a molecule and cartesian axes. It follows that for any it moment integrals to have a nonzero value. A nonzero integral leads to a nonzero value for integrated absorption coefficient. only if one of the integrals fMQ<)n*MQi) dr dr dr JMQi^vUQi) or SMQi)n*MQi) is nonzero. can occur with the absorption of electromagnetic radiation in the vibrational energylevel pattern corresponding to the normal coordinate Q. moment will not necessarily As was pointed out in Sec. components along the . Since this band must have the same value for all the molecule.
that all symmetry operations on under all ^„_o the ground vibrational state is invariant symmetry The wave functions responding to Qi ^„ i= i is for a molecule in which the wave function corthe form (58) excited to the v = 1 state is of = = (const) (Q 4 )(fc_) v which follows because the <p.Q t '+Qi'y was shown that a symmetry left operation on a degenerate set of vibrations the value of Q\ + Q] unchanged. yk . the nondegenerate terms a doubly degenerate of unchanged by the symmetry operation. [eTi«i ! e" Q » • • • .! ^ Tt.e~ro» •>«'«»•>] (57) where <y k = Now the effect of a can be investigated. and therefore ertQ^evM 2 A . For nondegenerate vibrations Qi symmetry operation on ^„_ —> + Qi. <A(„=o) = const 2wvk/h. lead to f„_o. the wave function for the molecule transforms Qi. for example. 64 it = e yt(. operations.. degenerate pair contributes terms of the type i n Sec. therefore. and since only the squares of Q's occur in the expression. in view of the wave functions of Eq. where = 1. (245).i = 1 wave function is of the form (const)' QieTe* v. i. YWO — *. In this case. It follows.. according to the normal coordinate vibrations are excited to the v Similarly..MOLECULAR SYMMETRY AND GROUP THEORY 197 one has <A (1>0) = <Po(Qi)<Po(Qz)<Po(Q3) " " " <Po(Qine) (56) For harmonic oscillator wave functions this total wave function has the form. if a degenerate pair of = 1 level.e. Degenerate vibrapair. These results can now be summarized according to the effect of the symmetry operation R — R — Hy yx * R as V Tx (R)n x r r!/ (it)/Xj/ (59) > p.(R)p. the vibrational wave function of the molecule will transform as does the degenerate pair of normal coordinates. in ^„_o are tions. have the same value of vt. R tf'rO ^ T Qi (R)i Vi =i result of the The R symmetry operation on the transitionmoment .
and therefore the symmetry. so forth. In a manner similar to that used to investigate transitionmoment inte can show that for a fundamental transition to be active the normal coordinate must belong to the same of symmetry species as one TX T Z Tx Ty and . be written as J^rtii dr » ?T.(R)?Q<(R)N'vwJ'v i dr i (60) The previous result that all symmetry operations must leave the integral unchanged if it has a nonzero value means that TtAR^qXR) must be unity for each and every symmetry operation. The argument can. C1 vibrations will lead to absorption of these will The Symmetry of Group Vibrqtions molecules have groups of atoms that vibrate rather inde Many pendently of the remainder of the molecule. normal coordinate coordinates With this result and an analysis of Tz how many normal belong td the various symmetry species. .. therefore. be nonzero. it should be mentioned. Which CH 3 bands corresponding to fundamental transitions. to v = 1 transition for a given normal coordinate can belongs to the be different from zero only same symmetry species as Tx Ty or . the transition It follows that moment integral for the v = . can be shown to transform as the product of the representations for the translations indicated in the subscript on a. and which be parallel and which perpendicular vibrations? 815. of a molecule is determined from the number of infrared absorption bands of infrared active of a given that are observed. (It will only be mentioned that whether or not a fundamental transition can lead to a The requirement of the for a Raman Raman shift can be deduced in a similar manner. since . product of T Tl (R) and T Qi (R) be unity for every R. The normalization and orthogonality characteristics of representations show that only if Y Tx and T Qi are identical representations can the and so forth. . sometimes be turned around so that the geometry. Furthermore. shift is that at least one of the integrals type J*o(Q<)«**i(Q«) dr SMQi)<**«MQi) dr etc. . one can determine the number fundamental transitions that can occur for a molecule symmetry.198 INTRODUCTION TO MOLECULAR SPECTROSCOPY integrals can. The hydrogen atoms. reason that .) of the Exercise 814. if the . . the polarizability components a xx a xy It is for this etc. character tables sometimes show the grals one symmetry species of TX T X Tx Ty Raman etc.
. This separation is of great aid in assigning the observed infrared absorption bands to normal vibrations of a molecule. Let us first calculate the types of the vibrations of the molecule without the two hydrogen atoms.™ = 2Ax + A + 2Bi + B 2 2 (64) (This same result could have been obtained tributions to the reducible representation that the by writing down the contwo hydrogen atoms make and then sentations that resolving this representation into the irreducible repreit contains.e. the normal coordinates can . The molecule probably belongs to the C 2„ point group. The characters of the reducible representation for tljis nineatom residue can be worked out. to be x 27 1 1 5 The number of vibrations of the different symmetry types atom fragment can now be calculated as a! of this heavy a2 a.) now be drawn to represent Exact indications of how the atoms move in Pictures can the six hydrogen vibrations.MOIECULAR SYMMETRY AND GROUP THEORY their 1W than that of the other atoms of the mass is usually so much less It turns out that in such a case molecule. at = ^[27 = #27 = H27 + = i[27 + 1 + 1 + 5] = 1 8 * ' 1 1 1 +  1 1  5] 5] + 5] =5 =6 = 8 Removal of the translations for the vibrations is and rotations shows that the representation composed of 2 X r„ h = 7 At + 4A + 4B + 6B 2 (62) i. A discussion of the molecule CC1 3 —CH —CC1 2 3 will illustrate the procedure. often lead to such vibrations. The same treatment with the hydrogen atoms. vibrations of the types PH.. with the aid of Table 84. therefore. can be molecule the vibrations of the terms of separate vibrations of the molecular skeleton and vibrations of the hydrogen atoms of the molecule. or at least visualized. in treated. gives entire molecule. including the two r vib = 9^! + 5A 2 + 6B! + 7B 2 (63) The additional two hydrogen atoms introduce..
1 = l. twisting) v 1.H» H^ $ H 7 «. helpful to recognize that each hydrogen atom introduces one vibration that is essentially a C stretching vibration and two vibrations that are essentially bending ones. With this approach the six diagrams of Fig.270cm 1 = Hi .950001' = 800cm. These vibrations can be expected to occur. more or less disturbed and altered. 810 can be drawn for the methylene vibrations.. The form of and the frequencies of the fundamental absorption bands depend the molecule containing the somewhat on the nature of CH 2 group. f 4 j(CH stretching) v .^(CH. — CH nized.1 .1 3 1.Jj .H rU** .H 200 INTRODUCTION TO MOLECULAR SPECTROSCOPY not be made without the lengthy calculations that will be introduced It is. however. in any molecule that contains a in the following chapter.460cm _1 .870cm. 2 group.(CH Z bending) v = 2. 810 Vibrations of the the actual vibrations CH2 group classified according to symmetry.(CH T stretching) I fljtCHj nuking) v : h * 2 (CHj 1> wagging) = 2.4. In a similar manner one can represent the vibrations characteristic of a it is CH 3 group. When such group vibrations are recog often possible to reach an initial understanding of some of the absorption bands of the absorption spectrum of quite complicated molecules containing CH 2 and CH 3 groups. FIG.330cm.
. Eyring. E. L... and draw diagrams to represent these vibrations.: New York. & New 3. of Bellamy.Inc.. —CH PRINCIPAL REFERENCES 1. Exercise 815. P." John Wiley & Sons. "The Infrared Spectra York. New York." Academic Press. Kimball: "Quantum Chemistry. Inc." John Wiley Complex Molecules. and G. J. H. Inc.. 1954.MOLECULAR SYMMETRY AND GROUP THEORY 201 3 Analyze the symmetries of the vibrations of a Classify the group.. E.: "Group Theory. 1948. Sons. Wigner. diagrams according to symmetry type and bending or stretching nature. 1959. Walter. 2. J.
. .. due to overtones and combinations of fundamentals. When all this can be accomplished for a particular molecule. . is For a diatomic molecule only one vibration occurs. an to v to a vibrational energylevel pattern and. its infrared absorption is spectrum. vibrations occur in a manyparticle system and that these normal vibrations are characterized by particle motions that are in phase and have the if same frequency. The frequency of this transition was. or normal. TVi V 'l \ CALCULATION OF VIBRATIONAL FREQUENCIES AND NORMAL COORDINATES OF POLYATOMIC MOLECULES It will has been shown in Chap. 21 and 71. what the form of the normal coordinate is. In a molecular system each normal vibration leads the selection rules permit. cule.e. weaker bands. related to the force constant and the reduced mass of the mole It is now necessary to see how the frequencies of the 3n — 6 fundamental transitions of a polyatomic molecule are related to the force constants and atomic masses of the molecule. tional energylevel pattern exists.*£££?^.V*£ "". in each one of the normal vibrations. 6 that certain natural. absorption band corresponding to the energylevel pattern will be observed.*' " ' R>v' 1 t:"'!"' J . i. 2. which 202 may look like those of Figs. in Chap. said to be analyzed or understood. We will also be interested in determining how the molecule vibrates. v = = 1 transition in that Furthermore. may be observed.. only one vibraand only one fundamental transition observed spectroscopically. 'J'' '': <# " '&i3> = * "^t™ .
*+.. will be worked through. simple examples the more systematic methods develwill be outlined.. in terms of these axes.CALCULATION OF VIBRATIONAL FREQUENCIES 203 The treatment of diatomic molecules in Chap. Zj. the form of the equations that result from substitution of Lagrange's equation rather than the of these functions in Lagrange's equation will be considered.(*. It is important to remember that a motion described as a vibration must not involve a net translation or rotation of the molecule. Finally. Following consideration of these general features. due to the vibrations This contribution to the kinetic energy of the molecule can be treated in terms of coordinates that measure the positions of the atoms relative to a set of axes that motion. oped by Wilson for performing such calculations 91. some typical. expressions for the kinetic energy T and the potential energy U of the After an initial discussion of how this can be done for a polymolecule. kineticenergy expression of Eq.) =0 • £ 3 mfcjXj • = (3) =1 . and the kinetic energy due to the vibrations of the molecule can be immedi down as r = * £ «. This requirement is imposed by requiring that no movement of the center of gravity of the molecule and that no angular momentum be imparted to the molecule by the velocity components Xj. yj and In this way the «. The Kineticenergy Expression for a Polyatomic Molecule The of the kinetic energy that is here of interest is that of the molecule relative to atoms each other. # + 3) (i) This expression tional problems.) ° ( 2> and t=i n X mi(Vfr  ziV. atomic molecule. by ately written rotational x. (1) can be used along with the relations j=i £ mjXj = ji 2 m&> = ° >=i X m& = XjZ. move with the molecule as it undergoes translational and If the jth atom is located. 2 anticipated the use f — ma relation in calculations With this approach it is necessary to set up for polyatomic molecules. yj. is the basis for the treatment of kinetic energy in vibra One should note that the orthogonality of the cartesian displacement coordinates results in an expression that contains squared terms and no cross products.
Thus. Many more parameters are introduced than can be evaluated.H'o(Sr H o)(ar H + kno. while complete. One approach would be nuclei i to base the description of the geometry on the changes. and the valencebond force more frequently used by chemists..H 2 there would be four force constant terms in the potentialenergy function.HH'(SrH 'o)(5r H H') where the fact that H o = Wo has been recognized.i X = K(*r«)' l (4) H— O— H'. except in favorable cases. (5).) The Potentialenergy Expression for a Polyatomic Molecule The of the potential energy of a molecule its is a function of the distortion It molecule from equilibrium configuration. as Eq. for. (The separation of vibration from translation and rotation suggested is treated in a more rigorous manner in ref. Two different points of view as to the dimensions of the molecule that are most impor tant in determining the potential energy have been used to approach this goal. V= for example. this would lead to the potential function 2 2 + !/c H o(SrVo) + IWOhh') + &HO..204 INTRODUCTION TO MOLECULAR SPECTROSCOPY here 92. one might write u = For • i i. to attempt to A. It is necessary. and these «ould not be determined from the 3w — 6 = 3 fundamental transition frequencies that are observed spectroscopically. from experimental data. field. generally favored by physicists. follows that displacement coordinates which show the change in the geometry of a molecule from its equilibrium configuration will form a suitable base for a potentialenergy expression. 1. of the potentialenergy function. therefore. Bry. (&r ho) (8r Bn + &Ho. These ideas led to the early use of the central force field. This might be done in a number of ways. are seldom satisfactory. With this coordinate system and the harmonicoscillator to keep only the quadratic approximation that assumes it is sufficient terms in the potentialenergy function. from the equilibrium value of the distances between the and j. The central force field assumes that the change in potential energy . ik B o(Srao)* 'o) an' fc ') (5) Such formulations. find a way in as the molecule which the change in the potential energy of the molecule is distorted can be described without recourse to this most general quadratic potential function expression.
With this approach one would attempt to describe the potentialenergy function of H2O by 2 the expression (7) its U = where 8a i/cHo(5r. H h'. equilibrium value and. In fact. that the valence bond field is usually the preferable approach to a satisfactory potential function. as before. k' a The two force constants k a and are related by k' a — k a /r\. When such calculations are performed it is clear that neither simplified potential function is com pletely satisfactory. will have the units of dynes per centimeter.. (7) as ^k' a (ro8a) so that like fc H o.CALCULATION OF VIBRATIONAL FREQUENCIES 205 that occurs when a molecule is distorted from its equilibrium position can best be described by terms in the potentialenergy expression. Thus. The success of the poten function can then be judged by how well the calculated value of the third frequency agrees with that observed. fc number of terms in the description of the potential energy of the H 2 molecule to two. 5r H 'o. reduces the The assumption i. the absorption band frequencies can be used to corresponding to the fundamental transitions of H2O deduce values of the constants in the potentialenergy expression. 2 [Frequently one writes the final term of Eq. librium internuclear distances. each of which involves a force constant and a square of an internuclear dis O tance. The results suggest. if the available spectroscopic data warrant. therefore. and 5r H H' represent changes from the designated equiof a central force field. add additional . however. for H H' the simple centralforcefield expression for the potential energy would be U= where ik B0 (8r B0 )* + ikvoiSrn'o) + itHH(5rHH 2 ') 2 (6) 5r H o.e. The valencebond force field assumes that the potential energy of a distorted molecule can best be described in terms of the changes in length of chemical bonds and changes in angles between chemical bonds from their equilibrium values. dr B0 and 8r S 'o are changes of the HO and H'O bond lengths from Again only two k' a) force constants appear in this simplified potentialenergy expression. H o and A.] As we will see later in the chapter. one can determine the two force constants in the previous potentialenergy expressions so that the calculated frequencies are close to two of the observed absorption tial band frequencies.HO ) is 2 + i/c H o(«r H'o) + 2 iMM the change in the angle between the two bonds from their equilibrium values. One can.
it should be mentioned that a potential function. based on what is called a UreyBradley field. Exercise 92. however. a should generally be small and need not always have Furthermore. such as k H o. one makes use of the spectra of isotopically substituted species. there is no way of verifying its success. Unless. one might investigate the force acting to restore the angle between the bonds of H 2 to its equilibrium value and obtain Restoring force on Sa = — ST I . terms involving the distortion of chemical bond lengths and angles and also terms depending on the distances between nonbonded atoms. not only as the result of these displacements. seems reasonable. central force field . 8r B > ) (9) One sees that the added terms allow for a restoring force to act on angular displacements. Draw of the H Finally. NH 3 Write expressions for the potentialenergy function for on the basis of (a) the simple valencebond force field and (6) the field. simple central force a diagram showing a displacement of the atoms from their equilibrium positions that would lead molecule 2 to a calculated increase in the potential energy that would be different for the simple valencebond and centralforcefield approximations. and turns out to be so. . but also as a result of changes in the bond lengths from their equilibrium distances. Exercise 91.206 INTRODUCTION TO MOLECULAR SPECTROSCOPY quadratic terms to improve on the simple force fields. Thus one might add a bond stretchingangle bending interaction constant to the simple valencebond potential and write U = i&Ho(SrHo) 2 + ik B o. a (Sa + iJc B0 (Sr B y + ^ a («<*) 8r BO + Sa Sr H 'o)  2 (8) This function will certainly provide a better description for the way in which the potential energy varies with molecular distortion. is now often used when fairly simple molecules are dealt with and several different isotopically substituted This field combines the features of both the and the valencebond field. = k a (8a) + %k BO a {8r BO + . that interaction terms. Thus. it positive signs. The significance of the various force constants can best be seen by determining the force that would act to restore a bond distance or angle to its equilibrium distance. therefore. It has. species can be studied.
.. This requirement creates considerable complexity in molecularvibration problems to set since. . . internal coordinates. necessary that both T and U be expressions involving the same set of coordinates. . more by Lagrange's equation. Use of Lagrange's Equation in Molecularvibration Problems The motion of particles constrained by certain forces. (11) and (12) one must be able to express either the R's in terms of the x's or the x's in terms of the R's. i.) To proceed from procedure. In the later method. Rt. either angle bending or Ri. is generally The first when an example is worked out in the following easier to set up since equations of the type . the coordinates internuclear distances which describe the displacements of the atoms from their equilibrium and the bond angles from their equilibrium values.. as pointed out in Chap. subject to a particular potentialenergy function.Xtn) Ri = fi(Xl. xi. one obtains the ith equation as For a vibrating molecule there tions. and if U in what are called . internuclear distance type coordinates. as we will see. .CALCULATION OF VIBRATIONAL FREQUENCIES 207 93. unknown at the beginning of the calculation. it is — 6 vibra tional frequencies In order to apply Eq. . we will see section. . can. . (10) to a system. They generally are evaluated such that the observed frequencies are obtained. it is convenient up T in cartesian displacement coordinates i. 3w — 6 such equa The set of these equations can be solved for the 3m and modes of vibration. conveniently. 2. bond stretching and Thus.e. If a typical coordinate is q. be determined by application of Newton's / = ma relation or. and the cartesian displacement coordinates are represented by x 3n one usually sets up the expressions . generally be (12) t\j'i (The values of &„• will. . as we have seen in the previous two sections. i. as Eqs. Lagrange's equation is applied to each coordinate needed to describe the system. of course. x^. Sn6 T = l yi I £ mfbj (11) U = i k/RiB.e.. will be. .e. Rz ne represent the internaldisplacement coordinates.
lem can be immediately written down as T = frn&l + \m x\ + £m x§ 2 3 (15) .. that even 3n — 6 equations become difficult to handle. solution 94.R 3n . even for fairly small molecules. will To illustrate . ®— ®—®. Transformation of Eqs. as U= More nates.t ) and to proceed to solve 6 internal coordinates. however. tion.e. the kinetic energy for the onedimensional prob respect to a set of axes fixed in space. as indicated in Fig. by one means or another. leads. (11) and (12) to cartesian coordinates. .how this can be done.)* + i/c 2 (Sr 23 ) 2 (13) generally we let is will use R's to indicate internal displacement coordi. and specific example by these methods becomes rather troublesome. be written. i*i(8n. which will involve stretching of the bonds. to 3n equations when Lagrange's equation is applied.208 INTRODUCTION TO MOLECULAR SPECTROSCOPY can often be immediately written down. x h x 2 and x 3 represent the positions of the three atoms along the onedimensional line of the molecule. however. differ general problem of a threedimensional molecule will only in complexity. 91. OCS. The potential energy is best expressed in terms of changes from the equilibrium values of the distances between the atoms of the molecule. The Stretching Vibrations of a Linear Triatomic Molecule A tions is suitable HCN. and one translatype i. In molecules of any complexity it is usually better to obtain. the problem in terms of the 3n — . More systematic ways of handling the problem will be introduced later. The The more example to illustrate the discussion of the previous secthat which considers motion of a linear triatomic molecule of the molecule. If the equilibrium bond lengths are denoted by r? and r§ and changes 2 3 from these values are denoted by dr u and 5r 23 the potential energy can . of which there are 3n. transformations of the type xs = fj(R h . along the axis of the degrees of freedom along this axis will lead to two 3 vibrations. and if we Ri represent 8r u and R 2 represent 8r 23 the potential energy function written as U = ihRl + faRl The kinetic energy expression can be If (14) immediately written with . in the simple valence bond approach. a be worked out in the following section by this approach. and so forth. . It is true.
(21) One cannot immediately solve Eqs. . Thus one has the FIG. one . 209 We i.).) r!j 3 r° 2 (17) and 8r i3 (x» x 2) rg. Eqs. as will be seen. for ±i. (15) can be written in terms of the internal coordinates . 91. (21). is required to obtain the desired expressions which are of the type x = f(RiR 2 . (19) and (20). that no net linear momentum be involved. must now proceed to obtain expressions of the type x2 x3 = fi&u&t) = /.GM0 = MRiA) Ri and jR 2 (16) so that the kinetic energy expression of Eq. The desired inverse relations can be obtained. (x 2 Br 12 = R = = R2 =  a:. therefore. In the convenient matrix notation. . some further effort. sees this in the impossibility of taking the inverse of the nonsquare matrix.CALCULATION OF VIBRATIONAL FREQUENCIES . or Eq. (17) and (18) are Ri Ri = = Xi x3 —  ii (19) (20) x2 The ease with which these equations. x 2 and x 3 in terms of Ri and Ri because of the occurrence of three cartesian coordinates and two internal coordinates. (18) Furthermore.). . The atoms of a molecule sional © —©—® in location of the type moving onedimen space by means of a set of axes fixed in space. however. . while . if one imposes the condition that the motions to be studied be vibrations and. The two types of coordinates are related by the t expressions. 1 1 (£)( 1 Ti v \x% 1 X. of the type is R = f(xi± 2l . the time derivatives of Eqs. since rj 2 and are molecular properties that are not time dependent. In the matrix expression. (19) and (20) can be written as are written typical.
