You are on page 1of 160

Copyright Hans J.

Reich 2017
All Rights Reserved
University of Wisconsin
5. Proton Nuclear Magnetic Resonance Spectroscopy

5-HMR-0 The NMR Experiment


5-HMR-1 Integration of Proton NMR Spectra
5-HMR-2 Chemical Shift
5-HMR-2.3 Calculation of Proton Chemical Shifts
5-HMR-2.6 Proton Chemical Shift Effects
5-HMR-2.8 Magnetic Anisotropy of Functional Groups
5-HMR-2.12 Aromatic Solvent Induced Shifts (ASIS)
5-HMR-2.23 Exchangeable protons (NH, OH, SH)
5-HMR-3 Spin-Spin Splitting - J-Coupling
5-HMR-3.7 Rules for Analyzing First Order Multiplets
5-HMR-3.14 Measurement of Coupling Constants
5-HMR-4.1 Geminal Proton-Proton Couplings (2JH-H)
5-HMR-5.1 Vicinal Proton-Proton Couplings (3JH-H) (Karplus)
5-HMR-5.3 Determination of Stereochemistry in Cyclic Compounds Using 3JH-H
5-HMR-5.10 Acyclic Stereochemistry Using 3JH-H
5-HMR-5.12 Allylic 3JHH
5-HMR-5.13 Olefinic 3JHH
5-HMR-6 Long-Range (4J and higher) Proton-Proton Couplings
5-HMR-7 Pople Nomenclature for Coupled Spin Systems
5-HMR-8 Symmetry in NMR Spectra - Homotopic, Enantiotopic, Diastereotopic
5-HMR-9 Second Order Effects in Coupled Systems
5-HMR-10 AX and AB Spectra
5-HMR-11 AX2 and AB2 Patterns
5-HMR-12 ABX Pattern
5-HMR-12.4 Solving an ABX Pattern
5-HMR-12.12 A Simple ABX Pattern as AB is Changed
5-HMR-12.13 Effect of Relative Sign of JAX and JBX on ABX Patterns
5-HMR-12.14 ABX with Accidental Coincidences
5-HMR-12.16 ABX of the Vinyl type
5-HMR-12.17 ABX Going to ABC
5-HMR-12-19 ABXYZ Patterns
5-HMR-13 ABX3 Patterns
5-HMR-14 AA'XX' Spectra
5-HMR-15 AA'BB' Spectra
5-HMR-16 Virtual Coupling

5-HMR-0.1
5.0 The NMR Experiment
Nuclear Properties
We are used to thinking of chemical properties in terms of elements (atomic number). For nuclear properties we
have to think in terms of isotopes (mass number) - different isotopes of the same element have different nuclear
properties. The main nuclear property we are interested in connection with NMR is the nuclear angular momentum
spin quantum number I, the "spin" of the nucleus.

I = 0 no spin, the nucleus has no magnetic moment and no NMR properties


I > 0 the nucleus has spin (I = 1/2, 1, 3/2, 2, etc) and a magnetic dipole , and thus may be suitable for NMR
observation.
x even 12 16 32
Ny Neven I=0 C6 O8 S16
odd
x = mass number Neither I = 1/2 1H1, 13
C6, 15N7, 19
F9, 31P15
y = atomic number 7
I = 3/2, 5/2, etc Li3 (3/2), 11B5 (3/2)
even
Nodd I = 1, 2, 3, etc 6Li3 (1), 14N7 (1), 2H1 (1),
Nuclei with I = 1/2 have especially advantageous NMR properties, and the vast majority of all NMR experiments
are done with such isotopes.
Nuclei with I > 0 have angular momentum P (spinning mass) whose direction is the spin axis. The angular
momentum is quantized, and can only have one value:
h
P = I (I + 1) 2
Nuclei with I > 0 also have a magnetic dipole (spinning charge). For the NMR experiment it is the ratio of to P
that matters (much in the way that m/e is what matters in mass spectrometry). We define , the gyromagnetic
ratio:

=
P

Interaction of Nuclei with a Magnetic Field


When we place a nucleus with spin in a magnetic field the nuclei tend to align with the field. The observable
component of the angular momentum Pz is also quantized, and can only have the following values:

h where m is the magnetic quantum number which can


Pz = m
2 have the values I, I - 1, I - 2, ... -I
Quantum restrictions prevent the nuclei from aligning exactly with Bo, since both the angular momentum (P) and
the observable component (Pz) are quantized. For spin nuclei there is a tip angle of 54.7.
z
cos () = Pz/P = 1/ 3
There are 2I + 1 allowed
B0 For m = + = 54.74 orientations of the nucleus in the
Pz P For m = - = 125.26 magnetic field
y For I = 1, m = 1 = 45, m = 0 = 90, m = -1 = 135
x
The nuclei precess around the direction of Bo, with a frequency o (Larmor precession frequency). The
frequency is a function of the magnetic field strength (Bo), the angular momentum and the magnetic dipole
(Gyromagnetic ratio ).
Bo
o =
2

5-HMR-0.1
Interaction with Radiofrequency
A radiofrequency (at the Larmor precession frequency 0) applied in the x-direction causes transitions between the
spin states if RF = o. These transitions are detected by the spectrometer and plotted as an NMR spectrum.

The NMR Spectrometer


An NMR spectrometer consists of a powerful magnet, and the associated electronics to control the properties of
the magnet and create and detect radiofrequency signals. In the first spectrometers (up to 60 MHz proton
frequency) permanent magnets were used, then electromagnets (to 100 MHz), and now most spectrometers use
superconducting magnets to achieve field strengths which give proton resonances from 200 to as high as 900 MHz.
The magnetic field must be very stable over a period of hours and very homogeneous over the sample volume
(better than 1 part in 109). A complex array of tuning coils is mounted in the magnet and probe to correct for
magnetic field inhomogeneities. The radiofrequency generators (at least two are required) must also be very stable,
and capable of providing frequencies accurate to 1 part in 109, and with short (microsecond range) very accurately
timed pulses. All of these very stringent requirements, together with the inherent insensitivity of the NMR
experiment, mean that NMR spectrometers have very complex electronics and are hence the most expensive of all
common analytic devices used by chemists, running from $100K to $3M (physicists, of course, use analytic devices
that cost billions of dollars).

At the heart of an NMR spectrometer is the probe, which is a removable cylinder inserted into the center of the
magnet. The probe contains: the sample tube holder and air spinner outlets; the radiofrequency coils for signal
detection, decoupler irradiation, and locking of the magnetic field; the electronics, dewar, gas inlets and outlets for
cooling and heating of the sample; the tuning coils for fine adjustments of the magnetic field, as well as (in advanced
probes) coils for producing precise field gradients. The very latest probes have the electronics for signal detection
cooled to liquid nitrogen or liquid helium temperatures (cryoprobes) to provide substantially improved sensitivity (at a
factor of 10 increase in cost). Just an ordinary probe can cost more than a low-end IR or UV spectrometer, and a
cryoprobe alone can cost as much as an entire mass spectrometer or X-ray diffraction instrument.

1/40
5-HMR-0.2
Detection of NMR Signals

The first generation of NMR spectrometers detected the NMR signals in much the same way as was done for the
earlier spectroscopic methods such as IR and UV/VIS - the instrument scans through the frequency region of
interest (or keeps the frequency constant and scans the magnetic field, a technically easier process used in the first
spectrometers). When there is a frequency match (resonance: RF = o) the transitions are detected by the coils in
the spectrometer probe, and, after signal processing, are plotted as an NMR spectrum.
Advances in microwave electronics made possible a much more efficient way of detecting NMR signals in which
frequencies are not scanned, but instead a very short powerful pulse is applied to the sample. The pulse is short
enough that its frequency is not well defined to within a few thousand Hertz, so it interacts with all of the nuclei of one
isotope in the sample. The pulse duration is accurately specified so that the precession of the nuclei around the
axis of the pulse corresponds to a well defined angle (say 90 degrees).

Spin nuclei in magnetic field B0

z z z
z

very short
B0 time 90 pulse time
x x x x
T1 (apply magnetic
y y field for a short time y y
in y direction)

Top:
Note the population difference
x (GREATLY exaggerated) between x
the aligned and anti-aligned
nuclei
y y
Bottom:
After the 90 pulse the population
difference in +z and -z directions
x has been converted into population x
difference in the +x and -x
directions.
Net Magnetization Vector
z z z

This net magnetization in the


x-y plane creates a fluctuation
radiofrequency signal that can
x x be detected. x

y y y

The pulse rotates the excess magnetization (resulting from the higher population of magnetic nuclei in the more
stable orientation aligned with the magnetic field Bo) from the z-direction to the x-y plane. This magnetization vector
rotates around the x-y plane at the Larmor precession frequency. The fluctuating magnetic field produced by these
nuclei is detected by the spectrometer. Each set of nuclei with a distinct chemical shift in the sample has its own
precession frequency (the chemical shift), and the spectrometer detects the sum of all of these oscillations (the Free
Induction Decay, or FID). The FID is then mathematically manipulated (Fourier transformation) to detect the
individual frequencies, which are plotted as a spectrum.

5-HMR-0.3 1/43
Chemical Shift
Circulation of electrons around the nucleus creates local magnetic fields which shield the nucleus from the
external field Bo. The extent of shielding depends on the local chemical environment. Thus NMR signals show a
chemical shift. The first NMR spectrometers used continuous wave detection, initially by using a magnetic field
sweep to scan through the spectrum (a technically simpler process), later frequency sweep electronics were
developed. All modern spectrometers use pulse techniques to detect NMR spectra.

Frequency sweep H2
7.012987 T 7.02000 T
300,003,000 Hz 300,000,000 Hz H1
H2 H1
m = - () Field sweep
E

0 Bo
Chemical shift is
m = + () greatly exaggerated

12 10 8 6 4 2 0
(ppm)
Proton NMR spectrum at 7.02 T (300 MHz)

The Larmour precession frequency o depends on the magnetic field strength. Thus at a magnet strength of 1.41
Tesla protons resonate at a frequency of 60 MHz, at 2.35 Tesla at 100 MHz, and so on. Although Hz are the
fundamental energy unit of NMR spectroscopy, the use of Hz has the disadvantage that the position of a peak is
dependent on the magnetic field strength. This point is illustrated by the spectra of 2-methyl-2-butanol shown below
at several different field strengths, plotted at a constant Hz scale.

Plotted on the same Hz scale c


CH3 b a
c
CH3 C CH2 CH3
OH
d Me4Si
220 MHz

400 300 200 100


c
0

d a
b
100 MHz

200 100 0

(Frequency shift from Me4Si in Hz) 60 MHz


=
(Spectrometer frequency, MHz)
100 0

For this reason, the distance between the reference signal (Me4Si) and the position of a specific peak in the
spectrum (the chemical shift) is not reported in Hz, but rather in dimensionless units of , which is the same on all
spectrometers. Note that in the above spectra the multiplet separations (doublet, quartet) are the same at all fields,
whereas in the spectra below the chemical shift separations are equal.

5-HMR-0.4
c
CH3 b a
Plotted on the same ppm scale c
CH3 C CH2 CH3
OH
d Me4Si
220 MHz

1.5 1.0 0.5 0.0

100 MHz

2.5 2.0 1.5 1.0 0.5 0.0


60 MHz

3.0 2.5 2.0 1.5 1.0 0.5 0.0


(ppm)
Coupling
Neighboring magnetic nuclei can also perturb the local magnetic field, so we observe J-coupling, which causes
multiplet structure for NMR signals. J coupling is mutual (JAX = JXA). Coupling constants are independent of the
magnetic field, and thus should always be reported in Hz.
Multiplet structure resulting from several couplings to a given nucleus, however, often depends strongly on the
chemical shifts between the nuclei, with larger chemical shifts usually leading to simpler spectra. Since chemical
shifts (in energy units like Hz) increase with magnetic field strength, higher field magnets typically give much simpler
and more easily interpreted NMR spectra.
Sensitivity
In sharp contrast to UV/VIS/IR spectroscopy, where essentially all molecules are in the ground state at room
temperature, in NMR spectroscopy the excited states are thermally populated, with population difference between the
spin states of only about one part in 105, so NMR signals are inherently very weak.
n+ At 500 MHz and 298 K E = -0.05 cal/mol, RT = 592 cal/mol, n+/n- = 0.999916
= e-E/RT
n-

The energy separation between the two spin states of a spin nucleus is directly proportional to the strength of
C
the magnetic field (E = Bo). This in turns affects the Boltzmann population differences of the and spin states.
Thus stronger magnetic fields result in large increases in the strength of the NMR signal.

Excess population: E = 0.05 cal/mol Excess population:


Bo = 7.02 T
1/98,000 E = 0.03 cal/mol 500 MHz 1/12,000
300 MHz Bo = 11.7 T
Bo = 1.41 T
m = - ()
E = 0.006 cal/mol
E 60 MHz

0 Bo

E = Bo m = + ()

Relaxation
Relaxation of spin for I = 1/2 nuclei is slow (T1 = 0.1 to 100 sec). This may further weaken NMR signals when the
RF field is applied repeatedly (as it usually is), since the population of the spin states can become equalized if nuclei
cannot fully relax back to their normal populations between pulses (saturation). See Section 8-TECH-1.

5-HMR-0.5
Copyright Hans J. Reich 2017
All Rights Reserved
5.1 Integration of Proton NMR Spectra University of Wisconsin

NMR is unique among common spectroscopic methods in that signal intensities are directly proportional to the
number of nuclei causing the signal (provided certain conditions are met). In other words, all absorption coefficients for
a given nucleus are identical. This is why proton NMR spectra are routinely integrated, whereas IR and UV spectra
are not. A typical integrated spectrum is shown below, together with an analysis.

270 MHz 1H NMR Spectrum in CDCl3

O O
S

22.8 mm
15.2 mm
36.2 mm

9 8 7 6 5 ppm 4 3 2 1 0

The vertical displacement of the integral gives the relative number of protons It is not possible to determine the
absolute numbers without additional information (such as a molecular formula). In the example above, if we add up
all of the integrals, we get 74.3. Dividing each integral by the smallest one (15.2) gives a ratio of 2.38/1.0/1.50 for the
three signals. Multiplying by two gives 4.76/2.0/3.03, which is close to the integral numbers (5/2/3) expected for a
pure compound. However, there is nothing in the spectrum that rules out 10/4/6 or higher multiples. If we have a
molecular formula (in this case C8H10O2S), dividing by the number of hydrogens gives 7.4 mm per H. We can then
determine the number of protons corresponding to each multiplet by rounding to the nearest integer. It is generally
possible to reliably distinguish signals with intensities of 1 to 10 or so, but it becomes progressively harder to make a
correct assignment as the number of protons in a multiplet increases beyond 10, because of the inherent inaccuracies
in the method.
The two parts of aromatic proton integral at 7.6 and 7.9 can be separately measured as a 2:3 ratio of ortho to
meta+para protons.

If given the molecular formula (C8H10O2S), we know there are 10H in molecule
Total area: 36.2 + 15.2 + 22.8 = 74.2 mm
Thus 7.4 mm per H
36.2 / 7.4 = 4.89 i.e. 5H
15.2 / 7.4 = 2.05 i.e. 2H
22.8 / 7.4 = 3.08 i.e. 3H

Reich, U.Wisc. Chem. 605 5-HMR-1.1 Integration


Accuracy of Proton NMR Integrations
The integration of NMR spectra can be carried out with high accuracy, but this is only possible if a number of
sources of error are properly handled. On a modern spectrometer accuracy of 5% can be achieved easily if
relaxation issues are handled properly. To get errors of <1% a number of factors have to be considered and
optimized.

1. Signal to Noise. The spectrum must have adequate signal to noise to support the level of accuracy required
for the experiment.

2. Saturation Effects. NMR spectroscopy has a feature unique among spectroscopic methods, that relaxation
processes are relatively slow (on the order of seconds or tenths of seconds), compared to milli, micro, and pico
seconds for IR and UV. In other words, once the spectrometer has perturbed the equilibrium population of nuclei by
scanning over the resonance frequency or pulsing the nuclei, it takes from 0.1 to 100s of seconds (typically several
seconds) for them to return to their original populations (T1 the spin-lattice relaxation time). If power settings are too
high (for CW spectra) or pulse angle and repetition rates too high (for FT spectra) then spectra can become
saturated, and integrations less accurate, because the relaxation rates of various protons in the sample are
different. Saturation effects are particularly severe for small molecules in mobile solvents, because these typically
have the longest T1 relaxation times.
To get reliable integrations the NMR spectrum must be acquired in a way that saturation is avoided. It is not
possible to tell whether a spectrum was run appropriately simply by inspection, it is up to the operator to take
suitable precautions (such as putting in a 5-10 second pulse delay between scans) if optimal integrations are
needed. Fortunately, even a proton spectrum taken without pulse delays will usually give reasonably good
integrations (say within 10%). It is important to recognize that integration errors caused by saturation effects will
depend on the relative relaxation rates of various protons in a molecule. Errors will be larger when different kinds of
protons are being compared (such as aromatic CH to a methyl group), than when the protons are similar or identical
in type (e.g. two methyl groups).
3. Line Shape Considerations. NMR signals in an ideally tuned instrument are Lorenzian in shape, so the
intensity extends for some distance on both sides of the center of the peak. Integrations must be carried out over a
sufficiently wide frequency range to capture enough of the peak for the desired level of accuracy. Thus, if the peak
width at half height is 1 Hz, then an integration of 2.3 Hz from the center of the peak is required to capture 90% of
the area, 5.5 Hz for 95%, 11 Hz for 98% and 18Hz for >99% of the area. This means that peaks that are closely
spaced cannot be accurately integrated by the usual method, but may require line-shape simulations with a program
like NUTS or WINDNMR to accurately measure relative peak areas.

4. Digital Resolution. A peak must be defined by an adequate number of points if an accurate integration is to
be obtained. The errors introduced are surprisingly small, and reach 1% if a line with a width at half height of 1 Hz is
sampled every 0.5 Hz.
5. Isotopic Satellites. All C-H signals have 13C satellites located JC-H/2 from the center of the peak (JC-H is
typically 115-135 Hz, although numbers over 250 Hz are known) Together these satellites make up 1.1% of the area
of the central peak (0.55% each). They must be accounted for if integration at the >99% level of accuracy is
desired. Larger errors are introduced if the satellites from a nearby very intense peak fall under the signal being
integrated. The simplest method to correct this problem is by 13C decoupling, which compresses the satellites into
the central peak. A number of other elements have significant fractions of spin nuclei at natural abundance, and
these will also create satellites large enough to interfere with integrations. Most notable are 117/119Sn, 29Si, 77Se,
125
Te, 199Hg. For more on satellites, see Section 7, Multinuclear NMR.
There is a bright side to 13C satellites: they can be used as internal standards for the quantitation of very small
amounts of isomers or contaminants, since their size relative to the central peak is accurately known.

Reich, U.Wisc. Chem. 605 5-HMR-1.2 Integration


6. Spinning Sidebands. These can appear at the spinning speed in Hz in spectra run on poorly tuned
spectrometers and/or with samples in low-quality tubes. They draw intensity from the central peak. SSBs are rarely
significant on modern spectrometers.
7. Baseline Slant and Curvature. Under some conditions spectra can show significant distortions of the
baseline, which can interfere with obtaining high-quality integrations. Standard NMR work-up programs have
routines for baseline adjustment.
8. Decoupling. When decoupling is being used, as is routinely done for 13C NMR spectra and occasionally for 1H
NMR spectra, peak intensities are distorted by Nuclear Overhauser Effects (NOE, see Sect. 8). Integrations of such
spectra will not give accurate ratios of peak areas.
Peak Intensities. Under certain conditions, peak heights can also be a quite accurate method of quantitation.
For example, if several singlets are being compared, and they all have identical line widths, and the spectra were
measured such that there are sufficient data points to define the lineshape of each singlet, then peak heights may be
useful, and under ideal conditions more accurate than integrations.

