Lecture notes on Quantum Mechanics

for the Course Principles of Atomic and Nuclear Physics
M.S. in Energy Engineering, Nuclear Curriculum

Stefano Atzeni
Dipartimento SBAI, Universit`a di Roma “La Sapienza”
Nov. 17, 2014

1

Last corrected: October 2, 2015

1

ii

Preface
Physics courses for students in Nuclear Engineering at Universit`a di Roma “La
Sapienza” were restructured following the reform of Italian University Curricula, as well as the transformation of the previous Nuclear Engineering M. S.
in one of the Curricula of the Energy Engineering M.S. As a result, two previous one-semester courses on Quantum Mechanics (actually named “Atomic
Physics”) and Nuclear Physics were merged into a single one-semester course
on Principles of Atomic and Nuclear Physics. Students attending this course
have previously attended General Physics and Calculus courses, but were not
thought any Modern Physics. Also, their Math courses do not include such
topics as Fourier transforms, vector spaces and complex analysis.
Principles of Atomic and Nuclear Physics should provide the students an
(at least) qualitative understanding of the main nuclear physics relevant to
fission (and fusion) reactors, as well to applications of nuclear radiations in
medicine and non-destructive analysis. The first part of the course introduces
modern physics concepts (in particular elementary quantum mechanics), the
second part concerns nuclear physics. This manuscript collects my edited lecture
notes on the first part of the course. Indeed, many, excellent textbooks on
quantum mechanics are not adequate, because meant for one- or even twosemester courses and often using too abstract formalism.
This manuscript (as well as my lectures1 ) deal with basics of special relativity and of quantum mechanics using wave mechanics formalism. It includes
simple applications of quantum mechanics to a few standard model problems,
which aim at providing the students examples concerning energy and momentum
quantization, bound and free states, tunneling and measurement uncertainty.
Note:
This is just a draft. Parts of the text are still in Italian. Parts are just scans of
handwritten notes. I hope to improve these Lecture Notes, in particular adding
the examples and excercises discussed in the classroom (and possibly many others). I also plan to add a few additional sections and appendices. Some of these
will deal with topics, such as Fermi golden rule and the proof of the Loren1 Actually the lectures also include elements of kinetic theory of gases, black body spectrum,
photoelectric effect, Compton effect and Bohr’s atomic model. They are not presented here
because they are treated by most General Physics textbooks at the level required by the
present Course.

iii

on the laser and on applications of quantum statistics. Other planned sections. . which may help the students to read introductory nuclear physics textbooks. may be useful to engineering students interested in non-nuclear applications of Modern Physics.iv PREFACE tian shape of a decay line.

. . . . . . 2. . . . 19 19 19 20 20 22 22 24 . . . . . . . . .4 Kinetic energy . . .2 Experimental confirmation . . . . . . . . . .1 de Broglie associated wave .4. .1. . . . . . .1. . . . . . . . . . . . . . . . . . .4 Four-vectors . . . 2. . . . . . . . .5. . . . . . . . . . . . . . . .5. . . . .2 Time and simultaneity . . . . . . . . . . . . 2. . . . . . . . . . . . 2. 2. . . . . . . . . . .1 Position and momentum . . . v . . 2. .2. . . . . . . . . . .2 Length contraction . . .2 Energy and time . . . . . . . . . . . .3 Energy . . . . . . .Contents Preface iii 1 Classical physics: success and crisis 1 2 Special relativity 2. . . mass. .2. 2. . . . . energy . . . . . . . . . . 4. . . . . . . . . . . .3. . . . . . . . . . . . . . . 2. . . . . . 2. . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . .5. . . .1 Recovering time dilatation and length contraction 2. . . 2. . . .2 Transformation of velocities . . . . . . . . . . .1 Momentum-energy 4-vector . . . . . .1 Time dilatation . . . . . . . . . .4. 3 Old Quantum Theory 3 3 4 4 5 7 8 9 10 11 12 13 13 13 14 15 17 4 Material waves 4.2 Heisenberg’s uncertainty principle . . . . . .3. 2. . . . . . . . . . . . . . 2. . . . . . . .2 Momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.3 Discussion . .5 Momentum. . . . . . . . . . . 4. . . . . . . . . . . .5. . . . . . . . . . . .4. . . .1 Justification of de Broglie’s hypothesis 4. . . . . . . .4 Lorentz transformations . . 2. . .1 The postulates . 4. . .3 Lorentz matrix . . . . . . . . . . . . . 4. . . . . . . .3 Consequences of the constancy of the speed of light . . . . . . . . . . . . . . . . . . . . . 4. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . Appendices 85 A Wave packets and group velocity 87 . . . . . . . . . 7. .3 Scattering from a hard sphere . . . . . . . 7. . . . . . . . . . . 75 75 76 79 80 84 . . .5 Absorption cross-section . . . 49 49 50 51 54 54 60 62 63 8 Quantum statistics 8. . . . .1 Descrizione del sistema . . . . . . . . . . . . . . . . . 5. 9. . . . . . . . . . . . . . . . . . . . 8. 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 34 37 43 44 46 47 7 Elementary atomic physics 7. . . . . . . . . . z-component . . . . . . . . . . . . . . . .6 Spin and Pauli exclusion principle . . . . . . . . . . . . . . . .6 Wavefunction normalization . . 5. . . . . . . . particelle . . . . 8. . . . . . . . . . . . . . . . 5.5 Interpretation of the wavefunction . . . . . . . . . . . 6. . . . . . . . .4 Limite classico: statistica di Boltzmann . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 Eigenvalues and eigenfunctions . . . . . . . . . .2 Statistica di Bose-Einstein per un sistema di N 8.1 Wavefunction . . . . . . . . . .3 Time-dependent Schr¨odinger equation . . . . . . . . . . . . . . . . . . . . . . . . 8. . . . . . . . . . . . . . . .5 Schr¨odinger equation for the hydrogen atom 7. .1 Infinite potential well . . . . . . . . . . .3 Total angular momentum . . . . . . . . .7 Electrons in the atom . . . . . .1 Partial wave expansion . . . . . . . .2 Wave equation for a free particle . . . . . . 7. . 7. . . . . 25 25 26 27 28 28 29 29 30 6 Quantum Mechanics: 1D problems 6. . . . . . . . . . . . . . . . . . . . 7. . . . . . . . . . .8 Operators and commutators . . . . . . . . . .4 Angular momentum and parity . . . . . . . . . .3 Statistica di Fermi-Dirac .1 Operators in spherical coordinates . 65 66 66 66 71 72 73 9 s-wave scattering 9. 5. . . . . . . . . . . . . . . . . . . . . . . . 6. . . . . . . . . . . . .2. . . . . . . . . . . . . .8 Indistinguishability . . . . . . . . . . . .2 Finite potential well . . . . . . . 9. . .3. . . . . . . . . . . . . . .4 Potential scattering and Breit-Wigner cross-section 9. . 9. . . . . . . . . . . . . . . . . .3 Potential step . . . . . . . .1 Statistica di Bose-Einstein per i fotoni . 5.3. .vi CONTENTS 5 Quantum Mechanics: Schr¨ odinger equation 5. 6. . . . .2 E < V0 . . . . . 6. . . . . . . 8.4 Potential barrier and tunnel effect . . . . . . . . . . .4 Time-independent Schr¨odinger equation . .2. . . . . . . . . .2 General expression of the cross-sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. . . . . . . .2 Angular momentum. . 5. . . . . . . . . . . . . . .2 Statistica di Bose-Einstein . . . 6. .1 E > V0 . . . . . . . . . . . .

CONTENTS vii B Fourier transform of a Gaussian 89 C A proof of the uncertainty relations 91 D Angular momentum operators 93 E Hydrogen atom: radial equation 97 F Potential Barrier and Tunnel Effect 103 G Metodo dei moltiplicatori di Lagrange 109 .

viii CONTENTS .

g. 1 . Maxwell’s electromagnetisms. thermodynamics and the kinetic theory of ideal gases. • Netwon’s mechanics – relativity principle – the three principles of mechanics – the law of gravitation – the laws of some non-fundamental forces (e. e.g. • Maxwell’s electromagnetism – the four Maxwell’s equations – Coulomb-Lorentz force explains electrical. optical phenomena. We should not be too much surprised. Classical physics was developed to explain (or to account for) phenomena observed by our senses. the existence of the planet Neptune (later found by astronomers). nor to bodies in motion at velocity close to the speed of light. and are only accustomed to velocities much smaller than the speed of light. predicted. the elastic force) explains fall of bodies. predicted the existence of electromagnetic waves (later discovered by Hertz). We very briefly summarise successes and elements of crisis of classical physics. etc. that do not allow us to see the atomic world. magnetical. Classical physics is neither applicable to the microscopic word.Chapter 1 Classical physics: success and crisis By classical physics we refer to Newton’s mechanics. Kepler’s laws.

specific heats of gases. chemical kinetics. CLASSICAL PHYSICS: SUCCESS AND CRISIS • thermodynamics – the three laws sets equivalence between heat and work explains limitation of heat engines • kinetic theory of gases. a bridge between mechanics and thermodynamics – Maxwell’s velocity distribution function – Boltzmann’s factor – Principle of equipartition of energy explains laws of gases. – atom stability. • Classical physics cannot explain – black-body spectrum. – photoelectric effect. – atomic spectra. but are covariant under Lorentz transformations.2 CHAPTER 1. Galileo’s law of velocity addition does not apply. . Brownian motion Element of crisis at the end of the XIX century • Maxwell’s equations are not covariant under Galilean transformations. – Compton effect (discovered in the 1920’s). – specific heats at low temperature.

special relativity tells us how electric and magnetic fields transform from one system to another. The special principle of relativity Physical laws should be the same in every inertial frame of reference.: in A. Here we only discuss kinematics and dynamics aspects of relativity. 891 (1905).2 Einstein showed that the inconsistency is due to the equations of kinematics. Galileans transformations are replaced by a new set of transformations (Lorentz transformations). when the concept of time is revised and the new postulate of constancy of the speed of light c is introduced (see below). Dover. In fact. Chapter XI. one also postulates that space is homogeneous and isotropic. 1 See the introduction to the original Einstein’s paper. 2. the inconsistency is removed. depending on the oberver’s frame. Indeed. 17. 1996). Napoli (1986. New York (1952) 2 A trivial example: the electrodynamic force q(E ~ +~v × B) ~ depends on velocity. 3 In fact. Galileo’s law of addition of velocity is not consistent with Maxwell’s equations. In this chapter. 3 . not to electrodynamics. so a different force acts on a particle. which in the limit of velocities v ≪ c reduce to Galilean transformations. as briefly mentioned in the preceding Chapter. Fisica I. Silvestrini. English tr.1 The postulates Special relativity is based on the following three postulates 3 1.Chapter 2 Special relativity Einstein developed special relativity in order to make the electrodynamics of moving bodies consistent with Maxwell’s theory applying to bodies at rest. Sommerfeld (Ed. Mencuccini and V.) The Principle of Relativity. Liguori. I urge the students to refer to this book for enlightening discussions of the postulates of special relativity. it was known that Maxwell’s equations are not covariant for Galilean transformations. and the resulting force does not change. Annalen der Physik. In other words. and description of a few experiments confirming the accuracy of special relativity.1 Indeed. critiques of simultaneity. we follow closely the presentation in the textbook C.

to any physical process. SPECIAL RELATIVITY 2. Events judged as simultaneous by an observer are not simultaneous for another observer moving with respect to the first one. Let us assume that the ruler and the source move with respect to another observer (see Fig. a and b. at the time of signal emission points a and b are equidistant from S. For an observer at rest with respect to the ruler.1). Also for this observer. in a sense. 891 (1905).4 CHAPTER 2. at distance ∆x/2 by both ruler ends. 2. 4 English translation of the original statement in A. Momentum and angular momentum conservation In an inertial reference frame. and light reaches the left hand side of the ruler (now in a′ ) before than the right hand side. Annalen der Physik 17. • postulate 3 is just the same as in classical mechanics. as light propagates. things are not so trivial.2 Time and simultaneity In classical mechanics we implicitly assume that time is universal. unexpected consequences of the constancy of the speed of light. momentum and angular momentum of an isolated system are conserved.4 3. the idea of absolute time is abandoned. 2. Constancy of the speed of light Light is always propagated in empty space with a definite velocity [speed] c which is independent of the state of motion of the emitting body.3 Consequences of the constancy of the speed of light In this section we show two simple. In the following Sec. Clocks run at the same speed in any reference system. stated by Galileo for mechanical phenomena. .4 we shall deal with the general transformations relating measurements in inertial systems in relative motion. Let us consider a source S of light pulses placed at the middle of a ruler. the ruler moves. replaces the assumption of absolute time implicit in classical physics). • postulate 2 is new (and. 2. They concern particular cases of measurements of the same quantity performed by observers sitting in different reference systems (in relative motion). However. the light signal emitted by S reaches a and b simultaneoulsy. In the special theory of relativity. it is postulated that the speed of light (in vacuum) is a universal constant. 2. Notice that: • postulate 1 just extends the relativity principle. Einstein. It is independent of the motion of the light source. Instead. However.

p 2 l2 + (V ∆t/2)2 to reach the detector. c (2. 2.1) ~ = (V.1 Time dilatation We refer to Fig. We are interested in measuring the time a signal emitted by the source takes to reach the detector. ∆t = c c (2. because the detector moves as light propagates . 2. with respect to a Let us assume that the device moves with velocity V frame S = (XOY ) (see Fig. a detector (just in the same position as the source) and a mirror. For an observer sitting in S light has to cover a longer path.2) . 2. showing a device consisting of a pulsed light source.2.1: Events which are simultaneous for an observer are not simultaneous for another observer in motion relative to the first observator. 2.2. CONSEQUENCES OF THE CONSTANCY OF THE SPEED OF LIGHT5 Figure 2.3. 0).3. He therefore measures a time (see Fig.3) s s  2 2  2 ∆t ∆t ∆t0 2 + V c 2 l2 + V 2 2 2 = . An observer sitting in the device’s frame measures a proper time interval ∆t0 = 2l . placed at distance l from the source.3).

(2. Figure 2.3: The primed frame moves towards the right hand side with velocity V .5) An important experimental evidence (one of many): the elementary particle µ is unstable. (2. Equation (2.6 CHAPTER 2. SPECIAL RELATIVITY Figure 2.3) V2 1− 2 c the observer moving with respect to the source-detector system measures a longer time interval than the proper time measured by an observer sitting on the device.3) is often written in the form ∆t = ∆t0 γ.2 µs. detector. A µ moving .2: Our measuring device: source. with measured (proper) mean lifetime of 2. from which we obtain ∆t = r ∆t0 : (2. For an observer in S=(yOx) a light signal emitted in O′ has to travel a path longer than 2l to reach the detector.4) with the function γ (known as the relativistic γ factor ) defined by γ= r 1 V2 1− 2 c ≥ 1. mirror.