Xi. the desired result (m 2 'zA Xi J M +m 3) —m M 3 1 M 3 ^ — mi xj M TOl M or one can obtain. with Eqs.m. T = ^ 2 (ml + 2m. the algebraic equations equivalent to this: . (15) for T gives.+ m^Rl + 2m m )EiiJ 2 + J^k (2m\ + + ^p (m\ + This expression for 2mim 3 3 2 2«iim 2 + m? +wm + 2 3 mitn 3 )R (26) T together with (27) U = ihR\ + fan* . (21). (19) and (20). This expression allows. f mf . after rearrange ment. with the notation M = mi +m + 2 wi 3 . M M mi + m M M 2 —m 1 (24) 1 by means of the necessary algebraic manipulations. Xl _ (m 2 mi $ M +m m 3) Rx > _ m M 3 _ Ri (25) X2 Ki ~M Kl ~M3J x *~M R i+ M R > Substitution of these expressions into Eq. the expresand x 3 in terms of Ri and ij 2 to be obtained. or Eq. In matrix notation one can add this additional relation to give (23) Now we can take the inverse of the square matrix to get.210 INTRODUCTION TO MOIECUIAR SPECTROSCOPY additional relation ffiiXi + nnxz + miXs = (22) sion for ±i.
on expansion. yield roots of the equation with frequencies corresponding to the observed values of 2.312 cm". when inserted in the determinantal equafrom Eq.) + kA ( Ai 3) ~ and "" I 4^ 2 " 2 2IJ2 2m l + 2w»im + 2m m 3 2 A2 = (31) 47rV I Jgp ^ 2m s (ml + 2m m + 2m m x 3 2 3) A. 3 + 4ttV ^1 +2mim + m? + m 2m 2 + w^ma) + ifc 2  A2 = Again. for exam one would know from spectroscopic studies that the frequencies of the absorption bands corresponding to the two stretching modes are 2. If.089 and 3. One can now choose numerical one were studying HCN. (28) and 2 (29) leads to the expressions I 4*V ^ 2 (ml + 2m 2m 3 + m\ + mm + mm. In this way one deduces that the force . Substitution of these trial functions and their second derivatives in Eqs.CALCULATION OF VIBRATIONAL FREQUENCIES leads.312 cm"'. to a quadratic in 2 The two roots of the equation (only 2X2 v) 1/ . give the desired frequencies. positive values are appropriate for It is ple. now perhaps better to consider a particular case. (31) along with the values of the atomic masses. nontrivial solutions will exist for the A's only if the determinant of the coefficients is zero. This determinantal equation leads. 211 when Lagrange's equation (ml 2 is applied successively to Ri and R it to jp + 2m m + m\ + miwt + mim )fli 3 2 3 + and 2j! (ml 2]^s ( 2m s + 2 2mi?»i + 2m m )S + &i#i = 2 3 2 (28) + 2mim 2 + m\ ( + m m + m mi)R\ 3 x + One can again Ri see if 22^2 2m 3 + 2mim 3 + 2m m )R + k R = 2 3 1 2 2 (29) solutions of the type = Ai = A2 cos 2Kvt ( 30) and Ri cos 25tj<< exist for this pair of equations. values of tion arising fc x and k 2 which.089 and 3.
212 INTRODUCTION TO MOLECULAR SPECTROSCOPY constants for the bonds in HCN are . formed by the coefficients of the A's. therefore. (31). atoms of given to the molecule translation is no net arrows to the atoms suoh that Eqs. can be immediately obtained by (More generally one takes the ratio of the cof acrearranging Eq.089(3 X 10 ) and v is 3. remains for us to see how a molecule. Sometimes. to the observed stretching frequencies of the HCN molecule. i.8 X 10 dynes/cm 17.312 cm' A A2 = x \ 1.312(3 X 10 10 ) cycles/sec. calculation was Exercise 97.e. One needs. Deduce relative displacement amplitudes and draw suitable diagrams for the two stretching vibrations of DCN. . (31). Consider the basis for the fact that some exercises. Eq. however. vibrates in the vibrational modes that have the frequencies used in the calculation. when v is 2. when substituted in Eq. HCN example one has For v = 2. (32) lead. = JW = fc HC 5.00 (33) and For v = 3. Calculate the positions of the absorption bands due to to be expected for DCN. deduce values of the k's to fit known frequencies. The relative values of Ai and A 2 and thus the form of the vibrations.) For the determinant the of tors .089 cm 1 A A i : 2 0. whereas in two previous the of diagrams the draw to necessary Exercise 63 the corresponding vibrations of the CS 2 molecule could be drawn with less calculation.137 (34) One can recognize that the first is vibration predominantly one of C N stretching while the second predominantly one of illustrating C—H stretching.. the relative values of A x and A 2 that solve the set of equations. (31). vibrations the stretching Exercise 94. Draw diagrams HCN and so obeyed. Verify that the force constants of Eq.00: is  0.405 : 1 . one can take force constants from some molecules and carry them over to other molecules and thereby reduce frequencies from a calculation of the It type illustrated here. the displacements of the Do this by attaching in the two stretching modes. again the HON example can 10 be considered. Exercise 96. given by (33) and (34) are amplitudes relative the that Exercise 95.9 X 10 dynes/cm 5 6 32  Exercise 93. Frequently one must make the calculation in this direction.
lead to these frequencies. The calculation is then perfectly feasible. and draw diagrams to represent the nature of the vibrations.1 (the former is not infrared active). Some of these frequencies and must determine what features will 95. and several 2 X 2 and 1 X 1 determinants. 94. the conditions of Eqs. benzene molecule. in the general when a more than onedimensional problem case. feasible only when the simplifications that result of the molecule are recognized. The fundamental stretching mode transitions of the C0 2 molecule correspond to frequencies of 1.CALCULATION OF VIBRATIONAL FREQUENCIES 213 The fundamental transitions of the stretching modes of lead to absorption bands at 2. what values of the stretching force constants for the and bonds are calculated? Exercise 98. as .050 and 863 cm 1 If a simple potential function like that of Eq. in Use of Symmetry Vibrational Problems of the relation between the force constants and the vibrational frequencies of all but the simplest molecules is frequently The determination from the symmetry vibrational problem of benzene involves 3w . (13) and (6) on the basis of a potential function that includes a cross term between the two bonds of the C=0 molecule. inear i. This is particularly troublesome when one knows the nately.349 cm. if full use is made of the symmetry of the (31). By the procedure outlined in Sec. when force constants Considerable simplification results. the solution will require the imposition of zero angular momentum as well as zero linear momentum. tioned that. The above examples. Thus one imposes. Exercise 99. obtain a value for the force constant from these data (a) on the basis of a potential function like that of Eq. the The basis for the simplifications that symmetry introduces is. now be considered. only needs to be men is considered. one 3 X 3. It should be apparent from the simple example treated here that the analysis of the vibrations of a molecule with many atoms will involve the solution of equations of high order. illustrate the proIt cedures that can be used in the general case. fortumolecules have elements of symmetry. OCS .e. For example. C=0 C=S Calculate the relative distortion amplitudes for the two bonds in each vibration. On the other hand.. the problem factors down to one 4 X 4.6 = 30 internal coordinates.388 and 2. (2) and (3). XYZ type molecule solution for the stretching modes of a in internal coordinates. (19) is assumed. and one would be led to a 30 X 30 determinant of the type that would arise from Eq.
reached conclusion would be contrary to the involving ( must not occur in U. 64. symmetry. The com if binations are such that each symmetry coordinate is symmetric or antisymmetric with respect to each symmetry operation. of different symmetry types means that the determinant resulting from will application of Lagrange's equation to the symmetry coordinates have blocks of zeros in the offdiagonal positions that correspond to the Thus crossterm positions between the different symmetry coordinates. and S p occur must not S S term cross the P ticular operation. Since the total determieach nant will be zero if the determinant of any diagonal block is zero.2U mentioned of INTRODUCTION TO MOLECULAR SPECTROSCOPY in Sec. If such cross terms did occur. then > { Si p —Sp in for a par T and a term S SP — symmetry operation would convert SiSp —* —<§. the determinant has the form a o o o o = a o (35) o where each shaded block be a subdeterminant of order equal to the number of coordinates of that symmetry type. the of energies potential the would alter the kinetic and symmetry 64 that a in Sec. These coordinates consist of linear combina tions of the cartesian or internaldisplacement coordinates. The 3n . vibrational block can be set equal to zero and can be solved for the will frequencies of that symmetry is determinants to be solved benzene. One recalls that and S are symmetry coordinates that behave differently with respect to — any symmetry operation. perhaps Si —> S. the operation cannot result in any change of the kinetic or potential energy The effect of the absence of cross terms between elements of a molecule. of high This reduction in the order of the very great in the case of molecules. The use of the H 2 symmetry coordinates can be illustrated by the example problem can molecule.6 = 3 internalcoordinate of . such as type.<§„ and SiSp » —SiSp and This molecule. the necessary absence of cross if and potentialenergy expressions these expressions are set terms in the kineticup in terms symmetry coordinates.
e. and so forth. T = im B (xl and + 2/) + \mo{xl + y%) + ^m a (x^ + 2 tf. displacement according to S 2 consists of motions of H and H' by S 2 units. 92. problem of the of the molecule in the plane of the molecule. to one involving two totally symmetric coordinates. for motion or rotation. The relative atomic motions that constitute S S 2 and S 3 are then h compatible with the symmetry requirements and involve no translation . according to the discussion of Sec. and S 3 that are convenient for the calculation of the vibrational frequencies and normal coordinates of the H 2 molecule. We consider the H . in the directions These coordinates are convenient for solving the vibrational H 2 type molecule.be designated by indicating that displacement according to Sj consists of motion of and H' by Si units H shown. 813. We can write.. Suitable coordinates are represented dia grammatically in Fig. of type A and one antih symmetric coordinate.CALCULATION OF VIBRATIONAL FREQUENCIES 215 be expected to reduce. S 2 . two Aj and one B 2 but also are such that they involve no translation or rotation of the molecule. and we set up coordinates which not only have the symmetry appropriate to the normal coordinates. 92 by the three figures. of type B 2 . 92 Three symmetry coordinates Si. 2 molecule of Fig. Again one must write the expressions for T and U in the coordinates in which they can be immediately formulated. V . The proper relative lengths of the arrows in each figure can. and by 2mu/m units as shown.) (36) U= PoH(5r OH ) 2 + i/b oa (SroH') + P«(M 2 (37) 2m H„ m *2 S2 \ H FIG. i.
. immediately the velocities that the atoms would have along the cartesian directions in terms of velocity contributions based on the symmetry coordinates as xB = = a Si Si  Si sin a.sin a (l ~ ^ +^ sin2 a ) Si sin* a) S 3 (43) sin a") S2 and These transformation equations between internal coordinates given V symmetry coordinates can be substituted into the expression for . = — Xw _ 2 * <!>i  S3 a . how displacements described by the sym metry coordinates Si.216 INTRODUCTION TO MOLECULAR SPECTROSCOPY it is Now expressed in terms of Si. + (l + + ^) S. can be used to convert Eqs. S t and <S 3 can lead to motions of the atoms equivInspection of Fig. relation 8f 0H 8r OH . j/h.S t ) (39) It is first necessary to ask . T = and t(Si. sin a (40) too ' .S 3 V*' = S + S3 cos a Substitution in Eq. . sin aSi aSi  8« = . a. . (40) for the time element dt. 92 allows one to write alent to xs. Vh . = = (dXo — &Xn) sin a + (Sy — (dy Q «J/h) (dXo J  Sxw) sin a + a  COS a Syw) cos a «2/h) sin (42) =The (Sxw  Sxk) cos + («{/h< + a  2«j/ sin a expressions for the effect of changes in the symmetry coordinates on the cartesian displacements. after rearrangement. T = m*S\ The + m H (l + ^) SI + m H (l + ^ sin' a) Sj (41) between the internal coordinates and the symmetry coordinates can best be found by first writing the cartesian displacements Thus of the atoms that result in bond stretching and angle bending. written in Eq. .cos + . and so forth. <S 2 and S s so that we can obtain . (42) to the desired relations 5r 0H = =  sin «S 1 cos a (l cos a (l + ^) S.S h S 3 ) (38) U= u(Si. + (l 6r OH. . (36) gives. .St. and Sr necessary to have the coordinates x B #h. y 6 • • = = 2m H <S — *** 3 sin a . x cos a.
obtained. 94. H o sin 4A.„ 5 . . by Eqs. Again three equations are and solutions of the 2irrf form Si Si Sz — Ai = Ai — A z cos COS %rvt cos 2icvt (45) can be sought.)]"' + Now of the that T and C/ are expressed. The determinant of the coefficients of the amplitudes A h A h and A 3 is set equal to zero to give the equation 6u — Xan 621 612 622 — Xfl 22 = &33 (48) — X<l33 It is clear that the symmetry coordinates have led to the desired breakblock. and S h Lagrange's equation can be applied to each of these coordinates. if Nontrivial solutions of this form are then seen to exist the determinant composed of the coefficients of the A's is zero. HO cos 2 A 2m H \ ^l+ j a cos \ — 2 +71 Wo / 2ka  . . (37) to obtain. the customary procedure of designating the term 4irV by X will be followed. and it is convenient to introduce the notation T = and a n &l + a 22 Sl + owSj l( (46) U = where the buS{ + 2& 12 S S2 + 6 22 SI + b n Sl (47) a's and b's refer to the coefficients of the terms of Eqs. after rearrangement. cos 2 a S% J . 2A. (41) and (44). (31) in the internal coordinate treatment of Sec. (41) and (44). 2fe. 217 u ' k BO sin 2 a + ~ r„ . Furthermore.CALCULATION OF VIBRATIONAL FREQUENCIES in Eq. fc „ . H sin a cos a (l+^o. in terms symmetry coordinates S h Si. down of the 3 X 3 determinant toa2X2andalXl The two . '2 + + . The elements of the determinant are rather cumbersome. and this leads to expressions comparable with those shown as Eq. sin2 H \ 1 „ (\ __ ^i + 2m smajjs 2 .
is seen immediately to be X = 633/033 The The above equations can be frequencies of an solved. for O bonds has been determined example. inter bond angle.A3 roots. noth Exercise 910. Repeat the calculations of Exercise illustrated here. B.. For the H 2 molecule. for the three if H 2 type molecule values of the atomic masses. Phys. corresponding to the antisymmetric coordinate.76 X 10 6 and —„ = 0. 99. but use the method of symmetry coordinates *96.69 X 10 5 dynes/cm the lead to the calculation of frequencies in quite good agreement with 1 and modes observed values of 3. now being done. is calculated.595 cm" for the symmetric 3. On I rearrangement one has &2 X (hi i = I (hi + hl\ + a 22 / ~ 2 \\an 2 \an A 2  8bl2 (49) anffl2 2 a 2 2/ third root of the determinantal equation. the angle between the two One can then verify that the force constants to be 105 deg. : This procedure vibrations is the basis for many of the calculations of molecular Wilson will will be approach Additional material on this therefore be elaborated. It should be pointed out that the on a familiarity with matrix methods. When this is done for each of the three modes can be described by linear combinations of Si and antisymmetric normal coordinate is. The form of each of the three normal coordinates can now be determined.652 and 1. the ratio At. in a particular case. 1. or v. One of the roots of Eq. (48) is substituted into the determinant. The original presentation of The Wilson method sets up both the kinetic and potential energies .At. Wilson [J.218 INTRODUCTION TO MOLECULAR SPECTROSCOPY roots corresponding to the symmetric coordinates result of the 2 from expansion X 2 subdeterminant. fcno H— = 7. and for this value of X. Solution of Vibrational Problems in Internal Coordinates by the Method of Wilson The treatment illustrated in the previous section can be systematized by an approach introduced by E. apart from a constant ing other than the antisymmetric symmetry coordinate <S 3 .756 cm1 for the antisymmetric mode. and force constants are inserted. one sees that the two symmetric <S 2 while the factor. 9 76 (1941)]. depends procedure found in ref. Chem.
and one can immediately write the general expression as 2U = ^k R k R kl t kl will The notation used by Wilson designates the now follow this notation and write force constants by /«. 94. (51) becomes 3n 1 23 = I « t=i in the (53) As we saw ©—©—© molecule example of Sec.CALCULATION OF VIBRATIONAL FREQUENCIES in 219 terms of the internal coordinates Rk .. Thus Eq. in matrix notation. As has been pointed out. the time set derivatives of the cartesian coordinates.'s. one must start with the cartesian coordinate expression 2T = It is I m<x\ (51) defining convenient to absorb the mass coefficients into the coordinates by new coordinates g( that are proportional to the cartesian coordi nates and are related to «« them by (52) = (mi)ixi With these coordinates Eq. or the are related by a . the potential energy can be easily formulated in these coordinates. is recognized to be \0 ki) set To up the 3n kineticenergy expression. 94 are set up to correspond to the general equations given here. for this example. The general discussion of this section ticular equations appropriate to the ©— ©— © may be clarified if the par problem studied in Sec. 94 can be written as and the matrix F. (14) of Sec. g. We 2U = ^fhiR k Ri kl or. (50) W = R'FR where R' is the transpose of R.
writes With Wilson's notation one 1=1 or. can be made square by writing. as six additional The matrix rows. the six conditions of zero translational and zero angular momentum. are 3n Rh = and J) D Mq\ where D =B ki ki (m.220 INTRODUCTION TO MOLECULAR SPECTROSCOPY of linear equations to the time derivatives of the internal coordinates. The general expression of Eq. x is a column matrix 3ft rows. but involving coordinates. comparable with the expresConsiderable manipulation necessary to impose the conditions of zero translation and zero angular momentum so that. (57) can be found and inserted into the kinetic energy expression. D In this way the matrix 2D where (58) © = (». (55) comis pares. (54) and (55). for instance. in matrix form.) _i (56) »=i R = Dq The principal problem now is to proceed from 2T = 1 t=i mi%\ = i=i X 9l and the transformation equation R = Dq to an expression for sion for is (57) T in terms of R that is U in terms of R given by Eq. compared with those which arose [Again is suggested that the expressions here be in Sec.) . 94 for the particular of the — ©—® example molecule. (55) R = Bx where with R is a column matrix with 3n — 6 rows. (21). with Eq. (50). and B is a transformation matrix consisting of 3m it — 6 and 3ft columns.] The the <ji expressions corresponding to Eqs. which appropriate to the ® —© — (f) problem. in effect. the inverse transformation of Eq.
dimensions. Furthermore. the last three terms are zero and one with (69) 2T = R'Q'QR The deduced. remains to see how Q can be conveniently .) In terms of the square matrix <R one has (60) = £>g where as q r = I>o9 is a column matrix consisting of six zero elements. — 6. and Q a 3n X 6 . Now the desired inverse of Eq. D and D . be written down = SCr'CR (62) With as this relation one can express T in terms of the internal coordinates 2T = q'q = ^'(XrO'SCr 1 ^ (63) Further manipulations allow the problem to be reduced to 3n rather than 3ra.CALCULATION OF VIBRATIONAL FREQUENCIES 221 could be formed.= 1 SLV (O'i>'o) (65) (QQo) (ori) ' to/ = (S) (66) the expression for 2T can be expanded as 2T = IT  (R'r') (<Q (QQ a) (f) (67) (R'Q' = Since r is ii'Q'QB + r'Q'„)(QB + Q f) + R'Q'Qor + r'Q' QR + r'Q'oQo* is left (68) a zero matrix. as also does Q . D 1 has been arbitrarily divided into Q. Since "(f) and and and «' = = (#'r') (64) GO ©. desired reduction to It 3n — 6 matrices has the matrix now been formally accomplished. and one could calculate. 1 matrix. (Note that Q depends on both 3>. formally. (60) can. a 3n X 3w — 6 matrix. ar = (QQ 1 ) (59) where © and 2D are square 3w X 3n matrices. in principle.
i. . Gki for k = 1. and the . . .. The first now be made to yield a valuable result. . 2. It is first relation can transposed to give D'Q' + D' Q' = 7 3 „ X 3„ left (72) and then multiplied on the by D and on the an right by Q to give (73) DD'Q'Q + DD' Q' Q = DQ = 7 3 „6 X 6 Now one must recognize that the internal coordinates will be the basis for descriptions of the pure vibrations of the molecule and that these coordinates will therefore be orthogonal to the overall translational and rotational coordinates. 3n — 6 rows and 3n columns.. . where the usually used symbol G = DD' has been introduced for this In view of the relation between the components of D and B given in Eq. 3n  6 and I = 1. . matrix.^ we can .e.e. is a 3w — 6 X 3n — 6 identity matrix.6 matrix Q. which is composed 3n X 3m 3n — 6 X 3n matrix D i. 2. term of Eq.222 INTRODUCTION TO MOLECULAR SPECTROSCOPY The matrix Q useful relationships arises in the inverse of the matrix 2X A number of become apparent if one writes ST 1© = (QQ and ) (£\ = {QD + QoDo) = J 3 „ X 3„ (70) wr> = (£) venient in that of the it ««•>  (£« S%) = '— < 71) where / represents a diagonal unit matrix. (56). It follows that D and D a are orthogonal matrices and that DD' is a zero matrix. . one can express the elements of G = DD' as an Q» = From 2. (73) and leaves This relationship eliminates the second DD'Q'Q = I The product Q'Q Q'Q = desired for 1 (74) T is then obtained as (75) (DDT* = G important matrix. (76) this result calculate each element of the G 6. 3n  (It will be . allows us to recognize that } The second result is conDQ.
96. (80) or (81). and the procedure obtained for setting up the F and G matrices.) One now has B matrix 2T = R'G'R and 2*7 (77) = R'FR is (78) When Lagrange's equation applied to each of the Rk coordinates and vibrationaltype solutions Rh = A k A k 's cos 2rvt are sought. '"Exercise 911. by the method of Sec. from Formation of the G Matrix by a Vectorial Method In a second paper by Wilson a convenient method for setting up the G matrix introduced in the previous. The . (80) or (81). section is presented.1 ) 2 iX F» ^22 (G^OiiX  (G'UX (79) where X = 47rV. The used to illustrate Eq. HCN molecule. In more complicated cases the systematic nature of "Wilson's FG method" is of real value. X's  ff»A = \G\ (80) it is sometimes more convenient to rearrange this determinant by to give multiplying through by \G\ \F  GiA = = \(G)(F A (7»A) \GF  = (81) With either Eq.CALCULATION OF VIBRATIONAL FREQUENCIES recalled that the 223 internal coordinates is the transformation matrix between the and the cartesian coordinates and that it can b e written down from a consideration of a diagram of the molecule. *97. Probably ®—©— ® most of the molecularvibration calculations that have been published have made use of this procedure. If A is used to represent a diagonal matrix with on the diagonals. Calculate the frequencies of the the given force constants. the secular determinant formed by the coefficients Of the has the form Fu F„  (6»)iiX ((?. we can solve for the 3w — 6 values of X that molecular type example can be satisfy the equation. this equation can be written as \F Finally.
there except for the notation based on x y is t.gi) • • form Rk = (B kXl ±i + B^yt + + + These (B kXl xt (B kXn x n +B kVl +B y + 2 ) + B kyJ n + B kz J„) products of components can be more neatly expressed by introducing the vectors Qt many with components x t. 3n  6 (82) B ki arise are determined by the relations 3ra Rk = £• i=l B*A (83) It is now suggestive to write out the atomic cartesian coordinates more specifically to show Eq. . the same contribution as t. . as will be illustrated.. I  1. y. (84) gives Gkt as (86) (79). is expected from Eq. — This is s*t • su — — (B B lXl +B zt ky . the n G matrix elements are simply (?*!=>— bm ti B« (85) This result for Gu can its be verified by considering atom contribution to kXl t and recognizing that Eq. The subscripts a and /3 indicate .) i. rather than on the running index if The formulation of Eq. zt and sjtt with components B kx„ B ky„ B kZl where t numbers off the n atoms of the molecule.Bi Vl + B/c^Bu. . (83) in the 4i . t helpful only is a convenient way of This can be done by introducing unit vectors e a/s along the bonds in the molecule. With this notation Rk is written as Rk = ^ < s*t • p( (84) =i and. . (85) expressing the s k vectors.224 INTRODUCTION TO MOLECULAR SPECTROSCOPY previous treatment required use of the relation e*« = / where the ZT BkiBli terms that k.
say between atoms labeled a and /3.e.. from the atom labeled a to that labeled e vectors. Sa for H 2 0. it is s*i With these possible to set up convenient recipes for determining and vectors. one draw the e vectors directed away from the apex angle thus If agrees to e OH// WCoh' More generally. In this Ska way one has simply = — 6<*0 (87) and Thus if molecule one were dealing with the internal coordinate Ri = H O H' and the « vector had been indicated as — — Sr on for the H one would have / H' and SlO (88) — — «BO Rk involves an angle bending. for example. the s* a and s^ vectors must show how the terminal atoms must move to lead to a bond stretching. and are directed agreed that the unit vectors 0. one might number the three atoms that fix the value of the interbond angle and draw the « vectors as . i.CALCULATION OF VIBRATIONAL FREQUENCIES 225 it is the terminal atoms of the bond. the s vector contributions of each atom (to a particular internal coordinate Rk. If Rk involves the stretching of a bond.