Determining Absolute Amounts by NMR Integration. Although NMR spectra in principle follow Beer's law, it is
difficult (although not impossible) to make effective use of the absolute intensities of NMR spectra for quantitation
(as is routinely done for UV, and sometimes IR). NMR integrations are always relative. Thus an internal standard
must be used to determine reaction yields by NMR integration. A commonly used internal standard for proton NMR
spectra is pentachloroethane -- it is a liquid, not too volatile, and appears in a region of the NMR spectrum ( 6.11)
where there are few signals. It is strongly recommended to avoid using volatile materials like CH2Cl2, CHCl3, C6H6
and others, since it is very difficult to avoid some evaporation losses during the transfer process of the standard,
leading to incorrect (high) concentrations of the substrate.
1/39

Reich, U.Wisc. Chem. 605 5-HMR-1.3 Integration


Copyright Hans J. Reich 2017
All Rights Reserved
5.2 Chemical Shift University of Wisconsin

Fortunately for the chemist, all proton resonances do not occur at the same position. The Larmor precession
frequency (o) varies because the actual magnetic field B at the nucleus is always less than the external field Bo.
The origin of this effect is the "superconducting" circulation of electrons in the molecule, which occurs in such a
way that a local magnetic field Be is created, which opposes Bo (Be is proportional to Bo). Thus B = Bo - Be. We
therefore say that the nucleus is shielded from the external magnetic field. The extent of shielding is influenced by
many structural features within the molecule, hence the name chemical shift. Since the extent of shielding is
proportional to the external magnetic field Bo, we use field independent units for chemical shifts: values, whose
units are ppm. Spin-spin splitting is not dependent on the external field, so we use energy units for coupling
constants: Hz, or cycles per second (in mathematical formulas radians per second are the natural frequency units
for both chemical shifts and couplings).

B = Bo - Be (magnetic field at nucleus)


Bo H e-
A o = B/2 (Larmor precession frequency of
Be HA - varies as B changes)

The Proton Chemical Shift Scale


Experimentally measured proton chemical shifts are referenced to the 1H signal of tetramethylsilane (Me4Si).
For NMR studies in aqueous solution, where Me4Si is not sufficiently soluble, the reference signal usually used is
DSS (Me3Si-CH2CH2-SO3-Na+, Tiers, J. Org. Chem. 1961, 26, 2097). For aqueous solution of cationic substrates
(e.g., amino acids) where there may be interactions between the anionic reference compound and the substrates,
an alternatice reference standard, DSA (Me3Si-CH2CH2-NH3+ CF3CO2-, Nowick Org. Lett. 2003, 5, 3511) has been
suggested.
Proton chemical shifts cover a range of over 30 ppm, but the vast majority appear in the region 0-10 ppm,
where the origin is the chemical shift of tetramethylsilane.
O-

H+ (naked proton - calculated) H + N+ Me4Si (TMS) H-Rh(CN)5-3


O Me

40 30 20 10 0 -10
ppm
most protons fall in this region
Downfield Upfield
Bo decreases "Field sweep" Bo increases
Deshielded Shielded
o increases "Frequency sweep" o decreases
High frequency Low frequency
In the original continuous wave (CW) method of measuring NMR spectra, the magnetic field was scanned from
left to right, from low to high values. We thus refer to signals on the right as upfield or shielded and signals to the
left as downfield or deshielded. Later spectrometers gained the capability of scanning frequency, which then had
to decrease from left to right during the scan, hence the "backwards" nature of NMR scales. units are defined
as follows:
[o(H) - o(TMS)]
=
[Spectrometer Frequency in MHz]

Chemical shifts of all nuclei should be reported using values, with frequency and increasing from right to
left. Many early papers on proton and multinuclear NMR used the opposite convention (not to mention other
references) - in particular the scale was used in the early days: = 10 - . Coupling constants are field
independent, and should always be specified in Hz.

Reich, U.Wisc. Chem. 605 5-HMR-2.1 Delta


The chemical shifts of protons on carbon in organic molecules fall in several distinct regions, depending on the
nature of adjacent carbon atoms, and the substituents on those carbons. The scale below should be used only as a
rough guideline, since there are many examples that fall outside of the indicated ranges. To a first approximation,
protons attached to sp3 and sp carbons appear at 0-5 ppm, whereas those on sp2 carbons appear at 5-10 ppm.

H
X=O,Cl,Br X=N,S
H X H X=O,N,C
H X H X Alkanes
O
H
H R
H

10 9 8 7 6 5 4 3 2 1 0
ppm

Within these ranges, for a given type of C-H bond (sp3, sp2 or sp) the chemical shift is strongly affected by the
presence of electronegative substituents as can be seen in the methyl shifts summarized below, which range from
-2 for MeLi to 4 for MeF.

1
H Shifts of MenX Compounds Hg Cd Zn Be Mg Li

Tl+ B Tl H Ga Al

Sn
C Pb Ge Si

N Sb Bi As P

O S Se Te

F Cl Br I

4.0 3.0 2.0 1.0 0.0 -1.0 -2.0


ppm
The 1H chemical shifts of protons attached to heteroatoms (H-X) show a very wide chemical shift range, with no
obvious correlation to the electronegativity of X or the acidity of HX.

Gas Phase 1H Shifts of H-X Compounds


H-CH3
H-SiH3 H-CN

H-OH
H-NH2 H-SH
benzene

ethylene
H-Hg-R

H-PH2

H-Br
H-Cl
H-F

H-I
H2

20.0 15.0 10.0 5.0 0.0 -5.0 -10.0 -10.5


ppm

Reich, U.Wisc. Chem. 605 5-HMR-2.2 Delta


Calculation of Proton Chemical Shifts

Parameters for the calculation of proton chemical shifts for many kinds of molecules have been tabulated (see
Section 9, Proton NMR Data). All of these work in the same way. We establish the base chemical shift for a
reference substance (e.g., ethylene for olefins, benzene for substituted aromatic compounds, methane for alkanes)
and tabulate Substituent Chemical Shift values () for the introduction of substituents into the reference
molecules. Thus for a vinyl proton (C=C-H) there will be parameters for the introduction of substituents cis, trans, or
gem to the hydrogen we are calculating, and this leads to reasonable estimations for most molecules, as in the
example below (parameters from Section 9-HDATA-6.1). Here 0.15 is the difference between calculated and
observed chemical shifts. However, when there are strong resonance or other electronic interactions between
substituents (as in the -aminoenone below, with 1.70), or strong conformational effects then the predictions
made by these calculations will be less accurate. NOTE: the chemical shift increments were determined in weakly
interacting solvents like CCl4 and CDCl3. They will work poorly for spectra taken in aromatic solvents like benzene or
pyridine (see later section on aromatic solvent shifts).
Calculate Calculate
H
CH3 H 5.25 (Base shift: CH2=CH2) 5.25 (Base shift: CH2=CH2)
0.45 (Zgem Me) Me2N
0.78 (Zgem COMe)
Ph Br 0.45 (Zcis Br) -1.26 (Zcis NMe2)
H O
-0.07 (Ztrans Ph)
Obs: 5.47, Calc: 7.17 4.77 ( calculated)
6.08 ( Calculated) 1.70 5.05 ( observed)
6.23 ( Observed) 0.28
0.15

For aliphatic (sp3) C-H proton chemical shifts we can use the Curphy-Morrison table (Section 9-HDATA-5.1). In
this system there are base shifts for CH3 (0.9), CH2 (1.2) and C-H (1.55) protons, and then corrections are applied
for all and substituents. The corrections for CH3, CH3 and CH protons are slightly different, and no corrections
are applied for alkyl groups.

Calculate Calculate Calculate


Br
H 1.55 (Base shift: tertiary CH) 7.36 (Base shift:PhH) 7.36 (Base shift:PhH)
Ph NO2 0.87 (o-NO2)
0.95 (-Ph for CH) 0.20 (m-NO2)
2.20 (-Br for CH) -0.71 (o-NH2) H -0.22 (m-NH2)
Me Br
0.25 (-Br for CH)
6.82 ( Calculated) 8.01 ( Calculated)
4.95 ( Calculated) H 7.93 ( Observed)
6.62 ( Observed)
5.00 ( Observed) NH2
0.20 0.08
0.05

Reich, U.Wisc. Chem. 605 5-HMR-2.3 Delta


Accuracy of Chemical Shift Calculations
Calculations using simple parameter lists such as in Section 9-HDATA-5.1 and Section 9-HDATA-6.1 will typically
give results accurate to within 0.5 ppm, but there are exceptions:

Multiple Substituents: The more parameters you are adding together, and the larger they are, the less accurate
the calculation is likely to be. This is especially true for electronegative substituents like O, N and Cl if they are
applied several times to the same proton as the examples below. This is perfectly reasonable, since electron
withdrawal from the C-H group becomes progressively more difficult as the C-H group becomes more electron
deficient.

Calculate
Cl H H Me2N
H 1.55 (Base shift: tertiary CH)
O2N C H
2.55 (-Cl for CH) Cl OMe Me2N
Cl MeO
3.05 (-NO2 for CH) Cl OMe Me2N
Me
Obs: 7.26, Calc: 9.2 Obs: 4.97, Calc: 8.9 Obs: 3.02, Calc: 5.60
7.15 ( Calculated)
5.80 ( Observed) 1.94 3.93 2.58

1.35

Cyclic Systems: Calculations are usually poor for cyclic systems, or otherwise conformationally constrained
compounds. The base shift for a CH2 group in an alkane is 1.2 ppm, and this would be the calculated value of any
methylene group in a cycloalkane. The actual shift for methylenes in cycloalkanes varies by 1.7 ppm, from 0.2 for
cyclopropane to 1.9 for cyclobutane, although if you ignore cyclopropane and cyclobutane, the range is only 0.5
ppm. One of the reasons is that in cyclic compounds conformational mobility is greatly restricted, so that less
rotational averaging of various chemical shift anisotropic effects occurs. At low temperatures the axial and C
equatorial hydrogens of cyclohexane differ by 0.5 ppm, the average shift at room temperature is 1.44, close to the
standard value of 1.2. Note especially that the protons on 3-membered rings of all kinds are strongly shifted to
lower frequency from the acyclic value.

Standard CH2
H 1.62
CH2
-103 C
1.20 0.22 1.94 1.51 1.44 H 1.14

Even more dramatic chemical shift effects are seen in polycyclic compounds. The Curphy-Morrison calculated C
values for all of the compounds below would be 1.55 (the base value for a methyne group), yet the actual values
vary by several ppm. Not sur[risingly, cubane and dodecahedrane are especially far from the typical values.

Standard CH H
H 1.74 H 3.38
C 4.0
CH3 H H
H CH3 H 1.13 2.50
CH3
1.55 2.19

Reich, U.Wisc. Chem. 605 5-HMR-2.4 Delta


Reproducibility of Proton Chemical Shifts
It is important to understand that the chemical shift of a given proton is not an invariant property of a molecule
(like a melting point or boiling point), but will change depending on the molecular environment. The variability is
especially large for NH and OH protons (several ppm), but even for CH protons reported shifts vary by a few tenths
of a ppm. This is in part due to changes in measurement conditions, but additional variability in chemical shift is
present in old NMR data (CW spectra) since spectrometer calibrations and spectrum referencing were not nearly as
accurate as they are today. Nevertheless, if conditions are rigorously controlled, very high reproducibility of chemical
shifts can be achieved. Databases of precise chemical shifts for many biomolecules have been created which
facilitate simultaneous detection by NMR in aqueous solution.
Solvent effects. The aromatic solvents benzene and pyridine cause changes in chemical shifts as large as 0.5
to 0.8 ppm compared to less magnetically active solvents like chloroform or acetone. Since the standard solvent for Cr
chemical shift parameters like the Curphy-Morrison ones is CCl4 or CDCl3, expect less accurate calculations for
spectra taken in aromatic solvents.
Concentration dependence. Chemical shifts of C-H protons can vary with concentration, especially if
intermolecular hydrogen bonding can occur, as for many amines, alcohols and carboxylic acids. The chemical shifts
of protons on oxygen (OH) and nitrogen (NH), which are often directly involved in hydrogen bonding are especially C
strongly dependent (several ppm) on concentration, solvent and temperature. Aromatic molecules can also show
significant concentration dependence because of the aromatic solvent effect mentioned above.
Temperature dependence. For molecules that are conformationally flexible, the populations of conformations
change with temperature. Since the chemical shifts of various conformations are different, the chemical shifts will
vary with temperature (the observed chemical shift is the weighted average of the shifts of the individual
conformations). Temperature will also affect the degree of intermolecular hydrogen bonding or other types of
aggregation, and this provides an additional source of shift changes.
Paramagnetic impurities (unpaired electrons, transition metals with unpaired spins) can cause very large
shifts (tens and hundreds of ppm) as well as large amounts of line broadening. Must avoid these alltogether if you
want to get high quality NMR spectra.

Reich, U.Wisc. Chem. 605 5-HMR-2.5 Delta


Proton Chemical Shift Effects
Cr
1. Electronegativity. Proton shifts move downfield when electronegative substituents are attached to the same
or an adjacent carbon (see Curphy-Morrison chemical shift table). Alkyl groups behave as if they were weakly
electron withdrawing, although this is probably an anisotropy effect.

CH3F CH3Cl CH3Br CH3I CH3CH3 CH4 CH3SiMe3 CH3Li


4.26 3.05 2.69 2.19 0.96 0.2 0.0 -2.1

The chemical shifts of protons attached to sp2 hybridized carbons also reflect charges within the system
(approximately 10 ppm/unit negative or positive charge).
+
CMe2 CH2Li
H 9.64 H 6.28
CH3
+ H H H H 8.80 6.09
H 13.50
8.97 + 2.46
CH3 7.97 6.30
H H H H
5.06
8.45 5.50
H 10.3 H 5.37 H 7.27 H 9.17

+ - +
Pr Pr
Even without formal charges, resonance interactions can lead to substantial chemical shift changes due to
polarization.
H H 5.25 EtO H 4.03 CH2 2.77 O2N H 6.55

N N
H H 6.32 H H 3.86 7.12 H H 5.87

This is especially useful in the interpretation of the NMR chemical shift of protons in aromatic systems. The
protons ortho and para to electron donating and electron withdrawing substituents show distinct upfield and
downfield shifts.

m o
NH2
p 3.13
2.00

OMe m o
p 3.00
2.00

CH3 5.00

NO2 o m
p 3.12
2.00

8.5 8.0 7.5 7.0 6.5


ppm Click for Spectra

Reich, U.Wisc. Chem. 605 5-HMR-2.6 Delta


2. Lone Pair Interactions. When lone pairs on nitrogen or oxygen are anti to a C-H bond, the proton is shifted
upfield (n --> * interactions). There is thus a strong conformational dependence of chemical shifts of protons to
heteroatoms. This interaction is one of the reasons that Curphy-Morrison chemical shift calculations work poorly
when multiple O or N substituents are attached to one carbon. This effect is also present in 13C chemical shifts. C-H
bonds anti to lone pairs also show Bohlmann bands in the IR spectra, as a result of weakening of the C-H bond by
hyperconjugation. For example, the = 180 compound shows IR absorption at 2450 cm-1, as well as at 2690-2800
cm-1.

:
:
EtO
Electron donation N
to C-H bond gives N H C H
EtO
a - (upfield) shift
H EtO
effect
Little C-M calculation: 7.85
interaction Observed: 4.96
Curphy-Morrison calculation would give 5.60 for all of these:
H 3.67
H 5.03 H 2.25
3.02

:
Me2N H
H N N
:

C H N N N N
:

H Me2N N
N N H N
Me2N

:
= 0 = 60 = 180
Weisman TL 1980, 3635 Stetter Ber 1973, 2523 Weisman TL 1980, 3635

3. Steric Compression. When molecular features cause a proton to be forced close to other protons, or to
various functional groups, the proton will in general be deshielded (dispersion interactions). Shifts of this type are
hard to distinguish from magnetic anisotropy interactions.
CH3 C(CH3)3
3.45 3.65 7.10 7.27
H H CH3 H H H
CH3 OH CH3 OH

2.91 7.73 8.53


H 1.32 2.68 0.92
H (CH3)3C H 1.03
H
H H
H O Ph
N N
5.44 H
N N
Ph
The N-H distance is 2.25 A JACS 1968, 3724 JOC 1967, 1304

These shifts are especially large in highly compressed compounds like the "birdcage" molecules. The inside
proton in the "out" alcohol A at 4.48 is downfield by 0.96 ppm from the model B. Even more striking are the shifts
in the "in" alcohol C, where the proton jammed into the OH group at 3.55 is downfield by 2.3 ppm from the model
D, and the gem partner at 0.88 is actually upfield by 0.5 ppm from its position in D, suggesting a migration of
electron density from the sterically compressed inside H to the outside H.

<2.4 4.48 3.52 3.55 H 3.92 1.4 1.2


H HH OH H OH 0.88 H HO H H H

A B C D
JACS 1965, 5247 JACS 1965, 5247

Reich, U.Wisc. Chem. 605 5-HMR-2.7 Delta


4. Magnetic Anisotropy. Whereas the local circulation of electrons around HA is a shielding effect (i.e., to the
right in the NMR spectrum, -), there can be both shielding and deshielding effects on HA from electron motion in
other parts of the molecule. We refer to such interactions as magnetic anisotropy effects, since they are caused
by anisotropic electron circulation (i.e., the electron circulation is stronger in some orientations of the molecule in the
magnetic field than in others).

The most dramatic examples of anisotropy effects are seen with benzene and other aromatic rings, which cause
very large shielding (-) effects for protons placed above the ring, and smaller deshielding (+) effects for protons to
the side of it. These chemical shift effects occur because electron circulation is stronger when the plane of the
benzene ring is perpendicular to the magnetic field than when it is parallel to it
The local magnetic field is higher here, so a higher
frequency or lower external magnetic field is needed to
achieve resonance. Signal is deshielded.

H When the benzene ring is oriented with the ring


parallel to the magnetic field, the electron
Ho H circulation is much weaker. The shielding effects
H in these orientations do not cancel the
deshielding effects in the other orientation.
H

The local magnetic field is lower here, so a lower frequency


or a higher external field magnetic field is needed to achieve
resonance. Signal is shielded.