2 Length contraction We consider a ruler of proper length L0 . For an observer moving with speed V the ruler length is L. i.6) with ∆t0 a proper time.7) + = c−V c+V V2 1− 2 c Time intervals ∆t and ∆t0 .3. Therefore they travel tens of km.4). 2.4: Experiment to evidence length contraction. at tens of km above sea level. and then ∆t = r ∆t0 2 V 1− 2 c 2L0 /c =r .8) . Figure 2. The ruler with the mirrors move towards the right hand side with velocity V.e. and the time interval is L L 2L/c ∆t = ∆t1 + ∆t2 = . as seen by an observer sitting on the foil. (2.3). CONSEQUENCES OF THE CONSTANCY OF THE SPEED OF LIGHT7 with velocity approaching that of light has an average path of 660 m. with mirrors at both ends (see Fig. We measure the time taken by a light signal to travel from mirror M1 to mirror M2 and back to mirror M1 . and is detected close to the earth surface. A large number of µ’s is produced by interaction of cosmic rays with the upper layers of the atmosphere. in turn. The figure shows the positions of the ruler at different times and the path travelled by light. (2.3. For an observer sitting on the ruler ∆t0 = 2L0 /c. are related by Eq. V2 1− 2 c (2. (2. This is possible because they move at a velocity close to that of light and an observer sitting on earth measures a strongly dilatated time. 2.2.

5: Reference frames S = (yxz) and S ′ = (x′ y ′ z ′ ) In special relativity. V t′ + 2 x′ c r . 2. parallel to the x-axis (see Fig. V2 1− 2 c (2.10) (2.7) and Eq. with velocity V .8). SPECIAL RELATIVITY We find the relation between the lengths by equating the right hand sides of Eq. (2. 2.9) the ruler is shorter for the observer moving with respect to it. which gives L = L0 r 1− V2 = L0 /γ : c2 (2. Lorentz transformations replace Galilean transformations. The relevant transformations. (2.4 Lorentz transformations Figure 2. Here we consider the simple case of measurements performed by observers in inertial frames.8 CHAPTER 2. V2 1− 2 c ′ y.13) . z ′.5). relating measurements performed in the S (OXY Z) frame and in the S ′ frame (X ′ O′ Y ′ Z ′ ) are x = y z = = t = x′ + V t′ r . translating with respect to each other.12) (2.11) (2.

(2. moving with velocity V with respect to S ′ . thus obtaining     V2 ∆t V ∆x′ ′ = ∆x γ 1 − + V ∆t. (2. (2. (2. measures a time interval ∆t given by Eq.1 Recovering time dilatation and length contraction We now check that the expressions for time dilatation [Eq. Time dilatation We consider a phenomen occurring in the origin of sistem S ′ .4. which we write compactly as ∆x = γ(∆x′ + V ∆t′ ). t = t′ ) as V /c → 0. An observer in S.e. V2 1− 2 c y. (2.9 2. A length measured in S is related to a length measured in S ′ by Eq. z′ = z.20) (2.3)] and length contraction [Eq. 2. For an observer in S ′ it occurs in a time interval ∆t′ .15) We immediately see that we recover the usual Galilean transformations (x = x′ + V t.21) . Length contraction We assume that an observer in frame S measures a given length ∆x in a time interval ∆t = 0. (2.16) = V t− 2x c r . V2 1− 2 c (2.17) x′ t ′ = (2.19). (2. analogously. i. y′ = x−Vt r . (2.4.18) ∆t = r V2 1− 2 c which is just Eq. (2. − γ c2 and insert this expression into Eq.10).9)] derived earlier are recovered from the Lorentz transformations.13) with x′ = 0. at x′ = 0.19) To eliminate the dependence on ∆t′ we rewrite Eq.14) (2. i.e.13) as ∆t′ = ∆t V ∆x′ . ∆t′ . (2. ∆x = γ∆x′ + γV − γ c2 c2 (2. LORENTZ TRANSFORMATIONS and.3).

31) dt = γ(dt′ + V ′ dx ).23) (2.13)] we obtain dx = γ(dx′ + V dt′ ). = dt dy . We now obtain the expressions for relativistic transformation of velocities.10)–(2. c2 (2.10 CHAPTER 2.26) (2. (2. unlike the classical case.22) which is just Eq.27) (2. time now depends on the reference system. 2.9) for length contraction. (2. c2 (2. dt′ ′ dy = ′.32) . 0) with respect Again.4.2 Transformation of velocities ~ = (V. (2. dt dy ′ = ′.28) Notice that. SPECIAL RELATIVITY Setting ∆t = 0 and using the definition of γ we eventually have ′ ∆x = ∆x r 1− V2 . By differentiating the Lorentz transformations [Eqs.25) and vx′ vy′ vz′ dx′ . (2. 0. dt dy .29) dy dz = = dy ′ . we assume that frame S ′ moves with velocity V to frame S. dz ′ . dt = (2. The velocity components in the two frames are defined by vx vy vz dx . = dt = (2.24) (2.30) (2. In classical mechanics the x-components of the velocities in the two frames would be related by vx = vx′ + V .

for any value of vx′ (including vx′ = c). x4 ) = (x. Equation (2. V γ(dt′ + 2 dx′ ) c (2. V γ(dt′ + 2 dx′ ) c dy ′ .3 Lorentz matrix If we consider four-component space-time vectors x = (x1 . V γ(1 + 2 vx′ ) c vz′ .4.38) (2.34) (2. ct) (2. 2.37) (2. LORENTZ TRANSFORMATIONS 11 and then we can write the velocity components in the S frame as vx = vy = vz = γ(dx′ + V dt′ ) . V γ(1 + 2 vx′ ) c (2.4. V 1 + 2 vx′ c vy′ . ct′ ).40) The observer in S measures a velocity vx = c. ct′ ) = (x~′ . z. we immediately see that we recover Galilean transformations as V /c → 0. assume that system S ′ moves with the speed of light. y.42) and analogously . x2 . y ′ .41) x′ = (x′1 . V γ(dt′ + 2 dx′ ) c dz ′ . Let us now take V = c. x 1 + 2 vx c c (2.39) As for the Lorentz transformations.e.36) from which we obtain the desired velocity transformation by dividing numerators and denominators by dt′ : vx = vy = vz = vx′ + V . z ′ . x′3 . i.35) (2.33) (2.2. x′2 . (2. ct) = (~x. x′4 ) = (x′ . x3 .37) yields vx = vx′ + c vx′ + c c ′ = c + v ′ = c.

for instance. = x2 . in short).48) (2.4  γ  0 =  0 βγ 0 1 0 0  0 βγ 0 0   1 0  0 γ (2. where A is Lorentz matrix.45) x′4 = −βγx1 + γx4 . (2.49) (2. Hence. ct′ )2 . T x = A−1 x′ . a′ = Aa. The inner product of two 4-vectors a and b is defined by a · b = a1 b 1 + a2 b 2 + a3 b 3 − a4 b 4 . (~x.46) where β = V /c.e. (2. In general.53) x2 + y 2 + z 2 − c2 t2 = x′2 + y ′2 + z ′2 − ct′2 . a 4-vector a is a four-component vector a = (a1 . where xT is the transpose of x and  γ  0  A= 0 −βγ 0 1 0 0 (2. SPECIAL RELATIVITY we can write the Lorentz transformations as x′1 x′2 = γx1 − βγx4 .50) Four-vectors The four-component vectors x and x′ introduced in the previous section are just two particular (and particularly important) relativistic four-component vectors (4-vectors. ct)2 = (x~′ . where A−1 is the inverse of matrix A. 2. (2. a2 .52) (Note the minus sign!) One can verify that the inner product of two 4-vectors is relativistically invariant.43) (2.44) x′3 = x3 . i. Analogously. (2.51) that transforms according to Lorentz transformations. (2. (2.4. In compact vector form x′ = AxT . a3 .12 CHAPTER 2. a4 ) (2.47)  −βγ 0   0  γ 0 0 1 0 is the Lorentz matrix.54) or . it takes the same value in any (inertial) frame.

It can easily be shown that total momentum ( mj ~vj ) is not conserved if the velocities follow the Lorentz transformations. energy In classical mechanics the momentum of a particle of mass m and velocity v is ~p = m~v . and m0 is the rest mass. m0 vy . i.5. m0 ) dt dt0 dt dt0 dt dt0 dt0 dt dt dt dt (m0 vx . (2. m0 γc). in a form resembling the classical one (with spatial part equal to mass times velocity).5.e. (2. i. The 4-vector p can also be written as p = = dx dt dy dt dz dt dct . mc) = (~ p. (2. m0 vz . as dt = dt0 /( 1 − v 2 /c2 ) or dt/dt0 = γ.59) 2. mc). p = (mvx . y.2 Momentum We have then found that the expression of the momentum is just the same as in classical mechanics.60) but with the mass m depending on velocity: m = m0 γ = r m0 v2 1− 2 c . z. 2. MASS. m0 γvz .5 Momentum.5. m0 ) dt0 dt0 dt0 dt0 (2. Total momentum of an P isolated system is conserved. mass. m0 . The reason for the name rest mass will be apparent soon. mvy . m0 . MOMENTUM. m0 .55) where t0 is the proper time. m0 γvy . mvz . the mass of the body in a system moving with the body itself.1 Momentum-energy 4-vector In analogy with the space-time 4-vector x = (x. mc) = (m~v . (2.57) The relation between the time differentials dt and dt0 is obtained from the p Lorentz transformation. . ENERGY 2. p~ = m~v .13 2. We can than write p = (m0 γvx . m0 c ) dt0 dt0 dt0 dt0 (m0 (2.56) (2. m0 . ct) we define the 4-vector p = (m0 dy dz dct dx .61) The mass tends to infinity as the velocity approaches the speed of light c. e. we have to modify the expression of the momentum. Since we want total momentum to be conserved. i.58) or. the time measured in the body’s frame.e.

69) . (2. E = mc2 = r v2 1− 2 c (2. i. (2.65) as F~ · ~v dt = c2 dm. the square of the modulus of the momentum-energy 4-vector is relativistically invariant.5.68) Equation (2. (2.64) from which ~v · d~ p = c2 dm. according to E = ∆m0 c2 .67) or m 0 c2 = m0 c2 γ.14 CHAPTER 2.68) also tells us that an energy release is associated to any reduction of rest mass of a system of particles (viceversa energy is required to increase the mass). SPECIAL RELATIVITY ~ = ~v /v and of γ. As for any 4-vector. (2. From the invariance of the momentum-energy 4-vector we have d|p|2 = d(|~ p|2 − m2 c2 ) = 0.e.62) Both the spatial part and the temporal-part (fourth component) of the 4vector are conserved simultaneously in an isolated system. We can then write dE = c2 dm.63) independent of the reference system. its energy increment.3 Energy We can find the relativistic expression of a particle energy by proceeding as follows. c2 (2. or 2~ p · d~ p − 2mc2 dm = 0. Hence mass conservation is replaced by mass-energy conservation. indeed |p|2 = |m~v |2 − m2 c2 = (m0 γv)2 − (m0 γc)2 = m20 γ 2 (−v 2 + c2 ) = −m20 c2 γ 2 1 − v2 = −m20 c2 . using the definitions of β written as ~ ~p = m0 cγ β.66) The left hand side of this equation represents the work done on the particle.. (2. 2.65) In classical mechanics the increment of momentum equals the instantaneous impulse of the force d~ p = F~ dt. (2. (2. We assume this relation still holds relativistically. and then write Eq. the momentum is often Note that.

(2.71) and using Eq. we can Taylor expand γ= r 1 v2 1− 2 c and T ≃ ≃1+ 1  v 2 3  v 4 − . 2.63) we obtain an important relation between energy. (2. and p = E/c = hν/c. F~ · ~v dt = c2 dm.4 Kinetic energy We go back to the Eq. (2.74) For V /c ≪ 1.5. E ). As v → c.5. c (2.76) We see that the classical limit T = (1/2)m0 V 2 is recovered when the velocity V v ≪ c.70) hence the name of momentum-energy 4-vector.73) where L is the work made by external forces on the body. according to the kinetic energy theorem. (2.66). form the definition of modulus of a 4-vector we can write |p|2 = (~ p)2 − E2 . thus obtaining L = E − E0 = mc2 − m0 c2 . (2.6. E = hν. 2 c 8 c 1 v4 3 m0 v 2 − m0 2 . c2 (2. namely E 2 = (pc)2 + (m0 c2 )2 . such a work is equal to the increment of the body’s kinetic energy T . 2 8 c (2. MASS. 2. T grows indefinitely: see Fig.15 2. Hence T = mc2 − m0 c2 = m0 c2 (γ − 1). momentum and rest mass.72) Note that for a photon (with zero mass). MOMENTUM. (2.75) (2. On the other hand. . Finally. and integrate assuming a body at rest at time t = 0. ENERGY Note that using Eq.68) the 4-vector p can be written as p = (~ p.

16 CHAPTER 2. . SPECIAL RELATIVITY Figure 2.6: Kinetic energy T vs v 2 .

The concepts of force and energy are just the classical one.292 × 10−11 m. . Light is composed of particle-like photons. . (3. The electron orbits and energies turn out to be quantized. of zero mass. introducing quantization of oscillator energy [E(ν) = nhν] and giving up the principle of equipartition of energy in favour of level occupation according to the Bolzmann factor exp(−E/kB T ). The concept of photon also explains Compton effect. rn En = = r0 n2 .) and that orbiting electrons do not irradiate.1) 2 E0 /n . (3. 2. • Atom emission and absorption spectra are recovered by Bohr’s atom model and its subsequent extensions. • The black body spectrum was recovered by Planck.3) . with n = 1.2) with r0 = 4πε0 ¯ h2 = 5. Bohr’s model assumes that orbital momentum is quantized (l = n¯h = nh/2π.Chapter 3 Particles. me e 2 17 Bohr′ s radius (3. . In the Hydrogen atom the electrons with quantum number n move in a circular orbit of radius rn and have total energy En . • The photoelectric effect was explained by Einstein assuming that electromagnetic waves carry energy in packets of finite quantity. photons and Old Quantum Theory Old quantum theory refers to quantum concepts introduced ad-hoc in order to overcome limitations of classical physics. energy E = hν and momentum p = hν/c.