Apply the Wilson FG method to the molecule HOD. two Si sections are applied to symmetric The symmetry coordinates can be set up in terms of the internal . . Some geometric manipulation is now necessary to arrive at the general expressions which result show how. S*2 r 32 sin (r 3 i (90) — . in terms of the vectors. in a very convenient way. which can be expressed as linear combinations of internal coordinates.76 X 10 6 dynes/cm and lc a /r 0H = 0. Calculate the frequencies predicted on the basis of the force constants kno = k D9 = 7. way cosines of angles between bonds of the molecule will arise and will.690 cm. S*l One _ = = cos y>€ 31 : — <p t 32 r» sin : (89) c3i cos ye 3 2 — <p .„„. that the use of symmetry coordinates. the three atoms must move to in an increase in the angle with no accompanying bond stretching.226 INTRODUCTION TO MOLECULAR SPECTROSCOPY Furthermore one lets r 3i and r 32 denote the equilibrium bond lengths and <p the equilibrium value of the interbond angle. 95. allow the secular determinant to be factored into a of determinants of lower order.69 X 10 5 dynes/cm. allow. and 3. Use of Symmetry Coordinates in Wilson's Method It has already been shown. finds that.719. 2. for very many G matrix elements to be readily calculated according to the expression = 7 G "'= ?w 7 — ZT s e  *< s" ( 92 ) The dot products between the various between the In this s vectors will lead to dot products vectors that are fixed along the directions of the bonds. if R k is an angle bending coordinate.1 *98. in Sec. Compare with the observed absorption band frequencies of 1. r 31 r 8 2 sin <p These relations between the molecules. introduce the molecular geometry into the G matrix. the latter intro ducing the molecular geometry into the problem. number This reduction can be performed when the methods of the previous molecules. Exercise 912. the s vectors and the e vectors. r 32 cos *>)«»! Sk3 + On —  r n cos <p)e 3 2 nn mji .402.
and the relation. In Wilson's procedure the kinetic energy is expressed as (98) 2T = R'GiR The corresponding expression involving symmetry coordinates can be obtained by inserting the identity matrices UU. one might write = = S3 = (S 2 + — 5r 2 Sa Sr% 8r 2 symmetric (Ai) symmetric (Ai) antisymmetric (B 2 ) 1 (95) or 0\ /«r. (97) = U' will hold. example will the up to be orthogonal to one another. since U and are known from the way in which the symmetry coordinates were set up and the elements of G V . as Jhis 0/\5a/ illustrates. three matrices are nonsingular. for example.1 = R'U' is recognized as S'. 2 0. and R'lJ. Thus. l7i which be used below.1 and C/_1 C7 to give 2T = R'UWG^UWR is (99) Now UR recognized as S.\ 1 )(*»») (96) 1 Since. we can l write (100) 2T = S'UGWS = S'&. again with U~ = l U'.S where the notation gi = UGW Since all (101) has been introduced. the symmetry coordinates can be set U matrix is orthogonal.CALCULATION OF VIBRATIONAL FREQUENCIES 227 coordinates Rk according to 3JI6 S = t ^ U R lk h (93) * =i (94) or 8 = For UR 5ri H S. the 9 matrix can be expressed by 8 = (£/')= UGU' W 1 (102) This matrix equation allows.
along with the appropriate atomicmass terms in Eq. (104) and sub stitutes these.228 INTRODUCTION TO MOLECULAR SPECTROSCOPY g. (104) and (105) the kineticenergy matrix g of Eqs. 97." except for numerical values sents no difficulty. R = uW = and R'U' U'S 1 (107) = R'U. obtained from Eq. the potential The energy expression 217 = R'FR (106) internal coordinate matrices R and R' can be replaced by symmetry coordinate matrices by use of the relations.kUi. (106) gives 2U = S'UFU'S or (109) 217 = S'5S (n°) . and The the relation of Eq. can be calculated from Eq. (92). as shown in Sec. (97). Thus one calculates the various S vectors from Eq. (102) shows that the elements of 8 are given by 7i Expansion of 3n6 u. If Sj is defined as 3n6 S= I Uun* 4= 1 (104) the elements of the kineticenergy matrix can be calculated from n 8 „. (105). the calculation of Eq.. = ^iSSS i=i S . of the components of the F matrix. to obtain the desired kineticenergy matrix based on the symmetry coordinates of the problem. (94).h 1 ' s "' = 2 (i X kik'= akt sk v (103) This cumbersome expression can be used more conveniently if symmetry displacement vectors S\ that are comparable to the s k t vectors are introduced. (105) With Eqs.= S' or R' = S'U (108) Substitution in Eq. (100) and (101) can be set up. potentialenergy expression based on these coordinates preOne can write down.
and (110) and the methods for obtaining g and ff given by Eqs.CALCULATION OF VIBRATIONAL FREQUENCIES 229 where the potentialenergy matrix been introduced as ff 'S based on symmetry coordinates has = UFU' (100) (111) With Eqs. the vibrations of a symmetric molecule can be analyzed in terms of symmetry coordinates and. Apply the Wilson FG method. New York. . McGrawHill Book Company. (105) and (111). (Compare with the observed frequencies of 519.97 X 10 6 dynes/cm and k' a = 0." Inc. to the 1.360 cm 1 . 95. making use of symmetry S0 2 molecule.. Wilson. B. Jr. The force. constants reported for a simple 6 valence force field are k a0 = 9. 1. Decius. the convenient factoring of the secular determinant will occur.. J. E. and P. C.81 X 10 dynes/cm.150. Cross: "Molecular Vibrations.. and *Exercise 913. as indicated in Sec.) PRINCIPAL REFERENCES 1. 98. 1955. C. as seen in Sec. The molecule is bent and has a bond angle of 120 deg.
In the course of such high energy transitions. that electronic transitions result in We broad absorption or emission bands structure. therefore. in fact. of these states can also be diatomic molecules can a reasonably complete diagram of the energies of the excited electronic states be drawn. as we will see. since all materials. will see. The that 230 is analysis of electronic bands of diatomic molecules is perhaps Follow best introduced by considering the vibrational and rotational structure found in typical electronic absorption or emission bands. and. . their electronic energy. and the energies deduced from studies of electronic transitions. cules exhibit Furthermore. radiation not only as a result of changes in their rotational and /ibrational energies but also as a result of changes in their electronic arrangement and. it must be expected that the vibrational energy considered in this chapter are gaseous. The lack of a simple. The energy changes involved in a transition from one electronic state of a molecule to another are usually relatively large and correspond to radiation in the visible or ultraviolet regions. and that these bands contain a of inertia large amount of fine Analysis of this structure often leads to a wealth of informa tion on the moments and the potentialenergy function for the electronic states involved in the transition. diatomic mole many different excited electronic states. or emit.10 I^^V ELECTRONIC SPECTRA OF DIATOMIC MOLECULES Molecules can absorb. the rota tional energy of the molecule will also change in the transition. the assignment of transitions to For only a few particular states a problem of considerable difficulty. generally adopted pattern for the energies of these excited states makes.
Furthermore. perhaps. . i. Much of the detailed analysis of the rotational structure observed in the gasphase spectra of diatomic molecules does not carry over to the analysis of these polyatomic solution spectra. The potentialenergy curves of Fig. the probabilities of the molecule being found at various internuclear distances. and. .ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 231 ing this analysis of the fine structure of typical bands. . treat only the principal features of the electronic spectra of diatomic molecules and will for those make reference to the very complete treatment given in ref. the square of the harmonic oscillator wave functions. 101 imply that for each of the two electronic arrangements the molecule will vibrate. is i. 2. be mentioned here that the analysis of the is electronic spectra of diatomic molecules a rather specialized study and that most chemists are likely to encounter only electronic spectra of polyatomic molecules. labeled v" = 0. We now consider an absorption or emission of radiation that changes the molecule from one electronic state to the other. lower electronic state and the various vibrational levels. . The Vibrational Structure of Electronic Bands in which the potential energy of a diatomic molecule might vary. a consideration can occur for a particular molecule mention will be made of how an observed band can be will be given. of the upper electronic question. as a function of the internuclear distance for two different The way arrangements of the electrons of the molecule. will see later that shown We potentialenergy curves occur variously placed of the allowed vibrational states. horizontal lines are drawn to represent the energies in Fig. therefore. . labeled 1. typical of a ground and an excited electronic state. The curves shown in Fig. states. . and assigned to a transition between two of these states.. As observation of the vibrational structure of electronic . 101. Three factors must be kept bands in mind to answer this 1. of the . We will. are shown on the energy levels to which they will refer.e.e. the visible and ultraviolet spectra of these species are likely to be obtained on solution samples. . The question immediately arises as to which transitions are to be expected between the various vibrational levels. of the kinds of electronic states that It should. v' = 0. as in Chap. 101 are. however. 2 who wish to pursue further the detailed study of the electronic spectra of diatomic molecules.. 101. Furthermore. 1. or ions. state. 2. 2. two different electronic on such diagrams and with various shapes for the different electronic states of a particular molecule.
Internuclear distance * . and vibrational prob ability functions ^2 for two typical electronic states of a diatomic molecule.232 INTRODUCTION TO MOLECULAR SPECTROSCOPY FIG. 101 Potentialenergy functions. vibrational energy levels.
there are 233 transition going no general restrictions on the changes in v for a from one electronic state to another. only the transition to high vibrational levels of the is spectrum of upper observed. Figure 103 shows absorption bands of CO and I2 which illustrate this behavior. With these three statements one can draw.) This characteristic of electronic and nuclear motions leads to the FranckCondon principle that an electronic transition in a molecule takes place so rapidly compared to the vibrational motion of the nuclei that the internuclear distance It can be regarded as fixed during the transition. Exercise 101. 101. follows that on a diagram. the expected transitions for the potential curves of Fig. to be most favored if it occurs while the molecule has an internuclear distance such that the transition connects probable states of the molecule. from the fact that an electron in a Bohr orbit of an atom completes a revolution around the nucleus in about 10 16 sec whereas a typical molecule vibrates with a period of about nuclear distance. 101. An electronic transition must be expected. in view of the transition moment integral defined in Eq. . as observed in the absorption electronic state Fig. to see a progression of absorption lines in the electronic absorp band due to transitions. In an absorption experiment the temperature is room temperature. often near 102. in view of the shapes of the probability functions sidering transitions to be relatively at the middle of the v = level in Fig. 103a. 10"" 13 sec. such as that of Fig. 101. Draw potentialenergy curves for ground and excited electronic states such that. by conmore probable if they begin or end or either end of any of the higher vi shown brational levels. This rule is usually simplified. 101.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES will confirm. The fact that the probability of the molecule being at a particular internuclear distance is a function of the distance. therefore. 3. It must be kept in mind that electrons can move and rearrange themselves much faster than can the nuclei move to alter the inter (The relative times for electronic and nuclear motion can be recognized. This result is in contrast to the rule of Av = +1 that is operative in vibrational transi tions within a given electronic state. as shown by the probability curves of Fig. and the v" tion v' — level is most populated. One expects. must be taken into account. (425). for instance. with v" = and v' varying perhaps from = to large values. as is done in Fig. 2. electronic transitions must be represented by essentially vertical lines connecting the initial and final states at some fixed internuclear distance.
. to various lower levels. various values. are to be expected. band are usually analyzed The components of such a vibrationalelectronic by arranging the frequencies of the compo nents in a table. and the difference in the frequencies in adjacent rows is likewise approximately constant and varies uniformly. that for a given value of v transitions to states with two different values of v are preferred.e. many from transitions from various upper i. lines arising v' In emission spectra. 102 relatively (o) Some of the most probable transitions. 1026. The FIG. for the potentialenergy curves of . generally lead to a band of exhibited in Pig..mission transitions Examples electronic of the most probable states. (b) low temperature. of the v' levels are appreciably populated. called a Deslandres table.e. as Table 101 shows. between vibrational levels of two (Note. in such a ence in the frequencies in adjacent columns is way that the differ approximately constant and varies uniformly. for an absorption experiment at a Fig. 104. state therefore. various v" The many probable transitions that occur such complexity. that without considerable experience one cannot recognize how the band structure is related to transitions such as those values. 101.) . as that of the molecule N^ shown in Fig. i.234 INTRODUCTION TO MOLECULAR SPECTROSCOPY In emission spectra the temperature is usually high enough so that many levels.
From the data on these vibrational spacings.) In favorable cases. and the Each frequency v' quantum number v" of the lower elecrows with the quantum numbers v' of the upper in the table can then be identified as a transiit tion between the its value of the row occupies and the v" value of column. therefore. Thus for a given value of v' two different v" levels to be preferred. This pattern is understandable in terms of the emission transitions of Fig.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 235 components of an electronic emission band of the molecule PN. for example. It is on the shape and often on the dissociation energy. can be . the pattern of vibrational levels almost up to the limit of dissociation can be obtained.000 an electronic transition of I 2 that leads to an absorption band at about A connecting the ground electronic state with an excited state. can be deduced by a suitable extrapolation. while the differences between the values in the columns give the vibrational spacing in the lower electronic state. 23. and the dissociation energy for the molecule in that electronic state particularly important to recognize that this information of the potential function. 1026 if the columns are labeled with the vibrational tronic state state. a This is shown welldefined potential energy function can be deduced. some of which are summarized in Table 102. bers approximation to the potential function of a diatomic molecule. gies of levels with high vibrational quantum numbers. much about the potentialenergy function. For example. For v' = 6. and one can learn. tend to follow a para curve as would be expected from the potential curves of Fig. the Morse curve that was introduced in Sec. are given in Table 101. in Fig. (The Morse curve is generally accepted as a good detail. as shown by the frequency differences also exhibited. 105 along with. It should now be clear that data from electronic absorption or vibrational energy emission bands provide much more information on the than can be obtained from direct One obtains information on the enerstudies of vibrational transitions. for comparison. preferred transitions are apparently those going to v" = 3 and boliclike v" = 9 or 10. The most intense lines. The one expects transitions to differences between the rows of the Deslandres table can now be recognized as giving the spacing of the vibrational levels in the upper electronic state. The vibrational structure of this band has been measured in considerable and transitions involving groundstate vibrational quantum numup to v" = 114 have been reported. there level pattern of a diatomic molecule is 2. 101 and the three statements given earlier in this section. arranged to conform to these requirements. moreover. such as the I 2 transition mentioned in the previous paragraph.
S <3 . u 3 to ja o t» » E £ — * .F= £ £ I 2 LI V 03 O I * E o < O £ u_ Ji _Q 236 .<k O S 8 •8 o o ° >! X 7 1 j £ *  1— io 1 1 o (3 o m o". 2 4 O _g c § z « C C C + 3 JQ O a.2 6 Kl2 £ P .
BeS 3UOJp33 jaddf) J3AU) o » — £ » J ' 2 — . s S08E O H 5 IS ** to LLS'% ° » a S E £ o a o —SSI'E • Of S* D e i £ •< i 4) ^ O 6 5 D (j £ S ZZ6Z II og ^ w N ii 3 n * o M H C O rr a) e O 237 .
does not consist of a single line but is rather a subband of considerable detail. studied in the previous section.0) 39. l'j component of an electronic the rotational energy of the excited molecule in the J' rotational state.8) 40.4) 3 41. 1 (1322.536.4) 1 1 2 3 (1322. .2 (1072. (Some molecules have been studied sufficiently so that a rather detailed diagram obtained not only for the ground electronic state but showing the variation is of potential energy with internuclear distance for various electronic states of the molecule can be drawn. in a lower electronic state and l'j the rotational energy of the molecule and the /" rotational state.786.7) 38. 859'.2 (1069.3 (1042. This additional detail can be attributed to changes in the rotational energy of the molecule that accompany the vibrationalelectronic transition.698.9) 6 42.476.8 (1087.082. in contrast to studies in the infrared region.1 (1015.068.9) (1319.066.2) 37. An example provided in Fig.4 (1309.) 102.288.861.2 41.5 (1090.5 (1294.3 (1071.3) 38. for excited electronic states. Rotational Structure of Electronic If Bands an electronic band is high resolving power.932.467.2 (1061.238 INTRODUCTION TO MOLECULAR SPECTROSCOPY also.6) 2 41. Van Nostrand Company.7 (1086.0) (1311. If i>o represents the frequency of a particular vibrational transition. it is studied with a spectrograph with sufficiently observed that each of the vibrational com ponents.2) 36. the rotational TABLE 101 The Components of cm Emission Band of PN Arranged in a )eslandres X 39.1) 40. Herzberg. "Molecular Spectra and Molecular Structure.0 7 8 9 * From _G.155." D.7) (1307.9) 40.331.5) 37.9 (1280. 1020.9) 4 S 41.597.
a double its rotational energy. N. which lead to considerable complexity.3 (1029.5) 41.798. the energy that the molecule would have if it were not rotat ing and remained at the state.387.446.6 41.3) 34.756. minimum of the potentalenergy curve for that The symbol G is molecule.0 (1042.239..662.5 (1060.1) 36.0) (1265.236. use is generally made of a different notation than has been used here to designate the various types of energy that a molecule can have.4 36. Princeton.4 Inc. state.712. prime represents a lower energy state..9) (1266. a higher energy Since detailed analyses of electronic bands will not be given here.) 4 6 7 36.3) 35.519.4 36. .7 (1225. particularly in studies of the rotational and vibrational structure of electronic bands.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 239 details of the subband can be discussed (i'j in terms of the expression (1) = h+ — e'J) (It should be mentioned that.027.4) 35.2) 37.e. a single prime..5 (1043.) Although many excited electronic states have electronic angular momentum contributions that couple with the rotational angular momenwill tum of the molecule. 1950. while F represents used to represent the vibrational energy of the Finally. it not be necessary to change over to this standard notation. of the spectral lines are given.3 41.4 (1252.4) 38. T is used to denote the total energy T is used to denote the electronic e energy. in 5 cm 1 . of a given state of the molecule.522.6) 38.2 (1059.J.194. it is here Table* (The wave numbers.6 (1238. i.998.
32: 92 (1960). J' = J (B' v = . one frequently has situa+1 are allowed. = J. Chem. on J depend on the types of electronic However.000 >" ObA* Ai (cm" 1 ) 213. 1 P branch: J" v = J. For these rotational branches is one has.31 1 10 11 200. D. Verma.0 0.. rigid rotor expressions.62 60 61 110. J.5 From R. ? In such cases the subband structure can be expressed as = 5. + B'J'{J' + 1)  B' V 'J"(J" + 1) (2) where B' v and B" are the rotational constants for the vibrational and electronic states involved in the transition.240 INTRODUCTION TO MOLECULAR SPECTROSCOPY sufficient to consider the simpler cases in which the rotational energies for the two electronic states can be expressed by the simple.33 100 101 114 115 * 9. b. further complica tions arise in that the selection rules states involved in the transition.09 80 81 52.68 40 41 154. Phys.. tions in which &J = 0. if the value of J" simply labeled as /. . J' + B' ')J + V {B' v  B' V')J* Q branch: J" i> = J c§: = h +  B'j)j + (5. In general.')/ 2 (3) TABLE 102 Some of the Vibrational Energylevel l Spacings for the Ground Electronic State of 2 tained from Studies of the Electronic Band at 2.
.000 8.000 2. can lengths. tional states equal and to neglect the terms equivalent to (B' v it apparent when the rotational structure of vibrationalelectronic transitions are analyzed is that B' v and B". 12. was a good approximation to set the B values for the different vibra— B"). + 25: + (35./ + (B' v  2 B' V')J These expressions are similar to those encountered in Chap. 7 where vibrationrotation bands were treated. The dashed line is a Morse curve for 2 R. 32: 738 (I960). There. 5:').] FIG. J' = J +  1 = . 105 is line l . be very the bond and therefore the moments of inertia and The principal difference that is different for different electronic states of the molecule. D. it will be recalled.000 6.000 10. The solid computed from the vibrational energy levels deduced from an electronic emission [From band that is observed at about 2.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 241 R branch: J" v = J. (As can be seen in Fig. Chem.000 .000 4. Verma. the minima of the potentialenergy curves for different The potentialenergy curve for the ground electronic state of h. 1020.000 A. Phys. J..
Such a parabola. i. In fact. 106 first recognized by Fort rat. if B'J is greater than B'„ the J* terms of the Q branch can dominate J increases. as this is particularly necessary when the branch is overlapped or forms a band head and moves back on itself. now apparent Insfjfufe lines o( Fig. an example of which shown in Fig.) It follows that the coefficient of the J 2 term of Eqs. as shown for the —» v' = 0. {Courtesy of J.7 increases. B'„ is greater than 8". instead of lower. at high J values. The parabolic behavior that is to be expected on the basis of Eqs. therefore. if. for which the rotational selection rule is A^ = ±1. (3). 108.e. in the rotational structure of each vibrational com A large difference in B[ and B'J has the shown by the second of Eqs. of spreading out the Q branch components and thereby preventing the sharp Q branch that is often a dominant feature of vibrationrotation spectra.805 A . and ponent of an electronic band. FIG.. (3) is then observed. for example. further effect. and the assignment — of J values to individual lines can be made so that all points fall is on a smooth parabola. the linear terms and the high J components of the band can turn around and move This behavior is to lower. known as a Fortral parabola. Similarly. 0. 106. move out linearly from the band origin. v" = band of a CN transition in Fig. and the components of the P and R branches will not. 107. Cose of Technology.242 INTRODUCTION TO MOIECUIAR SPECTROSCOPY electronic states can occur at quite different bond lengths.) 3. A. the J 2 terms of the P branch can. is shown in Fig. The components of the branches of such bands are usually sorted out. (3) can be quite large. dominate the terms linear in J and the branch components will "turn around" and move to higher. frequencies as . frequencies as leads to the often striking feature of a band head. Marquises. A vibrational component of an electronic transition of CuH. Ohio. Enlargement of a component of the emission bond of Ns showing thot the 104 are in fact band heads. by plotting the line frequencies against a quantum number. Cleveland. instead of higher.
280A.) energy electronic Nostrand Company. N. Hohn.740 _i_ _L 80 J_ 25. a component of the electronic band of Curl at 111: 484 (1939). Princeton. Physik. 107 The rotational fine structure of 4. "Spectra of Diatomic Mofecules. parabola formed by plotting the frequencies of the rotational lines against the rotational quantum number in the lower (From G." D. Inc. to values of the rotational constants E' v and B" for the two states involved in the transition. 108 of the A Fortrot 0~0 component of a CN band state. excited electronic states. Z. (3). and H. H. moreover.. In this way one obtains information on the equilibrium bond lengths not only for the ground but also for various These data. [FromH. Van 1950. Herzberg. L 25.900 60 40 jj(cm ) 60 80 10 40 .800 J_ 20 J_ 25. with Eqs.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 243 /j 9 Rn Kl3 FIG.J.] Analysis of the structure of the rotational branches leads.. are often obtained with FIG. GoHrrow. Schuter.