The consequence of magnetic anisotropy effects is to provide a stereochemical component to the chemical shift
of a nucleus: the chemical shift changes depending on the spacial relationship between a proton and nearby
functional groups. Such effects can be valuable for making stereochemical assignments. Some proposed
magnetic anisotropy shielding/deshielding cones are shown below:

- - -
- +
+ + O
+ O + + N
+ S O + O +
- C C -
- - -
-

Alkene Carbonyl Sulfoxide3 Alkyne Nitro4

- - -
+ H +
H
+ H
+ H + +
+ - O +
- H H
- - +
H
H
C-C Single Bond Cyclohexane Ax-Eq Cyclopropane2 Epoxide1

1. H. C. Brown, A. Suzuki J. Am. Chem. Soc. 1967, 89, 1933. L. A. Paquette, G. Kretschmer, J. Am. Chem. Soc. 1979, 101, 4655.
2. C. D. Poulter et al. J. Am. Chem. Soc. 1972, 94, 2291.
3. The Thiosulfinyl Group Serves as a Stereogenic Center and Shows Diamagnetic Anisotropy Similar to That of the Sulfinyl Group: Tanaka, S.; Sugihara,
Y.; Sakamoto, A.; Ishii, A.; Nakayama, J.; J. Am. Chem. Soc., 2003, 125, 9024.
4. Magnetic Anisotropy of the Nitro Group by NMR I. Yamaguchi, Mol. Phys. 1963 , 6, 103

Reich, U.Wisc. Chem. 605 5-HMR-2.8 Delta


Aromatic Chemical Shifts. The ring current in Huckel aromatic systems, i.e., those with 4n + 2 electrons (2, n
6, 10, 14, 18 ...) causes downfield shifts in the plane of aromatic ring.
5.76 5.77 7.75
H 1.64 2.36 H H
5.87 7.27 1.66 7.38
5.72
H H H CH3 CH3 H H

8.59
N N H
= 1.40 ppm = 0.72 ppm = 1.66 ppm
H
When protons are above or below the plane (or in the middle) of the aromatic ring then upfield shift effects are
observed.
9.32 9.82
-5.49 (d, J=13) H H (t, J=13)
H t
Bu But
H 8.58 (s)
14 e- H
H -4.03 18 e-
-3.64
H (t, J=13)

Pascall JACS 1987, 6878


H 8.50 (d, J = 7.5 Hz) t
Bu But
Boekelheide JACS 1970, 3511 Nakagawa TL 1973, 4743

H H -0.5 H -0.25 -1.74 10.31 14.19


H 2.55
H H H H

H 6.95 H 8.47 + 3.92


8.9 1.21
H 7.27 H 5.5 8.73 3.16
9.41 8.96 2.31 2.99
J = 9.5 Hz J = 1.3 Hz
10 e- Aromatic Nonaromatic
10 e- (aromatic) 12 e- (antiaromatic)
Vogel AC 1964, 145, 784 Vogel TL 1966, 655
Staley JACS 1973, 3384
When a cyclic conjugated system is planar and antiaromatic, i.e., 4n electrons (4, 8, 12, 16 ... ) then chemical
shift effects are in the opposite direction: downfield over the ring, and upfield in the ring plane. This is seen in the
Staley 10 and 12-electron methano annulene cation and anion above, as well as in the 14-electron dihydropyrene
below. The normal chemical shift effects are seen in the 10 and 14-electron systems. In the 12 and 16 -electron
anions the methylene bridge and propyl groups over the ring show very large downfield shifts as a result of the
antiaromatic ring current. The paramagnetic ring currents are a consequence of the small HOMO-LUMO separation
that is characteristic of 4n (antiaromatic) systems.

H 5.40 H 8.83
0.65 5.51 H 7.45
H
-3.95 1.87 21.24 12.59
H H
7.95 -2.56 to
-3.14 H H
10.43 -8.17

H H
-
14 e (aromatic) 16 e- (antiaromatic) 16 e- (antiaromatic) 18 e- (aromatic)
Boekelheide JACS 1969, 4931 Oth TL 1968, 6265

In the [16]-annulene the neutral compound has antiaromatic character. The shifts were measured at low
temperature, where conformational averaging has stopped. In the 18-electron dianion, large aromatic shifts are
reported.

Reich, U.Wisc. Chem. 605 5-HMR-2.9 Delta


Chemical Shift Effects of Phenyl Groups. The effects of a phenyl substituent are highly dependent on
conformation. For example, for styrenes the chemical shift effect of the phenyl is downfield when the phenyl is
in the plane of the double bond, but upfield when the rotamer with the phenyl group perpendicular is the more
stable one:

H H
H H
If ring is flat, get If ring is perpendicular,
downfield shifts (+) get upfield shifts (-)

CH3 CN CH3 CN
CN
CN
H H
5.46 5.22 CH3 H
CH3 H
5.48
5.31
H cis to Ph is downfield H cis to Ph is upfield - the ortho-methyl substituent
presumably rotates the Ph group out of the C=C plane.
TET 1970, 4783

The large differences in chemical shifts of the butadienes below can also be used to assign stereochemistry,
based on the effect of the "rotated" benzene ring when it is cis to the other vinyl group.

7.30 7.32
6.63
Cl H
Br H H
Br
Br
Br
H
H Br H
6.64

Reich J. Org. Chem. 1975, 40, 2248

If steric effects force a phenyl to adopt a face-on conformation (as in the lactone example below) then a cis
CH3 group will be shifted upfield compared to a trans group.
1.1
0.7
CH3 H
HO CH3 H

I JOC 1982, 3943


O O I
H H
O HO
O

Reich, U.Wisc. Chem. 605 5-HMR-2.10 Delta


Determination of Enantiomer Ratios and Absolute Configuration with Mosher Esters. Esters of
2-phenyl-2-methoxy-3,3,3-trifluoropropionic acid (Mosher esters, or MTPA esters) with secondary alcohol show
characteristic chemical shift effects in the alcohol portion which can be used to measure enantiomeric purity and
assign the absolute configuration of the alcohol. It is necessary to assign key protons, and to make both the R- and
S-Mosher ester to arrive at an unambiguous determination (Dale, J. S.; Mosher, H. S. J. Am. Chem. Soc., 1973,
95, 512; Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem. Soc., 1991, 113, 4092).
This method works because the principal conformation of MTPA esters is the extended one shown. The
anisotropy of the phenyl group then causes upfield shifts of the protons behind the plane of the paper, downfield
shifts for those in front. A typical procedure is to do a complete analysis of all assignable protons of the R and S
esters, and calculate the difference between the chemical shifts of the two diastereomers. Note that the t-Bu group
is upfield in the R,S ddistaereomer, whereas the Me group is upfield in the R, R isomer.

MTPA R,R
upfield R,S 0
MeO Ph H +15
- CH MeO O H
Me H
3
+65
O Ph C C O C CMe3 H
+ (CH3)3C R CF3 -15
R CF3 CH3 H H +30
H O -15 H OR
R,R -50 H
downfield +20
R,S H H H +10
-10
-60 -155

MeO Ph S - R (Hz at 500 MHz)


- (CH3)3C
O
+ CH3 R CF3
S
H O

1.5 1.0 0.5


ppm

For a related method using 1-phenyltrifluoroethanol, see Org. Lett. 2003, 5, 1745.

- Pr Even the remote


+ Me H Ph
O n-propyl CH3 group
CF3 splits nicely:
H Si O (R)

+ Pr
- Me H CF3
O Note 8% of other enantiomer
Ph
H Si O (S)

1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7


ppm
"Chiral Reagents for the Determination of Enantiomeric Excess and Absolute Configuration Using NMR
Spectroscopy." Wenzel, T. J.; Wilcox, J. D. Chirality 2003, 15, 256-70. " The Assignment of Absolute
Configuration by NMR," Seco, J. M.; Quinoa, R.; Ricardo, R. Chem. Rev. 2004, 104, 17. Parker, D. "NMR
Determination of Enatiomeric Purity" Chem. Rev. 1991, 91, 1441

Reich, U.Wisc. Chem. 605 5-HMR-2.11 Delta


Aromatic Solvent Induced Shifts (ASIS). Polar molecules have substantially different chemical shifts in aromatic
solvents (benzene, pyridine, C6F6) than in less magnetically interactive solvents like CCl4, CDCl3, CCl2D2, acetone-d6 N
and CD3CN. A typical result of going from CDCl3 to benzene is shown in the spectra of butyrophenone below. The
shifts are large enough that chemical shift calculations can be seriously in error when applied to molecules whose
spectra were taken in benzene (P. Laszlo Progr. NMR Spectrosc. 1967, 3, 231).

Click for full Spectra


O
Calculated
Curphy-Morrison

300 MHz
benzene
Source: Amanda Jones (HJR) 02-09
-0.47 ppm

C6D6

CDCl3

3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6
ppm
The origin of these chemical shift effects is believed to be a partial orientation of the solvent by the dipole moment
of the solute. For benzene, the shifts can be rationalized on the basis of a weak and transient complexation of the
electron-rich -cloud of the aromatic ring with the positive end of the molecular dipole, such that the protons spend
additional time in the shielding (-) region above and below the benzene ring. There is a strong correlation between
the dipole moment and the size of the solvent shift. With occasional exceptions, the benzene shifts are upfield
(-).
(CCl4 vs. C6D6):
0.82 4.91
Effect of dipole moment on ASIS: +0.03 C6F6 -0.00 C6F6
-0.20 C6D6 -0.18 C6D6
MeSnCl3 -1.43 3.6 -0.12 Py +0.06 Py
H
MeSnI3 -1.02 2.6 H
MeCCl3 -0.59 1.5 1.37 5.86
Me4Sn -0.09 0 +0.12 C6F6 -0.20 C6F6
-0.31 C6D6 +0.24 C6D6
-0.29 -0.52 O +0.22 Py
O -0.16 Py H
CH3CN -0.45 H 1.30
N 0.00 C6F6
-0.95 H -0.03
N -0.28 C6D6
-0.77
-0.14 Py
O O O K. Tori Tetrahedron Lett. 1975, 2199.
+0.02 O
-0.65
N -0.62 O

-1.05
S+ S+
-0.36
-0.35 O
O Me (CDCl3) 1.40 1.23
O O CH2
Me (C6D6) 1.18 0.74
S S
-0.21 -0.64 -0.35 -0.05 -0.22 -0.49
-0.45 -1.00 -0.62 -0.17 JOC 1970, 3655
JOC 1968, 2555 JOC 1968, 2555 JOC 1968, 2555 JOC 1968, 2555

Reich, U.Wisc. Chem. 605 5-HMR-2.12 Delta


When 1H NMR spectra are complicated by accidental superposition of coupled protons, as in the spectrum of
eugenol in CDCl3 below, then switching to benzene as solvent (or even just adding a few drops of C6D6 to the
sample) will often move signals enough that more interpretable (first order) spectra result. In the CDCl3 spectrum of
eugenol H2 and H6 are nearly superimposed, leading to a complex ABX pattern of the Solution 2 type. The spectrum
in C6D6 is essentially first order.

Eugenol C9H12O2
200 MHz 1H NMR spectra H6
Source: I. Reich H5

H5 HO H2
OMe H2

H2
H6
H6
H5
C6D6

CDCl3

7.0 6.9 6.8 6.7 6.6 6.5 6.4


ppm

Effect of benzene to simplify a strongly coupled NMR spectrum.

Reich, U.Wisc. Chem. 605 5-HMR-2.13 Delta


Anisotropy of Double Bonds. The magnetic anisotropy of C-C double bonds has generally been assumed to be
similar to that of aromatic rings, with a deshielding region in the plane of double bond. This explains both the
downfield shifts of vinyl protons, and the larger downfield shifts of the internal (which are affected by the anisotropy
of both systems) versus the terminal protons in conjugated dienes. It also explains the downfield shifts of allylic
protons.

- 6.26
5.25 0.92 1.60 H H
H H Me Me
+ + 5.05 H
H
- H H 5.16 H H
The shielding region above and below the plane of the double bond is more controversial. A number of
examples show the expected upfield shifts of protons above double bonds.

-0.17 3.53 3.75 1.17, 1.01 1.27 0.85


0.83
H OH HO H Me Me Me Me
H H
Me Me

0.70 0.79 1.23 0.72


H 1.44
0.22 CH3 H H CH3 Me Me
H H H
2.88 CH2
at -150 C

There is, however, one major exception. In norbornene itself, the proton shifts are in the opposite direction than
seen in the 7-substituted norbornenes above (J. Am. Chem. Soc. 1968, 90, 3721). Both the proton assignment and
the absence of a - region above the double bond are supported by high level ab initio MO chemical shift
calculations (J. Am. Chem. Soc. 1998, 120, 11510). Thus the deshielding region above double bonds shown in the
figure must be viewed with some skepticism.
1.32 H H 1.06

For this reason, assignment of stereochemistry in cyclopentanes based on an assumed anisotropy of double
bonds, as in the examples below, should be used with caution. Possibly the shifts are the result of C-C single bond
anisotropy of the C-vinyl bond.
0.99 4.06 4.01
O 0.82 O CO2CH2CH3
H
H 0.95 CO2CH2CH3 1.18

H H H H
J. Org. Chem. 1970, 35, 2688 Chem. Commun. 1988, 760

O 0.87 O CO2H
When the methyl and vinyl groups
1.17 are cis, the methyl group is shifted
CO2H
upfield.
H H
Vollhardt J. Am. Chem. Soc. 1980, 108, 5253.

Reich, U.Wisc. Chem. 605 5-HMR-2.14 Delta


Anisotropy of Carbonyl Groups. The magnetic anisotropy of C=O has a strongly deshielding (+) region in the
plane of carbonyl group. This accounts for numerous chemical shift effects in aryl ketones, ,-unsaturated
carbonyl compounds, and conformationally rigid ketones, and is reliable enough to be used for structure
assignments.
The effect is seen both when the proton is to the carbonyl group, as in the enones and acetophenones below,
or when there is a -relationship.

O
H 5.13 H 4.64
-
H 0.62
+ H 5.83 H 4.92
O +
O H 0.14
-
J. Org. Chem. 1992, 57,1970 H 0.21

The tetralone below shows a strongly downfield shifted ortho-proton, to 8.0. The ortho-methyl acetophenone,
on the other hand shows as smaller downfiled shift ( 7.7), probably due to some rotation of the carbonyl group out
of the plane, as well a preference for the conformation shown, with the smaller C=O group cis to the ortho-CH3
group. Acetophenone ortho proton appears at 7.9.
Ho O
Hm 300 MHz 1H NMR spectra
Source: Aldrich Spectra Viewer

Hp
Hm'

Ho Hp Hm Hm'
Ho
Hm = 0.35 = 0.09
O

Hp
Hm'
8.1 8.0 7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2 7.1

In the compounds below, the proton is to the carbonyl and close to same plane, leading to quite large downfield
shifts:

7.94 9.41
O 8.24 HO
H H H H
7.46 7.62 7.53
H H H O
H H O
1.1 3.68
Magn. Res. Chem. 1989, 27, 796

Reich, U.Wisc. Chem. 605 5-HMR-2.15 Delta


In the stereoisomer A below, one of the aromatic protons is close to the carbonyl, and is shifted downfield by 1.3
ppm, whereas in isomer B the carbonyl is remote, and the chemical shift is normal.

H
7.7 O 6.25 O O
H N H N

O O O
H H
N H N H
O O
A B
6.4 H O O 6.3 H O
O
J. Am. Chem. Soc. 1967, 89, 3600

These ,-unsaturated esters show a shift range of 1.7 ppm resulting from the various - and -carbonyl
interactions. In the most upfield shift ( 6.50 for the E,Z-isomer) there are no close interactions, whereas the most
downfield proton ( 8.20 for the same isomer) has a -interaction with one carboxylate function, and a -interaction
with the other:

5.88 7.87 5.80 6.50 O 7.34


H H O H H 5.90 EtO H 6.21
OEt H H
EtO EtO H
6.21 H OEt
O H H O H OEt
7.87 5.88 8.20 O 7.34 O

Amides also show these chemical shift effects. Thus, for the two rotamers of the formamide below, the -N
proton is 0.9 ppm downfield in the isomer with this proton close to the formyl oxygen (Buchi, G.; Gould, S. J.; Naf,
F. J. Am. Chem. Soc. 1971, 93, 2492 )
O 8.2 H 8.6
6.5 (d 5) 5.6 (d 5)
H H H O
N N

O O

N N
H H
There is some evidence that there is a shielding (-) region above the plane of the carbonyl group:

0.90 0.85
CH3 CH3

Reich, U.Wisc. Chem. 605 5-HMR-2.16 Delta


Anisotropy of Nitro groups. The NO2 group may have a a small anisotropic effect similar to that of C=O groups,
with a deshielding (+) region in the plane of carbonyl group. The ortho protons of nitrobenzenes are strongly
downfield, in part due to this interaction. For example the proton Ha between the NO2 and Br groups (the small
downfield doublet) has a very similar electronic environment in the two compounds whose spectra are shown below.
The upper one has this proton upfield in part because the ortho-methyl group turns the nitro group out of the plane.
Of course, turning the nitro group also causes reduced resonance interactions, which causes a shift in the same
direction, as seen from the change in the proton ortho to the Me group (Hb).

O
H
Br N+
300 MHz 1H NMR spectra Ha
Source: Aldrich Spectra Viewer O
Br NO2 CH3

H CH3
Hb
-
O
+ N O +
Ha
Ha Hb
O2N Br
-

CH3
Hb

8.4 8.3 8.2 8.1 8.0 7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2
ppm

A similar chemical shift effect in a naphthalene is illustrated below:

0.79 H NO2 0.02 H NO2


Me

0.17 H H

Reich, U.Wisc. Chem. 605 5-HMR-2.17 Delta


Anisotropy of Acetylenes. The magnetic anisotropy of CC bonds seems to be well-defined. Both the unusual
upfield shift of CC-H signals, and the downfield shifts of protons situated next to a triple bond as in the examples
below support a strong diamagnetic affect of electron circulation around the triple bond system. .

+
CH3 CH3 CH2=CH2 HCCH
- C C -
0.90 5.25 2.88
+
H

H
2.48 CH H 3.01 CH3
H 3
8.64 H 10.27 H

CH3 2.52 CH3


1.63 ppm
JOC 1984, 1323
0.49 ppm
JOC 1984, 1323

Anisotropy of Nitriles. The cyano group presumably has the same anisotropy as the alkynyl group, as shown by
the examples below.

N N
H H H C
C H 9.7 H
OH

H C
0.98 0.96 t, J=11.5 Hz 1.24 N
Ed Piers, J. Wai, private commun 1.0 ppm
ACIE 1975, 264

Anisotropy of Halogens. Protons positioned near lone-pair bearing atoms such as the halogens generally show
downfield shifts, as in the phenanthrene examples below. Interpretation of these values is complicated by the
close approach of the X and H atoms, which can cause geometry and orbital distortions and affect the chemical
shifts.
8.64 H H X
X
F 9.15 0.56
Cl 9.6 1.16
Br 9.83 1.39
MRC 1989, 13 I 9.9 1.46
ASV
Tet 1969, 4339

Reich, U.Wisc. Chem. 605 5-HMR-2.18 Delta


Single Bond Anisotropy. Because of the many single bonds in typical organic molecules, each with local NA
anisotropic effects, it has been hard to define single bond chemical shift effects, and even harder to make practical
use of them. Nevertheless, useful stereochemical effects have been identified in several situations, loosely based
on a magnetic anisotropy of C-C single bonds in which flanking hydrogens are shifted upfield, end-on hydrogens
downfield. N
Axial and Equatorial Cyclohexane Shifts. In cyclohexane itself, as well as in most substituted and heterocyclic
6-membered rings the axial protons are upfield of the equatorial ones. Unfortunately, there are a few exceptions, and
so this chemical shift effect must be used with caution. Below some e-a values:
e-
a

X
X
CH2 0.52 0.52 0.52 -
+
NH 0.48 0.12 0.45
H 1.62 H
NH2+ 0.47 0.16 0.34 +
H 1.14 O 0.50 -0.07 0.32 -

At -103 C (Garbisch, J. Am. S -0.19 0.38 0.50 H


Chem. Soc. 1968, 90, 6543)
SO2 <0.10 0.17 0.45

One explanation for this shift effect is based on the anisotropy cones shown in the figure, where the equatorial
protons reside in the deshielding (+-) region of the C-C anisotropy, and the axial in the - region. An alternative
explanation, or additional contributing effect, is based on the supposition that a C-H bond is a stronger donor than C
a C-C bond, which leads to increased electron density in the axial protons (anti to two C-H bonds), hence -. The
variation in 1JCH has also been interpreted in these terms.
A more complicated bicyclic ring system shows several shifts that are consistent with the chemical shift effect eq
> ax, and one exceptions:
5.22
5.25
4.17 5.66
2.78
H
H H 2.47
O H 2.41
H 2.32
H O MRC 1987, 843
O H 2.50
3.96 H 2.27
H
1.74 H H 2.38
2.04

Substituent effects on cyclohexanes (Anteunis Tetrahedron Lett. 1975, 687):

+0.08 +0.16 +0.07 +2.26 +0.45


H H H H H OH
+0.04 +0.01 -0.27
H CH3 H OH H H +2.29
-0.01 H H +0.06 H H -0.08 H H
+0.02 +0.20 -0.01

H H H H H H
-0.03 -0.27 +0.04 -0.04 +0.06 +0.27

Reich, U.Wisc. Chem. 605 5-HMR-2.19 Delta


Assignment of syn and anti Aldol Adducts. A similar type of single bond anisotropy has been used to rationalize
the empirical observation of a systematic variation in the chemical shift of the CHOH proton in syn and anti isomers
of aldol products (syn > anti) that can be used to assign configuration, although such assignments should be viewed
as less definitive than other methods, because of the usual problem with interpreting small chemical shift differences
(Kalaitzakis, D.; Smonou, I.; J. Org. Chem. 2008, 73, 3919-3921). The argument is that in the favored conformation
of the hydrogen bonded anti isomer the carbinol proton is in a pseudo-axial orientation subject to similar anisotropy
effects as an axial cyclohexane proton, whereas in the syn isomer the proton is pseudo-equatorial.