18 CHAPTER 3. and −e the electron charge. Treaments of • Maxwell-Boltzmann distribution • energy equipartition principle • black body • photoelectric effect • Compton effect • Bohr’s atom at the level of detail required in this Course can be found in many General Physics textbooks.60 eV = −2. When excited electrons in level m return to a lower energy state n they emit a photon with frequency νmn = (Em − En )/¯h). The reader is urged to refer to them. OLD QUANTUM THEORY and E0 = − 1 me e 4 = −13. me the electron mass.4) with ε0 the vacuum dielectric constant. 2 (4πε0 )2 ¯h2 (3.179 × 10−18 J. .

1) between wavelength and momentum is just the same as that holding for photons c h c = .3) On the other hand. and then v= dE . (4. dp (4.1) is as follows.2) λ= =h ν hν p Another argument in support of Eq. the group velocity of a wave packet is vg = dω/dk. molecules have confirmed de Broglie’s hypothesis. atoms. p (4. we have vg = − λ2 dE . The kinetic energy of a non relativistic particle of mass m and velocity v can be written as E = p2 /2m. If we write the angular frequency as ω = 2πν = 2πE/h and the wavenumber as k = 2π/λ. We associate a wave packet to a particle. nucleons. h dλ 19 (4. The wavelength associated to a particle of momentum p is λ= h .1 de Broglie associated wave Louis de Broglie suggested that just as light has both wave-like and particle-like properties.Chapter 4 Material waves 4.1) Experiments on electrons. which can be differentiated to yield dE = pdp/m = vdp. 4. (4.4) .1 Justification of de Broglie’s hypothesis The relation (4.1. as shown in Appendix A. particles also have wave-like properties. and impose that the group velocity (see Appendix A) of the wave packet coincides with the particle velocity. nuclei.

e. from which Eq. For an electron moving with velocity v in a circular orbit of radius r. electron diffraction by crystals (Thomson & Reid. 1927). 5th Ed.20 CHAPTER 4. (4. concerning. with a set-up analogous to that of Young’s experiment with light. Ch.7) 2π which is just Bohr’s postulate. In experiments with light 1 Free on-line edition: http://www. and in Feynmann. AddisonWesley (1962)..3 Discussion Just as for photons. In is interesting to observe that de Broglie assumption is. Krane. Chapter 46.1. A few words are in order to discuss the analogy between interference experiments with light and with electrons (see Fig. (4. The Feynmann Lectures on Physics. • atom interference.edu/ . Sands.caltech. • electron diffraction by powders (Davidson & Germer.feynmanlectures. Wave and corpuscular nature are complementary. l = mvr.1. • neutron interference. Physics II. in a sense.2 Experimental confirmation de Broglie hypothesis has been verified by a number of experiments. John Wiley & Sons(2002). If we associate a standing wave to an electron orbit or radius r we have (see Fig. 38. either particle behaviour alone or wave behaviour alone cannot explain all experimental evidence. consistent with Bohr’s postulated quantization of the electron angular momentum l. MATERIAL WAVES By imposing v = vg we obtain λ2 dE dE =− .5) and then −(h/λ2 )dλ = dp.1) is immediately obtained.6) 2πr = nλ = n mv or h l = mvr = n = n¯h. Leighton. (4. Resnick. 1927).1) h .g.2).1 4. 4. 4. Elementary descriptions of the above experiments can be found in Halliday. • electron interference (Tomonura 1989). dp h dλ (4. 4.

Figure 4. i. we have an analogy between the density of counts and the square of the wave amplitude. DE BROGLIE ASSOCIATED WAVE 21 Figure 4. This analogy somehow supports the physical interpretation of quantum mechanics wavefunctions. In experiments with particles.2: Interference patterns in experiments with light and with particles. we plot the number of counts. which is proportional to the square of the amplitude of the electric (and magnetic) field. the pattern shown in the figure represents the intensity of the electromagnetic wave.5.1.4.e. Therefore. which will be discussed in Section 5. the number of particles reaching the detecting screen at a certain position.1: Bohr’s electron orbit and standing de Broglie wavefunction. .

g. p ∆f = <f 2> − <f>2 . the principle concerns measurements of a coordinate (e.9) and analogously ∆y ∆py ≥ ∆z ∆pz ≥ ¯ h . px ). the localization of the wave-packet associated to a particle. . as such. defined as root mean square deviation of a sequence of measurements. (4. 2. 2 (4. whatever the instrumental accuracy. Heisenberg’s uncertainty principle then states that in any case. 2 In Appendix C.g. 4. cannot be proved. we proof the principle using solutions of the Schr¨ odinger equation.2.8) where the brackets denote average values. ∆x ∆px ≥ ¯ h . h bar. Such arguments lead to unequalities of the form ∆x ∆px ≈ h or ∆x ∆px ≈ ¯h. This requires having established (as a postulate) Schr¨ odinger equation itself.22 CHAPTER 4. In the following we indicate with ∆f the uncertainty in the measurement of quantity f . Uncertainties are unavoidable. x) and the homologous component of momentum (px ).1 Position and momentum In particular.11) where ¯h = h/2π is the reduced Planck constant or. the measurement of one quantity perturbs the status of the system.2 However it can be justified on the basis of arguments based on 1. Justifications The principle. induced by the measurement of the other quantity (x). 2 (4.10) (4. 2 ¯h .. simply. the perturbation of the measurement of a quantity (e.2 Heisenberg’s uncertainty principle The principle states that certain pairs of quantities cannot be measured simultaneously with arbitrary precision. so that the accuracy of the measurement of the conjugated quantity is limited. qualitatively in agreement with Heisenberg’s principle. MATERIAL WAVES 4.

hence ∆x = ∞ A wave packet of reduced spatial extension (as a packet representing a particle). instead has a broad spectrum. Uncertainties in a conceptual experiment Figure 4. (4. t) = Z −∞ −∞ g(k)ei(kx−ωt) dk.4. HEISENBERG’S UNCERTAINTY PRINCIPLE 23 1.2. (4. Wave packets and uncertainty Let us consider a plane harmonic wave ∼ ei(kx−ωt) .13) . Such a wave • has exactly defined wavenumber and wavelength: ∆k = 0 • occupies the whole space −∞ ≤ x ≤ ∞. Let us quantify the relation between spatial extension and spectral width. We can write such wave-packet as Ψ(x. 2.12) with wave number k and wavelength λ = 2π/k.3: Uncertainties in a conceptual experiments for the measurement of position and momentum.

t) (actually. for all processes concerning excited states. and then the energy of the emitted particle or photon is affected by an uncertainty ¯h ∆E = . t) is Gaussian in x (see Appendix B). with the equal sign applying when Ψ(x. Equation (4.2. Rather. If we apply this inequality to a de Broglie wave. we take the transform at t = 0) g(k) = 1 2π Z −∞ Ψ(x. MATERIAL WAVES Figure 4. we find ∆x ∆px ≥ ¯h. In principle such a time is the maximum length of the time interval during which we can perform the energy measurement. (4. (4. E and t are not two variables to be measured simultaneously. (4.17) τ This is the natural width of the emission line. 0)e−ikx dx. . more in general. with λ = h/p (and then k = 2πpx /h.24 CHAPTER 4.4: Wave-packet and its Fourier transform.16) In this case. with the k-spectrum g(k) given by the Fourier transform of Ψ(x. Eq.2 (4. (4. The energy of the excited level.16) is of great importance for all radioactive processes or. however. with mean life τ .16) concerns the uncertainty of a measurement of energy and the characteristic time of change of the status of the system being measured.14) −∞ One can show that the widths of the two functions Ψ(x. 0) and g(k) are related by ∆x ∆k ≥ 1. 4.15) Energy and time An uncertainty relation also applies to energy and time: ∆E ∆t ≥ ¯h.

We now outline the process leading to the establishment of the Schr¨odinger equation as the fundamental equation of non-relativistic quantum mechanics. 25 . For simplicity here we consider systems consisting of a single particle. namely • energy quantization. incorporating a few concepts/principles that have been introduced in the previous chapters. formal analogy between classical theory and quantum theory. Therefore the wavefunction will be a function Ψ(~r. 5. electron diffraction) a statistical interpretation is required. • the principle of complementarity (particle . In general.1 Wavefunction and general properties of its evolution equation • We postulate that a physical system is fully described by its wavefunction. E = hν = h ¯ ω. • following experiments on wave-like particles (e. t) of particle position ~r and time t.Chapter 5 Quantum Mechanics: Postulates and Schr¨ odinger equation Quantum mechanics is a coherent theory. • the principle of correspondence: classical theory as the large quantum number limit of quantum theory.wave duality. Ψ is a complex quantity. • the principle of superposition. p~ = h ¯~k).g.

Such a circumstance suggests us the expression of the operator of momentum. However. For instance. t) = pΨ(x. (5. once Ψ(~r. ∂x ¯ h ¯ h (5.4) We can rewrite this last equation as −i¯h ∂Ψ(x. E 2m (5. we consider a harmonic wavefunction in a one-dimensional space. (5.2 Wave equation for a free particle As a first step in the construction of an appropriate equation. QUANTUM MECHANICS: SCHRODINGER EQUATION • The equation describing the time-evolution of Ψ must be linear and homogeneous (for the superposition) and must be a differential equation of first order in t to guarantee the uniqueness of the solution for t > t0 . then 2 ˆ = pˆ . ∂x (5. we consider a wavefunction describing a free particle. if we indicate with p~ˆ and E of momentum p~ and kinetic energy E. t) = ei(kx−ωt) . • For the principle of correspondence. the relations between quantum operators should be identical to the classical relations between the correspondˆ the operators ing observables. anyhow. t). • We postulate that we can associate a linear operator Aˆ acting on the wavefunction to any measurable physical quantity a (an observable). respectively.6) . We can therefore also write Ψ(x.¨ 26 CHAPTER 5.e. and hence on superposition).3) Differentiating this expression with respect to x we obtain p p ∂Ψ(x. in three-dimensional space. Ψ(x. ∂x (5. on linearity. We postulate pˆx = −i¯h ∂ .1) since classically E = p2 /2m.2) representing a free particle of momentum p = h ¯ k and energy E = h ¯ ω. t) = ei(px−Et)/¯h . i. In general. t). we should consider a wave packet of arbitrary shape. t) = i ei(px−Et)/¯h = i Ψ(x. to make treatment simpler (and relying. t0 ) is known. 5.5) ∂ which states that applying the operator −i¯h ∂x to Ψ we obtain the product of p times Ψ itself.

(5. We kinetic energy operator replaced by the total energy operator H also postulate that the operator of the spatial coordinate x. ruling the evolution of the wavefuntion Ψ = Ψ(x. but with the ˆ = Eˆ + Vˆ .7) and Eq. t) with respect to time we have ∂Ψ(x. more compactly. particles are subjected to (conservative) forces. We then postulate that the evolution equation takes just the same form as Eq. In addition to kinetic energy we have to consider potential energy V (x). ∂t (5. i¯ h ∂Ψ ˆ = HΨ(x. (5. t). t). t). = − 2 2 Ψ(x.7) On the other hand. ∂t h ¯ ¯h (5.10) Equation (5. t) h2 ∂ 2 i¯ h + V (x) Ψ(x. for the correspondence principle and the definition of the momentum operator. (5.11) ∂t 5. We differentiate Eq.8) for the evolution of the wavefunction of a free particle can then be written as ∂Ψ(x.) We now look for the evolution equation for the wavefunction of the considered free particle.13) or. (5.8) we obtain i¯ h ¯h2 ∂ 2 Ψ(x.12) x) = − 2m 2m ∂x2 The time-dependent Schr¨ odinger equation. t) = − ∂x2 h ¯ ¯h ¯h2 (5. = − ∂t 2m ∂x2 (5. by differentiating Ψ(x. t) .3 Time-dependent Schr¨ odinger equation In real problems. We then write 2 h2 ∂ 2 ˆ = pˆx + V (ˆ H + V (x). TIME-DEPENDENT SCHRODINGER EQUATION 27 (We have added the index x because this is the operator of the x-component of the momentum. t) e = − Ψ(x. (5. is x ˆ = x. t). t) ˆ i¯ h = EΨ(x.4) again with respect to x: p2 2mE p2 i(px−Et)/¯h ∂ 2 Ψ(x. t).8) By comparing Eq. 2 ¯h2 ∂ 2 ˆ = pˆx = 1 pˆx (ˆ . (5.9) We now observe that. px ) = − E 2m 2m 2m ∂x2 (5.11). t) is then   ∂Ψ(x.14) . t) ∂Ψ(x. t) E E = −i ei(px−Et)/¯h = −i Ψ(x. =− ∂t 2m ∂x2 (5.3.¨ 5.

t) = ψ(x)e−iωt = ψ(x)e−iEt/¯h . t) ψ(x) 5.14) we obtain   E −iEt/¯ h ˆ e−iEt/¯h = Hψ(x)e . .15) Inserting this expression into the time-dependent Schr¨odinger equation (5. y. (5. i.18) are two equivalent forms of the 1-D time-independent Schr¨ odinger equation. y. by writing Ψ(x. z) = ψ(~r) (5.5 → ψ(x. we have discussed the equation that rules the behaviour of the wavefunction. by the following replacements   ∂ ∂ ∂ ∂ → ∇= .17) and (5.   h2 ∂ 2 + V (x) ψ(x) = Eψ(x). writing the operator and the dependent variable explicitly. Both time-dependent and time-independent Schr¨odinger equation are easily extended to three dimension. . Hψ (5.18) Equations (5. z.4 Time-independent Schr¨ odinger equation For a particle in a stationary state. angular frequency ω = E/¯ h. t) has a well defined.19) (5. QUANTUM MECHANICS: SCHRODINGER EQUATION 5.16) i¯ hψ(x) −i ¯h and then ˆ = Eψ. (5. − 2m ∂x2 (5. ∂x2 ∂x2 ∂y 2 ∂z 2 Ψ(x. ∂x ∂x ∂y ∂z   ∂2 ∂2 ∂2 ∂2 2 → ∇ = . We can then factorize the wavefunction as the product of the spatial part and of the time-dependent part.21) (5.20) (5. t) → Ψ(x. t) = Ψ(~r.22) Interpretation of the wavefunction So far.17) or. the wavefunction Ψ(x.e. but we have not yet assigned any meaning to the wavefunction. when the energy E is time-independent.¨ 28 CHAPTER 5. . time independent.