It has now been shown how typical electronic bands of diatomic molecules can be analyzed to yield information both on the shape of the potentialenergy curve. principal . . +1 magnetic quantum number An electronic state of a oneelectron atom is defined. . as application of the approprithree quantum numbers. . . the equilibrium internuclear distance. It is essential for the chemist. . 2.. however. used to describe electronic states. . 104 is. . as a function of the internuclear distance. are available to the electron. the potentialenergy functions can be assigned to particular electronic arrangements and a vast amount of information on molecular bonding is made available. 103. d. . . 2. be characterized. The remainder of the chapter will be devoted primarily to this in end. each identi fied by the values of three quantum numbers. . shows that various orbitals. a oneelectron atom or ion. by values for the Furthermore. minimum of this potentialenergy curve. . quantum number p. . . now necessary to see if the states involved an electronic transition can. i.'l. 3. . . 0.e. Electronic States of Atoms of A review of some features of the way in which the electronic states atoms are described is necessary before the corresponding descriptions Application of the Schrodinger equation to a hydrogenlike atom. that is. for various electronic states It is of a diatomic molecule. 0. and the implications of the notation. to learn something of the notation. These three quantum numbers are n I = = 1. if he is to have access to this information. the value of the quantum number I implies an .244 INTRODUCTION TO MOLECULAR SPECTROSCOPY an accuracy comparable to that provided by direct studies of rotational transitions in the microwave region. .e. the energy and probability function for the electron are given. When this can be done. in some way. of additional interest in that the molecular orbitals discussed there will be used again in connection with studies of the electronic states of polyatomic molecules in the following chapter. and on the internuclear distance at the i. I 1.) angular momentum quantum number and m= — — + I. The material of Sec. of states of molecules are attempted. .. n — 1 (or s. ate operator would show.
however (see. however. 124143. "Quantum Chemistry. p. 1944). and G. New York. our ability to predict a precise energy distribution for the outer electron is and spatial lost.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 245 orbital angular momentum of \/l{l + l)'(fc/2ir) and the value of the of quantum number m implies a component of this angular momentum m{h/2ir) along a specified direction. John Wiley & Sons. E. since becomes dependent on It matically in Fig. n = 4 tn i = * 3^ One electron Af Ad Ap 4« 3d 3p 3* = 2 n =l atom Is One outer electron atom . are shown and m in Fig. and d orbits. By such an extension. Inc. Kimball. an outer electron are not destroyed. for the lighter elements. Eyring.. for example. the important angular for momentum implications of shapes. Walter. 109. can be shown. closed inner shells provide spherically symmetric screening effects and no angular momentum I contribution." pp. that. With the theoretically and experimentally established result that inner shells do not upset the angular momenta associated with the quantum numbers I and m and. This procedure for describing the electronic state of a oneelectron system in terms of quantum numbers can be extended to atoms. 109 Schematic representation of the effect of filled inner shells on the allowed energies of an outer electron. s. The and magnetic quantum numbers m for the which are important in spectroscopic studies. The energy of the outer electron This is shown scheI as well as on n. H. possible orientations. consisting of one electron outside one or more filled inner shells by using an effective nuclear charge in place of the actual charge. J. or ions. 1010. do not greatly FIG.
i. Note that. and for the assumed to behave relatively independently. I. and m are moment they FIG. 1010 like the The shapes and orientations of the s. and d orbitals. p. the angular momentum along the in this direction. If to describe atoms in terms of their if there are several outer electrons.246 INTRODUCTION TO MOLECULAR SPECTROSCOPY upset the approximate energy and spatial implications of the three quantum numbers. rather classical rotation situation. one can begin to describe an electronic state by assigning each outer electron to an available electron orbit.. one can proceed outer electrons.=±2 2 d . m = ±l —^g m= 1 p < J3r mJCw m=±l m= « m= . one can assign values of n.e. field direction is greater the more the orbital projects out from the axis Symbol z r Orbitals (angular factor) Angular momentum in direction of field (<») .
. by [<U)»](2a)«(2p. as can be seen from a rather lengthy quantummechanical argument (see. . since each electron can be assigned a of quantum number +i or — £.)i(2p. "Quantum Chemistry. by the quantum number L which is the counterpart of the individual electron quantum number I. . (In the next section we will see that the writing of an electronic configuration can also be the in describing an electronic state of a molecule. for example. and then the counterpart of the individual electron quantum numstate of ber m will be introduced.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 247 to each electron. for example. presumed noninteracting outer electrons and the values of L for the atom as a whole. Such an abbreviation. and m. spin Furthermore. however. It turns out. (The effect of a magnetic field on a manyelectron atom will be treated later. orO . Eyring. h + h  2. 9). and as is customary. the supposition that the It is not even clear what is meant by the contribution of the individual electrons of the atom when they interact with each other to give some net effect. is said to specify the electronic configuration of the atom." chap. in part. For an actual atom. for example. The possible electronic states. or molecule. and Kimball. h + U  1. . L = h If + h. fluorine Write the groundstate electron configurations for the configuraelectronic tions atom and for the phosphorus atom.)«(2p. electrons are noninter acting cannot be made. 1. the Pauli exclusion principle allows two electrons to be assigned identical values of n.y The number of electrons in each orbit is indicated by the superscript.) first step Exercise 102. shells. that the electronic an atom is determined. Walter. of the individual. of an atom with two outer electrons in orbits with quantum numbers h and l 2 are described by values of L calculated according to . however. Suggest electron of these two atoms that would correspond to excited states. giving a description of an electronic state in terms of the orbitals occupied by the individual electrons. the inner closed shells are set off with brackets. I. which need not be given here. two p electrons outside the inner closed the possible values of L are L = 2. One would indicate the ground state of the oxygen atom. (h  l t) the electronic configuration under consideration has.) The quantummechanical treatment further shows that there is a simple relation between the values of /. for example.
bearing on the energy. D. In spectroscopy angular it is often helpful to focus one's attention on the momentum I implications of the quantum numbers L and S. designate the electronic states corresponding to L = 0. i. say. one Although the fact that the two electrons are p elecp and one d.. the values that S can have S = Si are calculated according to + s2 or S = Si — s2 and. for the two it would be expected that states with <S = 1 would have lower energies than those with S — 0. to indicate L = 0. P.. S = 1. . d. a general rule formulated by Hund. The relative energies of these states cannot readily be calculated but can be deduced from spectroscopic studies. 2. The net electron spin of the atomic state is described by a quantum number S and. This rule can be easily remembered in terms of the obvious electron repulsion and. since the spin quantum number has the fixed value of £. P. I values of 0. . that states with greater spin at lower energies than those with smaller spin.. . this atomic quantity can be deduced from the spins of the individual outer electrons. There is. 2. increased energy that would set in if. is of primary importance in determining the energy of the electronic state. states of the Thus. For the former case the spins must be outer electron example. trons. and for the latter they may be parallel. in this example. . <S = 0..e. . way that one uses s. one uses S. the way in which the two electrons are arranged in the p orbitals. and D p 2 configuration would have somewhat different energies. Just as the value of implies an orbit in which an electron has an angular 1) (h/2ir). the repulsion will be much less. for example. p. the possible S. paired. two electrons were paired up in a single p orbital. For the present example with two p electrons one would 1. or 2 by the for individual electrons. . so also momentum of \/l(l + does Limply an angular momen . however. rather than. . i. Exact calculations of the relative energies of states with different net spins cannot be made. P.248 INTRODUCTION TO MOLECULAR SPECTROSCOPY In the same . 1. i. For two outer electrons.e. also applies both to atomic lie and molecular systems. again. If they occupy different p orbitals. .. to indicate .e. 1. has some letters S. Thus. the value of L. with a given electron dependent on the net electron spin. . one can have S = for a 1 or S = state of an atom. The energy of the electronic is configuration. or D. which two outer electron system. therefore.
and not really correct. and in terms of the angular momentum contributions of the individual spinning electrons one has The previous discussion has assumed that the orbital and spin quantum numbers. momentum of l(h/2ir) rather than y/lQ. Similarly. In fact this is not so. by the vector combinations i which lead to vectors It L that correspond to integral values of L. Instead of using the previous equations to deduce the possible values of the quantum numbers L and S from the quantum numbers of the individual electrons. idea that the quantum number L implies an angular momentum of L(h/2ic) rather than ^L{L + 1) (A/2ir) and I an angular those which give turns out to be much L from the values of h. atom would make the a vector L represents the magnitude s/L(L + 1) (h/2ir) and the direction it is of the orbital angular momentum The of the atom when in the electronic state specified possible values of the vector L are correctly given by L. can be separately treated. + 1) (h/2w). Similarly. Further more. s = \ implies an electronic spin angular momentum of i(h/2ir) on the part of the individual electrons. Since it is the . one can ask for the electronic state of the molecule. and the corresponding angular momentum contributions. easier to draw the vector diagrams. such as if one returns to the early. although in some systems it turns out to be a satisfactory approximation. 1011 to represent the various possible combinations of 1 to form the net vector L. how the angular momentum con tributions of the individual electrons could lead to net angular To combine the momenta angular momen tum contributions. In this way one would draw vector diagrams such as those of Fig.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 249 turn of \/L(L + 1) (h/2w) for the electronic state of the atom. and S implies a net spin angular momentum of S(h/2ir). The vector model assigns it is an orbital angular momentum vector 1 to represent the magnitude + 1) (h/2w) and the direction of the orbital angular momentum if contribution that each of the electrons of the electrons were noninteracting. one describes the total angular momentum due to the spin of the electrons of the atom by a vector S. is what y/l{l known convenient to introduce vectors and to deal with as the vector model of the atom.
The quantum number / is used. 2ir ft i 2?r 2 — JL 27T '2~ D FIG. momentum of the atom that is really quantized.250 INTRODUCTION TO MOLECULAR SPECTROSCOPY "11 i — 2ir i * L2*A. This is illustrated in Fig. that the correct vectors J are obtained by vectorially combining FIG. and the value of the total angular momentum of the atom is given by s/J{J + 1) (h/2ir). In terms of t The vector diagrams that represent the ways to give resulting vectors. lOll PS is L=0= trons to give the total The vector additions of the angular momentum vectors of two porbit elecan angular momentum vector for the atom. It turns total angular ingful out again. 1012 can L and S. Again the vector model of the atom introduces a vector J to represent this magnitude and the direction of the total angular momentum.*. 1012. in which the and S vectors be combined J = 3Ss= 2 1—5._3 * = 1 fK L = 27 J *'2 The three components ot a 3 D term resulting from The two components 2 of a D term resulting from S = 1. L=2 1 L=2 .— 2 27T 2JT L=2. and this another illustration of the reason for using the vector model. . the meanquantum number of the atom is that which determines this total amount. L = 2 .
\L S\ electronic state of a free suitable for spectroscopic studies. or half integral if J is This is illustrated in Fig. *P}. a position to proceed to a similar description of possible electronic states of molecules and then to assign these states to the observed electronic transitions.L + S2. as mentioned earlier. if the total angular 5. . . . treated in Chap. diatomic molecules are of We will see that the electronic states of similar. . . that. atom can now be described. . multiple of Qi/%r). Just as for the case of a rotating momentum a quantummechanical treatment shows of the atom is y/J(J 1) (h/2ir). customary to indicate the values of these three angular momentum quantum numbers by a term symbol. + the component along the direction of the applied field will be M (h/2*). S. J Again the vector model is convenient in that it suggests that the angular momentum vector J can take up various orientations relative to the angular applied field and that these orientations are such that the component momentum along the field is an integral. D. To indicate that L has the value 0. 1013. S. . we indicate this number for L. of the atom must be quantized along the direction of the electric molecule. We are now in 104. . + S. forth. P. . results for such atoms indicate that the atomic orbitals must orient themselves so that the electric field. of ways (2S + 1) as a left superscript on the symbol for L. ways that the spin quantum number can be combined with the orbital angular momentum quantum number is often an Since the of number important feature (energy levels with a given value of L are sometimes split into sublevels. one writes. . . where M = J. . J = L The L. J  1. the value of J is written as a right subscript descriptions such as 2 Thus one encounters <Sj. . .ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 251 quantum numbers. . 1. depending on the coupling between L and S). on the symbol 2 Pj. and so Finally. component of the angular momentum field. 2. It is . . as well as theoretical. to those of atoms under the influence an Experimental. for the electronic state of an atom. . in a manner by the values of the quantum numbers and J. Electron Orbitals in Diatomic Molecules it is generally impossible to calculate the energies and detailed electronic arrangements for the possible elec As for polyelectronic atoms . half integral.L + Sl. the value of J is given correctly by . in many respects. 0.
not only the orbital angular momentum. with + 1)(h/27r) vectors. the internuclear distance diminished to zero. 1. although an orbit of the united atom is characterized by its total angular n momentum. . 1013 to Vector diagrams field illustrating that an applied by the angular three orientations can be taken up momentum vector corresponding to a J = J(h/2ir) is relative 1 state. first consider the united atom of some diatomic molecule and investigate how the electron orbits are to be described as the nucleus of the united atom is imagined to be subdivided and the separate nuclei that correspond to those of the diatomic molecule are formed. It becomes necessary. as indicated by the quantum number I = 0.. initial effect The on the atomic orbits can be understood on the basis of the fact that the divided nucleus presents a field of axial rather than spherical symmetry. whose quantization can be immediately recognized. 2. . 27T . Either diagram shows that the J = 1 state splits into three components. The effect on the orbits is comparable to that of an external electric field applied along the direction in which the bond is being formed. by means of values of the quantum numbers momenta. to specify. this can best be done by first considering the allowed individual electron orbits and then investigating principally angular the molecular states that result when certain of these orbits are occupied by electrons. if the orbits of such a deformed atom are to be described. i. therefore.. and the separated atoms.e. . On the left the simpler diagram based on vectors of length J[J shown. . On the right the correct diagram. FIG. axis. is given. The electron orbits of diatomic molecules are best described in terms of the orbits to which they would go in the limits of the united atom. Let us the internuclear distance increased to infinity.e. Just as for the case of atoms. M =1 2ir J = 1 < M=0 M = l 2ir = . states Again it is necessary to describe the allowed of the properties. i.252 INTRODUCTION TO MOLECULAR SPECTROSCOPY tronic states of molecules. but also the component of this along the internuclear Thus.
.) With this notation the step in the formation of a diatomic molecule from the united atom can be illustrated as in Fig. the orbital angular momentum quantum number is of major importance and is designated for each electron by I. the orbit projecting along this direction. 1. . . This distinguishes between orbitals of the given atom I that have the same value of internuclear axis that is but different orientations relative to the The three p orbitals. when the two separated atoms that correspond to the diatomic molecule are allowed to approach one another. 1. angular . the principal orbital angular is momentum quantum feature the component along the axis of the molecule. by the component of the angular momentum along the molecular For a given value of I of the initial atomic orbit. is when the diatomic molecule (It is heteronuclear. 8. . . is followed consists of using Greek letters for molecular quantities and Roman letters for atomic properties. . 1015a. . The when the separated atoms are different. 2. and are designated as irp . in the spherical field of an atom. the quantized momentum components along the bond direction can have the The symbol X is usually used to quantum numbers 0. .) A second atom affects the orbitals of a given atom by imposing a direction in space. 1014. . It is well to note the geometric arrangement of the electron orbit associated with each energy level as well as their angular momentum contributions. designate the quantum number for the component along the interhuclear are usually indicated by saying axis. momentum along this axis. number initial for this therefore. which has zero angular momentum in this direction and is therefore labeled as <rp. . for example. i. customary to represent the unitedatom limit at the left of the page and the separatedatom limit at the right of the page. Thus.e. . these orbits will be characterized also (and for large distortions better) axis. and values of X = 0. 2. which might be labeled p x pS) and pz are identical in the spherical symmetry of an isolated atom. is differently affected than being formed. have one unit of angular orbitals. When an axial direction is imposed. designated by X. For the cyclindrical that field of a diatomic molecule.d.ELECTRONIC SPECTRA OF DIATOMIC MOLECULES 253 or by the letters s. . . . the two orbitals that project perpendicularly from the axis. (The nomenclature principle that the orbit is a a. and the is..p. ir. I. . orbit. This order is maintained in the diagrams for initial effects in Figs. In a similar way. the electron orbits are affected perturbation that this causes by the imposition of an axial direction. as the atom is deformed toward a diatomic molecule. is illustrated in Fig. 1014 and 1015.
tations of the orbitals The notation based on angular momenta and the orien are shown. 1014 The splittings of the atomicenergy levels that result from the initial step in breaking apart a united atom.FIG. 3dS 3d 3/J7T 3p 2pir JF""£ 2p2p<r 2sa * UUnited atom lsu United atom "deformed" nucleus with * Angular parts of orbitals 254 .
3 O IB isg "I Hi ?§§ 111 b * N 6 o o a a o b fcfcfci 5 ° 2 I JE S £ o o .E n ~^"^ "^h"^"' O" o Ik CO .
1016. familiar As should be the H 2 molecule. 1016 states deduced for the perturbed limits of the united and A squarewell analogy to show the importance of the g and u character of wave functions resulting from the uniting of two similar systems. one additional feature must be recognized in order to see how the orbits of the separated atoms interact as they go over into molecular orbitals of the diatomic molecule.London description of is evident from the simple onedimensional illustration of Fig. designated by u. 10156. or as The FIG. designated by g.256 INTRODUCTION TO MOLECULAR SPECTROSCOPY In the case of homonuclear diatomic atoms. or antisymmetrically. as is done in Fig. from studies of the Heitler. Infinite separation Slight interaction Large interaction (comparable to bond formation) . These symbols must be attached to the description of the orbitals. two like wave functions can be combined either symmetrically. to higher Notice that the wave function with the node goes energy as the bond is formed while that without a node goes to lower energy. when molecular orbitals are formed from like atoms.
) orbitals of heteronuclear diatomic molecules. The two limits can be correlated with each other. The vertical broken lines indicate the approximate positions which give the correct ordering of the energies of the molecular orbitals for the indicated molecules. Herzberg. states with the same values of X. 1014. Fur thermore.KTRONIC SPECTRA OF DIATOMIC MOLECULES 257 separated atoms must now be connected to give the desired molecular orbitals appropriate to intermediate nuclear distances. a jt with a w state.H. as shown by the wavefunction sketches in Fig. and so forth. Inc.. 1017 A correlation diagram describing the energies of the electron 1950.e. 1017 and 1018. i. one joins up. 1017 and 1018 allow the electron orbits of a diatomic molecule to be described and the order of their FIG. when it is realized that the quantum number X is a good one even when the axial field is very important as it is at intermediate internuclear distances. as in Figs. "Spectra of Diatomic Molecules.J. Thus. for example. have the correct antisymmetry character. D. <rp states which. This can be done by joining orbitals designated by u for separated atoms with. for homonuclear molecules it is necessary to preserve the g and u character of the orbital.." Van Nostrand Company. The correlation diagrams of Figs. (From G. Princeton.. a a with a a state. to give a correlation diagram. SEPARATED ATOMS . N.
1019. "Spectra of Diatomic Molecules. 1018 A correlation diagram describing the energies of the electron Inc. D. Correlation diagrams are resorted to because of the relative ease with which the general shapes and energies of electron orbits of atoms can be described and the great difficulty of doing the same for molecules. Fig. (From G. The the nuclei is of lower energy (recall the ordering of particleinabox FIG. UNITED ATOMS . The correlation diagram gives immediately relative energies of molecular orbitals. Herzberg. in Fig.) Van Nostrand Company. 109 and provide essentially the tronic state of a molecule as They are. N. orbitals of homonuclear diatomic molecules. when the molecule is formed.. equivalent to same basis for describing the elecdoes the use of the hydrogenlike wave func tions for a manyelectron atom. are always the orbitals two of that to form a high electron density between one overlaps possible.258 INTRODUCTION TO MOIECULAR'SPECTROSCOPY energies to be roughly estimated.. therefore. Princeton.J. One should notice that two arrangements irp orbitals and that overlap. some information on the Something of the shape of the corresponding orbitals can also be deduced by looking at the shapes This is illustrated for the case of ap of the limiting atomic orbitals." 1950.
figuration for the and S aid of Fig. and w*p apply. is of higher energy and is known as an antibonding orbital. and 2ttpb . These shapes indicated far up. as a basis for describing electron changes that occur. The second arrangement. . particularly when polyatomic molecules are studied. and jr„2p in Fig. 1017 and c c 2p. xp. for example. 7r u 2p. in which the function must change sign and have a node between the nuclei. 1018. in Fig. 1017. to the energy levels labeled 1apA laps. cr„2p. wave functions according to the number of nodes) and is said to be a bonding orbital.ELECTRONIC SPECTRA OF DIATOMIC MOIECULES 25? FIG.) With the . In a similar way. 1019 Diagrams indicating something of the shape of some molecular orbitals. (Note that each w energy level can accommodate a total of four electrons. suggest an electronic conground state of the molecule NO. 2wpA. one can draw the bonding and antibonding orbitals that are formed from the coming together of other atomic orbitals. orbitals These qualitative pictures of electron are often used. a*p. Exercise 103.
260
INTRODUCTION TO MOLECULAR SPECTROSCOPY
Exercise 104.
With the
aid of Fig. 1018 suggest
2.
an electronic con
figuration for the molecule
105. Electronic States of
Diatomic Molecules
The arrangement
tions that are based
of the individual electrons in a given electronic
by the individual electron descripon the correlation diagram. Some examples are shown in Table 103. If the molecule is composed of like atoms, the separated atom designation for the orbitals is more revealing and is used. For heteronuclear molecules, the united atom designation is used. As for atoms, the electronic state of the molecule depends on the
state of a molecule can be described
net or total electronic arrangement. Various electronic states can arise from a given electronic configuration, i.e., from a given description of the electrons considered one at a time. The electronic arrangement of the molecule as a whole can best be characterized by the net orbital angular momentum component along the internuclear axis and the net electronic spin angular momentum along this axis. The orbital component along the molecular axis, designated by A since this quantity depends on the individual electron contributions designated by X, is easily obtained by summing up the contributions of the separate atoms, for a, 1 for w, 2 for S, and so forth. The value of A = 0, 1, 2, is indicated by writing a term symbol 2, II, A, The spin angular momentum along the axis is, as for the atomic case, indicated by a superscript giving the
. . . . . .
.
multiplicity of the
state.
If
the molecule
is
is
homonuclear, the net
symmetry,
i.e.,
the g or u property,
significant for the molecule as
is
well as for the individual orbitals.
(The electronic state
is
of.
even and
is
labeled with a g
if
the number of u electrons
if
even, and the state
is
odd
and labeled with a u
the
number
u electrons
odd.)