H
OR
OH O O
Me
H O H
O 3.81
Et
syn
H
OR
OH O O
Me
Et O H
O
3.57 H
anti

Cis-Substituent effect in Rigid Rings. Chemical shifts in rigid bicyclic or polycyclic systems can provide some
insights into general chemical shift effects, although care must be utilized because there are typically a number of
effects operating simultaneously. One example is the tendency for eclipsed or nearly-eclipsed cis-vicinal
substituents to cause upfield shifts relative to the trans proton (and also relative to the compound with hydrogen
replacing the substituent). In the dibenzobicyclo[2.2.2]octadiene system A the proton which is eclipsed (or nearly
so) with the R substituent is always upfield of the one trans to it, and upfield of the unsubstituted compound as well.
For the hexachloro bicyclo[2.2.2]heptane B this is also seen, although here the inherent shift difference is not known
since the compound with R = H has not been reported.
R c t X R c t X
HX H
R H 1.68 1.68 1.68
Cl6
Hc CN 2.15 2.93 3.40
Ht OAc 1.41 2.25 4.90 HX
CO2H 2.43 2.55 3.62
OH 1.18 2.19 3.94 Ht C6H5 2.38 2.83 3.87
OTs 1.50 2.08 4.88 R Cl 2.22 3.08 4.72
NH2 0.97 2.14 3.11
Hc OH 1.90 2.78 4.63
SPh 1.44 2.28 3.52
A B OAc 1.90 2.95 5.50
JOC 1966, 581 JACS 1963, 516

The upfield shift of cis substituents compared to trans is also seen in a series of succinic anhydrides:

X 3c 3t Jtrans Jcis

H 2.94 2.94 5.2 10.7


O Me 2.65 3.18 6.9 10.0 ASV
O O
AcO 3.03 3.39 6.3 9.6 ASV
AcS 2.97 3.48 6.9 10.7 ASV
H3c X HOCHN 2.87 3.26 6.2 9.9 ASV
H3t H2 CF3(O=)CHN 3.00 3.31 6.2 9.9 ASV
Ph 3.09 3.43 6.6 10.3 ASV

Reich, U.Wisc. Chem. 605 5-HMR-2.20 Delta


Stereochemical Relations in Cyclopentanes. Because coupling constants are not very reliable for determining
stereochemical relationships in 5-membered rings, chemical shift effects such as the one discussed above have
been utilized more extensively than in cyclohexanes. It has been observed that in cyclopentanes, -butyrolactones
(Ollis JCS-PT1 1975, 1480) and tetrahydrofurans the diasterotopic chemical shift effect of a ring CH2 group is
consistently larger when flanking substituents are cis to each other (when the anisotropic effects of the C-C or C-O
bonds are additive) compared to when they are trans (both protons see the effect). More specifically, protons with
cis-vicinal substituents are generally shifted to lower values (upfield) than those with cis hydrogens.

< 0.3 ca 1.0 < 0.2 > 0.5


H BzO H BzO H
H H
H H upfield
H

OR OR
C6H13 C5H11 C5H11 O
O C6H13 O O
JACS 1984, 2641
JACS 1992, 7318

2.0 1.25 0.75


2.65 2.32 2.73
H H H
2.0 H Ph H Ph
1.4 H 2.50 H 1.98 H

O O O
O O O O
O
JCS P1 1975, 1470 JCS P1 1975, 1470

Similarly, the chemical shift of a proton will be a function of the number of cis-alkyl substituents on the ring. To
use such chemical shifts it is necessary to have several members of a series for comparison.
3.9
SePh SePh H
H 2.8 H 3.5 SePh

O C6H13 C6H13 C6H13


O O

JACS 1992, 7318


(see also TET 1986, 3013)

Reich, U.Wisc. Chem. 605 5-HMR-2.21 Delta


Anisotropy of Cyclopropanes. The principal magnetic anisotropy of cyclopropane groups appears to be
shielding above the ring and deshielding in the plane of the ring, a ring current effect a little like that of benzene.

3.99 3.57 4.24


- HO H
HO H HO H H 0.67
H 1.47
+ + H
H H
-
H 2.32 0.44

JACS 1972, 2291 TET 1998, 337 TET, 1966, 2007 TET 1966, 2007 JACS 1972, 2291 JACS 1966, 5272
TET 1998, 337

O O
O O
H 4.95
O H O H
H 3.13 H 3.30
H 5.57 O O

JOC 1967, 939 H H


7.42 6.91 JACS 1965, 386 JACS 1965, 386

Reich, U.Wisc. Chem. 605 5-HMR-2.22 Delta


5. Hydrogen Bonding Effects on Chemical Shifts - OH, NH and SH Protons. The chemical shifts of OH and
NH protons vary over a wide range depending on details of sample preparation and substrate structure. The shifts
are very strongly affected by hydrogen bonding, with large downfield shifts of H-bonded groups compared to free OH
or NH groups. Thus OH signals tend to move downfield at higher substrate concentration because of increased
hydrogen bonding. Both OH and NH signals move downfield in H-bonding solvents like DMSO or acetone.
There is a general tendency for the more acidic OH and NH protons to move further downfield. This effect is in
part a consequence of the stronger H-bonding propensity of acidic protons, and in part an inherent chemical shift
effect. Thus carboxylic amides and sulfonamides NH protons are shifted well downfield of related amines, and OH
groups of phenols and carboxylic acids are dowfield of alcohols.

Recognizing Exchangeable Protons. In many samples NH and OH protons can be recognized from their
characteristic chemical shifts or broadened appearance. When this fails, the labile protons can be identified by
shaking the sample with a drop of D2O, which results in disappearance of all OH and NH signals. This works best if
the solvent is water immiscible and more dense than water (CDCl3, CD2Cl2, CCl4) since the formed DOH is in the
drop of water floating at the top of the sample where it is not detected. In water miscible solvents (acetone, DMSO,
acetonitrile, pyridine, THF) the OH and NH signals are largely converted to OD and ND, but the DOH formed
remains in solution and will be detected in the water region.

Hydroxyl OH Protons. In dilute solution of alcohols in non hydrogen-bonding solvents (CCl4, CDCl3, C6D5) the
OH signal generally appears at 1-2 At higher concentrations the signal moves downfield as a result of increased
fraction of H-bonded alcohols, e.g. the OH signal of ethanol comes at 1.0 in a 0.5% solution in CCl4, and at 5.13
in the pure liquid (from Bovey).

60 MHz NMR spectra of ethanol at various concentrations


(from Bovey, p 84).
O
H-O H CH2-CH3
Neat

10% EtOH in CCl4 H-O

H-O

5% EtOH in CCl4

0.5% EtOH in CCl4


H-O

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5
ppm

Reich, U.Wisc. Chem. 605 5-HMR-2.23 Delta


Dynamic Exchange. Under ideal conditions OH groups of alcohols can show sharp signals with full coupling to
neighboring protons even at room temperature, as in the spectrum of neat ethanol above, and in the spectrum of
1-phenyl-4,4-dimethyl-1-pentyn-3-ol below.

300 MHz 1H NMR spectrum in CDCl3 3


JH-OH = 6.2 Hz
Source: Olafs Daugulis/Vedejs H

1275.1

1268.9
O H
Ph H-O

4.25 4.20 1.85 1.80

7 6 5 4 ppm 3 2 1 0
More typically, signals for OH protons are subject to rapid (on the NMR time scale) intermolecular exchange
processes, which may result in broadening or complete loss of coupling to neighboring protons. Such exchange can
also broaden or average the signals of multiple OH, NH or SH groups in the sample, if more than one is present.
Any water present might also exchange with the R-OH protons. The rates of exchange are a complex function of
temperature, solvent, concentration and especially the presence of acidic and basic impurities. In CDCl3 the
presence of acidic impurities resulting from solvent decomposition often leads to rapid acid catalyzed exchange
between OH groups. In contrast, solvents like DMSO and acetone form strong hydrogen bonds to the OH group.
This has the effect of slowing down the intermolecular proton exchanges, usually leading to discrete OH signals with
observable coupling to nearby protons. Note the triplet and doublet for the HOCH2 group in the spectrum below
taken in DMSO.
300 MHz 1H NMR spectrum in DMSO-d6
H2N
Source: Aldrich NMR Library

HO

7.0 6.5 6.0 ppm 5.5 5.0 4.5

In the remarkable NMR spectrum of the OH region of sucrose below (Adams, Lerner J. Am. Chem. Soc. 1992,
114, 4828) all of the OH signals and their coupling are resolved in aqueous acetone solvent.
500 MHz 1H NMR spectrum of sucrose (2:1 acetone-d6/H2O at -20) 6g
H OH
4g
H
3f O 1f
4f OH
4 g HO
HO OH H
g HO H
3g 2 3 g
H HO
g
H HO
H 2 O f
4 OH
f
1 6f , 6g 6f

HO
f
H
3

6.7 6.6 6.5 6.4 6.3 6.2 6.1 6.0 5.9 5.8 5.7 5.6 5.5
ppm

Reich, U.Wisc. Chem. 605 5-HMR-2.24 Delta


Phenols. The OH signals of phenols are generally well downfield of those of alcohols, appearing at 5-7 in CDCl3,
and 9-11 in DMSO. The higher acidity of phenols results in faster exchange rates, so that polyphenolic compounds
will usually show only one OH signal.
In DMSO solution, even the exchange between carboxylic acid protons and other OH groups can be slowed
enough to allow individual observation, as in the spectrum of 2-hydroxycinnamic acid below.

300 MHz 1H NMR spectrum in DMSO-d6


Source; Aldrich NMR Library
O

HO

HO
C9H8O3

12 11 10 ppm 9 8 7 6
-Dicarbonyl Compounds. Especially dramatic shifts are observed for the strongly intramolecularly H-bonded enol
forms of -dicarbonyl compounds, o-ketophenols and related structures.

15.84 17.08 H 12.02


H H O O
O O O O O O

5.40 1.98 4.46 2.16


SMe
90% 10%
Favored by polar solvents OH 5

Carboxylic Acids. Most carboxylic acids are strongly hydrogen bonded in non-polar solvents, and the OH protons
are correspondingly downfield shifted. Acetic acid dimer in Freon solvent (CDClF2/CDF3) at 128 K appears at 13.04,
and the OH signals of acetic acid hydrogen bonded to a protected adenosine under conditions of slow exchange
appear at even lower field (Basilio, E. M.; Limbach, H. H.; Weisz, K. J. Am. Chem. Soc. 2004, 126, 2135).

CH3
CH3 O 8.5/8.6 O
13.04
H H
O H O N O
O
CH3 CH3 H H 14.9
16.3 N
O H O N

N N
SiiPr3

Reich, U.Wisc. Chem. 605 5-HMR-2.25 Delta


Amine and Amide N-H Protons. NH2 protons of primary alkyl amines typically appear as a somewhat broadened
signal at 1-2 in CDCl3. The broadening has several sources: partially averaged coupling to neighboring protons,
intermolecular exchange with other NH or OH protons, and partially coalesced coupling to the quadrupolar 14N nucleus
(I = 1), which usually has a short T1. In the example below, the CH2 group bonded to amino ( 2.82) shows little
indication of coupling to the NH2 protons, so NH exchange must be rapid on the NMR time scale. The amide proton at
7.1 is broadened by residual coupling to 14N, not by exchange, since the N-CH2 signal is a sharp quartet ( the vicinal
HN-CH2 and CH2-CH2 couplings are accidentally equivalant).

300 MHz 1H NMR spectrum in CDCl3


Source: Aldrich NMR Library
Note: triplet
Note: quartet
JHCCH = JHCNH

O 3.3 3.2 3.1 3.0 2.9 2.8 2.7

NH2
N
H

7 6 5 ppm 4 3 2 1

The N-H signals of ammonium salts are strongly downfield shifted, typically appearing at 4-7 in CDCl3 and 8-9 in
DMSO. If spectra are taken in strongly acidic solvents (e.g. trifluoroacetic acid), where intermolecular exchange is
slowed, the signals are sometimes very broad, and can show poorly resolved 1H-14N J coupling (1:1:1 triplet, JHN 70
Hz).

300 MHz 1H NMR spectrum in CDCl3


Source: Aldrich NMR Library CH2
NH2

NH2

NH3+ Cl

7 6 5 ppm 4 3 2 1 0

Reich, U.Wisc. Chem. 605 5-HMR-2.26 Delta


Aniline NH Protons. The NH protons of anilines are typically at 3.5-4.5 in CDCl3 solution, moving downfield by 1-2
ppm in DMSO solution. o-Nitroanilines (ca 5-6) and heterocyclic amines such 2-aminopyridines ( 4.5) have signals
downfield of this range.

300 MHz 1H NMR spectrum in DMSO-d6 NH2


Source: Aldrich NMR Library

OH

7.0 6.5 6.0 ppm 5.5 5.0 4.5


Amide NH Protons. Amide NH signals typically appear around 7, as in the example of N-acetylethylenediamine
above and N-methylpropionamide below. They are generally in slow exchange with other NH and OH signals. Thus,
neighboring protons will show coupling to the NH proton, as in the examples, where the CH2 bonded to the amide
nitrogen is a quartet and the N-Me group is a doublet. The amide N-H protons are typically broad from poorly resolved
coupling to 14N, so the coupling to neighboring protons is usually not resolved in the NH signal.

300 MHz 1H NMR


spectrum in CDCl3.
Source: Aldrich NMR N-CH3
Library O
Hz
40 20 0
N
C4H9NO H

2.8 2.6 2.4 2.2

1.201.151.10
7.0 6.8 6.6

7 6 5 ppm 4 3 2 1 0

Reich, U.Wisc. Chem. 605 5-HMR-2.27 Delta


Thiol S-H Protons. S-H protons of alkyl thiols typically appear between 1.2 and 2.0 in CDCl3. The position is not
strongly affected by hydrogen bonding solvents like acetone or DMSO, since SH protons are only weakly hydrogen
bonded. Coupling to nearby protons is usually seen, although broadened or fully averaged signals are not uncommon,
especially in molecules containing OH protons (or in impure samples).

300 MHz 1H NMR spectrum in CDCl3


Source: Aldrich NMR Library
SH
C3H8S2
SH

3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2
ppm

300 MHz 1H NMR spectrum in CDCl3-DMSO-d6


Source: Aldrich NMR Library

HO SH
OH SH
OH OH 3.03
1.99
C3H8O2S
1.04 1.00 0.96

4.5 4.0 3.5 3.0 2.5 2.0


ppm

Aryl thiol S-H signals are further downfield, typically 3.5-4.5, as a result of normal ring-currrent effects, and the
greater electron withdrawing effect of aryl vs alkyl groups.

300 MHz 1H NMR spectrum in CDCl3 SH


Source: Aldrich NMR Library

Br

7.40 7.35 7.30 7.25 7.20 7.15 7.10 7.05

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-2.28 Delta


Selenol and tellurol protons (SeH and TeH) behave like thiol protons, but appear further upfield -- around 0 for
SeH and -3 to -5 for TeH. Below a comparison of the NMR spectra of benzylselenol and benzylthiol. Note that both
the Se-H and S-H protons are coupled to the CH2 group (AX2 pattern).

SeH Me4Si

Hz
30 20 10 0

3.9 3.8 3.7 0.1 0.0 -0.1

SH

3.75 3.70 3.65


1.75 1.70

7 6 5 4 ppm 3 2 1 0

revised 43-12

Reich, U.Wisc. Chem. 605 5-HMR-2.29 Delta


Copyright Hans J. Reich 2017
5.3 Spin-Spin Splitting: J-Coupling All Rights Reserved
University of Wisconsin

There are two distinct types of magnetic interaction (coupling) between nuclei (A and X) with a non-zero spin -
the direct interaction (dipole-dipole coupling: D) and the indirect or scalar coupling (spin-spin splitting: J). The direct
interaction is about 1000 times as large as the scalar coupling (e.g. at 2 distance H-H dipolar coupling is ca 30,000
Hz). These direct couplings make the observation of high-resolution NMR spectra in solids and very viscous liquids
difficult, and make NMR spectra in liquid crystals (where molecules are partially oriented, and the dipolar coupling is
only partially averaged) very complex. In mobile isotropic liquids the random motion of molecules completely
averages the dipolar coupling, so no direct effects are seen. There are however, indirect effects, such as the
Nuclear Overhauser Effect (NOE) which have important consequences for NMR spectroscopy (see Sect. 8). In the
following sections we will be concerned only with J coupling.
The scalar coupling J is a through-bond interaction, in which the spin of one nucleus perturbs (polarizes) the
spins of the intervening electrons, and the energy levels of neighboring magnetic nuclei are in turn perturbed by the
polarized electrons. This leads to a lowering of the energy of the neighboring nucleus when the perturbing nucleus
has one spin, and a raising of the energy whenwhen it has the other spin. The J coupling (always reported in Hz) is
field-independent (i.e. J is constant at different external magnetic field strength), and is mutual (i.e. JAX = JXA).
Because the effect is usually transmitted through the bonding electrons, the magnitude of J falls off rapidly as the
number of intervening bonds increases. Coupling over one (1J), two (2J) and three (3J) bonds usually dominates the
fine structure of NMR spectra, but coupling across four (4J) and five (5J) bonds is often seen, especially through
bonds (double and triple bonds, aromatic carbons).

Sign of Coupling Constants

Coupling constants can be either positive or negative, defined as follows: coupling constants are positive if the
energy of A is lower when X has the opposite spin as A ( or ), and negative if the energy of A is lower when X
has the same spin as A ( or ).

JAX < 0 JAX > 0

X= A= X= lower in energy than


X= X=
The signs of couplings shows some consistency.
1JC-H and many other one-bond couplings are positive.
2JH-H in sp3 CH2 groups are negative, some others are positive.
3JH-H is always positive.
E
X= A= X= For first order patterns the signs of the couplings have no effect on the
X= X= appearance of the spectrum, and so cannot be determined by
observation. However, decoupling experiments (spin tickling) can
provide the relative signs. For second-order patterns (e.g. ABX or
X=

X=
X=

X=

AA'BB'), the relative signs of coupling constants often have dramatic


effects on the appearance of the spectrum, and relative signs can be
A-signal determined by proper analysis of the multiplets.

d d

Reich, U.Wisc. Chem. 605 5-HMR-3.1 Coupling


Mechanism of spin polarization: It is known from spectroscopy of the hydrogen radical (H ) that the more
stable orientation has the angular momentum vectors of the nucleus and the electron antiparallel. Since the
gyromagnetic ratio of the nucleus is positive, and that of the electron is negative, this means that the magnetic
vectors are parallel.
e
Magnetic moment of nucleus H
More stable
Magnetic moment of electron
configuration of H

For the Fermi contact mechanism of spin-spin coupling (there are other mechanisms), the bonding electrons for a
H- C fragment will become polarized as shown, so that the more stable orientation of the 13C-nucleus will be down,
13

when the proton is up. This corresponds to a positive one-bond C-H coupling.
J>0
J>0
e
H C


e e

e
H C

e
e J<0
C H


Positive one-bond e e

coupling constant
J>0

If we continue down the bond sequence, the polarization of the C-H electrons will cause polarization of the C-C
electron pair. Again, parallel spins are the more stable orientation (triplets are more stable than singlets -- Hund's
rule). Thus the two-bond coupling (2J) is predicted to be negative, and the three-bond coupling (3J) positive. This
alternation of signs is often (but by no means always) seen.