z)|. y.23) is proportional to the probability density of the particle at time t. the wavefunction ψ(x) = Aeikx represents a parallel beam of particles of mass m.e.z) is p(x. i.24) independent of time. z. i. t) = |Ψ(x. Such a factor is determined as follows. t)| (5. and hence current J = nv (actually.6 Wavefunction normalization Solution of the Schr¨odinger equation does not allow to determine a multiplying front factor. t)| (5. particles per unit surface per unit time)1 r mJ |A| = . the quantum mechanics probability current is i¯ h J~ = − (ψ∗ ∇ψ − ψ∇ψ∗ ) 2m (5. whole space When the wavefunction is normalized the probability density p(P ) at point P(x. z)|dxdydz = 1. 5.25) |ψ(x.29 5. z. and v = p/m = h ¯ k/m.28) ¯hk as obtained by observing that n = |ψ ∗ ψ| = A2 . z)∗ ψ(x. WAVEFUNCTION NORMALIZATION An interpretative postulate tells us that the quantity |Ψ(x. y. y. In the following we will only consider time-independent problems.y. instead.y. z. (5. z)∗ ψ(x. z.e. t)∗ Ψ(x. y. y. y. t)∗ Ψ(x. When the wavefunction represents a particle the probability to find the particle everywhere in space must be equal to 1. |Ψ∗ Ψ|dxdydz is proportional to the probability of finding the particle in an infinitesimal volume dxdydz around (x. (5. corresponding to physical eigenstates] for certain 1 In general.7 Eigenvalues and eigenfunctions The time-independent Schr¨odinger equation is an eigenvalue equation: it admits solutions [eigenfunctions ψ(~r). Therefore the wavefunction is normalized by requiring ZZZ (5. 5.26) If. current density. y.z) at time t. In such cases the probability density is |ψ(x. z.6. with density n and velocity v.27) . y. y.

i. particles in finite size systems (such as electrons in atoms and nucleons in nuclei) are characterized by a discrete spectrum. ˆ which cannot be measured simultaneously do not commute. e..g. ∂y (5.e. x and px . then ZZZ ˆ <A>= Ψ∗ AΨdxdydz.33) The operators of all other quantities are constructed using the correspondence principle.31) Operators and commutators We have already encountered the definitions of the operators of the x-components of position and momentum. cannot be measured simultaneously with arbitrary precision. Aˆ and B ˆ Bψ) ˆ 6= B( ˆ Aψ) ˆ A( . One can show that the result of a measument of an observable A can only be an eigenvalue a. of course. analogous. Typically.30) Ψ∗ Ψdxdydz If the wavefunction is normalized.32) ∂ −i¯h . The definitions of the other components are. One can show that the operators of two observables.g.8 (5. the average value of a series of measurements is given by ZZZ ˆ Ψ∗ AΨdxdydz Z Z Z <A>= . If it is not in an eigenstate. Aψ(~ (5.29) If a system is in an eigenstate ψn than the result of the measurement will be the corresponding eigenvalue an . QUANTUM MECHANICS: SCHRODINGER EQUATION values (eigenvalues) of E. We have already followed such a principle to write the kinetic energy and total energy operators. (5. 5.34) lz = (~r × p~)z = xˆpˆy − yˆpˆx = −i¯h x ∂y ∂x We have seen that according to the Heisenberg’s uncertainty principle certain pairs of variables.¨ 30 CHAPTER 5. Another example is given by the operator of the z-component of the angular momentum:   ∂ ∂ ˆ ˆ ˆ −y . e. Whether the eigenvalues have a continous or a discrete spectrum depends on the form of the potential and on the boundary conditions. satisfying the eigenvalue equation ˆ r ) = aψ(~r). such as. (5. yˆ = pˆy = y (5.

ˆlz ] = i¯hˆlx .37) It is therefore possible to measure simultaneously energy and position. it is not possible to measure simultaneously two or three components of the angular momentum (unless they are all zero).35) In general. (5. [A. pˆx ] = i¯h. energy and the modulus of the angular momentum.8. .36) (5. [lˆ2 . energy and single component of the angolar momentum. ˆ B] ˆ = AˆB ˆ −B ˆ A. Instead. [ˆlz . ˆlx ] = i¯hˆly . where [A. ˆly ] = i¯ hˆlz . H] x. H] while [ˆlx . OPERATORS AND COMMUTATORS or.31 5. H] ˆ = [ˆ ˆ = 0. (5. [ˆly . in quantum mechanics one can measure simultaneously only observables with commuting operators. ˆlz ] = [lˆ2 . [ˆ x. One can easily verify that ˆ = [ˆlz . in compact notation ˆ B] ˆ 6= 0. ˆ For instance.

QUANTUM MECHANICS: SCHRODINGER EQUATION .¨ 32 CHAPTER 5.

the wavefunction has to be normalized as discussed in Sec. (5. the solutions of Eq. appearing in Eq. In such cases Schr¨odinger equation becomes ∂ 2 ψ(x) 2m + (E − V )ψ(x) = 0. E < V.6. because otherwise ψ ′′ . • ψ(x) must be continuous. 33 . because the probability density must be continuous. (6.1) In the portion of the x domain where the potential energy takes the constant value V . with r 2m |E − V |. In the next sections we shall use the above rules to deal with a few simple problems of great physical interest. (6. We will only consider cases in which the potential energy V (x) is piecewise constant.1) take the form  A sin kx + B cos kx. because |ψ ∗ ψ| ∝ probability density. ∂x2 ¯ h (6. • ψ = 0 where V = ∞.2) A′ ekx + B ′ e−kx . • ψ ′ (x) = dψ(x)/dx must be continuous.2) are determined according to the following rules: k= • ψ(x) must be everywhere finite.18). (6. would be undefined. ψ(x) = (6.3) ¯h2 The values of the constants appearing in Eq. 5. E > V. because the product V ψ must be finite everywhere. In addition.Chapter 6 Quantum Mechanics: One-dimensional problems In this Chapter we will consider (time-independent) problems in one spatial dimension.

5) at x = 0 we have A sin 0+B cos 0 = 0.1) becomes ∂ 2 ψ(x) 2m + 2 Eψ(x) = 0. 0 ≤ x ≤ L.4) to be solved with the boundary conditions ψ(0) = ψ(L) = 0. It follows that k can only take values π kn = n .5) The general solution of Eq. Inside the well.7) By imposing the boundary condition (6. . hence B = 0. 2. QUANTUM MECHANICS: 1D PROBLEMS Infinite potential well Figure 6. ∂x2 ¯ h (6. kL = nπ. . i.8) which admits non-trivial solutions if kL is a integral multiple of π. Schr¨oedinger equation (6. with n = 1. The condition at x = L reads A sin kL = 0. As a first problem we consider the so-called infinite well. with k= r 2m E.1 CHAPTER 6.6) (6. ¯h2 (6.34 6.9) . (6. . (6.. (6.e. L (6. in which potential energy is zero in a region 0 ≤ x ≤ L and is infinite elsewhere. 0 ≤ x ≤ L.4) is ψ(x) = A sin kx + B cos kx.1: Infinite potential well of width L.

(6. 6. 1= Z 0 Using we obtain ∞ |ψ ∗ (x)ψ(x)|dx = Z nπ 0 Z 0 L A2n sin2 (kn x)dx = A2n sin2 ydy = n 2 Z 2π 0 r  L nπ π sin2 ydy = n . then the probability of finding the particle in an interval ∆x of finite size tends to a constant value. We then have nucleon energies of the order of MeV and me r02 Enucleon ≈ ≈ 6 × 105 . .1. 2m 2mL2 (6. m = mp = 1836me and L of the order of the nuclear radius rn ≈ 3 × 10−5 r0 . For an electron in an atom we set m = me and L of the order of the Bohr radius r0 . Eelectron mp rn2 (6. The shapes of the wavefunctions for n = 1–4 are shown in the left column of Fig. L (6. It is apparent that the probability density is not uniform.e.14) 2 .13) (6. An = .) The above result on the energy eigenvalues can be used for a rough estimate of atomic and nuclear energies.. However. 2. for a nucleon in a nucleus.2. x + ∆x] is therefore not uniform. (6. L].35 6. (6.11) Note that such eigenfunctions are just the standing waves on the interval [0. as in the classical case. The probability of finding the particle is an interval [x. INFINITE POTENTIAL WELL with quantum number n = 1. . (In fact the previous energy eigenvalues could have been simply found by imposing that the wavefunctions are standing waves with wavelength λn = 2L/n and with momentum given by the de Broglie relation. 2 Z 0 nπ sin2 ydy.12) The normalization constant An in Eq. The corresponding values of energy are En = π 2 ¯h2 ¯ 2 k2 h = n2 . i.10) and the corresponding wavefunctions (eigenfunctions) are ψn (x) = An sin nπ x. The figure also shows the corresponding shapes of the probability density.11) is determined by imposing the normalization condition (5.15) L independent on n.25). . as the quantum number n tends to infinite.

36 CHAPTER 6. QUANTUM MECHANICS: 1D PROBLEMS Figure 6. .2: Infinite potential well: wavefunctions and probability density for the first four quantum numbers.

symmetric potential well. In regions 1 and 3.20) (6. Schr¨odinger equation can be written in the form ∂ 2 ψ(x) − χ2 ψ(x) = 0.3: Finite.17) (6.18) and has solutions of the form ψ(x) ∼ e±χx (6.3.  x≤0 region 1  U1 0 0 < x < L region 2 U (x) = (6.19) Analogously. the relevant equation becomes ∂ 2 ψ(x) + k 2 ψ(x) = 0. We only consider the case E < U1 . namely. ¯h2 (6.2 Finite potential well We now refer to the symmetric potential well illustrated in Fig. 6.37 6. FINITE POTENTIAL WELL 6.21) and has solutions of the form ψ(x) ∼ sin(kx + δ).2. ∂x2 with χ= s 2m(U1 − E) .22) . (6.16)  U1 x≥0 region 3 Figure 6. In region 2. ∂x2 with k= r 2mE . ¯h2 (6.

32) . we impose continuity of the wavefunction and of its first derivative in x = 0 and in x = L: ψ1 (0) ψ1′ (0) ψ2 (L) ψ2′ (L) = = = = ψ2 (0) ψ2′ (0) ψ3 (L) ψ3′ (L) → → → → A1 A1 χ A2 sin(kL + δ) A2 k cos(kL + δ) = = = = A2 sin δ.24). (6. from the third and fourth equation (6. and then δ = −kL − arctan k + m2 π. and setting m2 − m1 = n. A3 e−χL.24) From the first two equations we obtain tan δ = k/χ. χ (6. By equating the right hand sides of Eq.23) ψ(x) =  ψ3 (x) = A3 e−χx x ≥ L. A2 k cos δ.25) where m1 is an integer number.29) ξ = U1 so that k = χ = with 0 < ξ < 1. We then have Kξ. χ (6.31) (6.26). tan(kL + δ) = −k/χ.27) with n an integer number. −A3 χe−χL . Analogously.30) (6. or δ = arctan k + m1 π.25) and Eq. (6. p K 1 − ξ2 . It is now convenient to express k and χ as functions of two new quantities. QUANTUM MECHANICS: 1D PROBLEMS By imposing that ψ(x) be finite for any value of x we can write the general solution as  x≤0  ψ1 (x) = A1 eχx ψ2 (x) = A2 sin(kx + δ) 0 ≤ x ≤ L (6. k ξ .26) with m2 integer. χ (6. = p χ 1 − ξ2 (6.28) K = ¯h2 r E . K and ξ r 2mU1 . (6. (6. (6. Next.38 CHAPTER 6. we can write nπ − kL = 2 arctan k .

n = 0. nmax = ⌊ KL + 1⌋. (n = 1. In this case K = 8. and Eq.2. 6.13ξ = 2 arcsin ξ. 1. . nmax (6. obtained as follows. FINITE POTENTIAL WELL Figure 6. instead. the unnormalized eigenfunction corresponding to the eigenvalue n (and then to ξn ) is obtained by . 6.13×1010 m−1 . but the number of bound states is finite.1 nm.13 and nmax = ⌊ KL π +1⌋ = 3. . 1. .e. 3). .34) with corresponding energy eigenvalues En = ξn2 U1 . we consider an electron in a well with U1 = 250 eV and L = 0. 2. n = 0.27) becomes ξ nπ − KξL = 2 arctan p = 2 arcsin ξ.4: Graphical solution of Eq. .4) that 2 arcsin ξ is an increasing function of ξ.33) We observe (see Fig.33). First.5 or by a straightforward numerical solution. π (6. As an example. 1 − ξ2 (6. . The figure also shows the three eigenfunctions ψn (x). (6. is a decreasing function of ξ. the quantity nπ − KξL. i.39 6. . .35) Energy levels are quantized just as in the case of an infinite well. KL = 8. which are easily found by the graphic construction shown in Fig. with maximum value ξmax = 1. The three values of ξ are the solutions of the equation nπ − 8. It follows that there are solutions ξn = ξ(n) for integers n such that nπ − ξmax KL ≤ π. (6.

We have δn A2.089. and therefore the probability density is not zero. (6.n is determined) by imposing the normalization condition Z ∞ |ψn∗ (x)ψn (x)|dx. 0. (6. the larger is the probability of finding the particle outside the potential well. The greater the eigenvalue. is apparent from Eqs.n = ξn3 + (1 − ξn3 1 ξn2 )1/2 2 KLξn + 1 4 sin(KLξn ) .n = A2.36) (6.n = 0. (6. respectively. 0.n / sin δn ] sin(kn x + δn ) ψn (x) =  [A1.38) and then   A1. The particle can therefore be found outside the well.0224.23) and (6.n eχn L sin(kn L + δn ) (6.42) In the case of Fig. . 3.6)we obtain pout.n sin(kn L + δn )/ sin δn ] e−χn (x−L) x≤0 0≤x≤L x≥L (6.n / sin δn (6.40 CHAPTER 6. QUANTUM MECHANICS: 1D PROBLEMS using the continuity conditions (6.41) −∞ With lengthy but simple algebra. (6. 6. in contrast with classical physics.39): in regions 1 and 3.289.n = = arcsin ξn A1. the greater the particle energy. with a probability Z 0 pout = 2 |ψn∗ (x)ψn (x)|dx. 2. where the energy of the particle is smaller than the potential energy U1 the wavefunction is different from zero. each eigenfunction is normalized (and the coefficient A1.40) −∞ A new feature. one finds pout.37) A3.n eχn x [A1.24). for n = 1.39) Finally.

FINITE POTENTIAL WELL 41 Figure 6.2. wavefunctions. quantized energy levels.5: Example of a finite potential well: eigenvalues. .6.