Finally,
TABLE 103 Some Examples of the Descriptions Used for Electronic States (K is used to denote 2 electrons in Is orbits, L to denote 8 electrons in the 2s and 2p orbits)
Molecule
Ground configuration
(<r„l«) 2
First excited configuration
H N
2
12+
l
(<r
Li 2
2
KK(<r,2s)'
KK
(<r„2s) 2 (<r„2s) 2 (w„2p) 4
^t XK(
12+
2
K
n
c ls)a„ls
^J, 3 2J
'2tf,
KK{<r g 2s)<r„2s
ffl
3S «
,2s) ! (<r„2*)»(T«2p)*
"n,, 'n„
(<r,2p)hr a 2p
LiH
iC(2«ff) 8
K(2s<r)2p<r
i2+,
4
s
2+
2
CH
£(2sa) 2 (2s<r) 2 2pir
K(2sv)*2p<r(2pw)'
2~, 2 A, 2 2+,
S
ELECTRONIC SPECTRA OF DIATOMIC MOLECULES
if
261
a molecule <— or
is
in a
2
state as a result of the opposition of
two w
or 5
angular
momentum vectors,
<
these vectors can have either the orientation
—
>
>.
It
turns out, by an argument not easily given, that these
alternatives correspond to molecular orbitals that are either symmetric
or antisymmetric with respect to a plane through the molecular axis.
this, and the discussion of Sec. 54, we label 2 states, which from two v or two S orbitals, as 2+ and 2 _ (It should be clear, in view of Fig. 1014, that antisymmetry with respect to any plane through the internuclear axis cannot be achieved from <r orbitals.) The
In view of
arise
.
notation used for molecular electronic states
is illustrated,
along with
the indicated orbital occupancy, in Table 103.
We are now at a stage where we can describe, in terms of angular momenta, the different electronic states that can be expected for a given molecule, and with the aid of the correlation diagrams, we can tell, although only approximately, the order of these states on an energy
scale.
106. Potentialenergy
Curves for Electronic States of
Diatomic Molecules
Although the energies
of states arising
from
different electronic
arrangements for diatomic molecules cannot yet be calculated, the
electronic states of diatomic molecules can, as indicated
by the lengthy
It
discussions of the previous sections, be described
and
labeled.
can
now be
pointed out that a given electronic transition can, in some cases,
be associated with a particular pair of states.
vibrational structure of the spectral
When
this
can be done,
the detailed information obtained from analysis of the rotational and
band can be used along with the
electronic configurations of the states involved so that diagrams, such as
that of Fig. 1020, can be constructed.
The way
here.
in
resulting in an electronic absorption or emission
which electronic states are assigned to the transition band cannot be treated
is
The
subject
covered in considerable detail in the references
It
given at the end of this chapter.
need only be mentioned here that
the principal basis for the assignment stems from a comparison of the
observed rotational structure with that expected for transitions between
various types of electronic states.
tions
is
strongly affected
The nature of the rotational transiby the coupling between the electronic angular
one observes such features as
momenta
and the rotational angular momentum P and R branches with or without a Q branch, P and R branches with each line
of the states involved of the molecule.
As a
result
262
INTRODUCTION TO MOLECULAR SPECTROSCOPY
FIG. 1020
The potentialenergy curves for various electronic states of the molecule Cj.
Inc.,
(From
G
is
Herzberg, "Spectra of Diatomic Molecules," D. Van Nostrand Company,
Princeton, N.J., 1950.)
(More recent data indicate that some revision
The notation
is
necessary.)
.
.
The
8,
symbol X
C,
is
inserted to identify the ground state.
a, b, c,
.
and A,
...
added
to identify the different states of the
same
multiplicity.
80,000
10
C( D)
60,000
1
+ C('S)
C( P)
3
+ C( \SJ + cr 'a)
cr'oj
£(cm
)
40,000
C( P)
3
+ C( D)
1
£(e.v.)
C( P)
3
+
C( P)
3
20,000
0K10~ cm)
1
ELECTRONIC SPECTRA OF DIATOMIC MOLECULES
split into
263
two or three
lines,
missing early
members
of the P, Q,
and
R
branches, and so forth.
For some diatomic molecules, a fairly complete analysis of the ground and lower excited states can be made. For many more molecules, only
specific
a few of the observed bands can be definitely attributed to
electronic states.
Much
of the wealth of chemical
bonding
is
information that has arisen from studies of electronic transitions
summarized
results
(1961)].
Appendix of ref. 2. A summary of some more recent has been given by P. G. Wilkinson [J. Molecular Spectroscopy, 6 :
in the
PRINCIPAL REFERENCES
1.
Herzberg, G.: "Atomic Spectra and Atomic Structure," Dover Publications,
New
2.
York, 1944.
Herzberg, G.: "Spectra of Diatomic Molecules," D.
Inc., Princeton, N.J., 1950.
Van Nostrand Company,
3.
Gaydon, A. G. "Dissociation Energies and Spectra of Diatomic Molecules," Dover Publications, New York, 1950.
:
4.
Mulliken, R. S.: Rev. Mod. Phys., 2(60): 506 (1930); 3: 90 (1931); 4: 3 (1932).
11
ELECTRONIC SPECTRA
OF POLYATOMIC MOLECULES
The wealth
of
experimental data and detailed analyses that characterize
is
the spectroscopy of electronic states of diatomic molecules
in sharp
contrast to the situation encountered with most polyatomic molecules.
Since the energy of an electronically excited polyatomic molecule can
usually be distributed so as to break a
cule,
weak chemical bond
of the mole
bound highenergy
electronic states are not so
diatomic molecules.
A
further consequence of this
abundant as in is that a sample
transitions
usually cannot be heated to a temperature at which emission spectra
can be obtained without risk of decomposition.
excited electronic state that
The only
that are normally observed are, therefore, those from the ground to an
show up
in absorption spectra.
Only a
few bands arising from such transitions are usually obtained, and since
the samples under study are often solutions or solids, the absorption
bands are often broad and relatively structureless. (Spectra of compounds that will be studied later in this chapter are shown in Figs. 112, 1110, and 1118.) The revealing rotational structure of diatomic absorption bands is, therefore, absent, and a band does not immediately
reveal the nature of the electronic states involved in the transition.
In spite of the relatively greater
difficulty
involved in the analysis
of the electronic spectra of polyatomic molecules,
much work
has been
and
is
being done in this area.
Spectral data on the nature and energies
than the ground state of the molecule provide important testing data for theories of chemical bondThe geometry of molecules in excited states is not necessarily the ing.
of electronic configurations of states other
264
ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES
265
same
as that of the ground state; formaldehyde
and ethylene
are, for
example, nonplanar in one or more of their excited states. In studies of reaction mechanisms, particularly those resulting from photochemical
excitation, such information
on excited electronic states
is
basic to a
detailed understanding of the mechanism. It should also be mentioned that the absorption bands due to electronic transitions, which usually occur in the ultraviolet or visible
spectral regions,
have long played an important role in the analysis and A number characterization of both organic and inorganic compounds. been developed which predict generalizations have of important empirical and intensity of the on the frequency effect structural changes the of absorption bands of a parent molecule. The relation between the electronic structure of the absorbing molecule and its absorption spectrum that will be introduced in this chapter provides some basis for the understanding of such empirical relations and for the use of ultraviolet and
visible spectra as a tool for the
study of electronic structures.
chapter
The material
falling within the scope of the title of this
can be conveniently divided into four topics.
The
first
topic concerns itself with molecules that absorb radiation
is
because of an electronic transition that
or group, of the molecule.
essentially localized in a bond,
Most studied
of such systems,
and suitable
to illustrate
the procedures used, are molecules containing a carbonyl
group
/
C=0.
The
analysis of the spectra of such systems
is
similar
to that for diatomic molecules, but important differences arise because
of the experimental limitations
mentioned previously and the
be quantized.
loss of the
axis along
which angular
momentum would
The second molecular type
to be studied consists of molecules con
taining conjugated 7relectron systems.
Aromatic molecules constitute
the largest and most studied group in this category, and after a brief
discussion of linear conjugated systems, the
way
in
which one attempts
to describe the various electronic states resulting from excitation, of the
jr
electrons in these molecules will be introduced.
The next topic deals with the absorption spectra produced by systems that contain a transition metal ion in a coordination compound. Many such systems are colored; i.e., they have electronic transitions that result in absorption in the visible region. The spectrum of such a compound is, therefore, an often referred to characteristic of the material. Analyses of such absorption bands in terms of the electronic states
the diagrams of the type used in the previous chapter. TRANSITIONS LOCALIZED (N BONDS 1 OR GROUPS 1—1  Electronic States of Localized Groups Chemists have devised a number of ways for describing the ground One of these. Finally. Thus. N.of the bonding in these coordination compounds. orbitals projecting in the direction of the and the bonding orbital resulting bond are said to be a orbitals. HjCO.266 INTRODUCTION TO MOIECUUR SPECTROSCOPY involved lead to further understanding.. or s and p hybrid. i. detail the spatial could be drawn. as . H H:C::0: where the placement of the dots represents the role of the outer electrons With the advent of quantummechanical descriptions based on the Schrddinger equation solutions for the hydrogenatom in the molecule. There will also be. these diagrams could be refined to show in somewhat more arrangement of the electrons. Lewis. Thus. from the overlap of such orbitals is can be seen by inspection of known as a a bond. bonding p. problem. In view of the notation used for diatomic molecules. this approach would give a diagrammatic description of the giound electronic state as electronic state of a molecule. for formaldehyde. uses dot diagrams to indicate the number of bonding and nonbonding electrons in a molecule. and some features of the processes of fluores cence and phosphorescence will be introduced. These atomic orbital type diagrams suggest a method for describing the available electron orbits in molecules of this type. the process electronic state can get rid of its excess energy state will by which a molecule or an ion in an excited and return to its ground be investigated.e. due to G.
the diagrams given in Fig. . 111 are important in electronic for the orbitals of the carbonyl bond that transitions of that group should be noted. If only the higherenergy 2p orbital transitions. It is enough. The ir bond between the carbon and oxygen atoms then is formed from the remaining carbon atom p orbital and one of the oxygen p orbitals.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 267 the correlation diagrams of Figs. as for diatomic molecules. and the behavior under the symmetry operations of the molecule will be seen to provide the best characterization of orbitals in polyatomic In anticipation of this. which will be occupied in the ground state of H 2 CO. orbits for a given molecule are recognized. momentum values can be used to charac We will see. which later abbreviated. They can. Furthermore. should be recognized that. [(lSc) 2 (lSo) 2 (2s 2 will ) (<r CH ) 2 (<r C H') 2 («rco) 2 ] (ttco) 2 («o) 2 The three <r bonds to the carbon atom are thought of as being formed from sp 2 trigonal hybrid orbitals on the carbon atom. 1018. 1017. the ground state of H CO 2 can be expected to correspond to the designation. no exact angular terize the orbitals. which are a lowlying 2s orbital and the remaining 2p orbital. a corresponding higher energy antibonding a orbital. Once the available individual electron electronic configuration of the molecule. The brackets enclose the lowlying inner and bonding electron orbitals that are not normally expected to be involved in the transitions studied. and 1019. that the shapes of the orbitals can be drawn well enough so that the symmetry of the orbitals can be deduced. if hybridization of the atomic orbitals on oxygen is not considered. not occupied in the ground state. Fig. and <r* orbitals as by a shown in *. This antibonding usually denoted is. and a higher energy antibonding orbital designated by ir*. however. produced by the overlap of the two orbital is <r atomic orbitals. the exact spatial distribution corresponding to these orbitals cannot be calculated. The atomic p orbitals that project perpendicularly from the bond direction and form a bond are said to be it orbitals. molecules. be located in the available oxygenatom orbitals. 1019. the The principal distinguishing feature of a node in the <r* orbital. The overlap of such atomic orbitals leads to a bonding w orbital. electrons can be assigned to these orbitals to specify an be Thus. The remaining electrons that are of interest in the study of electronic transitions of such groups are "lonepair" or "nonbonding" electrons on the oxygen atom. it is considered to be involved in the observed can be designated simply as a nonbonding orbital and the symbol n can be used It for it.
some of the excited states that might be expected to be impor tant can be indicated by [Inner electrons] (5r) 2 (m)(jr*) [Inner electrons] (x) (n) a (it*) [Inner electrons](x) 2 (?i)(ff*) FIG. . for the groundstate electron arrangement.2fi 8 INTRODUCTION TO MOLECULAR SPECTROSCOPY therefore. [Inner electrons] (jrC o) s (n or simply 2 ) [Inner electrons] (5r) 2 (n) s Similarly . 111 Some spectroseopically important orbitals of H2CO and their symmetry properties. to write.
111. A B % 1 1 1 1 1 Bi 2 1 1 1 1 1 Tx Ty 1 . symmetry classification of the individual orbitals. molecules.. Expressions such as those given above constitute electronic configurations for the various electronic states of the H CO 2 molecule considered here. 3 For some purposes the electronic state electron descriptions.e. for a given assignment of electrons to the individual orbitals. First one orbitals. a description. of a molecule is adequately described by such electronic configurations. as is customary. It is must consider the symmetry types of the The molecule being used as an illustration here. or any symmetric ketone.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 269 and so forth. by such individual more thorough understanding of the electronic transitions of polyatomic molecules requires. Write the electronic configurations for the ground and several excited states of the acetonitrile CH CN molecule. compatible with the symmetry of the molecular skeleton. It is this symmetry behavior that best characterizes electronic states of polyatomic along an axis. The character table for this group is shown in Table 111. As previously menon the basis of a net tioned. A i. momentum effectively H 2 CO. recognize that the total electronic wave function must have a symmetry that is. is obtained. this cannot be done for nonlinear molecules orbital angular One can. as for diatomic molecules. Assignment of the orbitals of Fig. Exercise 111. however. belongs to the point group C 2 „. one recognizes that. individual now necessary to assign the total electronic First wave function of the molecule to a symmetry type. shown in Fig. of the electronic state of the molecule as a whole. 111 to symmetry species can readily be made by investigating whether the orbital is left unchanged or changes sign as a In this way the result of the four symmetry operations of the group. to an approxi TABLE 111 CgB The C c2 2„ Character Table E 1 <r v Ai 1 1 1 1 T. In the above. a superscript 1 is understood and is not written when an orbital is singly occupied.
* = #>(1M2)*(3) where ¥>(2) <p(l) is • • • the wave function for the orbital occupied by electron the wave function results of The by electron 2. the net char usually acter for all the doubly occupied orbitals is the set of +l's that constitute of the the totally symmetric representation. to transform according to the symmetry type this result B 2. is The task greatly when recognized (1) that many (in the ground state. The effect of the given in the third row of the table. as we did in the symmetry behavior of the transitionmoment inteis equal to the fact that the character for such a product of functions direct product for the product of the characters of the individual functions. if electron 1 is in an orbital of type A 2 and electron 2 is in an orbital of type B h the result of symmetry operations on <p(l) and <p(2) could be tabulated as shown in Table 112. (In group theory one uses the term such a product of two functions.) The product With <p(l)(p(2) is found. investigations of gral. Presumed to Be of the Type A it and <p(2). can then be easily deduced from the symmetries of the singly occupied orbitals. Thus.1«>(1M2) 1«. The results are shown in Table 113.(2) mation sufficient for symmetry considerations. according to the results of Table 112. wave function wave functions Thus one writes (1) 1. one total electronic wave function is of the molecule with given electronic assignments. electronic states of The symmetry types H 2 CO. corresponding to the configurations previously displayed.(D 1*>(1) + M2) (+1)( + 1MD«>(2) ( 1„(2) + 1)(1)„(1M2) = 1«>(1M2) = + 1«. can calculate simplified and the symmetries the symmetry type of the it of the individual orbitals.(2) = MDt>(2) +i«. for example. the total for the molecule can be written as the product of the describing the individual behavior of the electrons.270 INTRODUCTION TO MOLECULAR SPECTROSCOPY 1 TABLE 12 The Effect of Symmetry Operations on <p(1). of the Type B u and on the Product ^(1)^(2) C. .(2) (DC1M1M2) = + M1)«. R V {1) = fl».(1)*>(2) + M2) (1X + 1M1M2) . the various symmetry operations on the <p's for that occupied are known. symmetry operations on the product ^(1)^(2) is We have used again. In view of this. and so forth. all) of the orbitals are occupied by two electrons and (2) that the product of one nondegenerate representation and itself necessarily leads to the totally symmetric representation.
Electronic Transitions and Absorption Bands A schematic absorption spectra of a carbonyl compound such as formaldehyde or methyl ethyl ketone. state.. the molecule can be in a singlet the molecule can be in a triplet state.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 271 Electronic States of TABLE 113 Some Electronic Configurations and Formaldehyde Configuration Symmetry of elec Transition moments from tronic state ground state [Inner electrons] (x) 2 (n) 2 [Inner electrons] (ir)*(n)(?r*) [Inner electrons] (ir)(n)*(ir*) [Inner electrons] (ir) 2 A. = Im«I = 0.. since all the electrons are paired and each pair must have opposite spins. this section is to see if is shown in Fig. the net spin can be either zero. howmentioned.. The designa tion singlet or triplet must. and known as Hund's rule. It follows that the ground state of H 2 CO must be a singlet state. A B 2 A. ^ * Exercise 112. ignored the relative directions of spin of the electrons. i. This rule states that. have a lower energy than the corresponding singlet Thus. The Pauli principle requires that the spins of two electrons occupying the same orbital be opposed but allows. when orbitals are occupied by single electrons. is also applicable to molecular systems. or unity. 2 W= = kl kl = \fy\ (n)(<r*) = M= *. ever.e. 0. The object of such absorption spectra can be understood on the . so far. One further feature of the electronic states of molecules must be The description has. either spin orientation. 112. the net spin must equal zero. lower energy electronic states are to be expected when the above excited configurations are occupied by electrons with parallel spins than when they are occupied with paired spins. therefore. the triplet state will state. if both singlet and triplet configurations are possible with the same set of orbitals. 1112.e. be added to the description of a molecular state. This qualitative discussion is as far as we need go must in attempting to describe the electronic states that are involved in the electronic transitions with which we will be concerned. /i. Verify the symmetry assignments shown in Table 113 for the various electron configurations. Now we see if the observed absorption bands can be correlated with transitions between the various electronic states.e. i. It can be mentioned in this connection that the rule based on atomic spectral results. „. i. For the excited states considered above.
FIG. if It is first necessary. in the previous section.272 INTRODUCTION TO MOLECULAR SPECTROSCOPY basis of the electronic states of molecules as described. the relative energies of the various electronic arrangements of the molecule can be deduced.000 2. for example. a S a E i « £ a § 1 1 1 1. According to the discussion of Sec.000 3. 40 it is necessary for one of the integrals transitions between the = Kl = Mx JVex»it«dM*f ground dr J^MiitodftdrVoimd A" or M* = MsTcitodM^ground *" to be nonzero for the transition to be induced. If assignments of the absorption bands to transitions between given states can be made. therefore.500 4.500 2. 112 Knowledge of the A schematic diagram for the ultraviolet absorption spectrum of a carbonyl compound.500 3. The excited states of a carbonyl compound that are usually sus pected of being involved in electronic spectral transitions are those described in the previous section.000 X in A . to see ground state and each of these excited states can be expected to be induced by electromagnetic radiation.
x . One mechanism that allows forbidden transitions to occur . Thus. It follows that the entire integrand has the symmetry Ai if Excited and x lilted and y„. and for all molecules with no unpaired electrons. transitions to an excited triplet state cannot be entirely ruled out. transitions that are predicted to be forbidden on the basis of the symmetry treat ment illustrated above can occur to such an extent that weak absorption bands are observed. Tx Ty . like all observable properties. has the symmetry type ili. in get fact. The selection rule prohibiting changes in spin that is operative in small atoms and diatomic molecules appears to be fairly well obeyed in The results of such simple polyatomic molecules. 49. belongs to a symmetry type that contains n y or . is expected to have an intensity fa(v) dv which corresponds to an oscillator strength.) by the symmetry operations which take the molecule from one orientaThus.e. Similar statements can be made for the transitionmoment integral in the y and z directions. to which ^WitedMi^ground belongs is A h the value of the integral must be zero. is sufficient to decide whether or not the integrals may (One can. Again. just as in vibrational transitions. we see that they play an interesting and important role in fluorescence and phosphorescence. altered. i. a completely allowed transition. type. must not be changed be. The / values for such bands will be much less than unity. such as formaldehyde. While i. an analysis applied to CH 2 are shown in Table 113.e. an electronic transition from a totally symmetric ground state is allowed if the excited state y. . Even in simple molecules. use approximate wave functions for the orbitals shown diagrammatically in the previous section and perform the integrations in order to a transition is some idea. triplet states are relatively unimportant in absorpwill they lead to very weak absorption bands. or / value.. such as that from the ground state to the A x excited state of formaldehyde. On the other hand.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 273 symmetry of the wave functions be nonzero. however. near unity. unless the symmetry species tion to an indistinguishable one. the presence of a nucleus with a large nuclear charge results in a strong coupling of the spinning electron with the orbital motion of the electron and allows the relative spins of electrons to be. \i z . comparison of the observed band intensities with those expected of the transitionmoment integrals provides the most frequently used approach to the assignment of electronic states to absorp A As mentioned in Sec. not only of whether allowed or not. or T z. tion studies. or ^ eicited and ju s belong to the same symmetry ii . on the basis tion bands.. The ground state for the molecules considered here. but also of how intense the absorption will The integrals.
or dielectric environment. principally on the basis of empirical correlations. McConnell. state involved in the transition. J. [Cf. i.. the nonbonding electrons on the oxygen.700 1. Two other general methods for aiding in the assignment of absorption bands are used. of the absorbing is frequently useful and has some H.500 Upper state symmetry 0. for example. »Bi Allowed re—> a* re—» x* re » i* 'A. analyses given in Table 113 indicate that a determination of the direction of polarization would identify the upper is.orption band that accompanies solvation of the molecule.02 0. and the prohibitions based on that symmetry group are not strictly adhered to. the skeletal symmetry to which the electronic wave functions must conform is. The first of these is concerned with the direction of the transition moment. •A. strength) (A) Configuration change 1.800 3. molecule is increased. for the H 2 CO example. It is found.(?) Symmetry forbidden Symmetry forbidden Symmetry and spin forbidden . a suitable crystalline form can be found that permits such studies. The generally accepted assignment for the observed carbonyl bands is shown in Table 114..e. a hydrogen bond to the carbonyl group to This tie up. Considerable experimental difficulty however. to some extent. This correlation theoretical basis. for example. Phys. the polarization of the absorption. encountered in align ing the absorbing molecules so that studies with polarized radiation can be used to determine the direction of the transition moment. In some cases. The that do not have the same librium configuration.] One expects. Another assignment aid comes from the shift in the wavelength of the ab.0004 io« 7T —» IT* M. TABLE 114 Assignments The Ultraviolet Absorption Bands of Carbonyl Compounds and Suggested (Values of X and / are Typical Values for a Number of Compounds) Assignment "band max / (osc. that absorptions due to n —> t* ir lengths while those due to — > transitions are shifted to shorter wavex* transitions are shifted to longer wave lengths as the solvation. 20: 700 (1952).800 2. Chem. not strictly C 2 ».5 0.274 INTRODUCTION TO MOLECULAR SPECTROSCOPY involves vibrations of the molecule that lead to nuclear configurations symmetry as does the molecule in its equiAs a result.
as we will see. is be dealt with in the the process by which molecules in excited electronic states lose their excess energy and revert to groundstate molecules. are responsible for the absorption found in the nearultraviolet or even the visible region. as is 1 l . those in the 2p orbitals systems that project perpendicularly from the plane of the trigonal sp 2 orbitals forming a bonds. often the case. when transitions between two electronic states are considered. or a shorter wavelength band. will pertinent to a discussion of local and to the conjugated and aromatic systems that final section of this chapter. i.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 275 would lower their energy and lead to a greater energy change. for a transition that promotes one of these electrons. * two configurations are usually evident. Thus.e. the relative energies can arise. becoming standard. concern ourselves with the possible orbitals that can arise from the atomic 2pw orbitals.. one would write 'Bi—» A x When. when the transition is discussed in terms of the configuration change of an electron of the molecule. for example. and.800 A would be referred to as due to a B X <— *Ai transition. if so. both absorption and emission spectra are dealt with. designations according to this system avoid much of the confusion that However. always to write the higherenergy state first.) of the 1 1 3. To what symmetry types do the ground and first excited (ir)(jr*) states belong? Would a transition between these states be allowed. the band in the H 2 CO absorption spectrum at about 1. therefore. and one writes the initial electron orbit first and then an arrow pointing to the new orbit reached by the electron. (It should be mentioned that. and assign them symmetry types in the point group to which ethylene belongs. If the same transition were studied in emission. An interesting and important aspect is of electronic transitions that has not been considered but which ized absorbing groups. A very simple but surprisingly successful approach. will be introduced for linear . can involve fluorescence and phosphorescence emission of radiation. This process. known as the freeelectron model. in what direction would it be polarized? Exercise to ELECTRONIC TRANSITIONS OF CONJUGATED AND AROMATIC SYSTEMS In molecules containing conjugated double bonds and in aromatic it is now clear that the k electrons. since the symbols for electronic it is states do not immediately reveal the relative energies of the states. In studying such systems we can. Draw the tt and it* orbitals of ethylene.
i. 1 13. This treatment puts obvious emphasis on the i. This suggests that one might treat the molecular skeleton as a region of roughly uniform potential throughout which. orbitals will will be systems section. enters into trie delocalized x bonding. CaHjo. Conjugated Systems by the Freeelectron Mode! at longer known that conjugated systems have absorption bands and longer wavelengths as the number of conjugated double bonds gets greater and greater. general features. shown os with fp' hybrid orbitals. the molecular orbit als that it leads In the following to are spread out over the entire molecular skeleton.e. . The structure. aromatic be set up for benzene by means of the classical molecular orbital approach which combines the 2px atomic orbitals into appropriate molecular orbitals. — .e. that these electrons are delocalized and are not confined to a given atom or even to a given bond position. long held by chemists. subject to the wave functions that can be calculated from the Schrodinger equation. as well as the wavelength and intensity of the absorption It is well band for a given molecule. Each carbon atom forms three Each remaining 2p orbital. long conjugated systems are These often colored. the electrons are free to move. In fact. A treatment of the x electrons can be based on the idea. The carbon atoms can be considered to form a bonds with sp 1 trigonal A total of eight orbitals and x bonds with the remaining p orbitals. derealization of the electrons. and molecular considered. is FIG. 113 <r bands. basis of a very simple Let us consider the specific case of the molecule octatetraene. Outside the molecule the potential energy. on the simplest model. consists of a planar nuclear array. electrons will occupy the x orbital system. 113. the absorption band is in the visible region. Indicated by The bonding and structure of octatetraene. can be understood on the model for the x electrons of the molecule.. an atomic orbital here..276 INTRODUCTION TO MOLECULAR SPECTROSCOPY conjugated systems. as indicated in Fig.