The principal mechanism for J-coupling is through bond polarization, but there are situations where a
through-space effect seems to be operative. These occur in molecules where spin 1/2 nuclei are forced into close
proximity. For example, in the compounds below, there is a substantial H - F J coupling even though the H and F
are separated by many bonds, where normally coupling is small or undetectable.

H F 7
JH-F = 7 Hz
F H
JACS 1972, 94, 2889

Reich, U.Wisc. Chem. 605 5-HMR-3.2 Coupling


A depiction of the perturbation of energy levels of a nucleus A by a neighboring magnetic nucleus X is shown
below (spin-spin splitting). The principal magnetic nuclei are other protons, the 100% abundant spin nuclei 19F
and 31P, and some spin 1 or greater (quadrupolar) nuclei such as 14N, 2H, 11B, and 12B. Although Br, Cl, and I all
have isotopes with spin >, coupling is not seen because of relaxation effects. This will be discussed in more
detail in Section 7.

H H H H

Note: vertical scale of the couplings is grossly exaggerated: the


C C C C vertical arrows are MHz, the couplings Hz

s d
E s d t s d dd

dd

Two equal couplings. t q Two different couplings.

Reich, U.Wisc. Chem. 605 5-HMR-3.3 Coupling


Two Different Couplings to one Proton

Consider the NMR spectrum of 3,4-dichlorobenzoyl chloride below.

R-19M C7H3Cl3O
Cl O
270 MHz 1H NMR Spectrum (CDCl3)

Cl
Hz
40 30 20 10 0 Cl

8.2 8.1 8.0 7.9 7.8 7.7 7.6 7.5


ppm

9 8 7 6 5 4 3 2 1 0
ppm

The proton-proton couplings in benzene are typically 7-9 Hz for Jortho, 2-3 Hz for Jmeta and <1 Hz for Jpara. The
substitution pattern can be derived from examination of each of the three aromatic protons. For example, the
doublet at 8.2 with J = 2.5 Hz is interpreted as follows: this proton has no protons ortho to it, and only one proton
meta to it. Structure A summarizes the information. For the doublet of doublets at 7.95 (J = 8.5, 2.3 Hz), formed
by coupling of one proton to both an ortho and a meta proton, the two structures B and C are possible. The doublet
at 7.6 (J = 8.5 Hz) defines the substitution pattern of structure D. In each case the position marked by ? is
undefined since the para coupling is usually too small to resolve.

X 8.20 7.95 X 7.95 X X


H H H H H X H ?
7.61
? X H ? H ? H X
X X H X
A B C D
J ca 2 Hz J ca 2 and 8 Hz J = 8 Hz
one meta H only one ortho and one meta H one ortho H

Reich, U.Wisc. Chem. 605 5-HMR-3.4 Coupling


A slightly more complicated case is 1,1,2-trichloropropane. A simulated spectrum is shown below.

WINDNMR Simulated Spectrum of 1,1,2-Trichloropropane


Chemical Shift (in ppm) Coupling constants (in Hz)
(1) = 5.850 2 3 4 5
(2) = 4.300 J(1:y) 3.80 0.00 0.00 0.00
(3) = 1.620 J(2:y) 6.60 6.60 6.60
(4) = 1.620 J(3:y) 0.00 0.00
(5) = 1.620 J(4:y) 0.00

Cl Cl H
H C C C H
Cl H H

8 7 6 5 4 3 2 1 0
ppm

The C-2 proton is coupled to one proton at C-1 and three protons of the methyl group at C-3. Naively, one might
expect a pentet (p), as shown in the left spectrum below. Although pentets are, in fact, often observed in such
situations, this occurs only if J1-2 and J2-3 are identical. When they are not (as is actually the case in this example),
then we get a quartet of doublets (qd). It is customary to quote the larger coupling first (q) and then the smaller
coupling (d). A proper text description of the multiplet is: 4.30, 1H, qd, J = 6.6, 3.8 Hz.

Cl Cl
6.6 Hz
Cl C C CH3
H1 H2
3.8 Hz

Hypothetical Hypothetical Actual


J12 = 6.6 Hz J12 = 5.0 Hz J12 = 3.8 Hz
J23 = 6.6 Hz
J23 = 6.6 Hz J23 = 6.6 Hz

4.5 4.0 4.5 4.0 4.5 4.0

Exercise: what would a dq, J = 6.6, 3.8 Hz look like?

Reich, U.Wisc. Chem. 605 5-HMR-3.5 Coupling


First Order Coupling Rules

1. Nuclei must be chemical shift nonequivalent to show obvious coupling to each other. Thus the protons of
CH2Cl2, Si(CH3)4, Cl-CH2-CH2-Cl, H2C=CH2 and benzene are all singlets. Equivalent protons are still coupled to
each other, but the spectra do not show it. There are important exceptions to this rule: the coupling between shift
equivalent but magnetically inequivalent nuclei can have profound effects on NMR spectra - see Sect. 5.7
2. J coupling is mutual, i.e. JAB = JBA always. Thus there is never just one nucleus which shows J splitting -
there must be two, and they must have the same splitting constant J. However, both nuclei need not be protons -
fluorine (19F) and phosphorus (31P) are two other common nuclei that have spin and 100% abundance, so they
will couple to all nearby protons (the other 100% spin 1/2 nuclei are 89Y, 103Rh and 169Tm). If these nuclei are
present in a molecule, there are likely to be splittings which are present in only one proton multiplet (i.e. not shared
by two multiplets).

3. Two closely spaced lines can be either chemically shifted or coupled. It is not always possible to distinguish J
from by the appearance of the spectrum (see Item 4 below). For tough cases (e.g. two closely spaced singlets in
the methyl region) there are several posibilities:
decouple the spectrum
obtain it at a different field strength (measured in Hz, coupling constants are field independent, chemical shifts
are proportional to the magnetic field)
measure the spectrum in a different solvent (chemical shifts are usually more solvent dependent than coupling
constants, benzene and chloroform are a good pair of solvents).
For multiplets with more than two lines, areas, intensities, symmetry of the pattern and spacing of the lines
generally make it easy to distinguish chemical shift from coupling.
For a simple example see the spectrum of 3-acetoxy-2-butanone below. Here it is pretty easy to identify one of
the doublets as the 4-methyl group, the other "doublet" (with a separation of 9 Hz, which could easily be a coupling)
actually corresponds to the two CH3C(=O) groups.

300 MHz 1H NMR in CDCl3


Source: Aldrich Spectra Viewer/Reich
O Hz
30 20 10 0
O

5.88

2.91
5.15 5.10 5.05 5.00

1.00
2.20 2.15 2.10 1.40

5 4 3 ppm 2 1 0

4. Chemical shifts are usually reported in (units: ppm) so that the numeric values will not depend on the
spectrometer frequency (field-independent units), coupling constants are always reported in Hz (cycles per second).
Chemical shifts are caused by the magnetic field, couplings are field-independent, the coupling is inherent in the
magnetic properties of the molecule. However, all calculations on NMR spectra are done using Hz (or, more
precisely, radians per sec).

Reich, U.Wisc. Chem. 605 5-HMR-3.6 Coupling


H

5. Protons two (2J, geminal) or three bonds (3J, vicinal) apart are usually coupled to each other, more remote
protons (4J, 5J) may be if geometry is right, or if -systems (multiple bonds) intervene. Long range couplings (4J or
greater) are usually small, typically <0.5 Hz, but up to 3 Hz in some cases where there are intervening bonds.

H H C H
H C 3
2
J = 2-15 Hz small J = 2-20 Hz C 4
C J = 0-3 Hz
Typical: -12 Hz 0-3 Hz C Typical: 7 Hz Usually 0
H H C H
H Occasionally 0!
geminal Note vicinal long-range

6. Multiplicity for first order patterns follows the "doubling rule". If all couplings to a particular proton are the
same there will be 2nI+1 lines, where I is the spin and n is the number of neighboring nuclei (n + 1 for 1H I = 1/2).
The intensities will follow Pascal's triangle.

n=0 1
n=1 1 1
n=2 1 2 1
n=3 1 3 3 1
n=4 1 4 6 4 1
n=5 1 5 10 10 5 1
n=6 1 6 15 20 15 6 1
Intensities triplet n = 2 quartet n = 3 pentet n = 4 sextet n = 5 nonet n = 8

7. If all couplings are different, then the number of peaks is 2n for 1H, and the intensities are 1:1:1: . . .. Thus a
proton coupled to two others by different couplings gives a dd (doublet of doublets, see Figure). This pattern is
never called a quartet. As the number of couplings gets larger, accidental superpositions of lines will sometimes
occur, so that the 1:1:1... intensity ratio no longer applies. The intensities are also often distorted by leaning effects
(see AB/AX patterns), as seen in several examples below.

dd n = 2 ddd n = 3 ddd n = 3 dddd n = 4 ddddd n = 5

8. More typically, some of the couplings are the same, others different, so get a variety of patterns. In
favorable cases, these patterns can be analyzed and all couplings extracted. The number and size of couplings
(J-values) provide important structural information.

Reich, U.Wisc. Chem. 605 5-HMR-3.7 Coupling


Second Order Effects

Protons or groups of protons form simple multiplets only if the chemical shift differences between the protons
() are large compared to the coupling constants between them (J). If /J (all in Hz) is <5 then second order
effects appear (see 5-HMR-9) which complicate the analysis.

Rules for Analyzing First Order Multiplets

A first order multiplet can be expected when both of the following criteria are met:
First, the chemical shift of the observed proton must be far away from any of the protons it is coupled to (far
away means >> J). In practice, multiplets can be treated in a first order fashion if > 3J, although the
substantial leaning distortions can complicate analysis. The leaning will have almost completely disappeared by
the time = 10J.
Second, if more than one proton is coupled to the observed one, then these protons must not be "strongly
coupled." In other words, if they are coupled to each other and very close in chemical shift then the observed
proton multiplet may not yield true coupling constants on analysis, even though it looks first order. See the section
on Virtual Coupling.
Structure of First Order Multiplets. The fundamental rule governing multiplet intensities for spin 1/2 nuclei
with all couplings identical is Pascal's triangle (n = number of equivalent couplings). A very characteristic and
diagnostic intensity relationship is that between the first and second lines - the intensity ratio is 1/n, where n is the
number of equivalent coupling partners.

n 2n Multiplet Intensities - Pascal's triangle


0 1 1Singlet (s)
1 2 1 1 Doublet (d)
2 4 1 2 1 Triplet (t)
3 8 1 3 3 1 Quartet (q)
4 16 1 4 6 4 1 Pentet
5 32 1 5 10 10 5 1 Sextet
6 64 1 6 15 20 15 6 1 Septet
7 128 1 7 21 35 35 21 7 1 Octet
8 256 1 8 28 56 70 56 28 8 1 Nonet

A first order multiplet consists of the product (not the sum) of several such multiplets. In other words, a single
line will first be split into one of the symmetrical multiplets (1:1 d, 1:2:1 t, 1:3:3:1 q, etc), then each line of this
multiplet will be again split into d, t, q, or higher multiplet.

Reich, U.Wisc. Chem. 605 5-HMR-3.8 Coupling


Recognizing a First Order Multiplet.
1. All truly first order multiplets are centrosymmetric - there is a mirror plane in the middle (in real spectra, this
is usually not strictly true because of leaning and other distortions). However, the reverse is not true: not all
symmetrical multiplets are first order.
2. If the small outermost peaks are assigned intensity 1, then all other peaks must be an integral multiple
intensity of this one (1x, 2x, 3x, 4x in height), and the total intensity of all peaks must be a power of 2 (2, 4, 8, 16,
32, etc). The intensity of each of the two outermost lines is 1/2n of the total multiplet intensity, where n is the
number of protons which are coupled with the proton signal being analyzed. There can be no lines smaller than the
outermost one. Note, however, that if n is large, the outermost peaks may not be distinguishable from noise.
Intensity assignments and determination of n cannot be easily made for such multiplets
3 3 6 6
2
2
2
3 33 3
1 11 11 1 2
1 1 1 1 11
2 1 1
1 1

=8 = 16 = 32
Coupled to 3 other protons Coupled to 4 other protons Coupled to 5 other protons
23 = 8 24 = 16 25 = 32

3. There is a strict regularity of spacing in a first order multiplet: if you have correctly identified a coupling
constant J, then every peak in the multiplet must have a partner J Hz away to the left or to the right of it.

3 3

This separation is 2
always a true coupling
2

1 1 1 1 1 1

This is also a coupling


This is not a coupling
This is not a coupling

4. Most first order multiplets integrate to a single proton, a few may be 2 or 3 protons in area. It is rare to have
more than 3 protons, unless there is symmetry in the molecule (e.g., (CH3)2CH- gives a 6-proton doublet for the
methyl groups). Thus a 4-proton symmetrical multiplet is usually not a first-order pattern (it is more likely to be the
very common AA'BB' pattern).
5. The symmetry and intensities of an otherwise first-order multiplet can be distorted by leaning effects (see
Section 5-HMR-9). Many such multiplets can still be correctly analyzed by first-order techniques, but you have to
mentally correct for the intensity distortions. However, the coupling constants extracted may not be perfectly
accurate.

Reich, U.Wisc. Chem. 605 5-HMR-3.9 Coupling


Analyzing a First Order Multiplet. First order multiplets are analyzed by constructing a reverse coupling tree,
by "removing" each of the couplings in turn, starting with the smallest.
1. "Take out" the smallest couplings first. The separation between the two lines at the edge of the multiplet
is the smallest coupling. Each time you remove a coupling you generate a new, simpler multiplet, which can then
be analyzed in turn. Remember that each line of the multiplet participates in each coupling.
2. Watch line intensities (i.e., peak areas or peak heights) carefully--when you "take out" a coupling, the
intensities of the newly created lines should be appropriate (i.e., each time you "take out" a coupling, also "take
out" the proper intensity). When a coupling has been taken out completely, all intensity should be accounted for.
Keep track of your analysis by using a "coupling tree".
3. The couplings may be removed one at a time as doublets, or as triplets, quartets and higher multiplets.
The intensity ratio of the first two lines signals the number of protons involved in the coupling: 1:1 means there is
only one proton, 1:2 means that there are 2 prtons (triplet), etc. Be especially careful to keep track of intensities
when you "take out" triplets (1:2:1) or quartets (1:3:3:1). Each time you completely remove a coupling you
generate a new simpler multiplet which follows first order rules, and can be analyzed in turn.
When you have finished your analysis, all peaks and all intensity in the multiplet must be accounted for. You
can check the analysis as follows: the separation of the two outermost peaks of the multiplet is the sum of all the
J's (i.e., for a dt, J = 8, 3 Hz the outermost lines are separated by 8 + 3 + 3 = 14 Hz).
Reporting a First Order Multiplet. Multiplets are reported starting with the largest coupling, and the
symbols must be in the order of the reported numbers: 2.10, 1H, qt, J = 10, 6 Hz means: a single proton q of 10
Hz, t of 6 Hz with a chemical shifts of 2.10 ppm.
Quartets. Keep clear in your mind the distinction between a simple q (one proton equally coupled to 3
others, with an intensity 1:3:3:1), an ABq (2 protons coupled to each other, see Section 5-HMR-10), and the
quartet formed by coupling with a spin 3/2 nucleus (e.g., 7Li, intensity 1:1:1:1, see Sect 7-MULTI-2). Only the first
of these should be referred to by just a "q" symbol. The early NMR literature (and even modern novices)
sometimes call doublets of doublets "quartets" (there are four lines, after all). Don't do this.

quartet AB quartet 1:1:1:1 quartet doublet of doublets

Reich, U.Wisc. Chem. 605 5-HMR-3.10 Coupling


Chem 605
Practice Multiplets Reich

J
J

50 40 30 20 10 0
Hz

Reich, U.Wisc. Chem. 605 5-HMR-3.11 Coupling


First Order Analysis
= 8 (3 couplings present)

1 1 11 11 1 1
The separation between the first and second lines is the
smallest coupling in the multiplet (J1). They are in an
intensity ratio of 1:1, so that coupling appears only once
(i.e., it is a doublet splitting).

"Remove" the first coupling. Keep track of line intensities,


and draw a child multiplet with positions at the center of
each doublet. Each unit of line intensity can only be used
once, and all line intensity must be accounted for

Repeat the process - the first and second lines now


represent the next largest coupling J2. Remove this in turn, J1 (4) ddd
and repeat the process until you get a singlet
J2 (10) dd
Report the couplings in order of size (reverse of the
analysis): J3, J2, J1
J3 (12) d
This is ddd, J = 12, 10, 4 Hz Hz
30 20 10 0 s

= 16 (4 couplings present)
3 3

As always, the separation of the first two lines is the smallest


coupling in the multiplet (J1). However, they are in an intensity
2 2
ratio of 1:2, so this coupling appears twice (i.e., it is a triplet
splitting). 1 1 1 1 1 1

"Remove" the triplet coupling. Keep track of line intensities,


and draw a child multiplet with positions at the center of each
triplet. Note that the central two lines of intensity 3 are the sum J1 (6)
of two lines - one of intensity 1 and the other intensity 2. The ddt
new multiplet is a dd, which can be solved in turn.

This is a ddt, J = 16, 10, 6 Hz - typically the way the J2 (10)


dd
central vinyl protons of an allyl group appears:
J3 (16) d
H
X Hz s
30 20 10 0

Reich, U.Wisc. Chem. 605 5-HMR-3.12 Coupling


Simple Multiplets

Exercise: Assign the protons where structure and chemical shift scale are given
Note the leaning in most of the multiplets, indicating that the coupled partner is not too far away.

Doublet of doublets (dd) Doublet of triplets (dt, td) Triplet of doublets (td)
O O
O O
Ph O H Cl
O C12H12O3 O
C5H5ClO2
2H

3.05 3.00 2.45 2.40 3.00 2.95 2.70 2.65 2.60


ppm ppm ppm ppm

Doublet of doublet of doublets (ddd) Hz


O 30 20 10 0 PhS
O Cl
PhS OH
O

5.85 5.80 2.55 2.50 2.45 2.15 2.10 2.05 2.00 1.95 1.90 1.85
ppm ppm ppm

Doublet of quartets (dq)


Exercise: The multiplet below integrates to two
O protons. Using the chemical shift and coupling
identify the structural fragment.
OMe

7.05 7.00 6.95 6.90 5.90 5.85 5.80


5.95 5.90 5.85 5.80 5.75 5.70 5.65
ppm ppm
ppm

Reich, U.Wisc. Chem. 605 5-HMR-3.13 Coupling


Symmetrical Multiplets which are NOT First Order

Exercise: Only ONE of the multiplets below is first order, find it. A second one is almost first order, but
ultimately can be ruled out because of a very subtle line position inconsistency.
Some criteria to use:
Pattern must be centrosymmetric (true of all of these)
Intensity of lines - patterns must be repeated, especially examine outer lines
Be wary if #H > 1, especially if 4H
Consider size of possible couplings

(a) (b)

Hz
50 40 30 20 10 0

(c) (d)

(e) (f)

Reich, U.Wisc. Chem. 605 5-HMR-3.14 Coupling


5-HMR-3.14 Measurement of Coupling Constants

The accurate measurement of J coupling constants requires that the multiplets be correctly analyzed. In the
following pages are described techniques for performing such analyses. The procedures are summarized
below.
For first order multiplets a simple "coupling tree" analysis as described in Section 5-HMR-3.9 can directly yield
coupling constants within the accuracy of the digital resolution of the spectrum. This includes AB spectra, where
JAB can be measured directly. See Section 5-HMR-7 for a description of the ABC... (Pople) nomenclature for spin
systems.