6.6: Probability densities for the three bound states of Fig. .5.42 CHAPTER 6. QUANTUM MECHANICS: 1D PROBLEMS Figure 6.

since. We assume that particles approach the step coming from the left. The solutions in each region will be oscillating if E > V . We now consider separately the cases E > V0 and E < V0 . particles coming from the left hand side) are represented by progressive plane waves Ψ(x.3. we are now dealing with beam of particles. exponential if E < V . and are reflected. concerning particles inside a well.7: Potential step. In the quantum case. we have to solve Schr¨odinger equation ∂ 2 ψ(x) 2m[E − V ] + ψ(x) = 0 ∂x2 ¯h2 (6.43) in the two regions x < 0 (a). as shown in Fig. Particles withe energy E < V0 . and x > 0 (b) and to properly match the solutions. Classically particles with energy E = (1/2)mv 2 > V0 will proceed to the right hand side of the step. (6. 6. we are simply interested in the spatial part.43 6. instead do not have sufficient energy to overcome the step. ∝ e−ikx . t) = const × ei(kx−ωt) . So. Notice that. incident particles (i.44) (6. and then we write incident wave : ψ(x) reflected wave : ψ(x) ∝ eikx . In particular.e. for energy conservation (1/2)mv12 = E − V0 . We now consider a potential step. unlike the previous problems.3 Potential step Figure 6. classically we have total transmission for E > V0 and total reflection for E < V0 . describing particles which move everywhere freely. This is of course an idealized situation. Since we are only considering timeindependent problems. POTENTIAL STEP 6.7.45) . and are subjected to a force (F = −dU/dx) directed toward the left only near x = 0. with lower velocity v1 .

8: Potential step. there cannot be a wave coming from the region x > 0.1 E > V0 In this case the solution is oscillating in both regions (a) and (b).44 CHAPTER 6.51) (6. Note. case E > V0 .3. The general solution has the form ψ(x) = ψa (x) ψ(x) = ψb (x) = = Aeika x + Be−ika x . that the classical limit is approached as E ≫ V0 . E (6. The other three constants are related by the matching conditions at x = 0: ψa (0) = ψa′ (0) = ψb (0) ψb′ (0) → → A+B (A − B)ka = C. in this limit kb /ka → 1. Indeed.50) and then B A C A 1 − kb /ka ka − kb = .47) with ka kb = = r s 2mE ¯h2 (6.49) Since we are considering a wave-particle coming from x < 0. the incoming wave is partially reflected (B 6= 0) even when E > V0 : transmission is incomplete. and then B → 0 and C → 1.48) 2m(E − V0 ) = ka ¯h2 r 1− V0 . = 1+ = A 1 + kb /ka = (6. ka + kb 1 + kb /ka B 2 . 6. . Ceikb x + De−ikb x . QUANTUM MECHANICS: 1D PROBLEMS Figure 6. and then D = 0.46) (6. (6.52) Surprisingly (at least if we were expecting the classical result). however. x<0 x>0 (6. = Ckb .

One may also be surprised that C/A > 1. and the relevant physical quantity in our problems are particle currents. ka (6. Asymptotically. . are Ja Jc = A2 ¯hka /m. we have total reflection! Note that this highly idealized problem qualitatively describes the scattering of a low energy neutron by a nucleus. again with a surprising result. e. it approaches 1 in the classical limit. According to Eq. 6.57) Contrary to classical physics. In this case. according to Eqs.51 and (6.3. a slow particle approaching a strongly attractive potential. the amplitude of the transmitted wave exceeds that of the incident wave. = C2h ¯ kb /m. we have to recall that the wavefunctions in this case represent free particles. (5. Let us consider a negative potential energy V0 < 0 and incident particles with E ≪ |V0 | (see Fig. Incoming and transmitted wavefunctions.53) (6.56) (6.e.54) respectively.9: Potential step. (6.45 6.52)) we have B C → −1 → 0 (6. incident and transmitted currents. ka ≪ kb and then. The ratio of transmitted to incident current is then given by 4kb /ka Jc =  2 . i. i.10). Another limiting case is of particular interest. (6.28).55) which is always smaller than 1. However. case E > V0 . Ja kb 1+ ka 0< kb < 1. POTENTIAL STEP Figure 6.

since now E − V0 < 0 the general solution in region (b) is a combination of exponentials. ¯h2 (6. we impose the conditions of continuity of ψ and ψ ′ at x = 0: ψa (0) = ψa′ (0) = ψb (0) ψb′ (0) → → A+B i(A − B)ka = C.46 CHAPTER 6. We can write ψ(x) = ψa (x) ψ(x) = ψb (x) = Aeika x + Be−ika x . (6.10: Potential step.58) (6.61) The coefficient D = 0 because ψ(x) must be everywhere finite.11: Potential step: case E < V0 6.59) with ka χ = s 2mE . However. As usual.3. . QUANTUM MECHANICS: 1D PROBLEMS Figure 6. case E > V0 . with V0 < 0 and E ≪ −V0 Figure 6.2 E < V0 We proceed just as in the previous case. = Ce = r −χx + De χx . ¯h2 (6. = −Cχ.60) 2m(V0 − E) .62) . x<0 x>0 (6.

Indeed |ψ ∗ (x)ψ(x)| |Ce−χx |2 P (x > 0) = e−2χx . Of course. from which we obtain B A C A 1 − iχ/ka . or penetrability of the barrier.63) (6. The case E < V0 is dealt with in Appendix F. (1999). the classical limits.12: Wavefunction for the potential step. ..g. inside the barrier and at the right hand side of the barrier. = 1+ = A 1 + iχ/ka = (6. Quantum Mechanics. and can then propagate freely again in the region at the right hand side of the barrier. e. (1966) or Quantum Mechanics . when E < V0 . Dover Publ. The rigorous calculation of the probability of tunneling T .1 However.65) Potential barrier and tunnel effect We now consider a potential barrier of uniform height V and thickness L. In quantum mechanics jargon we say the particle tunnels the barrier..6.4 (6. According to the results of the previous section we expect that a particle coming from the left can be found at x = L even if E < V . of full transparency and full reflection are recovered as E ≫ V0 and E ≪ V0 respectively. = = P (x = 0) |ψ ∗ (0)ψ(0)| |A + B|2 6.64) We see that there is a finite probability P of finding the particle at x > 0 in the classically forbidded region (b). The relevant solution also shows that even when the energy E is slighly greater than V0 there is a finite probability of reflection. POTENTIAL BARRIER AND TUNNEL EFFECT 47 Figure 6.4. Wiley. Messiah. matching the general solutions obtained in the regions on the left of the barrier. The full treatment can be found in several textbooks. requires solving the Schr¨odinger equation.Two volumes bound as one. 1 + iχ/ka B 2 . when χL ≫ 1 [with 1 see.

an accurate approximation is simply obtained by the ratio of the probability densities found for the potential step in the previous section at x = L and x = 0. # "  . QUANTUM MECHANICS: 1D PROBLEMS Figure 6.13: Potential barrier. respectively.61)). χ defined by Eq. (6.48 CHAPTER 6. with sketch of the wavefunction.

1/2 2m P (x = L) .

.

1/2 (V0 − E) L (6.66) = exp −2 T ≃ P (x = 0) .

In practical cases the potential is varies in space. The tunnel effect plays a central role in alpha decay and fusion reactions. theory shows that a good approximation of the barrier penetrability is obtained by still using Eq. can be found in Wikipedia: http://en. and replacing the argument of the exponential with an integral. with both historical and pedagogical references. V = V (r).wikipedia. 2 This is the so-called WKB approximation. 6.65. and is at the basis of devices such as tunnel diode and Scanning Tunneling Microscope. In this case. The probability that an alpha-particle with energy E = 5 MeV is transmitted through the barrier is e−85.5 × 10−14 m. A nice presentation. discussed in many books on mathematical physics or on quantum mechanics. let us consider a barrier of V0 = 20 MeV and width L = 2. according to2 −2  2m h2 ¯ 1/2 (V0 − E)1/2 L → −2  2m ¯h2 1/2 Z x2 x1 [V (x) − E]1/2 dx (6.3 ≈ ×10−38.step ¯h2 As an example.org/wiki/WKB approximation .67) where x1 and x2 are lower and upper bound of the region to be tunnelled.

θ. is in the form V (r). φ) is also convenient to deal with angular momentum.1 Operators in spherical coordinates Atomic physics problems are most easily dealt with by using spherical coordinate systems. We recall (see Figure 7. φ) are related to Cartesian coordinates 49 . as is the case of Coulomb potential. with r the distance from the centre of symmetry. i.Chapter 7 Elementary atomic physics 7.1) that spherical coordinates (r.e. θ. 7. The use of spherical coordinates (r.1: Sperical and Cartesian coordinates. Fig. Such coordinates are indeed convenient when the potential has a central symmetry.

50 CHAPTER 7. θ.9) as shown in Appendix D. (7. = r cos θ.6) and writing the derivatives using chain rules as ∂ ∂r ∂ ∂θ ∂ ∂φ ∂ = + + .4) y z = r sin θ sin φ.6) = (7. Expressions of the angular momentum operators in spherical coordinates are derived in Appendix D.7) Before solving Schr¨odinger equation with a Coulomb potential. φ) = lz ψ(r. The possible values of the z-component of the angular momentum are obtained by solving the operator equation ˆlz ψ(r. (7.4)-(7.2) x = r sin θ cos φ. p x2 + y 2 = arctan . ELEMENTARY ATOMIC PHYSICS (x. y.3) and The expressions of the operators in spherical coordinates are obtained from those in Cartesian coordinates by using Eqs.11) ψ(φ) = const ei(lz /¯h)φ . θ. z) by r θ φ p x2 + y 2 + z 2 .12) −i¯h which has solutions . x (7. we consider angular momentum. ∂φ (7. ∂φ (7.5) (7. z y = arctan .8) Its expression in spherical coordinates is ˆlz = −i¯h ∂ .2 Angular momentum. 7. or (7. (7. z-component Let us consider the operator of the z−component of angular momentum ∂ ∂ ˆ − y ). ∂z ∂z ∂r ∂z ∂θ ∂z ∂φ (7.1) (7. φ). (7. lz = (~rˆ × ~pˆ)z = x ˆ × pˆy = −i¯h(x ∂y ∂x (7.10) ∂ψ(φ) ) = lz ψ(φ).

The corresponding eigenfunctions are ψ(φ) = const eimφ . (7.3 (7. Θ(θ) ∂θ ∂θ Φ(φ) ∂φ2 | {z } | {z } function of θ only (7. . φ). we can measure one component (e.g. m = 0. (7. (7. while the right hand side only on φ.18) by separation of variables. (7. TOTAL ANGULAR MOMENTUM Imposing the the boundary condition ψ(φ) = ψ(φ + nπ). 5.8 that we cannot measure two components of the angular momentum simultaneously. we factorize the wavefunction as the product of a function of θ and a function of φ Y (θ. (7. Instead.18) and multiply by − sin2 θ/(Θ(θ)Φ(φ)). φ) = l2 Y (θ. ±1. . It can take the values lz = h ¯ m.13) we find that lz /¯ h must be an integer. Since the equality holds for any value of θ and φ. (7. in spherical coordinates. which has operator lˆ2 = lˆx2 + lˆz2 + lˆz2 .51 7. . this quantum number m is called magnetic quantum number. the . We also set l2 = h ¯ 2 β.18) with eigenvalues l2 and eigenfunctions Y .3.17) We then have to solve the eigenvalue equation lˆ2 Y (θ. (7.e. and obtain   sin θ ∂ ∂Θ(θ) 1 ∂ 2 Φ(φ) sin θ + β sin2 θ = − .20) where β is a dimensionless quantity. 7. . the z−component) and the square of the modulus.19) We substitute this last expression into Eq. φ) = Θ(θ)Φ(φ). i. (7.16) or.14) For historical reasons.15) Total angular momentum We have seen in Sec. We solve Eq.21) function of φ only We observe that the left hand side of this equation depends only on θ. Therefore the z-component of the angular momentum is quantized. ±2.     ∂ 1 ∂2 1 ∂ sin θ + lˆ2 = −¯ h2 sin θ ∂θ ∂θ sin2 θ ∂φ2 (7.

7.23) by Θ(θ)/ sin2 θ we have   1 d dΘ(θ) m2 Θ(θ) sin θ + βΘ(θ) − = 0. It is found that the solution P (w) is everywhere finite only for β = l(l + 1). (7. .22) = (m∗ )2 .1 .2. Multiplying Eq. we set w = cos θ and call Θ(θ) = P (w) and obtain      dP (w) m2 d + β− P (w) = 0. It can be found in most textbooks on quantum mechanics. (7.15)]. and then equal to the same constant.27) The solution of this equation is rather long and requires considerable manipulations. (7. 1.26) sin θ dθ dθ sin2 θ Next. 2.25) We now consider the equation for Θ. 1 − w2 dw dw 1 − w2 (7.30) 1 The solution procedure is analogous to that for the radial equation of the Hydrogen atom. (7. . ELEMENTARY ATOMIC PHYSICS two sides have both to be constant. .22). (7. which we will describe in detail later .52 CHAPTER 7. We can then write two separate ordinary differential equations 1 ∂ 2 Φ(φ) Φ(φ) ∂φ2   sin θ ∂ ∂Θ(θ) sin θ + β sin2 θ Θ(θ) ∂θ ∂θ − = (m∗ )2 . and the angular momentum is quantized: |~l| = h ¯ p l(l + 1) (7. (7. The angular momentum eigenfunctions are the associated Legendre polyno|m| mials Pl (w) defined by |m| Pl (w) 2 |m|/2 = (1 − w )  d dw |m| Pl (w). (7. (7. (7.24) This is just the same solution found in the previous section for the wavefunction of the z-component of the angular momentum [Eq.23) We first consider Eq. and then Φ(φ) = const eimφ (7.29) The relation between angular momentum and the possible values of its zcomponent are illustrated by the diagram of Fig. and l ≥ |m|. l = 0. Here we simply report and discuss the solution.28) The eigenvalues of lˆ2 are then h ¯ 2 l(l+1). which we call (m∗ )2 . with m the magnetic quantum number. which has solution ∗ Φ(φ) = const eim φ . We recognize that m∗ = m.

2: The angular momentum diagram for angular quantum number l = 3 where Pl is the Legendre Polynomial of order l.36) (7. Pl (w) = 1 2l l!  d dw l l w2 − 1 .3.34) P1 = cos θ 1 (3 cos2 θ − 1) = 2 1 = (5 cos3 θ − 3 cos θ) 2 (7. Plm (w) = Pl0 (w) = Pl (w). (7. φ) = Θl (θ)Φ(φ) = Al.53 7. The expressions of the first four Legendre polynmials are as follows P0 = 1 (7.37) . (7.32) with normalization constants Al.m Pl (cos θ)eimφ . φ) = Ylm (θ.m = il  2l + 1 (l − |m|)! 4π (l + |m|)! 1/2 ×  (−1)m 1 m>0 m≤0 (7. (7.33) Note that when m = 0. TOTAL ANGULAR MOMENTUM Figure 7.31) The normalized angular momentum wavefunctions turn out to be |m| Y (θ.35) P2 P3 The l = 0 eigenfunction is therefore spherically symmetric.