ELECTRONIC SPECTRA
OF POLYATOMIC MOLECULES
277
assumed to be
infinitely high.
This freeelectron or partieieinabox
for octatetraene
is
model, although clearly ratheT crude, leads to quite satisfactory results.
The potentialenergy function The exact width
usually assumes that the
jt
is
shown
in Fig. 114.
set.
of this potential well
not unambiguously
electrons are free to
move about
is
half a
One bond
length beyond each end carbon atom.
Likewise, one usually assumes
that an endtoend measurement of the molecule
appropriate rather
than one that follows the zigzag of the chain.
With these assumptions,
and ignoring repulsion between the ?r electrons, one can deduce the allowed orbitals and their energies. The quantum mechanical problem has already been solved in Sec. 15. There it was found that the wave functions for the particleinabox
problem are
I = A
sin
—
n =
1, 2, 3,
.
.
.
(2)
and that the energies
of these solutions are
e
"
8ma*
(3)
The Pauli exclusion stipulation that no two electrons can have the same quantum numbers requires that no more than two electrons, one with spin +i and one with spin — ^, be assigned to a wave function with a
FIG. 114
Oecuponcy of molecular orb
transition
i
tali
by
to
the
electrons of octatetraene In the ground state.
The
Hie
arrow indicates the
that
leads
observed absorption band.
278
INTRODUCTION TO MOLECULAR SPECTROSCOPY
n.
given value of
The occupation
of the available
it
molecular orbitals
that which
for octatetraene can then be illustrated as in Fig. 114.
The
longest wavelength transition that
is
expected
is
takes one of the
arrow in Fig.
n = 4 electrons to the n = 5 114. The energy involved in
(52
orbital as indicated
this transition is given
by the by
46
_
£4
=
.4>)
8
^
would cause
(4)
The frequency
with a
v
of radiation that
this transition is calculated,
=
=
9.5 A, to be
27,000
cm
1
(5)
This calculated result
is
to be
compared with the absorption band
maximum found
at 33,100
cm
1
.
The agreement
is
surprisingly good.
that, as discussed in
A
further advantage of this simple model
is
Sec. 46, the intensity with
which a sample absorbs radiation can be
calculated.
The
prediction, for octatetraene, that the integrated absorp
tion coefficient be, according to Eq. (58) in Chap. 4,
32iVe
!!
aV
10 19 sec 1
3ir/icl,000
=
1.9
X
cm 1 mole"
1
liter
(6)
can be compared with the experimental value of approximately
fa(v) dv
=
1.1
X
10 19 sec 1
cm mole"
1
1
liter
(7)
Again the agreement
is
satisfactory.
A
number
of elaborations of this simple
method
for calculating the
electronic states of conjugated systems
have been introduced.
The
chief
merit of this approach, which clearly oversimplifies the potential experi
enced by the
ir
electrons,
is,
however,
its simplicity.
Exercise 114. Note the similarity between the orbitals of Exercise 111 and the n = 1 and n = 2 particleinabox wave functions that would be drawn for the ground and first excited orbitals for ethylene using the
particleinabox model.
Exercise 115. What are the calculated and observed / values for the longest wavelength irelectron transition of octatetraene? Exercise 116. Lycopene is a linearly conjugated hydrocarbon conIt is responsible sisting of all double bonds alternating with single bonds.
for the red color of tomatoes.
Calculate the wavelength expected for
ELECTRONIC SPECTRA OF POIYATOMIC MOLECULES
279
—* v* transition and compare with the observed band which occurs at X = 4,700 A.
the w
114. Electronic States
maximum
and Transitions
in
Aromatic Systems
description of the ground electronic state of aromatic goal that occupied the attention of many early organic was systems a Following the advent of quantum organic chemists. physical and
An adequate
mechanics, attention could be extended to the excited electronic states
and the spectral transitions that occur between these
states,
A number
of approaches to the description of the excited 7relectron states of aromatic systems have been used. An introduction to the electronic spectra of aromatic systems can, perhaps, best be given in
terms of a simple
classification
molecular orbital treatment, based on 2pr atomic
the example benzene.
orbitals, applied to
The importance
of the
symmetry
of the molecular orbitals and the electronic states will be brought out,
and
it
will
be shown that some assignments of observed absorption bands
to the states involved in the transitions can be made.
The treatment
can be extended to other aromatic systems, but this will not be done here. As in the previous section, attention will be centered on the 2prr
orbitals,
shown
for
benzene
in Fig. 115.
It is first
necessary to see
how
these atomic orbitals can be combined to provide suitable molecular
orbitals.
The guiding
principle here
is
that an electronic wave function,
must conform to the That is, by an argument analogous to that used in discussing the symmetry of normal coordinates in Sec. 64 and illustrated for locabzed orbitals in Sec. 111, each molecular orbital must transform, under the various symmetry operations, according to one
or molecular orbital, of the benzene molecule
symmetry
of the molecule.
of the irreducible representations of the group.
Let us
now
see
if
such
molecular orbitals can be obtained
2prr
by
suitable combinations of the six
atomic orbitals.
FIG. 115
The 2p atomic orbitals of
the carbon atoms of benzene.
280
INTRODUCTION TO MOLECULAR SPECTROSCOPY
TABLE 115
The
D
6
Character Table
D,
A,
E
I I
2C e
2C1
C
3
3C
2
3c;
I 1
1
i
A,
B,
1
I
1 X
—1
1
1
B2
Bi E,
2 2
1 1 2
2
i i
i
T,
1
—1
—1
T*,T y
The
point group to which the nuclear skeleton of the benzene mole6fi
.
is D The treatment can be somewhat simplified if the symmetry about the plane of the molecule is, for the time being, ignored. It is enough to treat the wave function above the horizontal plane of the molecule or below that plane. Later the fact that all px orbitals are antisymmetric with respect to this plane can be added in. With this
cule belongs
problem can be analyzed in terms of the simpler which the character table is given in Table 115 and the symmetry elements are illustrated in Fig. 116.
simplification, the
point group
D
8,
for
The
six
2px atomic
i.e.,
orbitals
form a basis
are
for a reducible representa
tion of the group,
the transformation matrices that are obtained applied to these atomic
when the various symmetry operations Thus
orbitals will lead to representation matrices that can later be reduced.
to obtain the reducible representation, one investigates the oper
ations on the individual s orbitals
and
writes the transformation equations
FIG.
1 1
6
The symmetry elements of the
Da
point group.
Note
that, unlike the
is
ease for
the Dbi point group, symmetry with regard to the plane of the molecule
Cfin
not considered.
C gT Cj
ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES
281
as
p' /l
p"
E
=(010000
0^
character for
E =
6
^000100^
10
character for
d
—
and
so forth.
In this
way one
finds that the character of the reducible
is
representation corresponding to the six 2pir orbitals
tl
2C6
2C 6
C7 6
0C2
0C2
As
it.
in the
normal vibration problem,
this representation
can be
analyzed in terms of the irreducible representations that correspond to
Thus, application of Eq. (51) in Chap. 8 shows, as can readily be
checked, that the reducible representation contains representations of the types
Ah B E
x,
x,
and
E
l
2 , i.e.,
= A
l
+B +E +E
l
t
(8)
We
type
Ei,
see
from
this that the
atomic orbitals should be combined to give
one pair according to the degenerate class
.
molecular orbitals, one of which transforms according to the symmetry
A
1,
one according to
Bh
and another pair according to E 2 We must now see what combinations of the atomic orbitals have these transformation properties.
A
systematic group theoretical procedure
is
available for obtaining
Rather than develop this, we will present the results and merely check that they have the correct transformation properties. (The two nondegenerate orbitals can, in fact, be constructed
the suitable combinations.
by
inspection.)
The molecular
orbitals that
have the correct symmetry
properties are given in Table 116.
(The degenerate pairs can be written
282
INTRODUCTION TO MOLECULAR SPECTROSCOPY
TABLE 116
Molecular Orbitals for the
tt
Electrons of
Benzene
Symmetry
type
Molecular orbital*
fa^
iAi>i
Al
Bi Ei El
*
= = <An = ^e, = H = ^e =
<ls
2
+ <Pb + <Pc + <Pd + <Pb + <Pf <Pa — <Pb + <Pc — .<Pd + <Pe — <Pf 2^>x + <Pb — <Pc — 2<p D — (ps + <Pf <PA + 2<PB + <PC — f>D — %PB — <PF 2<PA — <PB — <PC + 2<fD — <PE — <PF + +<PD — 2<PE + <pF <PA —
<Pa
2(fiB
<fiC
Lowercase letters are used as subscripts to identify the molecular orbitals, just as lowercase letters were used for individual electron orbitals in atoms and diatomic Capital letters are reserved for the symmetry type and the net electronic molecules.
state of
an atom or molecule.
and it may be more revealing to consider the sum and the difference of the two functions given in Table 116.) The form of the molecular orbitals can be seen by drawing the wave One should recognize that a corresponding functions as in Fig. 117. function, but with opposite signs, will project out on the other side of the benzene ring. Also one should keep in mind that each molecular orbital can accommodate two electrons.
in various ways,
The order
of the energies of these individual molecular orbitals
can be recognized by recalling that, as can readily be seen from the
particleinabox problem, the
the higher
is its
energy.
a,i
more nodes there The wave functions
are in a
wave
function,
of Fig. 117
ordered so that the
function at the bottom of
have been the diagram has the
lowest energy and the &i function at the top has the highest energy.
Calculations of the energies of these orbitals can be made, and the
simplest calculation leads to the spacing
shown
in Fig.
118.
energy spacings are usually expressed in terms of a quantity
represents an integral that
is difficult
These which
to evaluate.
The symmetry
are also
designations for the orbitals on the basis of the
is
De» point group, for which the character table
given in Table 117,
shown in Figs. 117 and 118. The fact that the pr wave functions, or any linear combination of them, have opposite signs on
opposite sides of the plane of the molecule allows, after inspection of the wavefunction diagrams of Fig. 117, these designations to be given to the
orbitals previously labeled a\,
ei,
e 2,
and
&i in
accordance with the De
group.
Exercise 117.
117, verify the
classified
By
D
6»
inspection of the molecular orbital diagrams of Fig. classifications of a lg , e lg , e iu , and b ig for the orbitals
on the basis of
D
6
symmetry
as
oi, ex, e 2 ,
and
fei,
respectively.
tions are given The relative energies of the individual Symmetry designaon the basis of the De the. 118 molecular orbitals of benzene. FIG. Da) w 1 z» FIG. in parentheses. negative outlines gray light of outlines positive parts $. . in parentheses. <ai«) Dark gray Schematic representation of the molecular orbitols for bemene. symmetry type is [The parts. 117 given in terms of D 6 and. point group 283 and. on the basis of D$s group.bi(bi„) ei(e iu ) Cl("«l B J a.
furthermore. The direct product of the char. acters of the Ei s and 2C« E iu representations is 2C\ 3C 2 3C' 3 oh 3<r„ 3<r d 2S e 2S 3 XJS la * XB«. situation encountered with nondegenerate representaa reducible representation is obtained. The symmetries of the electronic states of these configurations can be derived from the products of the characters for the symmetry species of the orbitals occupied by each electron as described for localized groups in Sec.e. i i _1 I —1 2 2 2 —I _1 _1 „1 2 2 2 2 1 —1 —i —i 2 2 T T —1 —1 2 2 . 1 4 4 01 Now. unlike the tions. that are involved in spectroscopic transitions.i Alg TheDe* Character Table 6 E 2C 1 1 2C\ CI 3C 2 3C 2 (Th &T V 3tTd 2S e 2£ 3 i 1 A iu At. of orbitals of given resulting electrons. from the occupancy The electronic configurations that need to be considered (remembering that two electrons can occupy a given orbital and four a doubledegenerate orbital) are that of the ground state and that of the excited states which result when one electron These are indicated in Fig. 119. It is necessary to analyze TABL Ell7 D. Aiu Bi„ Biu 1 1 1 1 1 1 —1 —1 _1 _1 _J j 1 1 1 Bt a Bi„ —1 —1 —1 —1 —1 —1 1 Ela E lu Eig Eju 2 2 2 2 1 1 1 1 2 —1 —I _1 —I _1 J —I I J i i i T.284 INTRODUCTION TO MOLECULAR SPECTROSCOPY Now it is necessary to consider the assignment of the six 2prr electrons to the available orbitals in ways that correspond to the states. to determine the symmetry of each of i. since the ground state involves pairs of electrons occupying the same orbitals. ground and excited. It will be important. the total symmetry of this electronic state must be A lg The symmetries of the excited states can be obtained by analyzing the direct product of the representations of the orbitals e lg and e 2u occupied by the lone electrons. 111. the symmetry symmetry by the six w these electronic states.. Thus. is promoted to an e 2 „ orbital.
P*5*= Groundstate configuration Excitedstate configuration . 8. + Xtf„ two (9) The electronic states of benzene corresponding to the lowest excited electronic configuration are. since they involve occupancy of the same set of individual orbitals. 119 Distribution of the 7r electrons of benzene in the ground and first excited elec tron configurations. therefore. This can be done by application this in of Eq. the relative energies of these states and the ground state. therefore. and as suggested by Hund's lie rule cited previously. as can be readily checked.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 285 terms of its component irreducible representations in order to determine the symmetries of the states arising from singleelectron occupancy of the e ig and e 2„ orbitals. two electrons in each of these excited states need not have their spins paired. that /CBtg A. and the result is obtained. in a purely theoretical way. and triplet states. to have both singlet It is possible. Furthermore. correspond to the same It is a matter of considerable difficulty to total electronic energy. calculations indicated that the triplet states lower than the corresponding singlet states. expected to consist of states with symmetries Bm and B iu and a degenerate pair of states with sym metry E\u . To a first approximation these four states. deduce.E 2t Xb. (52) in Chap.„ + xs*. Many attempts have been made to calculate the energy pattern for the excited states of the benzene FIG.
however. metries ij.J. S. would be allowed. 1110. . 1953. transitions to the states with sym.. changes in multiand one expects only singletsinglet transitions to lead to strong absorption bands." PrenticeHall. and just as for localized groups.286 INTRODUCTION TO MOLECULAR SPECTROSCOPY it has not been possible to make.500 A. Englewood Cliffs.500 3. around 1. but so far basis. 1 Rydberg series 31.. about the states involved in the various absorption bands.000 3. on a theoretical a certain assignment of energies to the various electronic states.'A. The absorption spectrum tion bands than the single l benzene shows. make what deductions one can from the observed spectrum.000 X in 2. One must. The absorption spectrum of benzene in the ultraviolet spectral region. 1110 K.) 5r \B ta or %. Furthermore. (From "Quantum Chemistry. more absorpis most intense band that assigned to the E lu *— A 1 lg transition. which upper states .500 A . Thus the jnost intense band in the absorption spectrum of of 1 benzene should be assigned to the completely allowed transition to the E Xu state. Some assignments of the remaining absorption bands can be made. Inc. molecule. and E lu plicity are expected to be forbidden. is attributed to a FIG. for a D 6h molecule. N. belong to a symmetry type that contains Tx T or Tz One sees from y Table 117 that. one needs to ask . It is first necessary to consider which transitions will be allowed. shown in Fig. The very short wavelength band.. therefore. Pitzer.500 2.
67: 2127 (1945)]. 1 The This band near 2.. aromatic systems can be approached have been discussed by R. Jones [J. . . transition. Am.l A it highly 3. \H H \H in a similar manner to that in which benzene was treated above. On the basis of the assignments of the absorption bands draw an energydiagram showing. the effect of solvents.100 A is due to either the transition lB^i—iAig or 1 E ig <^ 1 A u These forbidden transitions could occur as a result of "borrowing" intensity from the nearby intense band at . Similar attempts can be made to assign the ultraviolet absorption bands of other aromatic systems to transitions of certain symmetries. included in the spectrum of Fig. 1.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 287 "Rydberg" type transition one of the 2pr electrons is in which the principal quantum number of increased. It has the in included transitions have not been been suggested that the band near 2. N. It should be apparent from these comments that. by means of horizontal lines. ing groups. prohibited on the basis of the selection rule derived from the electronic states. Exercise 1 1 9. Treat the molecule cyclobutadiene of Fig. 1110. can again be allowed as a result of coupling with a The very weak absorption at about vibration of 'suitable symmetry. although the electronic it states of benzene can be classified as to symmetry. As in between electronic states studies of molecules with localized absorb by orientedmolecule and the Some of the ways in which more complicated effect of substituents. Exercise level 1 1 8. 1110 to electronic states of the benzene molecule.300 A is attributed to the i . treatment given here. Chem. forbidden transition *B u *.800 A or by coupling with a vibration of appropriate symmetry.500 A is attributed to the transition B iu <— A ig 1 . Attach electron volt and kcal per mole energy scales to the diagram. Soc. the energies of the various electronic states of the molecule. aids to the assignment are provided spectra studied with polarized radiation. as in the Rydberg series usually Such highenergy discussed in connection with the hydrogen atom. remains a task of considerable difficulty definitely to identify each absorption band with a The assignments suggested here are transition to a particular state. H \c— / c— II II H H  \c=c/ / c=c I I H H /.
(c) Arrange these orbitals in order of their energy. to these systems. and extended. (d) Assign the * electrons to 'these orbitals in accordance with the lowest and the first excited electronic configurations.) These chargetransfer The treatment of the d orbitals that has been used and that allows the absorption spectrum of a coordination compound to be related to the properties of the metal ion and the ligands is known. to some extent. in its simplest form.288 INTRODUCTION TO MOLECULAR SPECTROSCOPY form. that occur are attributed to what are called chargetransfer transitions. transferred from the ligands interpreted. when a more detailed treatment the of . colorful ions gives an intense blue color. by inspecting the way in which the orbitals must trans What transitions would be allowed? between the ground and these excited states ELECTRONIC TRANSITIONS OF COORDINATION COMPOUNDS The absorption spectra of compounds or ions containing transition metals have long been used to characterize these species empirically. transitions will not be dealt with here. In recent years an earlier theory developed by Bethe and by Van Vleck has been applied. the familiar. (Other absorptions. as the crystalfield theory and. In these transitions electrons are. usually more intense and farther out in the ultraviolet. (b) Draw. all bands in the visible. for the most to the metal or from the metal to the ligands. for example. these molecular orbitals. coordinate to the metal species. : NH H 3. called ligands (such as ion. What are the symmetry species and multiplicities of the electronic states that arise from these configurations? (e) (a) Deduce the number of molecular orbitals of the various symJ metry types. and it has ime) 2 which forms a red crystal. comprise the most extensively studied type of transition metal In this class one has. : CI : ). Almost the transition metals have absorption coordination and Ni (dimethylglyoxcompounds of become clear that the principal absorption of these compounds can be part. as these familiar examples do. [Cu(NH 3 ) 4 ]++. or in the nearinfrared or nearultraviolet spectral regions. in terms of transitions involving the electrons in the incomplete outer d shell of the metal ion. which . 2 : . Coordination compounds. in which a number of electronrich groups.
i.e. 1 i 6S 4 1 1 8S e 3o* 6avj 1 1 A lu A ig Azu 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 —1 —1 —1 1 1 1 1 1 1 2 1 2 E. on the repulsion which those electronrich groups (The coordination compound exert on the d electrons of the metal ion. some of the d To analyze these changes we must proceed. the relative energies of these orbitals can be investigated and the available electrons can be assigned to these orbitals. the theory supposes that the ligands impose on the central metal ion an electric field whose symmetry depends on the number and arrangement of the ligands and whose strength depends on the electrical nature of the ligands and.. It will such as exists in the free ion the i. When this is . In its simplest form.) Coordination Compounds the effect of the ligand field on the be recalled that in a spherical potential 115. Le.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES is 289 participation of the ligands made. adequately illustrated by treating only the imporAn illustration of the geometry such complexes is shown group in Fig. more particularly. E 1 6C 2 1 1 6C 4 1 1 3c.. with the arrangement of the ligands. Electron Orbitals of It is first necessary to investigate d orbitals of the metal ion. The procedure of is tant example of octahedral complexes. and some of the 1 1 8 O h Character Tabl e 8C„ 1 o» Ai. d orbitals have identical energies. 2 2 3 3 3 3 Eu Tu Ti u —1 —1 i l l 1 2 2 1 2 3 1 Tn Tzu l 1 1 1 1 1 1 2 1 1 1 1 1 J 3 3 1 1 1 X) 1 yi i z 1 1 3 1 .done.e. The character table for this TABLE The given as Table 118. as the Ugandfield theory. 1111. such ions have the symmetry is of the point group O*. to find orbitals that have symmetries compatible with that of the system. Since all six ligands are equivalent. The effect of the ligand field is to break this degeneracy. to change the energies of orbitals relative to other d orbitals.. which is all that will be introduced here. as in the benzene analysis of the previous section. by the attractive interaction primarily held together being pictured as is of the charge of the metal ion with the charges or dipoles of the ligands. they constitute a degenerate five set.
benzene were combined to give this molecular orbitals consistent with the skeletal symmetry of the benzene Again the details of the way in which is done five will be passed over. *T NH. For considerations of enough to concern ourselves with angular factors.. arise in this way were shown in terms of J = 2 in Table 51. All six ligonds are equivalent. Treatments similar to that given here for octahedral complexes can. and only the results will be given. ^ c 4 . O. of course. They provide. 1 111 (a) The geometry of an octahedral complex. I Five functions occur since. and the complex belongs to the symmetry point group symmetry elements of the Oj. of course. (b) Some representative NH. It is found that the reducible representation that the d orbital wave functions lead to under the operations of the octahedral point FIG.c. define only the angular factors of the orbitals. it is to be located about the central ion. ± 1. basic functions from which orbitals can be constructed that have symmetries compatible with the octahedral arrangement of the ligands that are now imagined symmetry. The d orbitals of an I atom or ion are those corresponding to the solution functions with = 2 for an electron of a hydrogenlike atom. (a) to Cj and C 4 . point group. A radial factor that depends on the principal to determine the quantum number must also be introduced wave function completely.) The dorbital functions of Table 51 are appropriate to a system with spherical symmetry. ±2. furthermore. the magnetic quantum number m can have the values 0. The five d orbital wave functions that (These..290 INTRODUCTION TO MOLECULAR SPECTROSCOPY symmetry elements are shown in Fig. The procedure in is similar to that in which the separate 2yw atomic orbitals molecule. 1111. for = 2. also be given for complexes with other symmetries.