For AB2 spectra both the coupling constant JAB and the chemical shifts can be obtained by simple arithmetic
manipulations, provided that line assignments can be made correctly. For ABX spectra JAB is accurately
measureable by inspection. An approximate analysis, which treats the peaks as AMX, will give values for JAX and
JBX that will be in error by varying amounts, depending on the relative size of JAB and AB (the smaller AB the
larger the error), and the relative size of JAX and JBX. To get accurate values for the JAX and JBX coupling
constants a proper ABX analysis as described in Section 5-HMR-12 is required.
For many simple compounds the symmetry is such that protons are homotopic or enantiotopic, and no coupling
constants can be measured directly (e.g., the 2J coupling in methane or dichloromethane; the ortho, meta, and
para couplings in benzene; the cis, trans and gem couplings in ethylene, etc). For such compounds the following
techniques are used to measure JHH:
Analysis of Complex Spin Systems. In molecules where the chemical shift-equivalent protons are of the AA'
type (part of an AA'XX', AA'X3X3' or similar system), complete analysis of the coupling system can, in favorable
circumstances, give the value of JAA'. An example is 1,3-butadiene, an AA'BB'CC' system in which all protons are
compled to all other ones. Analysis of the complex NMR spectrum gave, among numerous others, values for the
following couplings between chemical shift equivalent nuclei: 3JAA', 5JBB' and 5JCC' (Hobgood, R. T., Jr.; Goldstein,
J. H. J. Mol. Spectr. 1964, 12, 76).
A C B
HA' HC 60 MHz
HB' Solv: cyclohexane
HB
HC' HA

3 3 4
JAA' = 10.41 JAB = 10.17 JAC' = -0.83
5 4 2
JBB' = 1.30 JAB' = -0.86 JBC = 1.74
5 3 5
JCC' = 0.69 JAC = 17.05 JBC' = 0.60
6.6 6.4 6.2 6.0 5.4 5.2 5.0 4.8
Isotopic Substitution. Replacing one of the protons by deuterium (or even tritium) breaks the symmetry of the
coupled system and allows measurement of JHD (or JHT). The value of JHH can then be calculated from the
gyromagnetic ratios. In the example below, the 60 MHz NMR spectrum of a mixture of undeuterated (s),
monodeuterated (1:1:1 triplet, the spin of D is 1, see Sect. 7-MULTI-2) and dideuterated (1:2:3:2:1 quintet)
acetonitrile is shown. Note the isotopic shifts (Grant, D. M.; Barfield, M. JACS 1961, 83, 4727)

CH3CN
JHH H
= = 6.488
JHD D

JHH = 6.488 JHD


JHD (CH2D-CN) = 2.58 Hz
CDH2CN JHH (CH3-CN) = 16.75 Hz

CD2HCN

120
115
Hz
Reich, U.Wisc. Chem. 605 5-HMR-3.15 Coupling
Analysis of 13C Satellite Spectra. Vicinal couplings between homotopic or enantiotopic protons 3JHH can often
be obtained by analysis of the 13C satellites. The 1H NMR signal for the vinyl protons of dimethyl maleate is a
singlet. However, the 13C satellites are doublets, with a splitting that is equal to 3JHH. In effect, the A2 spin system
of the 12C isotopomer has become an ABX pattern in the mono-13C labelled compound, where X is the 13C
nucleus, and A and B are the two vinyl protons, one on 13C and the other on 12C.

Spinning
side bands O C6H8O4 CDCl3
H 200 MHz
12
C OMe
13
13
C Satellite C Satellite 12
C OMe
1 H
JHC 97.8% of the molecules
O have 12C at both vinyl
3 X
JHH carbons

O
H
13
1300 1250 1200 1150 C OMe
12
C OMe
H
O
2.2% of the molecules have 13C
at one of the vinyl carbons

8 7 6 5 4 3 2 1 0
ppm
1879.1
1875.8

1698.2
1694.9
H H

Cl Cl
Cl Cl

1
JHC = 180.9 Hz

3
JHH = 3.3 Hz

6.3 6.2 6.1 6.0 5.9 5.8 5.7


ppm

Reich, U.Wisc. Chem. 605 5-HMR-3.16 Coupling


Below is an example of the measurement of a 4JHH in a symmetric tricyclic system using the 13C satellite
method (Masamune, S. J. Am. Chem. Soc. 1964, 86, 735)

Ph Ph
60 MHz NMR spectrum
4
H H J = 14 Hz

1
JHC = 190.1 Hz
4
JHH = 14 Hz

2.88 (190)

8 7 6 5 4 3 2 1 0
ppm

For systems of the X-CH2-CH2-X type, the mono-13C isotopomer is an AA'BB'X pattern (X = 13C), which can be
solved to obtain JAA' (= JBB') as well as JAB and JAB'. Note that when both protons are on the same carbon the
value 2J cannot be determined by this method. Thus for the O-CH2-O signal, the 13C satellites are singlets.

C3H6O3
H H The BB' protons
3.97
O O 12 are hidden under
C
O O the central peak
CDCl3, 300 MHz
W. S. Goldenberg O O
13 12
C C
HA HB
1
JHC HA' HB'
AA'BB'X
H H 2.00
13
C 1
O O JHC

5.2 5.0 4.8 4.6 4.0 3.8 3.6

9 8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-3.17 Coupling


Copyright Hans J. Reich 2017
All Rights Reserved
5.4 Geminal Proton-Proton Couplings (2JH-H) University of Wisconsin

Two-bond H-H couplings (Review: Tet, 1969, 25, 4681) vary in a complicated way with structure, and they can
only be understood if both magnitude and sign is taken into account. Some extreme examples are given below.

H H H H H
H
O O O O
O O
H H
+40.2 -15.8 -21.5 ~0
2 3
Most J couplings fall into two well-defined groups. For unstrained sp CH2 protons with innocuous substituents,
the coupling is typically around -12 Hz, whereas the 2-bond coupling of sp2 (vinyl) protons is much smaller, typically
2 Hz. The molecular orbital perturbation theory of Pople and Bothner-By (J. Chem. Phys. 1965, 42, 1339) predicts
the electronic effects of substituents on these coupling constants based on the interaction between filled and empty
orbitals of the CH2 fragment. Excitation between orbitals of the same symmetry has a negative effect on J, between
orbitals of opposite symmetry has a positive effect. The 1/E term is largest for the HOMO-LUMO transition, so
coupling effects are dominated by the 2 3 transition. Substituents which reduce the energy gap between 2 and
3 (i.e. raise 2 or lower 3) will increase the size of 1/E and thus have a (+) effect on the coupling, whereas those
which increase the energy gap (i.e. lower 2 or raise 3) will have a (-) effect. There are also changes in the orbital
coefficients which affect the magnitude of the coupling.

H
C R
H

2
a 4 JHH' = h (4H/3)2Sh4(0)hh'

occ unocc cih cih' cjh cjh'


s 3 hh' = -4 (i - j)
i j
+ +
1 + ++
a 2 23 = 23 (c2c2'c3c3') (+)

s 1

-Acceptor substituents (electronegative atoms like F, O, N) interact mainly with 3 and 1 because of symmetry
restrictions. The most important effect is to lowerr 3, and thus have a (+) effect on the coupling, whereas -donor
substituents like Si or other metals will raise 3 and thus have a (-) effect.

-Acceptor -Donor
2
JH-H coupling H
2
JH-H coupling H
becomes more C SiMe3
becomes more C F
3 H
3 H negative
positive

Reich, U.Wisc. Chem. 605 5-HMR-4.1 J-gem


Remarkably, -donors and acceptors have the opposite effect -- symmetry requires that these will interact mainly
with 2. Thus -donor substituents (directly attached atoms with lone pairs, or adjacent electron rich bonds) will
raise 2, and result in a (+) effect, and -acceptor substituents (carbonyl groups and related functions, or adjacent
electron poor bonds) will lower 2 and have a (-) effect.

H H O
-Donor C N -Acceptor C C
- H 2
+
H
2
JH-H coupling JH-H coupling

:
becomes more H Li becomes more H F
positive C C negative C C
2 2
H H

Gem coupling in Saturated Carbons (sp3): In acyclic and unstrained ring systems the gem coupling is typically
from -10 to -13 Hz. Substituents will change these couplings as described above: when the CH2 group is substituted
with a -acceptor like a carbonyl or cyano group, the coupling becomes more negative, i.e larger in magnitude,
ranging from -16 to -25 Hz. This is a reliable and important effect which can help with structure assignments.

O O H H -21.5
Cl
O -14.3 -14.8
-18.0 -19.5 O O t
CH4 CH3-C-CH3 CH3CN Bu
-12.4 -14.9 -16.9
-12.1 -19.0 -13.5 H O
JACS 1961, 4727 JACS 1961, 4727
OMR-77-92
Conjugating aryl, alkene and alkyne substituents also make the coupling more negative.

H H H H H
CH4 CH3-Ph H
-12.4 -14.5

-16.0 -20.8 -22.3

Substituents like the halogens, alkoxy and amino groups are both -acceptors and -donors. Both are (+)
effects, so the couplings become more positive (i.e. smaller numbers), in some cases they are close to zero.

H H H H H H
CH3SiMe3 CH4 CH3OH CH3F CH2Cl2
O O O
2 -14.1 -12.4 -10.8 -9.6 -7.5
J:
-13 -8.0 ~0
Ring strain has a (+) effect on gem coupling. Thus in cyclopropane the coupling has increased from -12 to -4 Hz.
The additional (+) effects of oxygen bring the coupling to +5.5 in ethylene oxide
H
H H H
N O

2 H H H
J: -4.3 +2.0 +5.5

Reich, U.Wisc. Chem. 605 5-HMR-4.2 J-ge


Gem coupling in Unsaturated Carbons (sp2): The gem coupling in ethylene itself is +2.5 Hz, and most terminal
alkenes have small couplings in the range of 1-3 Hz. Electronegative substituents (F, O) on the double bond behave
as -acceptors, with a (-) effect on the coupling, which can become close to zero for weakly accepting substituents
(as in methyl vinyl sulfide). Electropositive substituents on the neighboring carbon (Si, Li) behave as -donors with a
(+) effect on the coupling. For -trimethylsilylvinyllithium both substituents have a (+) effect, and result in an
exceptionally large coupling, whereas in -ethoxyvinyllithium the two substituents have opposite effects, and the
coupling was too small to observe.

F H F H MeS H H H Me3Si H Li H Li H Li H

F H H H H H H H H H H H Me3Si H EtO H
2
J: -4.6 -3.2 -0.3 +2.5 +3.8 +7.4 (+)9.8 0
-acceptor (-) -donor (+)
The large positive coupling in formaldehyde, and large negative coupling in ketene can be understood in these
terms as well. For formaldehyde the oxygen substituent behaves as a strong -acceptor as well as a strong -donor
from the -lone pair, both (+) effects, rendered especially large because of the short bond distance. Imines also
show large positive 2J.
In a similar vein, for ketene the carbonyl substituent behaves as a strong -acceptor, giving an usually large
negative coupling.
H tBu H H
O N O
H H H
:

2
J: +40.2 +16.5 -15.8
-donor (+) -acceptor ( +) -acceptor (-)

Reich, U.Wisc. Chem. 605 5-HMR-4.3 J-gem


Geminal Coupling - Electronic Effects

-Acceptor -Donor
-
2 2
JH-H coupling JH-H coupling
+

:
becomes more becomes more
negative positive

H H O H H
C CH2 C C C CH2 C N
H H H H
4 coefficients
4 4 smaller
a 4 coefficients
larger
3 3
s 1/E larger
+ 1/E smaller +
2 coefficients
2 2 larger
a 2 coefficients
smaller

1 1
s

-Acceptor -Donor
2
2
JH-H coupling JH-H coupling
becomes more becomes more
positive negative
H H H
H
C CH2 C CH2 C SiMe3
C F H
H H
H

4 4

3 coefficients
smaller
3 3
3 coefficients
+ larger + 1/E smaller
1/E larger
2 2

1 coefficients
1 larger
1
1 coefficients
smaller

Reich, U.Wisc. Chem. 605 5-HMR-4.4 J-gem


Geminal Proton-Proton Couplings Summary (2JH-H)
Geminal couplings between protons vary widely in sign and magnitude. There are both positive and negative
substituent effects on the coupling, as summarized below. The remarkable feature is that and acceptors have
opposite effects on the coupling, as do and donors.
Inductive -Acceptor +
O Oxygen and
H
CH4 CH3OH CH3F fluorine
H substituents are
O
-12.4 -10.8 -9.6 also -donors
~0

Inductive -Donor -
CH4 (CH3)4Si
-12.4 -14.1

-Acceptor - O
H H F H H H Gem-coupling constants of
O protons to carbonyl groups
H
are unusually large
H H F H H
O
+2.5 -4.6 -15.8 -21.5
-Donor +
H H Gem-coupling of vinyl protons is H The oxygen in formaldehyde
usually small (<3 Hz). The coupling O is both a donor and a
Li H can be both positive or negative, and H -acceptor - both positive
is sometimes too small to detect. effects.
+7.1 +40.2

High s-Character + H
H H H
H O N
H H H H
-13 -6 +5.5 +2.0

H 2
C R JHH' = h (4H/3)2Sh4(0)hh'
H occ unocc cih cih' cjh cjh'
hh' = -4 (i - j)
a 4 i j

1 + ++
23 = 23 (c2c2'c3c3')
s 3
Excitation between orbitals of the same symmetry has a
+ + negative effect on J, between orbitals of opposite symmetry has
a positive effect.
a 2
The 1/E term is largest for the HOMO-LUMO transition, so
coupling effect are dominated by the 2 3 transition.
s 1 Pople, Bothner-By J. Chem. Phys. 1965, 42, 1339

Reich, U.Wisc. Chem. 605 5-HMR-4.5 J-gem


Copyright Hans J. Reich 2017
3 All Rights Reserved
5.5 Vicinal Proton-Proton Coupling JHH University of Wisconsin

The single most useful H-H coupling relationship is that between vicinal protons. The size of 3JH-H is predictable
and provides detailed information about the spacial orientation between the two protons. Almost all 3JHH values are
positive (a rare exception is the -2 Hz 3JH-H in cis-1,2-difluoroethylene), but their magnitude varies widely (from close
to 0 Hz up to 25 Hz) depending on structural and conformational details.
Three-Bond Coupling across Single Bonds. In acyclic systems with small conformational preferences, vicinal
couplings are generally in the range 6-8 Hz, with electronegative substituents causing smaller J values (see the
pauling electronegativies E in the graphic). Note in particular the reduced 3J for protons on carbons bearing oxygen
substituents (as well as F), which is seen for all types of 3-bond couplings (Review: Bothner-By Adv. Magn. Reson.
1965, 1, 195.)

CH3 CH2 X
For a methyl group, the observed coupling is
X 3
J E Ref. the average of the three couplings, since
Li 8.9 -1.0 TL 1963, 767 these will be fully averaged by methyl
H rotation:
SiEt3 8.0 1.9
Hg Hg Jg + Jg + Ja
H 8:0 CJC 1963, 41,2114 3 4 + 4 + 13
Jobs = = = 7
X Y 3 3
I 7.45 2.5
Ha
Br 7.33 2.8
CH3 7.26 2.5 CJP 1960, 32, 67
X Jtrans Jcis Ea
Cl 7.23 3.1
H 3.81 9.73
NEt2 7.13 3.0 JPC 1964, 68, 3430 Cl6 CN 4.6 9.3 2.49
H CO2H 4.4 8.5 2.60
OEt 6.97 3.5 JPC 1964, 68, 3430 Hcis C6H5 4.2 8.9 2.75
OH 6.97 3.5 JCP 1956, 25, 362
X Cl 3.2 8.0 3.25
Htrans OH 2.4 7.4 3.43
F 6.9 4.0 JACS 1961, 83, 4473
+ OAc 2.5 7.7 3.80
JACS 1962, 84,516
OEt2 4.7 JACS 1959, 81, 3826
a
JCP 1961, 34, 1099

When there are two electronegative substituents the vicinal coupling is reduced further:

CH3-CHF2 CH3-CHCl2 CH3-CH(OH)2 CH3-CH(OMe)2


3 3 3 3
J = 4.5 Hz J = 5.9 Hz J = 6.2 Hz J = 5.4 Hz
JCP 1962, 37, 2907 ASV JCP 1956, 25, 362 ASV

The Karplus Equation


The Karplus equation is based on the observation, supported by theoretical considerations, that vicinal H-H
couplings will be maximal with protons with 180 and 0 diheral angles (anti or eclipsed relationship results in
optimal orbital overlap) and that coupling will be minimal (near 0) for protons that are 90 from each other. The
equation gives us approximate values for 3JHH as a function of dihedral angle between the protons. It should be
remembered, however, that this relationship strictly applies only in unstrained hydrocarbon systems, and that
electronegative substituents and ring constraints may cause substantial perturbations (in both positive and negative
directions) to the values predicted by this equation. Nevertheless, the Karplus curve (together with more
complicated variants) is the mainstay of conformational analysis for all ring systems, and has generally proved
reliable if care is taken. The constants Jo and K are used to correct for substituent effects in more sophisticated
uses of the Karplus equation, different Jo values are also used for the 0 to 90 and the 90 to 180 sections of the
curve.
The Bothner-By equation provides an empirical "Karplus" curve that does not require different J0 values for the
0-90 vs 90-180 sections: 44-01

Reich, U.Wisc. Chem. 605 5-HMR-5.1 J-vicinal


18 H
HH H H

16 H
Karplus Equation H
14
H
12 = 0 = 60 = 90 = 180
3
Bothner-By equation J = 7-11 Hz 3
J = 2-5 Hz 3
J = 0-2 Hz 3
J = 8-15 Hz
10
J / Hz

8 Karplus Equation
3

3
6 JHH = Jo cos2 - K
Jo = 14 (90-180), Jo = 10 (0-90), K = 0
4

2 Bothner-By equation
0 3
JHH = 7 - cos + 5 cos 2
180 150 120 90 60 30 0

A convenient graphical form of the Karplus relationship is given in Figure 5.5.1 below. Here two curves, separated
by 120, represent the predicted coupling constants for a proton H1 coupled to an adjacent methylene group (Hcis
and Htrans), as a function of the dihedral angle.

1-c /
60 30 0 30 60 90 120
18 18
H1

16 Hc H1 16
Hc
Ht
14 Hc 14
H1 axial
Ht
J1-t Hc eclipsed H1 Bothner-By equation
12 with H1 Ht 12 3
J = 7 -cos + 5 cos 2
H1 equat.
10 10
J / Hz
J / Hz

J1-c J1-t
8 8

6 6
J1-c

4 4
J1-t

2 2
J1-c

0 0
180 150 120 90 60 30 0
1-t /
Figure 5.5.1. Double Karplus Curve for Vicinal coupling in Cycloalkanes.

Reich, U.Wisc. Chem. 605 5-HMR-5.2 J-vicinal


5.5.3 Determination of Stereochemistry in Cyclic Compounds Using 3JHH

Cyclohexanes. It is often straightforward to establish stereochemical relationships among substituents, provided


that the spectrum can be analyzed. In chair cyclohexanes, the relationship among vicinal protons is restricted to the
narrow regions for 1-c = 40-60 on Figure 5.5.1 (i.e. to the left of the H1-eq crossing point at 60, and to the right of the
H1-ax point). These regions correspond to flattening of the cyclohexane, which is energetically easy. The opposite
distortion (1-c = 60-85 ) cannot occur to any significant extent. Jaa is usually much larger (9-12 Hz) than Jee or Jea
(each usually 3-4 Hz).
Below is reproduced the 100 MHz NMR signal of the H1 proton of iodocyclohexane at -80C (from F. R. Jensen, C.
H. Bushweller, Beck JACS 1969 91, 344, 3223). Under these conditions the ring inversion is slow on the NMR time
scale, and separate signals are seen for the two conformational isomers. The couplings are not always this well
resolved, but the axial proton multiplet will almost invariably be much wider than the equatorial one (remember that the
separation of the outer two lines of a first order multiplet is the sum of all the coupling constants). At room
temperature, the ring inversion will be fast on the NMR time scale, so an average spectrum will be observed. It will
look much like that of the axial proton, since the equatorial isomer is the major one.