(7. of beta-deacy. φ) + 2µ [E − V (r)] ψ(r. [7.42) Schr¨ odinger equation for the hydrogen atom We now consider a Hydrogen atom.39) Then. φ′ ) = Ylm (θ′ . ¯h2 (7. φ′ ) = Pl ′ |m| (cos θ′ )eimφ = −Pl (cos θ)eimφ (−1)|m| (7. the polar coordinates change as r → r′ = r.g.4 CHAPTER 7. θ.32)] are polynomials of even powers of x (and then symmetric in x) if l − |m| is even. θ. or a 1-D potential for which V (x) = V (−x)] then the solutions (eigenfunctions) of the Schr¨odinger ~ ′ ). θ → θ′ = π − θ. . φ) = 0. (7. e. aside from normalization constants. (It is not necessarily conserved in the weak interactions responsible.5 (7. 7.e. a property which has no classical analogue.38) One can show that when the potential is not changed by such a reflection of coordinates [as in the case of a spherical potential V (r). (7. y → y ′ = −y.43) . Parity is an important property because the total parity of a system (i. φ′ ) = (−1)l Ylm (θ. equation are either symmetric ψ(~x) = ψ(~x′ ) or anti-symmetric ψ(~x) = −ψ(−x In the first case we say that the eigenfunction has positive parity (Π = 1). ELEMENTARY ATOMIC PHYSICS Angular momentum and parity Electron wavefunctions in atoms. Parity concerns coordinate reflection about the origin x → x′ = −x. It follows that Ylm (θ′ . are also characterized by their parity. (7. because it enters in the determination of the so-called selection rules. as well as nucleon wafefunctions in nuclei.e. the product of the parities of all its components) is conserved in electromagnetic and strong nuclear interactions.41) The parity of angular momentum eigenfunctions is then only determined by the angular quantum number l and is simply Πl = (−1)l .38). φ → φ′ = φ + π.54 7. in the second case it has negative parity (Π = −1).) It is useful to consider the parity of angular momentum wavefunctions. the motion an electron in the field of a proton. z → z ′ = −z. When the coordinates are reflected according to Eq. i. We determine the energy levels by solving Schr¨odinger equation ∇2 ψ(r. ruling possible (and forbidden) atomic transitions. φ).40) We also observe that the associate Legendre polynomials Plm (x) appearing in the angular momentum eigenfunctions Ylm (see Eq. while are polynomials of odd powers of x (and then anti-symmetric in x) if l − |m| is odd. |m| Y (θ′ .

in the previous section we have found that lˆ2 Y (θ.46) r2 ∂r ∂r r ¯ h ¯h Furthermore. (7.5.45) Here r is the distance of the electron from the nucleus. This suggests to factorize ψ(r. SCHRODINGER EQUATION FOR THE HYDROGEN ATOM 55 where ∇2 = 1 ∂ r2 ∂r       ∂ 1 ∂ 1 ∂2 1 ∂ r2 + 2 sin θ + ∂r r sin θ ∂θ ∂θ sin2 θ ∂φ2 and V (r) = − e2 . and multiplying by r2 /Y we have   ∂R(r) 2µr2 ∂ (7.43)] as 2 ( )   1 lˆ2 1 ∂ 2µ 2 ∂ r − 2 2 + 2 [E − V (r)] ψ(r.49) r2 + 2 [E − V (r)] R(r) − l(l + 1)R(r) = 0. and me and mp are the electron and proton mass. using the identity „ « 1 ∂ 1 ∂ 2 (rψ) 2 ∂ψ r = r 2 ∂r ∂r r ∂r 2 . φ) = h ¯ 2 l(l + 1)Y (θ. ∂r ∂r h ¯ Since we have obtained an equation in the single variable r we can replace partial derivatives with ordinary derivatives. φ). 4πε0 r (7. (7.48) where R(r) is a function depending on r only. µ = me mp /(me + mp ).¨ 7. (7. φ) = R(r)Y (θ.44) (7. respectively. (7.51) + + dr2 r dr r2 h2 ¯ 2 Note that ∇2 can be written in a slightly different form. the other terms only depend on the radial variable r. (7. (7. to obtain   2 l(l + 1) 2 d 2µ d [E − V (r)] − R(r) = 0 (7.47) in Eq. φ) = 0. µ is the reduced mass of the electron in the Hydrogen atom. we can write Eq.44) is just equal to −lˆ2 /¯ h2 [see Eq. (7. φ).48) and Eq.47) We also notice that while the angular momentum term depends on θ and φ. Substituting Eq. (7. To proceed further.49) and divide by r2 . θ. (7.50) + dr dr dr2 r dr We substitute this expression into Eq. θ. we note that  2    d d 2 d 2 d 2 r =r . and Y (θ. φ) is the momentum eigenfunction discussed in the previous section. (7.17)]. (7. Observing that the term in square brackets in Eq.46).

+ −ε+ − dρ2 ρ dρ ρ ρ2 (7. r0 = E0 = − 4πε0 ¯h2 . However.56) (7. e.3)] and E in units of first Bohr’s energy level E0 [Eq. µe2 1 µe4 .3 Veff (r) = V (r) + l(l + 1)¯ h2 2 2µr (7. we can easily study the limiting cases for ρ → ∞ and ρ → 0.57) (7. a combination of in-going and out-going spherical waves.59) dρ2 √ which has solutions R ∼ e±ρ ε . In the limit ρ → 0.51) becomes  2  d 2 d 2 l(l + 1) R(r) = 0. ELEMENTARY ATOMIC PHYSICS Note that this last equation can also be written as  2  2 d 2µ d + + 2 [E − Veff (r)] R(r) = 0 dr2 r dr ¯h (7.58) reduces to   2 d − ε R(r) = 0.60) (7. E = εE0 . we instead have  2  d 2 d l(l + 1) R(r) = 0. r = ρr0 .45) for the potential energy. + − dρ2 ρ dρ ρ2 (7. (7. where one can neglect both the Coulomb potential and the centrifugal potential.61) 3 It is also interesting to observe that at large distance from the nucleus. The solution with the plus sign has to be discarded because it diverges for large ρ.e.52) with an effective potential equal to the sum of the potential of the central force and a centrifugal potential due to the angular momentum.4)].53) ∂r 2 h ¯ with solutions e±ikr R(r) ∼ . (3. k = (2µE/¯ h)1/2 (7. As ρ → ∞ Eq. (7. respectively. Schr¨ odinger equation becomes 2 2µE ∂ (rR) + 2 rR. .55) We write r in units of Bohr radius r0 [Eq.54) r i.56 CHAPTER 7.58) which cannot be solved in closed form. Equation (7. (3. i. (7. (7. 2 (4πε0 )2 ¯h2 and use Eq. We are then left with √ R(ρ) ∼ e−ρ ε = R(∞) (ρ).

i. (7. but elementary algebra (detailed in Appendix E) leads to the equation    √ df √ 2 2(l + 1) d2 f + − 2 + 1 − ε(l + 1) f = 0 ε (7.65) i=0 and then substitute ∞ √ X ai xi . . ε (7. we find that the coefficients ai satisfy the recurrence relation √ i + l + 1 − (1/ ε) ai+1 = ai . SCHRODINGER EQUATION FOR THE HYDROGEN ATOM 57 We look for a solution of the form R ∼ ρb . Discarding the solution ρ−l−1 . Since the solution has to remain finite.e. . must be a positive integer n = 1. (the principal quantum number). We therefore search a solution.63) where f (ρ) is a function to be determined. valid everywhere. because ai+1 → ai /i. . We substitute Eq.5. which has solutions b = l and b = −l − 1. f (ρ) must be a polynomial of finite order σ. From Eq. Again with lengthy algebra (see Appendix E)). that diverges in the origin. which occurs if aσ+1 = 0.69) . and then f (ρ) ∼ eρ . we notice that the product of the solutions R(0) (ρ)R(∞) (ρ) still has the correct behaviour in both limits. better. (7.68) √ with σ and l non negative integers (σ = 0.66) i=0 into Eq. of x = 2 ερ. .67) (i + 1)(i + 2l + 1) The series so obtained diverges. 2. (7. (7. 2. 1. .62) Now. (7. R(ρ) = e−x/2 R(x/2 ε) (7. (7.64) 2 dρ ρ dρ ρ √ We assume f (ρ) can be written as a series of powers of ρ or. 2.58). . (7.) It follows that 1/ ε. in the form √ R(ρ) = R(0) (ρ)R(∞) (ρ)f (ρ) = ρl e−ρ ε f (ρ). n2 n = 1. f (ρ) = ∞ X ai xi .¨ 7. (7.63) for R in Eq.61) we have b(b + 1) = l(l + 1).67) we see that this requires 1 σ + l + 1 − √ = 0. (7. .58). the series must terminate. for large i. (7. . we are left with R(ρ) ∼ ρl = R(0) (ρ). The electron energy E = εE0 can then only take the discrete values E(n) = En = E0 . . By Substituting this last expression into Eq. Lengthy.

θ. and in absence of a magnetic field. which can now be written as σ + l + 1 − n = 0. . m) = (1. (7.4 From Eq. . n − 1. The energy levels of all other atoms also depend on the angular momentum. . . 0).72) For the fundamental state of the hydrogen atom. l. . . φ) = ψ(r) = An00 e−ρ = const e−r/r0 (7.3. 1. .0. . with a0 the Bohr’s radius. and then l = 0. . we have ψ(r. 0.73) Exercise: For the fundamental state of the Hydrogen atom (n. and the wavefunction is |m| Ψ(r. (7.0 = 1/ πa30 and rmax = a0 .. 1. ELEMENTARY ATOMIC PHYSICS in agreement with Bohr’s model. n i=0 (7. 2. [Answers: A1. 1. φ) = Ψnlm = Anlm Rnl (ρ)Pl (cos θ)eimφ . θ. (7. we also see that l = n + σ − 1. 7.70) The radial eigenfunction then is l −ρ/n Rnl (ρ) = ρ e σ X  i 2 ai ρi .71) with l = 0. 4 This is the case for Hydrogen only. . n = 1 (and then l = σ = m = 0).68).58 CHAPTER 7. find the normalization constant and the radius rmax at which the probability p density of finding the electron is maximum. . . Notice that the energy levels only depend on n. n − 1 and σ = n − l − 1 = 0.] The spatial distribution of the probability density of the (electron in the) hydrogen atom are shown in Fig.

org/wiki/Wave function .5. from http://en.wikipedia.3: False colour probability density plots of the Hydrogen wafenctions.¨ 7. SCHRODINGER EQUATION FOR THE HYDROGEN ATOM 59 Figure 7.

ELEMENTARY ATOMIC PHYSICS Spin and Pauli exclusion principle .60 7.6 CHAPTER 7.

SPIN AND PAULI EXCLUSION PRINCIPLE 61 .7.6.

7 CHAPTER 7.62 7. ELEMENTARY ATOMIC PHYSICS Electrons in the atom .

7.8 Indistinguishability 63 .8. INDISTINGUISHABILITY 7.

ELEMENTARY ATOMIC PHYSICS .64 CHAPTER 7.

tramite le propriet` a della configurazione pi` u probabile. che un sistema termodinamicos isolato tende a raggiungere. le fluttuazioni (relative alla configurazione pi` u probabile) sono estremamente piccole.5. con cui anzich`e determinare la distribuzione media. cio`e alle particelle con spin intero (fra cui i fotoni. Per quanto riguarda quest’ultima. Ci`o giustifica il procedimento che seguiremo. Si verifica comunque che. § XVII. che si applica ai fermioni. abbiamo visto (cfr. gli atomi di elio e le coppie di elettroni nei superconduttori). Mencuccini-Silvestrini. In condizioni di equilibrio termodinamico. che si applica ai bosoni. con ottima precisione. lo stato che corrisponde al valore massimo della probabilit` a termodinamica W .1) . Ricaveremo le espressioni della funzione di distribuzione di BoseEinstein. Sette-Alippi. determineremo la pi` u probabile. I. fluttuando in prossimit`a della configurazione pi` u probabile.12). 65 (8. definita come numero di microstati distinguibili con cui pu` o essere realizzato un dato stato macroscopico. all’equilibrio. poich´e i sistemi considerati sono costituiti da un grandissimo numero di particelle. Vedremo imoltre in quale limite la distribuzione (classica) di Boltzmann approssima le distribuzioni quantistiche. Le propriet` a fisiche medie di un sistema macroscopico si possono allora descrivere. I. la probabilit` a termodinamica `e legata all’entropia tramite la relazione S = kB ln W. cio`e alle particelle con spin semintero (fra cui gli elettroni). § 17. dove kB = 1. Vol.Chapter 8 Quantum statistics In questo capitolo descriviamo il procedimento per ottenere le funzioni di distribuzione di equilibrio per particelle che obbediscono alle leggi della meccanica quantistica.4 e XVII. vol. e della funzione di distribuzione di Fermi-Dirac.3807 × 10−23 J/K `e la costante di Boltzmann. Le configurazioni di un qualsiasi sistema in equilibrio macroscopico in effetti variano nel tempo.

Scopo finale `e ottenere il valore pi` u probabile della popolazione ns (gs ) di ciascuna shell e quindi la funzione di distribuzione f (εs ) = ns gs (8. e un elevato numero di particelle.1) raggruppamenti di livelli (o shell).2) s E X = ns ε s . Ai fini del calcolo della funzione di distribuzione. il procedimento fornisce quindi la funzione di distribuzione f (ε) per qualsiasi valore di ε. I passi successivi del procedimento consistono nella determinazione del peso e della probabilit` a termodinamica di ciascuna shell.1 CHAPTER 8. in numero non assegnato e non vincolato. con energia e quantit` a di moto legate alla frequenza tramite le relazioni ε = p = hν. 8. QUANTUM STATISTICS Descrizione del sistema Il primo passo nello studio del sistema di cui si vuole determinare la funzione di distribuzione consiste nella descrizione dei livelli energetici e della loro occupazione da parte delle particelle (particelle nel senso classico o fotoni) che costituiscono il sistema stesso. c (8. le energie dei vari stati della shell possono essere considerate coincidenti.66 8. 8. con εs l’energia di ciascuno degli stati e con ns il numero di particelle contenute nella shell.2 Statistica di Bose-Einstein In questa sezione descriviamo la statistica di Bose-Einstein. Osserviamo che i sistemi oggetto di studio sono costituiti da un grandissimo numero di particelle.5) (8. hν .2. in equilibrio alla temperatura T .1 Statistica di Bose-Einstein per i fotoni Consideriamo particelle (fotoni) indistinguibili. distribuite su un gran numero di livelli energetici (o stati).4) Si pu` o dimostrare (non lo facciamo qui) che il risultato ottenuto `e indipendente dal modo in cui si raccolgono i livelli nelle shell. Indichiamo con il pedice s la generica shell. con gs il suo numero di stati (o peso). piuttosto che considerare i singoli livelli. Inoltre. abbastanza grandi da contenere un elevato numero di stati.3) s dove le sommatorie sono estese a tutte le shell del sistema. conviene considerare (vedi la Fig. (8. (8. Il numero totale di particelle N e l’energia totale E del sistema possono allora essere espresse rispettivamente come X N = ns . Iniziamo considerando un sistema costituito da fotoni. 8.6) . con energie molto prossime fra loro.