. . unique and that the One must. These are indicated diagram matically. . . the symbols tig and and <r. to describe individual electron e g will be used to describe orbits that transform FIG.") d <*i. however. so that Eg two orbitals result. along with the d zv d a dg. jt. corresponding to the notation d^gi axis is and dn. notaDiagrams can be drawn to represent tion usually used. It is customary. orbits. the two orbitals that transform according to the E g representation by drawing three pictures and imagining linear combinations of these to be made the als. orbitals by combining two in Fig. . . . in Fig. the of use (To be consistent with the cartesian axis. . to illustrate of these diagrams to yield the two orbit shown 1112.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 291 can be recognized as containing the doubly degenerate irreA set ducible representation E g and the triply degenerate one 2V latter representathe according to transform of three orbital s that group O h tion can be directly constructed from the original wave functions. 3. . such as s. d. 1112. keep in mind that no orbitals that project along orbitals with E g symmetry consist of a pair of lowercase symbols. 1112 Diagrammatic representations of the angular factors of the d orbitals. however. 3£ 2.¥ z> d („2_ z Z) . p.
therefore. The extent of splitting is customarily discussed either in A or lODq. 1113. Both notations are included . that a metalion electron in an e g orbital will since such orbitals point in the direction of the ligand electrons. be immediately recognized. 1112. .) We now consider the and e„ orbitals that must be occupied by the ion forms a outer d electrons of a transition metal ion when the metal sets of octahedral complex. It follows effect along these experience. (This step corresponds. The amount of splitting must be deduced empirically. to designate the energy splitting between the and is t 2g orbital For bivalent ions one generally finds A = 10 Dq to be of the order of 10.292 INTRODUCTION TO MOLECULAR SPECTROSCOPY according to the symmetry types relative energies of the l 2g T 2g and E g . lie Thus the two e g orbitals can be expected. between the electron of the mentioned that it has become customary to use either the quantity e„ A or lODq sets. 118. might be mentioned. to at a higher energy than do the sets of orbitals orbitals. Finally it should be spectroscopic studies often allow this to be done. while for trivalent ions this splitting increased to FIG. more repulsion than one in a fa g orbital. to the deduction that the energies of the molecular orbitals of benzene are arranged as in Fig. The energy difference between the two d orbitals can. in a it qualitative way.000 cm 1 . produce an electronrepulsing axes. and as we will see. depends on the details of the interaction metal ion and those of the ligands.) One supposed that the six ligands occupy sites along the cartesian axes of Fig. here. The exact amount of splitting of the It cannot be satisfactorily calculated. t 2g as indicated in Fig. These ligands present electronfilled lonepair orbitals in the direction of the metal ion and. 1113 The removal of d orbital degeneracy by an octahedral terms of field.
therefore. and thus 2S +1 = is 2. Now we must consider the electronic states that arise when the d electrons of the transition metal ion occupy the available orbitals.) 1 two states are indicated as and (e. For no crystal 2 therefore. 2 ^» <— 2 ?V .. indicates the spin and writes the two states as 2 T 2„ and One sometimes indicates the effect of the crystal field for this one d electron example by a diagram like that of Fig. be 2 and the net spin S must be field.400 cm" 1 Alterna. The quantity A.) 1 . configurations corresponding to these The electronic (fo. the is degenerate d orbitals. For example. This observation allows the vertical dashed line to be drawn in Fig. The effect of the crystal field on the states of such an ion will.400 cm. 1114 at a position where the splitting is 20. in which the single d electron split to give e g orbitals. A can be made between this quantity and chemical properties of both the transition metal ions and the ligands that have been studied. 116. or ionic. Furthermore. one d electron species must be expected to give a 2 D state. 1114. In a sense. must. a set of lowerenergy Ug orbitals and a set of higherenergy Now. one *E„. the electronic state of the Ti*+ ion a D state. or 20. however. Electronic States of Coordination Compounds d electron. be expected to be characteristic of both the metal ion and the ligands. the electronic states are letter designations 2 similarly described and are labeled with the capital T ig and E„.000 cm1 . electron states indicated as Ti 3+ . 103. The electronic states of an ion with only one outer are in Sec.900 A. of interesting correlations number We splits have now obtained the symmetrycorrect orbitals for a metal ion in an octahedral complex and have recognized that the ligand field the original fivefold degeneracy. depending on which set of orbitals the electron is assigned to. the aquated Ti ! + ion [Ti(H 2 0) 6 ] s+ has an absorption band at 4. i. When ligands are now placed in octahedral positions about such an ion.e. or lODg. be discussed first. which will be more evident later. there are two different electronic states. In view of the survey of atomic. such most easily treated. Since there is only one electron.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 293 around 20. the net orbital angular momentum quantum number L must £. Various ligands will have crystal fields of various strengths and will correspond to positions along the abscissa. this is a correlation diagram between the situation for the spherical field of the ion at the left and the strong crystal field at the right. accommodated. as for the original D atomic state.1 which can be identified with the transition .
give L values of 4.294 INTRODUCTION TO MOLECULAR SPECTROSCOPY tively we can say that for [Ti(H 2 0)e] 3+ the value 20. i. so that the energylevel pattern for any strength of crystal two I = 2 field be combined to 103.000 10. P. F. 3. 0. 1. in Sec. 1114. like that of complexes. 2. or lODq. showing the effect of increasing crystal field The posiarbitrarily assigned the value zero for the ground state of the undisturbed ion. [Ti<OH 2 ) 6 P State 'E g configuration (e„) State 'T^ 1 configuration (t 2g ) 5.e. For the zero crystalfield limit. The figuration.000 15.000 25. G.000 20. is example since it has two outer d electrons and We wish to draw a diagram. discussed can. d conis energy scale effect..e. An illustration of how 1 The energylevel diagram for a metal ion with one d electron. FIG..400 of the crystalfield splitting factor A. there are two d electrons. the assignment of the two electrons to the available d orbitals must be made in can be picked off. these features were not mentioned in the earlier discussion. the accordance with the Pauli exclusion principle. octahedral forms The Now let us consider ion V 3+ serves as an the cmsituation when 1 . 1114 tion of [Ti(OHj) 6] 3+ is located and the value A determined by the frequency of 20. Fig.400 cm 1 of the observed absorption band. or i. Now. S D. as electronic contributions and or states.000 A (cm Zero crystal field ) Large crystal field .
to a strong crystal The three possible electronic configurations (t 2g ) 2 . D.. One finds by this means that the d 2 configuration can only lead to states P. respond to Now. "Atomic Spectra and Atomic Structure" (pp. illustrated in Analysis of the 2 (<2») configura Table 119. and these results for all nine configurations of an incomplete d shell subject to octahedral symmetry TABLE 119 Illustration of the Determination of the Electronic States Arising from the Configuration (hg)* Oh E Ct C\ Se Formation of direct product XT. that must be considered are. to find what electronic states these configurations corwe must. + + XT. drawn for the V 3+ ion... F. Similarly.. X XT. The energy 3 Z 1 1 pattern of these states 1115 can be is not easily predicted but can be determined from In this analyses of atomic spectra. S. Herzberg. 1 1 1 1 1 1 1 XA U + XB.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 295 these restrictions operate to allow only certain spin arrangements of the two electrons for certain values of ing to the detailed discussion given L can be followed through accord by G. XT. = XT U 9 1 1 1 9 1 1 1 . and (e„) 2 . Dover Publications. as in the previous treatments of localized groups direct product of the representations and conjugated systems. or *G. the electronic states that arise from other electronic configurations can be found from the appropriate direct products of irreducible representations. New York. and the results of these analyses are shown in Table 1110. form the tion is of each electron in a given configuration. way the pattern at the origin of Fig. would be expected for a d 2 ion subjected Let us now move over to the far righthand side of the diagram and consider the states that field.. 1944) and will not be repeated here. in order of increasing energy. In a similar way the states for the other two configurations can be found. 130135. (k g ) 1 (e g ) 1 . 3 9 1 1 1 1 1 1 3 1 1 1 1 1 1 9 Analysis in terms of characters of irreducible representations XAjg 1 1 1 1 1 1 1 1 1 1 XB. 2 3 3 1 1 1 1 2 2 3 3 1 1 1 2 XT la XT.
. . 2 4 £„ fol'fe)'. («. *T 1( «T„ . >£„ T.. (<i. .. . 'Ai.)» (*•. one assignments included in the electronic states listed in Table 1110. as in the 2 (£2„) and (e„) 2 configurations. .».„ IT.. ...„..„. >A.„.. 2 2 !T„„ 2T 2o r 2„ are given in Table 110. 2»£„.„. . 4T. T„ Ti. •r. . 'Aip. 3 T 2o . (We will see later that the configurations 10.. 1( . two electrons have the would arrive at the spin same orbital description..296 INTRODUCTION TO MOLECULAR SPECTROSCOPY Electronic States Arising TABLE 1110 Symmetry from an Electron System Subject to Octahedral Electronic configuration Electronic states Free ion Ion subject to octahedral symmetry *£„ ! d>. as listed in Table 1111... *£„.) 1 «?." and it is states for d n systems are similar to those for d systems. .„ 2l.. 2 2 r l0 2«r l( .„. 3>T ig 2'T 2e 2iA. d« (e„) 4 (W(«„) 3 (WW (WW (W d« TU .„ 2^„. necessary to obey Pauli's exclusion principle in assigning electron spins when. .) r s «. *A 2 „. 3'£ t T.) Again.. »j?„ >r 2 „ (WW (WW (WW (WW (W 1 T 2 r 28 l0 . 2*T 2 „ 2'A lg . Exercise 1110. »£„. . *A. >T 2I *E g . T 4»r. . 3T. A 2o 2^ 2'7 M.„ ia.„. Tip. 2T s 1( „ 2T 2„ »r. d» <«. 'A. 2T. 2»r lln 2»r 8 „' 'A.. M„. 3 A l5 »A 2o 23£„. (*i»)* <A 2 „. *T U *r lln .„. For large crystal mined by the number the energy of an electronic state t ig is deter of electrons in the lowenergy orbitals and the . Verify that the (e„) 1 2 1 configuration leads to the states configuration leads to two states A ig Ai g and Eg and each of the type that the (ti g ) (e g ) T ig and T 2g fields. >£„. 3 *T i0 *A ln M. >T. ir... •Ai. ) .)' T 2„ >£.». »A. *T lg *T lg 2 . VE B. By a procedure similar to that which allows spins to be assigned to the atomic states. «r. 2T„ d*. d\ d 8 (U* d\d? (e») 3 WW («») 2 ..„.. T„.„ ! A 2o .
on this basis. On the left is the energylevel pattern for the free ion. with the states of the d* configuration. 1115 can be drawn. The energylevel pattern for such situations can be drawn FIG. 1115 A schematic correlation diagram lines for the states of a is d2 ion. the pattern for very strong crystal multiplicities The correlation are drawn with regard to the of the levels and the correlations of Table 1111. The righthand limit of Fig. are important. which produces the different states of the free atom.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 297 number in the highenergy e„ orbitals. For intermediate field strengths the tendency for the electrons to occupy the lowenergy t ig orbitals as well as their tendency to avoid one another. Free ion Strong crystal field \3™ '» (No repulsions between the two d electrons) . on the right fields.
schematic energylevel diagram that would be drawn on this basis for Ni++ is shown in Fig. consequence. Furthermore like those drawn for the d example of V . Phys. When this is done. the crystalfield configurations will be the same except that we will be As a dealing. It follows that.e. is in many respects identical to that for an atom lacking 8 two d electrons from a completed d shell. such as the d configuration shown in Fig. 1115 is reversed. the correlation of atomic states and the octahedral crystalfield descriptions are determined by the atomicstate symbol. 1115.e. for Which Spherical Symmetry Exists. one must also preserve the spin.298 INTRODUCTION TO MOLECULAR SPECTROSCOPY by joining up the appropriate in states of the right and left sides of Fig. P. In practice one does not encounter ligands that have a sufficiently strong crystal field completely to overshadow the electronrepulsion factors that operate in the atom to produce the different atomic states TABLE 1111 Correlation of Atomic States. Thus s Ni++.. •It is now appropriate to point out that the often used spectroscopic trick of recognizing that the treatment for 2 an atom with two d electrons. as it were. with the States Occurring in a Field of Octahedral Symmetry Atomic state symbol Octahedral symmetry states S Alg P D F G Tu Ea + T H A 2a + Ti. 1114. The appropriate correlations are given in Table 1111. i. 1114 can be drawn. 57: 3.. of the electronic states. the ordering of the configurations with regard to energy The at the right of a diagram such as that of Fig. with positive holes in the completed d shell. Bethe [Ann. the correlation lines shown in Fig. for any atom. 133 (1929)] by means of a grouptheoretical derivation of the relation between the atomic states of the spherically symmetric problem and those states described by the crystalfield notation appropriate to the potential field of octahedral symmetry. a d system. i. and so forth. 2 In a particular case.e. The derivation shows that it is the net angular momentum of the atomic state that determines the correlation. 1116. a d configuration. + T 2g A le +Ea + T lg + T lg . 1115. i. or multiplicity. The way which such correlations must be made was first shown by H. which has the configuration d has a set of atomic states arranged 2 3+ in Fig. D. S..
One usually. Free ion Strong crystal field (No repulsions between the d electrons) . Ni ++ in Fig. focusses one's attention on the lefthand part of the diagrams that have just been discussed.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 299 for a given number of d electrons. 1117. and to form somewhat modified from one another. The lefthand part for the is coordination chemistry. therefore. shown taken of the fact that like which is often studied in Here account is also quantummechanical states tend to "mix" ion. The energy FIG. 1116 states whose energies remain different levels of like states are said to repel The schematic correlation diagram between the free ion and the strong octahedral field complex for d8 ions.
1117.. "Methuen & Co. Arrange according to energy. in a field of octahedral The energies of the important states with the same multiplicity as the ground lines. £.000 ««. shown as heavy The position of [NilHjOle]" (From 1. (hgYieg) 1 {t2.Metal Chemistry. "An Introduction to Transition.) (Ni(H 2 0) 6 ] « 2! /<V4 (A = 8. T2 S and the two . such as Ni ++. to the higherOne can immediately carry over the rule that forbids FIG. 1117 symmetry. E energylevel lines. this effect the two 1 is 1 noticed in the curvatures of g T lg .) 20. rules governing transitions it remains for us to consider the selection energy states. such as that of Fig. possible configurations of a d 1 ion in an octahedral complex. London. 1960. 1116. fields of . Orge/. Before the predicted energylevel pattern of a Ni ++ complex. state are The energylevel diagram for a d 8 ion. as in Exercise 1111. 3 A 2g state.000 10. 1 "" 1 is located along the abscissa to fit the observed spectrum. Exercise 1111. can be used as a basis for the interpretation of the observed spectral transitions.000 . What are the differences in the limit of strong crystal . Ltd.500 cm" ') 20. the two 3 In Fig..g) l {e g y.000 E u c i  10.300 INTRODUCTION TO MOLECULAR SPECTROSCOPY one another.000 <v 5.000 15. the energies of the configurations {h Y. and 3 Illustrate the answer by an energy diagram and show the energy (e g ) scale in terms of both the parameter A and the parameter 10D 9 Exercise 1112. the . from the ground.
The spectrum of the aquated ion [Ni(H 2 0)«]++ is shown in Fig. well known in atomic spectra and referred to as the Laporte This rule. 45. all symmetric and. For a coordination compound this rule is broken down. is different from the previous examples field. and since the dipolemoment term itself is antisymmetric. as d orbitals are symmetric with respect to their origin. is based on the symmetry result that. in this case. therefore. and this configuration has an energy that independent of the value . Fig. because some of the vibrations of the ligands with respect to the metal ion destroy the symmetry i.e. the integral would necessarily be zero. 1117 if one assumes a value of A = 8. and the positions of these bands match up rather well with the spectrum. much more intense A<S = bands occur. that dd transitions. consideration of the transitionmoment integral leads to the conof the other triplet states. transitions between energy tion levels such as those of Fig. expect the important spectral bands to be interpretable on the basis 'T ig state and the two 3 Ti g states. be mentioned that the A<S = selection rule can also be violated.500 cm 1 for the [Ni(H 20)6]++ ion.) the energylevel pattern for a given octahedrally coordinated metal ion can be used as a base for the interpretation of the observed Now The example of Ni++ is still suitable. found in Mn++ and Fe"11". however. 1118. with 7 and these are not generally observed when the / values of perhaps 10 sitions involving spin . One finds. Typical values of is at the maximum of the absorp band for such absorptions "Wm^ = ylog ( j ) =5 cm1 mole 1 liter is (10) The corresponding typical oscillator strength about 10~ 4 . The three absorption bands can be attributed to the states indicated above each absorption band. according to the discussion of Sec. the clusion. the product of the terms in the transitionmoment integral would be anti1112 shows.. to very weak bands. (It should Tranchanges lead. to some extent. The d* configuration. in two regards. Furthermore. however.Y. that no transitions between states arising from a given configuration such as d 2 are allowed. One final illustration of the crystal field interpretation of the delectron energy levels and the spectrum resulting from transitions between these levels can be given. spacing between the energy levels of Fig.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 301 changes in the net spin of the system and can therefore. for not too strong a crystal the groundstate configuration is (U l ) (e<. of the system. lead to relatively weak a(i>) absorption bands. is First. 1117.
Chem. this ground state configuration greater than can be achieved has a spin of 5/2. 6] .) The number and positions ] of the absorption bands of various Mn++ coordination of to the suggestion that a value of A about 7.000 20. to interpret the electronic spectrum.000 . 1119.000 12. The spectrum of the predictions that [Mn(H 2 0) ++ shown in Fig. one would now normally focus one's attention primarily on the higher energy states with the same spin as the ground state. J. Although. have noticed the generally pale colors displayed by manganese solutions. e This leads to the horizontal groundstate energy level for the S state shown in Fig. 1119.. [From Holmes and McClure. states when radiation is absorbed low crystal is shown in the region of relatively field energy in Fig. are and this situation produces The energy levels for these states.000 28. and this again can be attributed to the spin change involved in the visible region transitions. is Secondly. 1118 The absorption spectrum due to the octahedral ion [Ni(H 2 0)6] ++. 26: 1686 (1957). 1119. in fact.000 24. The oscillator strengths are less for Mn++ by a factor of about one hundred and this can be attributed to the violation of the AS = rule that must accompany the dd transitions of the [Mn(H 2 0) 6 ++ ion.000 16.302 INTRODUCTION TO MOLECULAR SPECTROSCOPY of A. the low extinction coefficients. a net spin of 3/2 and quartet states. which now must serve as the excited in a dd transition. Phys.800 cm" 1 complexes have led for [Mn(H 2 0) 6 ++ ] . in line with First. as compared for example with the cor responding spectrum of the aquated Ni ++ ion. and this by any of the other possible electron configurations. 1120.] 8. (The reader may. FIG. such states do not exist. should be noticed. would be made on the basis of Fig. The highest spin found in the excited configurations corresponds to two electrons being paired with opposite spins.
1119 state. K.000 1 4. E.000 1 5.000 12." Methuen & Co. Acta. Jorgensen. Only the ground designated as 6 S and e T ig and quartet states are shown. 1960. (From L.. [From C.500 i 0. ltd. Introduction to TransitionMetal Chemistry.000 An 1 30. London. 1120 The absorption spectrum due to the ion [Mn(H 2 0) 6 ] ++ .ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 303 50. in cm 1 1 1 The energylevel diagram for a a* 5 ion such as Mn" "" ".01 FIG. Orgel. "An .000 <"V<V 2. 8: 1502 (1954).000 i 3.500 5.000 25.500 A FIG...] A^ 20.02 0. Scand.03 0.500 10.) XinA 6.000 7. Chem.000 ><(cm _1 ) .
of the chapter. Now we will investigate the ways in which a molecule. occur with coordination compounds. from comparisons of the observed In a similar evaluated. measuring the magnithe interaction between the ligands and a metal ion. EMISSION AND DECAY PROCESSES Various types of electronic transitions that lead to the absorption of electromagnetic radiation and raise the absorbing molecule or ion to some highenergy electronic state have been dealt with in the preceding parts ion. In one type. energy.304 INTRODUCTION TO MOLECULAR SPECTROSCOPY which predicts a group of three absorptions around 25. be dealt with here. This ligand field approach will not. and the parameter A. however. square planar.000 cm1 and another group of three around 30. can be Other spectra of other octahedral ions can also be underof A determined. tetrahedral. and so forth. The positions of the principal rfelectron absorp tion bands are understood. energylevel diagrams for these geometries can be constructed. The above examples illustrate the chief goals and results of the dis. i. or can get rid of its excess energy and return to the ground state after Two general it has been raised to such a highenergy electronic state.. it might be added. be mentioned again that a more complete and detailed study must take into account the orbitals of the ligands as well as those of the metal ion. contribute to the characteristic colors often observed with transition metal complexes. They do not. leading to chargetransfer bands. and the spectra of ions with these geometries can also be treated. energy is transferred to molecules of this energy that collide with the excited molecule and carry as translational.000 cm 1 gives the best correlation between the energylevel diagram and the observed spectrum. tude of stood. and the value spectra with the energylevel diagram. can be studied. No emission of . which we will call nonradiative processes. or rotational away some and vibrational. Although these bands are generally very intense. cussion of this section. types of energydissipation processes can be recognized.e. Again it should be pointed out that other electronic transitions. Then one constructs suitable orbitals from these two types of localized orbitals. d electrons and can understand the principal feaFurthermore. other geometries. however. they usually occur farther out in the ultraviolet region than do the dd bands treated here. struct energylevel diagrams for different way one can con octahedral complexes of ions with numbers of tures of the spectra of these ions. It should.