H
H H
H H
H I
H
H H
H I

180 160 140 120 100 80


Hz
Axial and equatorial conformations of iodocyclohexane at equilibrium, -80 C (60 MHz)

The coupling constants in cyclohexane itself were determined by analysis of the AA'BB' pattern of
1,1,2,2,3,3,4,4-octadeuteriocyclohexane at -103 C (Garbisch, J. Am. Chem. Soc. 1968, 90, 6543). The bottom spectrum
(deuterium decoupled) is the experimental one, the top one is a simulation with the parameters listed.

D D JAB = -13.05 (Jgem)


D HA'
-103 C JAA' = 13.12 (Jaa)
D D
HB JBB' = 2.96 (Jee)
D
Calcd. D JAB' = 3.65 (Jae)
HB'
D HA
A = 1.12
HA' B = 1.61
Exp. HB
HB'
100 90 80 70 60 50 HA
Hz

The spectra of iodocyclohexane and cyclohexane itself also illustrate another feature common to many axial and
equatorial cyclohexane protons: the chemical shift of the axial proton is usually upfield of the equatorial one, in the
case of cyclohexane by 0.5 ppm.
The near identity of the magnitudes of the gem (2JAB = -13.05 Hz) and axial-axial (3JAA' = 13.12 Hz) couplings
seen in cyclohexane is a common feature of substituted chair cyclohexanes and half-chair cyclohexenes. Note that
the couplings do have opposite signs, although this is not detectable in first-order spectra. In molecules with
electronegative substituents (e.g. pyranose sugars) the vic axial-axial couplings are smaller than these, with typical
values between 8 and 11 Hz.

Reich, U.Wisc. Chem. 605 5-HMR-5.3 J-vicinal


In an idealized cyclohexane, Jee and Jae would be identical, since each corresponds to a dihedral angle of 60.
However, cyclohexanes are typically slightly flattened, presumably due to axial-axial repulsions. This moves the
dihdral angle for Jee to slightly higher than 60, hence smaller coupling, and that of Jae to slightly below 60, resulting
in larger coupling (see the shaded areas in Figure 5.5.2). The dihedral angle in cyclohexane itself is 57, and this
leads to the slightly smaller value for Jee (JBB' = 2.96) compared to Jae (JAB' = 3.65). Similar effects are also
commonly seen in substituted cyclohexanes which are conformationally homogeneous, especially if there are axial
substituents of any size. If the flattening is substantial, Jee can become too small to detect (as is the case for some
bicyclo[3.3.1]nonanes with 1-t = 90), and Jae can become substantially larger than the normal values of 3-4 Hz,
reaching valuesof 5 or even 6 Hz. Thus you cannot always rely on getting an exact count of vicinal neighbors to a
proton from its multiplicity.
1-c /
60 30 0 30 60 90 120
18 18
60 H1
Ha 16 16
Idealized He H1 axial Hc Hc
cyclohexane 60 14
He 14
Ht H1
Ha J1-t Ht
12 12
H1 equat.
ring 10 10

J / Hz
J / Hz

flattening J1-c J1-t


8 8

30 (J larger) 6 6
Ha J1-c
He Flattened
Flattened 4
90 (J smaller) J1-t cyclohexane 4
cyclohexane
He
Ha 2 2
J1-c
0 0
180 150 120 90 60 30 0
1-t /
Figure 5.5.2. Karplus Curve (using the Bothner-By equation: 3J = 7 -cos + 5 cos 2) for vicinal coupling in
cycloalkanes. The shaded area represents the conformational space of chair cyclohexanes, showing ring
flattening.
The near identity of Jee and Jea has the unfortunate consequence that the couplings to an equatorial proton do not
provide information about the stereochemistry of neighboring protons (i.e. whether they are axial or equatorial)
although they will usually provide a count of the vicinal neighbors.
Exercize (R-05I). The multiplet below corresponds to two of the protons of compound R-05I (C13H16O2). Analyze
and assign the multiplet, report couplings and values, and determine the stereochemistry and conformation of the
compound (add appropriate substituents and protons to the structure on the right). Briefly explain your reasoning.

Click for answer

50 40 30 20 10 0
Hz
O
O O O

Ph
1.85 1.80 1.75 1.70 1.65 1.60 R-05I
ppm

Reich, U.Wisc. Chem. 605 5-HMR-5.4 J-vicinal


Exercise: Analyze the NMR spectrum of the mixture of 3,5-diphenylbromocyclohexanes below (assign signals):

270 MHz 1H NMR spectrum (C18H19Br)


10.00 CDCl3, Mike Bowe Br H

Ph H Ph Br

Ph Ph

4.95 4.90
3.37

Hz
30 20 10 0
1.63 1.63
4.30 4.25 4.20
1.20
0.80 0.52 0.56
0.28

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5
ppm
Exercise: Examine the 220 MHz spectrum of proto-quercitol reproduced below, and analyze the couplings and
chemical shifts (McCasland, G. E.; Naumann, M. O.; Durham, L. J., J. Org. Chem. 1968, 33, 4220).

H H
6 OH OH
1 2
H OH
H OH
5 4 60 MHz
H 3
HO H H

proto-Quercitol 4.5 4.0 3.5 3.0 2.5 2.0 1.5


(C6H12O5)
J1,2 = 9.0
J2,3 = 9.0
J3,4 = 3.0 100 MHz
J4,5 = 3.1
J5,6e = 3.1 4.0 3.5 2.0 1.8
J5,6a = 2.9 ppm
H5 H4 H3 H2
J6a,6e = 13.8 40 20 0
J1,6e = 5.0 Hz H6e H6a
J1,6a = 11.5 H1

220 MHz

4.0 3.8 ppm 3.6 2.0 ppm 1.8

Reich, U.Wisc. Chem. 605 5-HMR-5.5 J-vicinal


Exercise: The 1H NMR spectra of two isomers of methyl 2-(N-benzoylamino)cyclohexanecarboxylate are show
below. Determine which isomer corresponds to spectrum A and B, and which conformation is the major one for
each. Focus on an assignment and complete analysis of the three downfield protons corresponding to the N-H,
-carbomethoxy and -aminobenzoyl protons.
H CO2Me
NHBz
CO2Me
H NHBz H
CO2Me
CO2Me NHBz
H
H H trans NHBz
H cis H

A 300 MHz 1H NMR spectrum


Dan Apella/Gellman
Hz
30 20 10 0

2.25 2.20 2.15 2.10


4.40 4.35 4.30 4.25

2.95 2.90 2.85


7.35 7.30 7.25

8 7 6 5 4 3 2 1 0
ppm

B 300 MHz 1H NMR spectrum


Dan Apella/Gellman
Hz
30 20 10 0

2.50 2.45 2.40 2.35 2.30 2.25 2.20 2.15 2.10 2.05 2.00 1.95 1.90

6.35 6.30 6.25 4.25 4.20 4.15 4.10 4.05

8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-5.6 J-vicinal


The pyranose (6-membered ring) forms of pentose and hexose sugars provide many examples where vicinal
proton coupling constants allow complete assignment of stereochemistry. Analyze the 1H NMR spectrum of glucose
pentaacetate reproduced below, assuming that you don't know the stereochemistry. Analysis of this type always
begin with the specific assignment of one or more of the protons, either from chemical shift information or the
number of couplings. In this example, the best place to start is H1, which can be recognized both from its chemical
shift (at 6.6), as well as from the fact this it will be the only proton in the molecule coupled to just one other
proton.
Exercise: Examine the 300 MHz spectrum of glucose pentaacetate reproduced below. Assume you don't know
the stereochemistry and use the spectrum to assign it at each carbons.

H4 OAc
6 H2
O
AcO
H1
6.606.58 AcO
5.85 5.80 5.75 5.35 5.30 5.25 5.20 H5 AcO
ppm ppm ppm 3 OAc
H
Glucose pentaacetate
(C15H22O11)
10 0
Hz

4.30 4.25 4.15 4.10 4.05 4.00


ppm ppm

7 6 5 4 3 2 1 0
ppm

Boat Conformations. In boat and twist-boat cyclohexanes there are multiple conformations, each of which have
available several C-C-C-C dihedral angles. In an idealized twist-boat there are four kinds of hydrogens, with eight
dihedral angle relationships (ca 30, 30, 50, 50, 70, 90,150, 170 degrees). In addition, there are six different twist
boats possible for a multiply-substituted cyclohexane so stereochemical assignments are very difficult.
Cyclohexanes in twist-boate conformations are quite rare, since twist-boat cyclohexane is ca 5 kcal higher in energy
than the chair form. They are commonly seen only in bicyclic structures, or in 6-membered rings with multiple
heteroatoms or those containing multiple sp2 carbons. Even cyclohexanes with a tert-butyl groups forced to be axial
can adopt modestly distorted chair conformations.

Twist-boat cyclohexane

Reich, U.Wisc. Chem. 605 5-HMR-5.7 J-vicinal


Cyclopentanes. The conformational analysis of substituted cyclopentanes is much more complicated than that
of cyclohexanes. The energy differences between various envelope and twist conformations in five-membered rings
are generally small, and there are as many as ten different envelope and ten different twist conformations, and each
conformation has multiple dihedral angle relationships. Several of the 20 possible conformations may be populated
in an individual structure. Thus the vicinal couplings in 5-membered rings are highly variable. For cyclopentanes in
envelope conformations Jcis > Jtrans in the flat part part of the envelope, whereas in twist conformations the tendency
is for Jtrans > Jcis. In general, no firm assignments of stereochemistry can be made using the size of couplings alone
unless a specific substitution pattern or heterocyclic system has been carefully investigated, or if substitution
patterns allow prediction of the conformation.
1-t
180 160 140 120 100 80 60 40 20 0

Jcis > Jtrans H1


Hc
16 Hc 16
H H1 H1
Hc
H Ht Ht
H1 axial
Ht Hc
J1-t H1
12 Hc eclipsed with H1 12
Ht
Envelope
J H1 equat. J

Jtrans > Jcis J1-t


8 J1-c 8
H J1-c
Jcis > Jtrans
Jtrans > Jcis

H 4 4
Twist
J1-t J1-c

0 0
60 40 20 0 20 40 60 80 100 120
1-c
Inspection of the double Karplus curves indicates a significant difference between the typical behavior of adjacent
CH2 groups in cyclohexanes and cyclopentanes. In a chair cyclohexane only one of the four vicinal couplings can be
large (> 7 Hz), whereas in a cyclopentane it is common for 2 or even 3 of the 3J couplings to be large.

HA HA JAD > 7 Hz
JAD > 7 Hz
JAC > 7 Hz
HB JAC < 6 Hz
HC HC HB
JBC < 4 Hz
JBC < 4 Hz
HD JBD 7 Hz
JBD < 6 Hz HD

In most cyclopentanes, the C-C-C-C dihedral angles are significantly smaller than the 60 found in cyclohexanes.
Cis protons will tend to have H-C-C-H dihedral angles close to 0, and trans near 120. The cis couplings (8-10 Hz)
are usually larger than trans (2-9 Hz). However the Karplus curves for cyclopentane have a region where the cis and
trans lines cross (Figure above, at ca 20 dihedral angle), so there is a small region where Jtrans > Jcis. There are
alsoe cases where cis and trans couplings are identical, as on the compound below, where the allylic proton is a
quartet of doublets, arising from accidental equivalence of three vicinal couplings.

Reich, U.Wisc. Chem. 605 5-HMR-5.8 J-vicinal


The three vicinal couplings to the allylic hydrogen (cis
and trans in the 5-membered ring, and the coupling to
20 10 0
the vinyl H) are accidentally equivalent.
Hz O O O
This signal provides a word of warning about jumping
OEt
to conclusions during the interpretation of coupling
H patterns - most would assume that the proton
responsible for this multiplet is definitely coupled to a
methyl group and one other proton by a small
5.6 5.5 5.4 coupling.
ppm

If the ring puckering is strong enough, then Jtrans > Jcis. In bicyclo[2.2.1]heptanes the endo-endo and exo-exo 3J are
always greater than endo-exo couplings. Thus stereochemical relations among vicinal protons in 5-membered rings
cannot be reliably determined by simply measuring coupling constants, except in cases where the substitution pattern
of the specific ring system has been carefully investigated. For example, in the benzodihydrofurans below, changing
the size of the substituent R causes a reversal in the size of Jcis and Jtrans.

In rigid cyclopentanes, Jcis > Jtrans

Ph R= Me Et iPr
H
Y Hexo Hexo 3
Jcis 7.2 6.4 4.8
X Hexo X H 3
Jtrans 6.8 7.1 8.4
Hendo Y Y R
Hendo X Hendo O
JHH = 8-9 Hz JHH = 7-9 Hz JHH = 2-3 Hz

Cyclobutanes. Cyclobutanes are even flatter than cyclopentanes, so that cis couplings are almost always larger
(6-9 Hz) than trans (2-8). However, if structural features which promote strong puckering of the ring such as a trans
ring fusion, large or electronegative substituents are present, then trans couplings can become larger than cis, as
shown for cis-1,3-dibromocyclobutane and cyclobutanol below.

Br O O
O H H 9.7 10.7
cis
H trans
7.0 H
Br HO H
3
Jcis 10.4 7.1 4.7 10.0 6.9
3
H 8.1 H
Jtrans 4.9 8.8 1.8 6.4 4.6
JACS 1969, 91, 5124 JACS 1969, 91, 5124

Reich, U.Wisc. Chem. 605 5-HMR-5.9 J-vicinal


Cyclopropanes. Dihedral angles in cyclopropanes are rigidly fixed by the geometry of the ring system. We
therefore find that Jcis (7-10 Hz) is always larger than Jtrans (2-6 Hz), and this can be reliably used for structure
assignment.

H CO2Et 10 Hz 4.5 Hz
H H
H CO2Et
CO2Et H Ph H Ph
H
7 Hz
CO2Et H
Ph Ph 13.3 Hz
5 Hz H 6.5 Hz O H
O
3 3
Jcis = 8.7 Hz Jtrans = 5.4 Hz
JOC 1970, 35, 1600

The same relationship holds for the 3-membered ring heterocycles, although the range of observed couplings is
wider.

cis H Y = CH2 O NH S S=O


H 3
Jcis 9.0 4.45 6.3 7.15 11.5
Y trans 3
Jtrans 5.6 3.1 3.8 5.65 10.5
H
Tet. Lett 1970, 2877
H

Reich, U.Wisc. Chem. 605 5-HMR-5.10 J-vicinal


Summary: On the double Karplus curve below are indicated the dihedral angles and hence the cis and trans
3-bond couplings that can be observed for various rings. Chair cyclohexanes are conformationally well defined, with
a relatively small range of 3J couplings possible (Jeq-eq and Jeq-ax typically 3-4 Hz, and Jax-ax typically 8-13 Hz). With
5 and 4 membered rings a wider range of couplings are seen depending on the extent and type of puckering
present. Cis couplings will typically be larger than trans couplings. Unfortunately for both cyclopentanes and (less
commonly) cyclobutanes, Jtrans can occasionally be larger than Jcis for pseudoaxial protons, if the conformation
places the dihedral angle to the left of the crossing point at ca 20. For such systems both Jtrans and Jcis will be
relatively large (8-10 Hz). Cyclopropanes are rigid, and Jcis (eclipsed, = 0) is always greater than Jtrans ( =
120). With this in mind, the appearance of only well defined large (ca 10 Hz) and small (ca 3 Hz) in a CH coupled
vicinally to one or more CH2 groups is quite characteristic of cyclohexanes. Cyclopentanes and cyclobutanes, on the
other hand, tend to more frequently have intermediate size couplings (5-9 Hz), and often nearly equal and large
coupling to cis and trans vicinal neighbors.
1-t
180 160 140 120 100 80 60 40 20 0

H1
Hc
16 Hc 16
H1
H1
Ht Hc Ht
H1 axial
Hc
J1-t Ht H1
12 H1 eclipsed with Hc 12
Ht

J H1 equat. J

J1-t
8 J1-c 8

J1-c

4 4

J1-t J1-c

0 0
60 40 20 0 20 40 60 80 100 120
1-c

Chair cyclohexane

Cyclopentane

Cyclobutane

Cyclopropane

37-03

Reich, U.Wisc. Chem. 605 5-HMR-5.11 J-vicinal


5.5.10 Acyclic Stereochemistry using 3JHH

The use of 3J for conformational analysis in acyclic systems can be more difficult than within rings because of
the larger number of conformations typically possible (Review: "Determination of Relative Configuration in
Organic Compounds by NMR Spectroscopy and Computational Methods" Bifulco, G.; Dambruoso , P.;
Gomez-Paloma, L.; Riccio, R. Chem. Rev., 2007, 107, 3744 DOI: 10.1021/cr030733c). Basically, a single
coupling constant cannot usually distinguish which of two diastereomers might be present since there are 3
possible staggered conformations for each diastereomer, two of which will typically have very similar predicted
coupling constants for a pair of vicinal protons. Assignments become possible only when one can make some
reliable predictions on which conformation predominates. One such situation is encountered with the
diastereomeric products of an aldol condensation, as shown below:

O O HO HB O HO HB
O
Ph + Ph
+ H Ph
Me HA HA Me
syn (erythro) anti (threo)
In non-polar media, a hydrogen bond between OH and the carbonyl group is expected. Since in the syn isomer
both hydrogen bonded conformations have a gauche relationship between HA and HB, we expect a smaller 3J for the
syn isomer than for the anti, where one of the H-bonded conformations has an anti relationship between HA and HB
(Stiles-House rule: Stiles J. Am. Chem. Soc. 1964, 86, 3337; House J. Am. Chem. Soc. 1973, 95, 3310; Heathcock,
JOC, 1980, 45, 1066; Mukaiyama JACS, 1974 96, 7503).

H H H H
H
O O O O O O O O O
HA CH3
HA CH3 HA CH3

Ph HB Ph HB Ph HB Ph HB Ph HB
CH3 HA CH3 HA
O
JAB = 3-4 Hz JAB = 3-4 Hz JAB = 10-12 Hz JAB = 3-4 Hz
syn (JAB = 6.5) no H-bond anti (JAB = 8.5)

Syn Anti Jsyn < Janti


O OH O OH
H H Jsyn Janti syn anti
Ph Ph
2.5 9.0 5.40 4.83

O OH O OH
H H
Ph Ph
6.0 6.5 3.51 4.00

O OH O OH

6.1 9.5 3.50 3.65

Ph Ph

Note that the chemical shift of the CH(OH) proton is not a relible indicator of stereochemistry.

Reich, U.Wisc. Chem. 605 5-HMR-5.12 J-vicinal


This method will only work if the intramolecular hydrogen bonded conformations are the principal ones for both
diastereomers. Thus it sometimes fails in situations where the and/or -substituent is large, as in the -t-Bu
aldols below. Here gauche interactions destabilize the hydrogen bonded six-membered ring of the syn isomer,
leading to a large coupling because of a high population of the non-hydrogen-bonded conformation with t-Bu and Ph
anti periplanar in the syn isomer (Heng, Simpson, Smith J. Org. Chem. 1981, 46, 2932). Similarly, in more
complicated systems additional conformational constraints can overwhelm the hydrogen bond effect. For example a
3-alkyl substituent in a cyclohexanone aldol has Jsyn > Janti (Kitamura, Nakano, Miki, Okada, Noyori J. Am. Chem.
Soc. 2001, 123, 8939).
O HO H O H OH
O HB Ph O HO Ph H H
large
Ph Ph
OH groups HB
t anti
Bu HA t
Bu HA
Jsyn = 10.0 Hz Janti = 3.5 Hz Jsyn = 4.4 Hz Janti = 3.9 Hz

For use of 13C shifts to assign stereochemistry see: Heathcock, J. Org. Chem., 1979, 4294.