5 e 21. Indichiamo gli stati con (8. Torino.9) z1 .15) 4πV 2 ν dν. New York. . . trad. `e dato da (Mencuccini-Silvestrini. II. inoltre. . Pesi . Vol. shell s z1 z2 z3 zg -1 s zg s stati Figure 8.1: Stati e raggruppamenti. Dover. dove h `e la costante di Planck e c `e la velocit` a della luce. XII. 1976. . Atomic Physics. Sette-Alippi. it. (8. Eq.Modi di vibrazione Per esprimere il peso gs della shell s sfruttiamo un risultato che abbiamo gi`a utilizzato nella trattazione classica del corpo nero. . Vol. § 16.: Fisica Atomica. 8th Ed. 7. Consideriamo la s-sima shell. Bollati Boringhieri. zgs 1 Seguiamo l’elegante e semplice presentazione di M.2. occupati da ns particelle. I.7) c3 Il peso gs della shell relativa a fotoni con energia compresa fra ε = hν e ε + dε = h(ν + dν) `e allora 8πV gs = 2N (ν)dν = 3 ν 2 dν (8. Cap. 1989. Born. contenente gs stati.. che ha peso gs e contiene ns particelle. STATISTICA DI BOSE-EINSTEIN energia ε + dε ε Raggruppamento (o shell). Abbiamo infatti visto che in una cavit`a di volume V il numero di modi di vibrazione indipendenti con frequenza compresa fra ν e ν + dν.1 Ricordiamo che per i postulati della meccanica quantistica i fotoni sono indistinguibili.8) c dN (ν) = N (ν)dν = Probabilit` a termodinamica Calcoliamo ora la probabilit` a termodinamica W .11. per ciascuna delle due possibili polarizzazioni delle onde elettromagnetiche. non c’`e limite al numero di fotoni che possono occupare un dato stato. z3 . 2a Ed..67 8. z2 .

zgs −1 zgs ans −1 ans rappresentiamo la configurazione della Fig. e pu` o quindi essere scelto in gs modi diversi. lo stato z3 le particelle a4 . lo stato z2 contiene la particella a3 .. . . le configurazioni ottenute l’una dall’altra per permutazione degli stati o per permutazione delle particelle sono indistinguibili. lo stato z4 non contiene nessuna particella. lo stato z5 contiene la particella a7 .12) Inoltre. Per esempio. (8. . . Una volta scelto il primo elemento. ans . QUANTUM STATISTICS e le particelle con a1 .11) z1 a1 a2 z2 a3 z3 a4 a5 a6 z4 z5 a7 . zgs-1 zgs ans-1 ans Figure 8. essendo per ipotesi ns ≫ 1 e gs ≫ 1..10) Per valutare il numero di configurazioni possibili.15) . ovvero W = Y s e quindi si ha ln W ≃ X s Ws ≃ Y (gs + ns )! s gs !ns ! . . Il primo elemento `e sempre una z. con (8. in cui lo stato z1 contiene le particelle a1 e a2 . . gli altri gs + ns − 1 possono essere ordinati in (gs + ns − 1)! modi diversi. .2: Esempio di configurazione di una shell e stringa che la rappresenta.13) Il numero di configurazioni possibili dell’intero sistema `e pari al prodotto del numero di configurazioni di tutte le shell. a3 .. (8. possiamo scrivere Ws ≃ (gs + ns )! . procediamo nel modo seguente. il numero di configurazioni distinguibili di una shell risulta Ws = gs (gs + ns − 1)! (gs + ns − 1)! = . Osserviamo che tutte le possibili stringhe a1 a2 z1 a3 a4 a5 a6 z2 z3 an a7 z4 z5 zg s-1 s-1 zg an s s z1 a1 a2 z2 a3 a4 a5 a6 z4 z5 a7 . . a2 . gs !ns ! (8.14) (8. che descrivono la shell considerata sono costituite da gs + ns elementi. per i postulati della meccanica quantistica. 8.68 CHAPTER 8. Poich´e il numero di queste permutazioni `e gs !ns !. lo stato zgs −1 non contiene nessuna particella e lo stato zgs contiene le particelle aan s−1 e ans .2. Tuttavia. Indichiamo una possibile configurazione con una stringa costituita da sequenze stato-particelle nello stato. a5 e a6 . . gs !ns ! (gs − 1)!ns ! (8. [ln(gs + ns )! − ln gs ! − ln ns !]. .

(8.20) Il massimo di F si ottiene imponendo che si annullino le derivate parziali rispetto a tutte le ns : ∂F = ln(gs + ns ) + 1 − 0 − ln ns − 1 − βεs = 0. X ns εs = E.16) s Configurazione pi` u probabile e funzione di distribuzione Cerchiamo ora la configurazione che rende massima la funzione ln W . Riscriviamo l’Eq.2.69 8.18) s A tale scopo usiamo il metodo dei moltiplicatori di Lagrange (vedi Appedice G). con il vincolo. imposto dalla conservazione dell’energia.16). (8.19) esprimendo ln W tramite l’Eq.23) β = 1/kB T La relazione fra il moltiplicatore di Lagrange β e i parametri macroscopici del sistema (in realt` a un solo parametro. ns 1 ns = βεs . (8.21) (8. come si vedr` a fra poco) si trova nel modo . Cerchiamo cio`e i massimi non condizionati della funzione X F = ln W + β(E − ns εs ). ∂ns da cui si ha ln e quindi g s + ns = βεs . ln n! ≃ n ln n − n. possiamo infine scrivere X ln W ≃ [(gs + ns ) ln(gs + ns ) − gs ln gs − ns ln ns ]. (8.17) s ovvero E− X ns εs = 0. (8. (8.22) (8. STATISTICA DI BOSE-EINSTEIN Usando la formula di Stirling.19) s dove β `e un moltiplicatore di Lagrange. −1 gs e (8. (8. ottenendo X X F = [(gs + ns ) ln(gs + ns ) − gs ln gs − ns ln ns ] + β(E − ns ε s ) s = X s s [(gs + ns ) ln(gs + ns ) − gs ln gs − ns ln ns − βns εs ] + βE.

(8.27) s D’altra parte. applicato a una trasformazione infinitesima reversibile (δE = δQ − δL). S = kB ln W . (8. il secondo addendo a primo membro dell’Eq. Differenziando la condizione X ns ε s = E (8. se i livelli energetici non dipendono dalla temperatura (come ragionevole pensando ai modi di vibrazione in una cavit`a). possiamo scrivere X δ ln W = β εs δns (8. (8. . per il primo principio della termodinamica.70 CHAPTER 8. si ha X s ns δεs = X s ns ∂εs δV = −pδV = −δL. B gs e −1 dove abbiamo scritto la frequenza ν omettendo il pedice s.31) = hν/k T . confrontando le equazioni (8.30) Introducendo quest’ultima relazione e la condizione di quantizzazione [Eq. quindi.24) s otteniamo X ns δεs + s X δns εs = δE.28) δ ln W = βδQ. QUANTUM STATISTICS seguente. si ha β= 1 .23) otteniamo la funzione di distribuzione di Bose-Eintein per i fotoni 1 ns (8. (8.5)] nell’Eq.25) rappresenta il calore δQ scambiato dal sistema: X εs δns = δQ (8. ∂V (8.19). (8. utilizzando l’Eq.25) s Il primo addendo a primo membro di questa equazione rappresenta il lavoro −δL compiuto sul sistema a seguito dello scostamento dall’equilibrio. nell’intorno di uno stato di equilibrio δF = 0.29) Esprimendo la probabilit` a termodinamica W tramite l’Eq. Infatti.27) e (8. (8. kB T (8. Consideriamo una variazione infinitesima dell’energia del sistema attorno all’equilibrio. (8.28) s e ancora.26) Allora.1). e la quantit` a di calore scambiata nella trasformazione infinitesima tramite la definizione di entropia (δS = δQ/T ).

(2). Si ha quindi 4πV p2 dp . 2 cfr.20.2. .34) In questo caso il peso gs va calcolato tenendo conto che [vedi Eq. di numero non assegnato. ma imponendo il vincolo aggiuntivo che N = s ns sia costante. (8. (8.35) gs = h3 La quantit` a α = α(N/V. Eq. 8. pag. vedi pi` u sotto. Eseguendo calcoli analoghi a quelli della precedente sezione si giunge all’espressione della funzione di distribuzione di Bose-Einstein: f (ε) = 1 ns = α ε/k T . ε + dε] (e frequenza [ν. II. mentre gs .2. si ottiene procedendo P in modo del tutto analogo.8).71 8. XII. T ) si determina imponendo la condizione di conservazione del numero totale di particelle.33) s s dove α `e un altro moltiplicatore di Lagrange3. 3 Si Mencuccini-Silvestrini. L’espressione della funzione di distribuzione nel caso sia assegnato il numero totale N di particelle (bosoni). dove si usa il simbolo f in luogo di u. La densit` a di energia (energia per unit` a di volume) dei fotoni con frequenza fra ν e ν + dν `e quindi data da2 u(ν)dν = 8πhν 3 dν ns hν gs 1 . In questo caso la funzione da massimizzare `e X X F = ln W + α(N − ns ) + β(E − ns εs ). STATISTICA DI BOSE-EINSTEIN Lo spettro del corpo nero Ricordiamo che nelle formule precedenti il pedice s si riferisce all’intervallo di energia [ε. quindi ns rappresenta il numero di fotoni con frequenza compresa fra ν e ν + dν. noti che alcuni testi definiscono α con segno opposto a quello usato in questo capitolo.32) che fornisce il noto spettro del corpo nero.2 Statistica di Bose-Einstein per un sistema di N particelle La dimostrazione illustrata nella sezione precedente si riferisce a una popolazione di fotoni. rappresenta il numero di stati compresi nello stesso intervallo. 50] il volume minimo di una cella nello spazio delle fasi d3 xd3 p `e pari a h3 . ν + dν]). espressa dall’Eq. Utilizzeremo questa espressione nel seguito del corso anche per studiare il laser e il rumore termico nei componenti circuitali ohmici. gs e e B −1 (8. (8. vol. = hν/k T = hν/k T B B V V e e −1 −1 (8.

72

CHAPTER 8. QUANTUM STATISTICS

8.3

Statistica di Fermi-Dirac

Mentre per i fotoni (e, pi`
u in generale per tutti i bosoni, cio`e le particelle con
numero quantico di spin intero) non esistono limiti alla popolazione degli stati
quantici, per gli elettroni (e, pi`
u in generale per tutti i fermioni, cio`e le particelle
con numero quantico di spin semi-intero) vale il principio di esclusione. Quindi
due fermioni con uguali numeri quantici non possono occupare uno stesso stato.
La funzione di distribuzione dei fermioni si ricava in modo del tutto analogo
a quella dei bosoni, ma tenendo conto del principio di esclusione. In questo caso
la probabilit`
a termodinamica della s-sima shell, cio`e il numero di modi in cui si
possono disporre ns particelle nei gs stati della shell stessa, `e pari al numero di
disposizioni di gs oggetti in gruppi di ns oggetti, ovvero
Ws =

gs !
.
ns !(gs − ns )!

(8.36)

Utilizzando la formula di Stirling possiamo scrivere
ln Ws ≃ gs ln gs − ns ln ns − (gs − ns ) ln(gs − ns ).

(8.37)

Procedendo come nel caso della statistica di Bose-Einstein, applichiamo il
metodo dei moltiplicatori di Lagrange per determinare il massimo della W , con i
vincoli posti dalla costanza del numero totale di particelle e dalla conservazione
dell’energia. Cerchiamo quindi i massimi della funzione
X
F =
[gs ln gs − ns ln ns − (gs − ns ) ln(gs − ns )]
s

+α(N −

X
s

ns ) + β(E −

X

ns εs ).

(8.38)

s

Derivando rispetto a ciascuna ns otteniamo
∂F
∂ns

= − ln ns − 1 + ln(gs − ns ) +
= ln

g s − ns
− α − βεs .
ns

g s − ns
− α − βεs
g s − ns
(8.39)

Imponendo l’annullamento di ciascuna derivata, ∂F/∂ns = 0, si ha
ln

g s − ns
= α + βεs ,
ns

(8.40)

da cui `e immediato ricavare l’espressione della funzione di distribuzione:
f (ε) =

ns
1
.
= α+βεs
gs
+1
e

(8.41)

Rimuovendo il pedice s e utilizzando la relazione β = 1/kB T (che si dimostra
valere anche in questo caso), possiamo finalmente scrivere della funzione di

8.4. LIMITE CLASSICO: STATISTICA DI BOLTZMANN

73

distribuzione di Fermi-Dirac:
f (ε) =

8.4

1
ns
= α ε/k T
,
gs
e e B +1

(8.42)

Limite classico: statistica di Boltzmann

L’espressione (8.42) differisce da quella relativa alla distribuzione di Bose-Einstein
solo per il segno che precede l’unit`a nel denominatore. Possiamo riassumere le
espressioni delle funzioni di distribuzione quantistiche tramite la relazione
f (ε) =

1
ns
= α ε/k T
,
gs
e e B ±1

(8.43)

dove il segno + si riferisce alla statistica di Fermi-Dirac e il segno − alla statistica
di Bose-Einstein. Nel caso di una popolazione di fotoni di numero non vincolato,
α = 0. Altrimenti,
P α si determina imponendo la conservazione del numero di
particelle, N = s ns = f (ε)gs . Per esplicitare questa relazione sostituiamo
la sommatoria sulle shell con un integrale sull’energia e scriviamo il peso gs in
funzione di ε, ottenendo
Z ∞
g(ε)
dε,
(8.44)
N=
[α+(ε/kB T )]
e
±1
0
con g(ε)dε dato dall’ Eq. (3), pag. 50.
L’equazione (8.44) definisce implicitamente α = α(N/V, T ). Si dimostra che
essa pu`
o essere posta nella forma
α = −ε0 /kB T,

(8.45)

dove ε0 `e un’energia caratteristica, che dipende dalla densit`
a di particelle N/V
e, molto debolmente, dalla temperatura. Si ha allora
f (ε) =

1
e(ε−ε0 )/kB T

±1

.