One solvent often used is boric acid. can be expected that restricting process. 101 and illustrated in Fig. The net . to of dissolving the material sets to a rigid glass. calculations. for reasons which will be made clear in the remainder of the chapter. effect a slowing down of this process consists under study in a solvent which. such analysis leads to the further classifications of fluorescent emission and phosphorescent emission. the excess higher vibrational states of the solution it vibrational energy. Such radiation can be detected and analyzed. is observed in such processes. motion of the molecules of the solvent as emitted radiation. 1020 is more typical. and for polyatomic molecules the potential energy would have to be represented by a surface in this manydimensional space. The second type of dissipation on the other hand. when cooled. vibrational. involves an emission of radiant energy. it is seen that many the excess vibrational energy Since the principal being lost. and is not detected from ultrasonic dispersion measure ments and theoretical of 10~ 10 sec. Furthermore. another is a mixture of ether. 102. probably by one vibrational level step at a time. and perhaps It appears. This energy goes into the thermal. 117. The simple diagram of Fig. isopentane. the potential energy for a given electronic in the region covered state will be a function of all the internal coordinates of the molecule. do occur while is 10~ 13 sec.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 305 radiation processes. referred to as EPA. say a thousand. In this regard Fig. or eliminating such collisions will slow down the A procedure that has been often used. 101 showing the potential curve of the ground electronic configuration configuration is and that of a single excited state not typical of the situation that exists for polyatomic molecules and is inadequate for a complete discussion of radiationless energy dissipation. that a typical time required for the dissipation of excess vibrational energy When this is by such a process is of the order compared with a typical vibrational period of vibrations. the transi tion associated with an electronic absorption band produces a higher energy electron configuration and often leaves the molecule in one of the new electronic state. For molecules in appears that the molecular collisions are very effective in removing. and as we will see. Nonradiative Processes As was discussed in Sec. A number of different electronic configurations must be expected to have energies that lead to potentialenergy curves by this diagram. and ethanol. mechanism for the vibrational deactivation of a it molecule involves molecular collisions.
to change over into the configuration corresponding to the curve labeled S'. moreover. surfaces of different electronic states will cross one another. FIG. initially The occurrence of a crossing. 1121 Energy dissipation by vibrational deactivation and internal conversion. 1121. (The dashed curve suggests the course of these nonradiative processes.306 effect of these INTRODUCTION TO MOLECULAR SPECTROSCOPY two features is that potentialenergy curves or. two potential curves cross as in Fig.) Ground state potential curve Representative internal coordinate . is that shown in Fig. it is When two potential surfaces or. This type can. more properly. originally in the excited electronic configura tion corresponding to the curve labeled S. The crossing of the potential curves facilitates this process. known as internal conversion. lead to the return of the excited molecule to ground to simplify the discussion. reached by the absorption process its potential surface with that leads to a second type of radiationless process. state. possible for the molecule. and a representative arrangement of potential curves 1121. or nearby.
. potential curves often exist with suitable relative positions so that combinations of vibrational deactivations and internal conversions can return the molecule to its ground state Other molecules. than the absorption band suggests that vibrational deactivation within the potential curve of the upper electronic state is essentially FIG. some of the The observation that the emission band generally appears at longer wavelengths. The simplest mechanism that in Fig. various emission processes can occur. curves or have potential however.e. lower energy. before any emission process has had a chance to occur. For polyatomic molecules. In such cases emission does occur.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 307 because at the crossover point the potential energies of the two electronic states are equal and the vibrational kinetic energy of the molecule in would be zero. either do not have such crossing either state which the molecule becomes trapped. is consistent with observed emission is illustrated 1122. Representative internal coordinate  . and we will now investigate how this happens. electronic states in 118. Fluorescence If the efficiency of internal conversion processes that return the molecule to its ground state is not too great. i. 1 122 Fluorescent emission from the same electronic state as initially reached by the excitation process.
although other conventions are used. Absorption and emission bands which illustrate this behavior are shown in Fig. 1123 The absorption and fluorescent emission spectrum of anthracene dissolved in dionane showing the occurrence of the emission band at longer wavelengths.] XA 5. Kortum and B. 263 (1942). Z. B52. Since most fairly intense absorption bands that are observed occur from the ground state to an excited state with the same multiplicity as the ground state. will fluorescence. therefore.000 1 3. [From G. usually triplet to singlet transitions. as suggested by the transition arrows of Fig.000 .000 30. often continue for periods be defined as phosphorescence and.000 1 4.000 22. therefore. rather than the exact reverse of the absorption is to be drawn. can and generally do occur within about 10~ 4 sec after excitation. 1122. This definition includes. emissions from states reached from the initial excited state by internal conversion as long as the second state has the same multiplicity. much emission occurs and.000 32. that the emission transition of Fig.000 1 Absorption \ / Emission J i i i i I 1 I 20. 1122.000 26. Finckh. as defined above.) FIG. to use the term fluorescence for emission transitions that occur between states of the same multiplicity. physik Chem. in contrast to up to seconds after the absorption process is ended. the emission process of Fig. 1123. It is now becoming customary.000 24. We will now see that fluorescent transitions.308 INTRODUCTION TO MOLECULAR SPECTROSCOPY complete before transition. usually a singlet state. 1122 is a transition with no multiplicity change. (Transitions connecting states of different multiplicities. as the ground state.000 28.
h introduced in Eq. about the process of spontaneous emission. for an emission process the type illustrated in Fig.g The nature (We know. derived in Chap. however. Molecules can be excited from state g to state h by transitions induced by the radiation in the cavity. will be determined by relations established here. one writes where hv g h Now we . We will then be in a position to use the measurable absorption coefficient to deduce the behavior of the spontaneousemission process.e h. . g . or absorption. induced emission. = e». Let us consider molecules with a ground electronic state designated as g and a higher electronic state designated as h.iO +A h > g ] (14) where Ah* g of is Einstein's coefficient of spontaneous emission.lkT _ 8whv" 1 i 11A (") . been turned Information is needed. h = B h. from the discussion of Chap. per molecule.B*. 4. downward can occur by either induced emission or sponOne can write the net rate of these processes as h Rate of — > g transitions = iV\[. or probability. (35) of Chap. coefficient of A brief derivation will show that a it is coefficient for spontaneous emission can be deduced and that related to the 4. . At equilibrium the ratio of the number g of molecules N h in the higher energy state to the number N in the ground state will be given by Boltzmann's distribution as Ei. must investigate the mechanism by which this equilibrium is established and maintained. Consider the sample containing these molecules to be in a blackbody cavity in which the radiation density p(v) is related to the frequency and the temperature by Planck's ..rt /*r (12) is the energy difference. 4 that B g .) A h.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 309 Let us now consider the rate. between the two states. p(>'. The experimental results that of might try to understand would consist of measurements of the one change of intensity of the emitted radiation with time after the radiation beam causing the excitation to the higherenergy electronic state had off. Recalling the Einstein coefficient of induced absorption B g. 1122. Rate of g —* h transitions = N B ^hp{y g g g h) (13) Transitions taneous emission. radiation law P(») = —0. therefore.
. (17) A h^g can be calculated from the observed integrated intensity of the corresponding absorption band.integrated intensity of the /<*(?) dv. the value of g(f) d v . Thus " _ . with the desired relation . 4.000)cft0 ' J With this result.310 INTRODUCTION TO MOLECULAR SPECTROSCOPY At equilibrium the two opposing rates must be equal. absorption band. . the rate equation takes for the number of molecules in the excited state to decrease to half its value. i. 64r«(iv»)'M. One can write the or (18) d In N h = .(l. by Eqs. Rearrangement of the resulting equation gives. (11). and one can obtain from the above rates the relation N Now tity.e.(A k^g ) dt I Integration over the time interval = to t = t\ in which N h goes from N h to iNh gives or '* (19) = In 2 AT~ .»» 3h 8. _ = 8rh(v gh )* C3 8irh(v gh ) 3 B ^h g (16) We can further relate #„_* to the observed. The time it lifetime of is exciting radiation an excited state that spontaneously emits after the turned off can be calculated as a half life. (36) and (46) in Chap. g B^ gP ( Vgh ) + A^ g ^ one can replace N h /N a by the Boltzmann expression for this quanand one can also replace p{vhg ) by the blackbody expression of Eq.
(20). in halflife. collisional deactivation. 4. according to the above analysis. Lewis. one has <21 > «<S? A typical strong absorption band due to an allowed electronic band maximum at. perhaps. as by nitrogen temperature. than that cited for typical completely allowed transitions. or shorter than.55 4 halflife. Eq. with a correspondingly longer One finds in practice that fluorescent emission can be prolonged to halflives of as much as 10~ 4 sec in cases where the value of Ah. i. Phosphorescence is Phosphorescent emission here defined as resulting from transitions i. occur with a smaller integrated absorption coefficient. typical nonthat any decrease in the efficiency of freezing in a glass or cooling to liquid radiative processes involving internal conversion treated in the previous It is clear.e.000)(P„») 2 /a(P) dv 0. have a to the emission process. absorption should. (78) of Chap. N.55 2 X 10" (iVJk) /a(P) dv (20) Alternatively one can relate the halflife to the oscillator strength / If one inserts the relation of of the transition discussed in Sec..000 cm = 10 9 and a . 49. (17) gives h = 2Vln2 8irc(l.. the early work of G. substitution of the expression for A h ^g obtained in Eq. is small enough and nonradiative processes are relatively ineffective. will allow the fluorescent emission to occur to a significant extent. that connect electronic states of different multiplicities. that the most . 3. Emissions corresponding to these bands occur. or 30.e. 119. It is now recognized.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 311 Finally.. view of Eq.000 transition has }a(v) dv 1 The emission from the excited state reached by this A. due X 10" 9 X 10 8 (10 9 ) Q 6 x 10 _9 gec (22) Such section. lifetimes are comparable with. of 0.. Many absorption bands are less intense. therefore. different net electron spin angular states with primarily as a result of momenta.
tripletstate molecules further vibrational energy will be lost and the molecules will occupy the lowlying vibration levels of the triplet electronic state.e. the spontaneousemission process will FIG.. Although internal conis version between states of different multiplicities this process can occur apparently not easy. singlet Absorption of radiation from the ground singlet to an excited can occur to an appreciable extent. Emission from this phosphorescence. ^___ ' — S' T \ \ V >v \ \ *.y / \/ / >^_^/\s. 1124. 1 124 Arrangement of potentialenergy curves of singlet S and triplet T states that can lead to phosphorescence. The potentialenergy diagram that represents a situation that can lead to phosphorescence is shown in Fig. triplet state to the ground state will constitute Since such transitions will be forbidden by the pro hibition against spin changes. Collisional deactivation can then drop the energy past the point where the potential curve of the singlet state crosses that of the triplet state. ^——— ' S '/ Representative coordinate  . and can lead to collisional deactivation does potential crossing point.312 INTRODUCTION TO MOLECULAR SPECTROSCOPY tions between a triplet excited state important phosphorescent bands of organic compounds arise from transiand a singlet ground state. if the not carry the molecule too quickly past the Once a triplet electronic state is formed. the transition can be allowed. i.
and the phosrelatively long periods of time.) The state is process of internal conversion between a singlet sufficiently difficult. a will be very small.e. and parand phosphorescence. The occurrence of a longlived highenergy species. such as the triplet states that lead to phosphorescence. (One should recognize that the relation between the £„. 1122. often provides the means whereby photochemical reactions can occur. such as that of Fig.ELECTRONIC SPECTRA OF POLYATOMIC MOLECULES 313 occur with low probability.& and Aj_„ coefficients and the relation between A h + g and ij require a roundabout mechanism. generally referred to as inter system crossing. and photochemistry are intimately related. With this procedure many organic compounds undergo appreciable phosphorescence excited singlet to triplet internal conversions. and a triplet even if the potential curves cross.. or longer. . It should be mentioned that the lifetimes of excited states is a subject of great importance in photochemical studies. Again the process of forming a glass with the material as a solute and cooling the system to liquid nitrogen temperatures is often resorted to. and are observed to emit radiation as phosphorescence. are not A^ uncommon. in order to populate a longlived excited state. that is seldom observed in liquids at ordinary temperatures. continue over phorescent emission will Halflives for phosphorescence of the order of seconds. ticularly In this fluorescence way the study of electronic spectra. i.
of. 80. 100. 70 Benzene. 204.. 811 Bohr orbits. electronic states of.. 288 Classes. 247 in molecules. 8 Operators for. 116122 B. 116122 for particleinabox model. 109112 112114 114 Character table. 267269. 207 normal. in atoms. 106 statistics. 205 248251 Angular velocity. 114 7476 for vibrational transitions. quantum mechanical. 9 Boltzmann distribution. 108 energy levels structures of. rotational spectra of. 27 internal. 50. 6 Coordinates. 2123 Conversion factors. 290 Davisson and Germer. displacement. Centralforce field. 24 for rotational for vibrational energy levels. 186188 Configurations. 242. 32 for spring systems. 183186 Chargetransfer transitions. 33. 9 Broglie. 11 Angular momentum. 257259 Coupling of rotationvibration. 7680 Amplitudes of vibration. Born. 216 Coordination compounds. in assymmetric top mole for electronic transitions. 86. de. 288312 Correlation diagrams. 41 d orbitals. 22 quantum restrictions on. 243 Beer's law. 284 Conjugated systems. 53 energy levels. rotation Centrifugal distortion. 81 cules.INDEX Absorption coefficient. 205. 5760 in symmetric top molecules. 11 Degenerate vibrations. 215. 129. table of. 48 Anharmonicity. 130 315 . 11 Degeneracy. M. 204. 4346 Applications of spectroscopy. 276278 Average value. definition 52 Band head. of. 279286 286 symmetry. in linear molecules. 8 Bohr hydrogen atom model. L. 87. 6 Asymmetric top molecules.B. Bohr condition on. in group theory. N. 251260. ultraviolet spectrum of. 23. 138141 Crystalfield theory. 12 BoseEinstein 145 Angstrom unit. references to. 70^71 Bond lengths. 57 in symmetric top molecules. 168171 Characters of representations. electronic. 288 Bohr.
31 Hund's rule. intensities of. 4760 rotationvibration bands of. 100. 116 Deslandres table. 5060 rotationvibration bands of. 2546 Dipole moments. 8489 . 231244 Electronic states. 92 Displacement coordinate. 42 204206 for polyatomic molecules. 251260 Electronic spectra. 209 Dissociation energy. 271 cal solution for. 270. Induced absorption. 208 Laporte rule. 234 Experimental methods. 39 Hexatriene. 266269 electronic spectrum of. 35. vibrational. 272 of coordination compounds. 16 Einstein. 310 Electric field. 303 of diatomic molecules. 25 Interaction force constants. 279282 276279 in coordination compounds. 109 of a vibrating molecule. 142149 rotational selection rules for. 76 Infrared spectra. table of. of a rotating molecule. 31. 269271 of coordination compounds. 279286 of carbonyl compounds. 206 Internal conversion. 172201 Group vibrations. 29 Emission. 29. 289 Linear molecules. 115. of atoms. 242. 302. 230263 rotation of. 76 Induced emission. 233 Freeelectron model. 104. 306. 134142 vibration of. for atoms. 293304 of diatomic molecules. from rotational spectra. references to. 237. 173 of. interaction with molecules. 286 of carbonyl compounds. symmetry 198200 Electromagnetic radiation. Degrees of freedom. 6 Lagrange's equation. 27. quantum mechani 3739 Hermite polynomial. 28 of HC1. 289293 in diatomic molecules. 284 for carbonyl group. 260261 Electronic transitions. 301 Ligandfield theory. on energy levels. 42 Formaldehyde. effect 92. 263 of a vibrating particle. 238 oscillator strength. 251 89 222 Gerade states. 271275 electronic states of. 243 FranckCondon Frequency. 267269 for diatomic molecules. 3 Electronic configurations. 100 Group. FermiDirac 145 Fluorescence. 234238 Diatomic molecules. electronic configurations of. 309. 13. 17 Hooke's law. 8081 in conjugated systems. 307 Intersystem crossing. 5 Electron spin resonance spectroscopy. 307311 Force constant. 9092 table of. definition of.. 276278 69 G matrix. 208. Harmonic oscillator. 49. of electromagnetic radiation. 8 Einstein's coefficient. 6670 regions of. 313 Kinetic energy. principle. A. 207. 244251 of benzene. 81 statistics. 27.4 316 INTRODUCTION TO MOLECULAR SPECTROSCOPY /. 28. 271 Fortrat parabola. 304313 Emission spectra of diatomic molecules. electronic states of. 4 Group theory. of induced absorption. 247 for benzene. 12. in aromatic systems. 248. of spontaneous emission. rotation of. 27. Eigenfunction. of benzene. 251260 Electronic orbitals.
5. 28 for springs. of a linear system. 3 Nuclearspin coupling. 17. 238244 . 212 symmetry of. 3536. 7 Oblate symmetrictop molecules. 22 Ortho states. 46 Para states. 1417. 307312 Raman spectroscopy. 115130 Population. 142149 of symmetric top molecules. of rotational energy levels. 146149 of symmetric top molecules. 235 Multiplicity. 11. 3437. vibrations of. of linear molecules. 204206 relation to Hooke's law. of diatomic molecules. 12 Nonbonding orbitals. 5 Reduced mass. 175183 irreducible. of asymmetric top molecules. 2123 table of. 4. 47 Rotation. 8489 Rotational structure of electronic bands. 125130 Normal modes. 53 of vibrational energy levels. 32. 9295 Quadrupole coupling. 81 Radiative processes. 102 Orthogonality of wave functions. 267 irelectron transition. 142 146 of symmetric top molecules. 188190 Overtones. 24 of rotational levels. for chemical bonds. 19 Oscillator strength. 53 Newton's equation. 28. 276278 Operators. symmetry of. 29 . 48.. 138142 Morse. rotational Stark effect for. 100 Perpendicular bands. 51 of symmetric top molecules. 2 Reciprocal centimeters. 7 Planck's constant. 281282 Moment of inertia. 50. 7 Point groups. 311313 tt bonds. 134142 of linear molecules. 95 Quantum numbers of electrons in atoms. 102 Parallel bands. 8992 Phosphorescence. 267 Nonradiative processes. 43 for polyatomic molecules. 94. 267 76 Telectrons. 152158 Rigid rotor. 162 Polyatomic molecules. 49 Representations. 244251 Quantum restrictions. 149158 Rotational energies. 238243 Planck. 41 Potential energy function. 179183. 149152 Particleinabox problem. 179183 reducible. 4852 of symmetric top molecules. M. 37 for I 2 241 Maclaurin series expansion. 12. 276278 Pauli exclusion principle. 304307 Nonrigid rotor.. 27. 116122.INDEX 317 Linear molecules. M. 5760 Normal coordinates. 194195 Normal vibrations. 103109 Rotational selection rules. 105 Normalization. 36. 18 Nuclear magnetic resonance spectroscopy. 113 of linear molecules. Lonepair electrons. 13 Prolate symmetric top molecules. P. 290 Octatetraene. 261263 for HC1. 75. intensity of. 49 for excited electronic states. from rotation spectra. 202 of a linear triatomic molecule. 45. 16 Molecular orbitals for benzene. 36. 103109 Rotationvibration spectra. 116122 Probability function. 105 Octahedral symmetry. 160163 table of. 18. 57 from rotationvibration spectra. 305. of linear molecules. 35 Morse function.
2 Spontaneous emission. 10 Thomson. 76 symmetric top molecules. 12 functions. 205 Vector model of the atom. 4 Wilson's method for vibrational calcula tions. of diatomic molecules. 203206 Structure. 39 for rotating linear molecules. 106. 165168 Transition Schroedinger. 86 harmonic oscillator. 109 of. 252. 2634 of rotation of. 244 226. of asymmetric top molecules. E. classical. 100 United atom. 8 Rydberg equation. 8 Timedependent Schroedinger equation. 8489. 307312 Ungerade rotational. 114 of linear molecules. 57 Vibrational structure of electronic bands. application to tions. 249251 Vibrational energies. of linear triatomic molecules. J. 253 <r bonds. timedependent. 69 Transitions. 266. 208212 H 2 vibra 213218 of degenerate vibrations. 106 states. 103 of 80 Vibrations. 98103 Valenceforce field. 1217 Selection rules. 86. 160163 Symmetry operations.318 INTRODUCTION TO MOLECULAR SPECTROSCOPY Rutherford. E.. 149158 structures of. 6163 Transformation matrices. 2546 of polyatomic molecules. 96103 Symmetry elements. 7274 electronic. 252. 130 of electronic states. 129. 103109 H 2 0. Wave Wave for equation. 122124. 253 vibrational. 41 Separated atoms. 238 Vibrational transitions. 109 Symmetry. intensity 106. induced.. 8992 Statistical weight of rotational levels. 267 Spectrometers. 214218 rotationvibration bands of. J. 231. 6366 spontaneous. 6163 timeindependent. 11 Sehroedinger equation. 5 Wavelength. 213218. 273 moment. 125130 in vibrational calculations. 309311 Stark effect. 229 wave functions. 40. 160163 of 245 Wave numbers. 218229 . Symmetric top molecules. for atomic orbitals. basis of.. 269271 of normal coordinates.
Engineers and 2nd Ed. 2nd Ed. 2nd Ed Fundamentals of Optics. 2nd Ed I Electronics. 6th Ed Physical Chemistry of Metals Instrumental Methods of Chemical Analysis. n 9 99 SYNGE WEBER WHITE O cn o o Production to Atomic Spectra Principles of Biochemistry. 2nd Ed Fundamental Concepts ol Inorganic Chen iiiclear Pliv IMING Numencal Methods For Scientists And Engini Ed ilculations of Analytical Chemistry.Other McGrawHill INTERNATIONAL STUDENT EDITIONS in 4> 99 99 Related Fields O Physical Chemistry for PreMedical Students. 4th Ed. 4th Ed rmciples ol Modern Physics w o c > 99 (It hemodynamics. 5th Ed. 3rd Ed. PIPES Applied Mathematics itroduction to . PAULING Introduction to Quantum Mecha for •0 Physicists.cs. 2nd Ed. 3rd Ed fundamentals of Electricity and Magnetism nemistry of Engineering Materials. 2nd Ed Electromagnetism Modern Physics Mathematics of Physics and Modern Engineering Principles of Mechanics. College Physics. 2nd Ed 3 as Physical Chemistry Organic Chemistry. 2nd Ed he Modern Theory of Solids Chemistry. Principles of Mathematical Analysis T Quantum Mechanics. 3rd Ed. 2nd Ed. Modern Physics. 2nd Ed 'hys. ieat and Thermodynamics. 4 . 6th Physical Organic Chemistry. Eiperimental Physical Chemistry.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.