(8.46)

Nel caso della statistica di Fermi-Dirac, ε0 prende il nome di energia di Fermi
e si indica usualmente con il simbolo εF . La statistica di Fermi verr`a discussa e
utilizzata nel seguito del corso per studiare alcune propriet`
a di metalli e semiconduttori (vedi Sette-Bertolotti, Vol. II, § 16.3–16.5 e Mencuccini-Silvestrini,
Vol. II, § XII.9.2).
Quando il termine exp[α+(ε/kB T )] nel denominatore dell’Eq. (8.43) `e molto
maggiore dell’unit`a, l’Eq. (8.43) stessa pu`
o essere scritta
f (ε) ≃ e−α e−ε/kB T .

(8.47)

Ritroviamo cos`ı la funzione di distribuzione di Boltzmann, che quindi
costituisce un’ottima approssimazione delle distribuzioni quantistiche quando
l’occupazione dei livelli energetici `e molto bassa; vedi la Fig. 8.3.

74

CHAPTER 8. QUANTUM STATISTICS

Bose-Einstein
Boltzmann

2
f (ε)
1

Fermi-Dirac
0
ε0/kBT

0

20
ε/kBT

Figure 8.3: Confronto fra le funzioni di distribuzione di Fermi-Dirac, BoseEinstein e Boltzmann, per ε0 = 10kB T .

Osserviamo che all’Eq. (8.47) si giunge anche applicando il procedimento
variziazionale seguito per ricavare le espressioni delle funzioni di distribuzioni
quantistiche. In questo caso nel calcolare la probabilit`
a termodinamica W si
deve tenere conto della distinguibilit`
a delle particelle. Si scrive quindi
Y g ns
s
W = N!
.
(8.48)
n
s!
s
Nell’Eq. (8.48) fattore N ! rappresenta le permutazioni delle N particelle; il
fattore gsns i modi diversi in cui ns particelle contenute in una shell possono essere
distribuite nei gs stati della shell stessa, il fattoriale ns ! a denominatore tiene
conto dell’equivalenza delle configurazioni realizzate permutando le particelle
contenute in una stessa shell. Per calcolare la distribuzione pi`
u probabile si
massimizza quindi la funzione ln W rispetto alle occupazioni ns , imponendo la
conservazione del numero totale di particelle e dell’energia totale. Utilizzando,
come in precedenza, il metodo dei moltiplicatori di Lagrange, si massimizza il
funzionale
X
X
F = ln W + α(N −
ns ) + β(E −
ns εs ), .
(8.49)
s

s

Inserendo l’espressione (8.48) nell’Eq. (8.49), usando la formula di Stirling per i
logaritmi dei fattoriali e uguagliando a zero la derivata rispetto a ns , si ottiene,
per ogni s,
gs
− α − βεs = 0
(8.50)
ln
ns
e quindi l’espressione della funzione di distribuzione
f (εs ) =

ns
= e−α e−βεs = e−α e−εs /kB T .
gs

(8.51)

1). . the Breit-Wigner resonant cross-section.g. . . The goal of our treatment is to recover qualitative features of important nuclear reaction cross sections. General scattering theory is extremely complex and outside the scope of the present introductory course.1 Partial wave expansion Figure 9. in a semi-classical picture (see ¯h l(l + 1) ≃ ¯ Fig. We consider a beam of particles with momentum p = h ¯ k interacting with a particle (a scatterer) at rest in the origin of a coordinate system. Therefore. elastic scattering far away from resonances). According to quantum mechanics angular momentum can only take values L = p hl. Particles in the beam with impact parameter b have angular moment L = pb = h ¯ kb. 9. and the 1/v behaviour of many absorption cross-sections. 9. If we indicate with a the distance 75 . 1. with l = 0. particles with angular quantum number between l and l + 1 have impact parameters l/k ≤ b ≤ (l + 1)/k..Chapter 9 s-wave scattering In this Chapter we present a few basic elements of scattering theory. we shall find the nearly constant behaviour of potential scattering (e.1: Geometrical interpretation of partial wave expansion. Indeed.

e. i. It follows that if ka ≪ 1 even particles with l = 1 are unaffected by the potential: only l = 0 particles matter.e. S-WAVE SCATTERING from the scatterer at which the interaction potential vanishes (i.31)] and jl is the spherical Bessel function of order l. Sec.1) l=0 Here Pl is the Legendre polynomial of order l [Eq. with l > ka are not affected by the potential and do not suffer any scatter (nor other reactions). or s-wave) of the expansion. Wiley (1988). Krane. We then have jl (kr) ≃ Ae−ikr Aeikr + . dependence on the other angle φ is omitted. This wavefunction can be expanded in a series of angular momentum wavefunctions 1 2 ψ0 = Aeikr cos θ = A ∞ X (2l + 1)il jl (kr)Pl (cos θ). (7. and the second one an outgoing spherical wave. We can then consider the interaction as due only to such particles. ψ0 (r) = − 9.76 CHAPTER 9.e. (9. (9. assuming cylindrical symmetry around the z-axis. ψ0 (r.e.8. in the spherical coordinate system centred around the scatterer.2) kr Each term of the expansion. θ) = Aeikr cos θ . each partial wave corresponds to a specific angular momentum l. i. for kr ≫ 1 approaches sin(kr − lπ/2) . S.g. 2 The . 11.e.2 General expression of the cross-sections After the interaction with the particle in the origin. We start by observing that the beam of incident particles is represented by a plane wave ψ0 (z) = Aeikz or. the s-wave component at large distance from the scatterer will again be given by the sum of an entering 1 See. i. Following the discussion above. K. Introduction to Nuclear Physics. which asymptotically. (9. to waves with l = 0 (angular momentum s-waves). we only retain the first term (l = 0.2: We study scattering in a reference frame centred on the scatterer.3) 2ikr 2ikr where the first term on the right hand side represents an entering spherical wave. V (r) = 0 for r > 0) particles with b > a. Figure 9.

By definition σsc = particles scattered per unit time (dN/dt)sc . While in general η depends on θ and φ.2. scattering is fully characterized by the differential cross-sectionR σ(θ) = dσ/dΩ.3 and can then be written as e−ikr eikr + Aη . However s-wave scattering is symmetric in the centre-of-mass system. GENERAL EXPRESSION OF THE CROSS-SECTIONS 77 Figure 9. m (9. (5. and then we simply have σsc = 4πσ(θ).e.5) scattered or absorbed which highlights the difference between incoming wave and scattered wave. ψ(r) = −A ψ(r) = A | eikr − e−ikr + 2ikr {z } ψ0 (r) eikr A(η − 1) . It is useful to write such a wave in a slightly different way. ¯k 2 h |A| . (9. = incident particles per unit time and area Jinc (9.28) Jinc = 3 See footnote 2 of Sec. | {z 2ikr} (9.9. spherical wave (left unchanged by the interaction) and and an outgoing spherical wave.5. 7. The knowledge of η is sufficient to compute both the scattering cross-section and the reaction cross-section. in the l = 0 approximation η is a (complex) constant. In general.4) 2ikr 2ikr where η is a complex quantity.3: The wavefunction at large distance from the scatterer is the sum of an entering spherical wave and an outgoing one.6) where the current of the incident beam is (see Eq.7) . with the (integrated) scattering crosssection given by σsc = Ω dΩσ(θ). i.

where v = p/m = h ¯ k/m is the particle velocity.78 CHAPTER 9. Therefore the scattering rate is equal to the product of the particle density (i.e. S-WAVE SCATTERING The scattering rate can instead be written by observing that the same number of particles will cross any sphere centred in the origin. V = 4πr2 × v × 1.e. i. the square of the modulus of the wavefunction) at distance r multiplied by the volume V spanned by the particles in the unit of time. We can then write  dN dt  = sc = = .

.

ikr .

2 .

.

(η − 1)A e .

× 4πr2 × ¯hk = .

2ikr .

Particles can also be absorbed or scattered inelastically. (9. (9. m 1 ¯hk A2 |η − 1|2 2 2 4πr2 4k r m ¯h 2 2 .11) − where we have introduced λ = λ/2π = 1/k. Such processes are taken into account by the reaction cross-section σreaction . (9.13) .12) and the total cross-section is −2 σtot = σsc + σreaction = 2πλ [1 − ℜ(η)].10) and finally − 2 σsc = πλ |η − 1|2 . So far we have considered elastically scattered particles.8) (9. A |η − 1| π mk (9. It can be shown that −2 σreaction = πλ (1 − |η|2 ) .9) (9.

.

.

.

.

.

.

Appendices 85 .

.

Appendix A Wave packets and group velocity 87 .

88 APPENDIX A. WAVE PACKETS AND GROUP VELOCITY .

so that σt σω = 1. with standard deviation σt = 2. We now use Eq.1) F (ω) = −∞ of a function f (t).Appendix B Fourier transform of a Gaussian We show that the Fourier transform of a Gaussian is still a Gaussian and that the product of the standard deviation of a Gaussian and of its Fourier transform is 1.4) 2 π √ which is a Gaussian in t. according to Z ∞ 1 f (t) = F (ω)eiωt dω.3) √ i. 89 . (B. We consider the Fourier-transform F (ω) Z ∞ f (t)e−iωt dt (B. = (B. it is a Gaussian exp (−ω 2 /2σω2 ) . (B. with standard deviation σω = 1/ 2. Z ∞ 2 1 f (t) = e−ω +iωt dω = 2π −∞ Z ∞ 2 2 1 = e−(ω−it/2) −t /4 dω = 2π −∞ Z ∞ 2 2 2 e−t /4 √ 1 π e−x −t /4 dx = = 2π −∞ 2π 2 1 √ e−t /4 . (B. which can be recovered by anti-transforming F (ω).2) 2π −∞ Let us assume that F (ω) is given by 2 F (ω) = e−ω .2) to obtain f (t).e.

90 APPENDIX B. FOURIER TRANSFORM OF A GAUSSIAN .

Appendix C A proof of the uncertainty relations 91 .

92 APPENDIX C. A PROOF OF THE UNCERTAINTY RELATIONS .

1) (D. related to the Cartesian coordinates x. We introduce spherical coordinates r.3) .1: Spherical and Cartesian coordinates.Appendix D Angular momentum operators in spherical coordinates Figure D. θ. z by r = θ = z = p x2 + y 2 + z 2 p x2 + y 2 arctan z y arctan x 93 (D. φ.2) (D. y.

94

APPENDIX D. ANGULAR MOMENTUM OPERATORS
x =

r sin θ cos φ

(D.4)

y
z

r sin θ sin φ
r cos θ

(D.5)
(D.6)

=
=

We want to obtain the expressions of the angular momentum operators

ˆlx

=

ˆly

=

ˆ
lz

=



−z )
∂z
∂y


−x )
zˆpˆx − xˆpˆz = −i¯h(z
∂x
∂z


−y )
x
ˆpˆy − yˆpˆx = −i¯h(x
∂y
∂x
yˆpˆz − zˆpˆy = −i¯h(y

(D.7)
(D.8)
(D.9)

in spherical coordinates. To express derivatives with respect to Cartesian coordinates in terms of derivatives with respect to spherical coordinates we have to
use expressions such as


∂r ∂
∂θ ∂
∂φ ∂
=
+
+
.
∂z
∂z ∂r ∂z ∂θ
∂z ∂φ

(D.10)

As an example, we detail the derivation of the expression of ˆlz :

ˆlz

=  


∂θ ∂
∂φ ∂
∂r ∂
+
+
+
−i¯h x
∂y ∂r ∂y ∂θ
∂y ∂φ  

∂r ∂
∂θ ∂
∂φ ∂
−y
,
+
+
∂x ∂r ∂x ∂θ
∂x ∂φ

(D.11)

and reordering the terms on the right hand side,

ˆlz

=  

∂r
∂r ∂
x
−y
∂y
∂x ∂r  

∂θ ∂
∂θ
−y
+ x
∂y
∂x ∂θ   

∂φ
∂φ ∂
+ x
,
−y
∂y
∂x ∂φ

−i¯h

(D.12)

95
where, using Eqs. (D.4)–(D.6), we can write
∂r
∂y
∂r
∂x
∂θ
∂y
∂θ
∂x
∂φ
∂x
∂φ
∂y

=
=
=

=
=

=

y
1
p
2y =
2
2
2
r
2 x +y +z
x
r
zy
y
1
1
1
p
2y = p
= 2
2
2 z
2
2
2
2
2
r tan θ
x +y
2 x +y
r x +y
1+
z2
y
r2 tan θ
y
1 −y
2 x2 = − x2 + y 2
y
1+ 2
x
1
x
= 2
x
x + y2

Substituting Eqs. (D.13)–(D.18) into Eq. (D.12) we finally obtain    


y


x
ˆlz = −i¯
h 0·
,
+
y
+0·
+ x 2
∂r
∂θ
x + y2
x2 + y 2 ∂φ

(D.13)
(D.14)
(D.15)

(D.16)
(D.17)

(D.18)

(D.19)

and then
ˆlz = −i¯h ∂ .
∂φ

(D.20)

By proceeding analogously, one obtains

and  



ˆ
− cot θ cos φ
lx = i¯
h sin φ
∂θ
∂φ

(D.21)  



ˆly = −i¯
h cos φ
.
− cot θ sin φ
∂θ
∂φ

(D.22)

96

APPENDIX D. ANGULAR MOMENTUM OPERATORS

1) + + 2 [E − V (r)] − dr2 r dr r2 h ¯ 97 . (7.51)].solution of the radial equation We give here all details of the solution of the radial equation for the electron wavefunction in the hydrogen atom [Eq.Appendix E Hydrogen atom . which we write again here:  2  d l(l + 1) 2 d 2µ R(r) = 0 (E.

98 APPENDIX E. HYDROGEN ATOM: RADIAL EQUATION .

99 .

100 APPENDIX E. HYDROGEN ATOM: RADIAL EQUATION .

101 .

102 APPENDIX E. HYDROGEN ATOM: RADIAL EQUATION .

.

.

.

.

.

.

x2 . n. m. i = 1. Soddisfano quindi le condizioni ∂F/∂xi = ∂F/∂αk = 0. ) soggetta alle m condizioni gk (x1 . . . x2 . . . . . Si dimostra che i massimi condizionati di f si trovano fra quelli non condizionati di F = f + α1 g1 + α2 g2 + α3 g3 + . . . 109 . . . xn . . ∂xi I moltiplicatori di Lagrange αk si trovano poi imponendo che i vincoli gk = 0 siano soddisfatti. . . + αm gm . 2. . .Appendix G Metodo dei moltiplicatori di Lagrange Il metodo dei moltiplicatori di Lagrange consente di trovare i massimi di una funzione y = f (x1 . k = 1. . . k = 1. . m. 2. . Le xi estremanti si trovano allora risolvendo le i equazioni ∂F = 0. xn ) = 0. 2. . . .