Conformations of CH2 Chains. Adjacent CH2 groups in acyclic molecules (X-CH2-CH2-Y) typically show
apparent triplets, or higher multiplets if X and/or Y contain vicinal protons coupled to the CH2 groups. These are
actually AA'BB' or AA'XX' systems, and thus are inherently non-first order. It turns out that if X and Y are sterically
small, then the gauche conformation is sufficiently populated (anti/gauche ca 3:1) that nearly equal JAX and JAX' are
seen, leading to the apparent triplets. If X and/or Y is sterically large, then more complicated patterns are seen. See
Sect. 5.15.

Reich, U.Wisc. Chem. 605 5-HMR-5.13 J-vicinal


5-HMR-5.12 Allylic 3J

Couplings of vinyl hydrogens to vicinal protons across single bonds follow Karplus relationships similar to those of
other vicinal couplings. The size of J is maximal at dihedral angles of 180 and 0, and minimal when the C-H bonds
are perpendicular ( = 90), although the coupling does not go to 0.
60 30 0 30 60 90 120
15

H
H
H H H
H H H
H
10 H H H
H H
3 3 3
H
J = 6.6 Hz J = 11.2 Hz J = 2.6 Hz

J / Hz
3
J = 6.6 cos2 + 2.6 sin2 (0 < < 90)
3
J = 11.6 cos2 + 2.6 sin2 ( > 90) 5

H C
0
90 60 180
30 0 150 120
/
In acyclic systems without strong conformational restrictions, rotational averaging produces couplings of 5-8 Hz,
very similar to those observed in aliphatic chains.
For cyclic olefins, the 3J coupling decreases as the ring size gets smaller. In cyclohexenes the couplings of an
adjacent CH2 group to the vinyl hydrogens are typically 4-5 Hz for the equatorial H, and 1-3 Hz for the axial H, as
shown in the figure above. In cyclohexene itself the average of these is observed.

H H H
3
H
J = 5.7 Hz 3
3
J = 3.1 Hz 3
J = 2.1 Hz J = 1.0 Hz
H
H H
H
Dienes: The central 3J coupling in acyclic dienes is typically 10 Hz, very similar to the 3Jcis across double bonds,
provided that steric effects do not prevent the diene from achieving a near planar conformation. The coupling is
again reduced in cyclic dienes, both because the dihedral angle is now 0 instead of 180, and because of inherent
reduction in the coupling because of angle distortions.
3
H J = 1.94 Hz
MeO O
3
J = 11.6 H
H H
PhS H O H 3
3
J = 5.06 Hz
J = 6.9
H OMe H
3 H
3
J = 10.4 Hz H J = 11.3 Hz
Aldehydes: In unconjugated aldehydes the 3J coupling is typically small (1-3 Hz). The coupling becomes
considerably larger in conjugated aldehydes like acrolein, where the dihedral angle will be either 0 or 180 to
maximize overlap of the systems. 37-03

H H H H H
O O O Ph O
3 3
H 3
J = 1.4 Hz H
3
J = 1.2 Hz H J = 8 Hz H J = 7.7 Hz

Reich, U.Wisc. Chem. 605 5-HMR-5.14 J-vicinal


5-HMR-5.13 Olefinic 3J

The cis and trans couplings across a double bond are very reliable indicators of stereochemistry. With virtually no
exceptions, 3Jtrans > 3Jcis, typical vaues are 17 and 10 Hz. However, the ranges do overlap for very strong electron
donating (J increases) and withdrawing groups (J decreases).

X Jc Jt EX

H Y H H Li 19.3 23.9 -1.0


H X SiMe3 14.8 20.4 1.9
X H X Y H 11.6 19.1 2.1
H H
Jt = 12 - 24 Hz Jc = 3 - 19 Hz CO2H 10.4 17.2
OMe 7.0 14.3 3.5
F 4.7 12.7 4.0

The coupling varies with bond order. Thus the cis coupling in benzene and other aromatic six and larger
membered rings is typically below 10 Hz (one empirical equation is: 3J = 8.65 ( bond order) + 1.66):

H J = 8.6

H H H H
J = 11.6 J = 7.6 J = 6.0
H H H H

The tropone shows larger bond-alternation effects than the aromatic tropylium ion or the azulene.

12.0 9.5
8.4 10.4
9.7 10.0
10.7 O Ph
9.8 +

JACS 1969, 5287 JACS 1969, 5287

Cycloalkenes smaller than cyclohexene show substantially reduced 3J values (Chem. Rev. 1977, 77, 599). Thus
cyclopentenes can be easily distinguished from cyclohexenes and larger rings if this coupling can be identified.

H H H
H H H

H H H H H H
3 3 3 3 3 3
J = 10.4 J = 11.0 J = 10.1 J = 5.6 J = 2.9 J = 1.3
(11.8 - 12.8) (9.7 - 12.5) (8.8 - 10.5) (5.1 - 7.0) (3.0 - 3.5) (0 - 2.1)

Heterocycles also generally have smaller 3J values than hydrocarbon systems.

6.9 - 9.1 1.94 3.1 - 3.8 3.4


O
5.06 1.3 - 2.0 2.6
4.0 - 6.0 O 1.47
N O N O

O H

2 Hz
6 Hz 6.3
O O O

Reich, U.Wisc. Chem. 605 5-HMR-5.15 J-vicinal


Copyright Hans J. Reich 2017
All Rights Reserved
5.6 Long-Range (4J and higher) Proton-Proton Couplings University of Wisconsin

Proton-proton couplings over more than three bonds are usually too small to detect easily (< 1 Hz). However,
there are a number of important environments where such couplings are present, and can provide useful structural
information. Coupling across -systems are the most frequently encountered 4J couplings: the meta-coupling in
aromatic compounds, and the 4-bond allylic, propargylic and allenic couplings. 4-Bond couplings across saturated
carbons (sp3) or heteroatoms are rarer, and are usually seen only in cyclic compounds when there is a favorable
geometric alignment along the H-C-C-C-H chain ("W-Coupling"). Longer range couplings (5J and higher) are also
observed, particularly in acetylenes and allenes (Chem. Rev. 1977, 77, 599).

H H H
H H H H
H H H

Allylic Propargylic Allenic W-Coupling


4 4 4 4 4
J = 2 to 3 Hz J = -3 to +3 Hz J = 2 to 4 Hz J = 6 to 7 Hz J = -1 to +3 Hz

H H H
H H

H
Homoallylic Homopropargylic Homoallenic
5 5 5
J = 0 to 8 Hz J = 2 to 4 Hz J = 3 to 6 Hz

W-Coupling in Saturated Systems. Normally long-range couplings across saturated carbons (or O and N) are
too small to detect easily (<1 Hz). However, if there is proper orbital alignment between C-H bonds and the
intervening C-C bonds then 4-bond and higher couplings can be observed. The most favorable alignment is the W
arrangement of the connecting bonds ("W-coupling"), in which the H-C-C and C-C-H fragments are close to
coplanar in an anti-arrangement. Thus coupling between 1,3-equatorial protons in cyclohexanes is frequently seen.
However, couplings across U-shaped HCCCH fragments can also sometimes be detected. Long-range couplings
can become quite large in rigid strained bicyclic ring systems and/or when there are multiple coupling pathways
available.

H H
H H H

H H
"W-Coupling" "U-Coupling" H
J = +2 Hz J = -1.1 Hz H
H
4 4 4
J = 7 Hz J = 18 Hz J = 10 Hz
JACS` 1966, 88, 4437
Do not expect to see such large long-range couplings
in unstrained (5,6,7-membered) rings.

H H H
H H
H HO H
O H H H

H H H H H H
H J = +5.2
4
4 4
Jcis = +4.8 4
J = 2-3 Hz J = 1.2 Hz J = -0.9 Jee = 0-2 Hz
4
Jtrans = -2.8 The "U" couplings
JPC 1967, 1555 are very small (<0.1 Hz)
JACS 1969, 5124

Reich, U.Wisc. Chem. 605 5-HMR-6.1 J-long range


Exercise: Examine the 300 MHz spectrum of two of the protons of 3-bis(phenylthio)methylcyclohexanone
reproduced below. Assign the protons and analyze the couplings (Sikorski/Reich).

O
Hz
30 20 10 0

SPh

SPh
C19H20OS2 2.70 2.65 2.60 2.55 2.50 2.45
ppm

Exercise: Examine the 300 MHz 1H NMR spectrum of O-benzyl rhamnal below. Assign all protons, and extract
couplings

300 MHz 1H NMR Spectrum of O-benzyl-L-rhamnal (C13H16O3)


Source: ASV/Reich
Hz
30 20 10 0

Ph O
O

HO

4.90 4.85 4.80 4.75 4.70 4.65 1.95 1.90

5.16

6.35 6.30 6.25 4.40 4.35 4.30 4.25


3.083.95 3.90 3.85 3.30 3.25 3.11

1.00 1.01 1.02 1.04 1.06

1.40 1.35

8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-6.2 J-long range


Cyclobutanes generally show substantial cross-ring 4J couplings, with 4Jcis, which has the proper orientation for
a W-coupling, greater than 4Jtrans. In fact, Jcis > 0 and Jtrans < 0 in almost all cases (A. Gamba, R. Mondelli
Tetrahedron Lett. 1971, 2133), so this coupling can be used to assign stereochemistry in cyclobutanes. The figure
below illustrates the effect of the long range couplings in a cyclobutanone (a simple AB quartet would be expected if
there were no long-range couplings - top simulation). An AA'BB' simulation gives the parameters shown on the
figure. The pattern is not completely centrosymmetric because there is a small long-range coupling from the
side-chain CH2 ( 1.75) to one of the cyclobutane protons.

200 MHz 1H NMR Spectrum O


AB = 45.20
Source: Art Cammers/Vedejs
JAA' = 6.50
JBB' = 5.70
JAB = -18.70 1.75
JAB' = -2.00
OMe

Decoupling at 1.75 symmetrizes the


AA'BB' pattern (there is a small rong-range
coupling between the pentyl CH2 group
and one of the ring protons).
640 620 600 580 560
Hz

7 6 5 4 3 2 1 0
ppm

O
JAA' = 0 AB = 45.20 Simulation
B B' JBB' = 0
JAB = -18.70
A A' JAB' = 0
R OMe
JAA' = 0
JBB' = 0 Simulation
JAB = -18.70
JAB' = -2.00

JAA' = 6.50
JBB' = 5.70
Simulation
JAB = -18.70
JAB' = -2.00

Spectrum

3.2 3.1 3.0 2.9 2.8


ppm

Reich, U.Wisc. Chem. 605 5-HMR-6.3 J-long range


Stereochemistry of cis/trans Decalins. A useful application of long range couplings for the assignment of
ring-fusion stereochemistry in decalin ring system bearing an angular methyl group has been developed
(Williamson, K. L.; Howell, T.; Spencer, T. A. J. Am. Chem. Soc. 1966, 88, 325). In the trans-decalins, there are
usually several ideal W-pathways for long range coupling between the methyl group and axial protons. In
cis-decalins, there are fewer or no such pathways. Thus in a pair of cis/trans isomers, the methyl group in the trans
isomer will usually be broader (or will actually show splitting), whereas the cis isomer will have a sharper (unsplit)
methyl group.

H H
H H H
w1/2 = 0.90 Hz C H H
H H
C (TMS: 0.38 Hz) OH C
Me4Si H OH
H
H
H H O H H
H H
trans-Decalin O

H 0.3 Hz 0.63 Hz 1.34 Hz


H w1/2 = 0.39 Hz
C H (TMS: 0.33 Hz)
H
H
H

cis-Decalin

W-Coupling Across Heteroatoms. In conformationally well-defined systems significant 4J couplings can be seen
to OH and other XH protons. In the example below, the long-range W-coupling between the OH proton and the
axial proton at C6 was used to assign configuration to the major isomer formed in the reaction. In the minor isomer
the OH proton was not detectably coupled. The well-defined cyclohexane 3Jax-ax and 3Jax-eq at C2 in both isomers
shows that the ring-flip isomer shown predominates (Bueno, A. B.; Carreno, M. C.; Ruano, J. L. G Tetrahedron
Lett. 1995, 36, 3737).

broad s
H H
O d, J = 1.9 O
H O OH
H OO O S
6
AlMe3 O S H Tol
O S : Tol +
Tol H O H
H 2 ZnBr3
O H 96 : 4
O H
dd, J = 13.3, 4.0
dd, J = 13.1, 3.6 (this is major conformation)

In a related system, the observation of an unusually large 4J across the sulfone sulfur was interpreted in terms of
the conformation shown, in which the methyl group is over the ring, rather than alternative conformations in which
the sulfone oxygen is over the ring (Kaloustian, M. K.; Dennis, N.; Mager, S.; Evans, S. A.; Alcudia, F. Eliel, E. L. J.
Am. Chem. Soc. 1976, 98, 956-965).

O
CH3 CH3
S O O
S O
O
i
Pr H O
O i H
Pr O
4
J = 1.14

Reich, U.Wisc. Chem. 605 5-HMR-6.4 J-long range


Allylic Coupling. 4-Bond coupling of vinyl to allylic hydrogens is usually easily observable. We can think of the
coupling as having two components, the usual W-coupling transmitted through the -system, which is positive and
is maximized for the trans proton when the allylic C-H bond is in the plane of the vinyl C-H group ( = 0 , J > 0), and
a -component, which is negative, and whose magnitude is maximized when the allylic C-H bond is perpendicular to
the double bond ( = 90 , J < 0) (Garbisch, J. Am. Chem. Soc. 1964, 86, 5561). The positive -contribution added
to the larger negative -contribution normally results in a numerically slightly smaller (negative) coupling to the trans
vinyl proton, but the effect is small, and not reliable enough for the unambiguous determination of double bond
stereochemistry (note the marked entry below in which Jtrans > Jcis) (Barfield, M.; Chakrabarti, B. Chem. Rev. 1969,
69, 757).

Allylic Coupling (4J)

H H H

H H H
H H
H 4
J = 1.3 cos2 - 2.6 sin2 ( = 0 - 90)
= 0 = 180 = 90 4
J = 2.6 sin2 ( = 90 - 180)
4
J = +1.3 0.0 0.0
4 Can one distinguish E and Z isomers?
J = 0.0 0.0 -2.6
R' R'
X X
H X R X
H H CH2 CH2
CH3 CH3
R H
H H
Jtrans Jcis Jtrans Jcis
X = CH3 -1.25 -1.25 X = CH3 -1.2 -1.7
X = Cl -0.7 -1.3 X = Cl -0.93 -1.41
X = Ph -0.8 -1.5 X = OH -1.25 -1.55
X = Ph -1.47 -1.47
X = CH=CH2 -1.50 -1.30

A typical example of allylic coupling is shown by methyl crotonate, where the 4J between the CH3 and the -CH is
1.8 Hz.
477.1
475.3
470.1
468.4

NMR R-7L, C5H8O2 O


CDCl3 300 MHz 1H NMR
Source: A. Fiedler/HJR CH3 O
J = 1.8 Hz H 30 20 10 0
Hz

3.17
3.04

7.05 7.00 6.95 6.90 5.90 5.85 5.80


1.00 1.01

3.70 1.90 1.85

7 6 5 ppm 4 3 2 1 0
Reich, U.Wisc. Chem. 605 5-HMR-6.5 J-long range
Benzylic Coupling. Coupling between benzyl protons and ortho hydrogens on aromatic rings are typically <1 {
Hz, and thus often not resolved, but almost always cause significant broadening of both the aromatic and benzyl
protons. This can be clearly seen in the spectrum of p-methylacetophenone below, where the Ar-CH3 is wider and
thus noticeably shorter than the C(=O)-CH3, and the protons ortho the CH3 are broader than those ortho to the
acetyl group. The coupling is related to -bond order, so it is usually smaller than allylic coupling.

300 MHz 1H NMR spectrum in CDCl3 (C9H10BrO)


Source: Aldrich Spectral Viewer/Reich Hz O
40 20 0
CH3
CH3 O Ar
CH3 Ar

CH3

7.9 7.8 7.7 7.6 7.5 7.4 7.3 7.2 2.6 2.5 2.4 2.3

8 7 6 5 ppm 4 3 2 1 0

CH3 Note the much broader aromatic peaks


(almost no meta-coupling detectable) in
the bromoxylene than in the tribromide
due to coupling of the CH3 protons to the
CH3 aromatic C-H
Br

7.4 7.3 Br 7.2 7.1 7.0 6.9 6.8 6.7

Br
Br

7.8 7.7 7.6 7.5 ppm 7.4 7.3 7.2 7.1


Benzylic Coupling (4J)

CH3 4
J = -0.75 Hz H
CH3 4
J = -1.11 Hz H
H 4
H J = -1.5 Hz
H
4
J = -0.45 Hz
CH3
4
CH3 J = -1.75 Hz H
H 4
J = -1.06 Hz

Reich, U.Wisc. Chem. 605 5-HMR-6.6 J-long range


Homoallylic Coupling. Couplings across 5 bonds are unusual, but can be seen under favorable circumstances.
Coupling is optimal when both C-H bonds are aligned with the -orbital of an intervening double bond (perpendicular
to the plane of the double bond). Especially large long-range couplings are seen for 1,4-cyclohexadienes and related
structures where there are two paths for the coupling.

Homoallylic (5J)
H H
H
H H


H
CH3 O
H O
H H
4
5
J = 2-5 Hz J = 1.6 Hz
5
Optimal alignments for J 5
J = 2.7 Hz

Some extreme examples of large homoallylic coupling:

H H H H H H H Ph
H H N
H H Ph3C Ph Ph3C H
5
Jcis = 9.63 Hz Ts
5
5 Jcis = 3.0 Hz 5 5
Jtrans = 7.5 Hz
Jtrans = 8.04 Hz Jcis = 11 Hz
5
JACS 1968, 3590 Jtrans = 7.2 Hz

A typical eample of homoallylic as well as allylic coupling is shown below:

540.2
539.1

533.2
532.1
2059.6
2058.2

Hz
5
JHH = 1.1 Hz 30 20 10 0

CH3
CH3 O

H O
4
JHH = 1.4 Hz

6.9 6.8 1.85 1.80 1.75

Reich, U.Wisc. Chem. 605 5-HMR-6.7 J-long range


Long range Couplings in Acetylenes and Allenes. No special structural features are required to observe 4- and
5-bond couplings across acetylenes and allenes - such couplings are always present. Even couplings across 5, 6, and
more bonds are detected across polyacetylene or cumulene chains

H
H CH3 CH2-R CH3 CH2-OH
9
4
5
J = 2.5 Hz J = 0.4 Hz
J = 3 Hz
H CH3 H
C H2 CH2 CH3

t
Bu H CH3 CHO
4
J = 6.5 Hz 6
J = 0.4 Hz
5
J = 3.3 Hz

753.7
751.1
748.5
R-22B C3H4O
300 MHz 1H NMR spectrum in CDCl3
Source: ASV/Reich
triplet (J = 2.6 Hz)
OH
H
H
H

Hz
30 20 10 0

4.30 4.25 2.5

R-22I (C4H5Br)
300 MHz 1H NMR spectrum in CDCl3 5
J = 2.5 Hz
1178.2
1175.6
1173.1
1170.8

Source: Aldrich Spectra Viewer/Reich Br

Hz
30 20 10 0

3.95 3.90 3.85


1.9

Long range couplings like this are also observed across nitrogen as in the nitrilium ion below:
+
CH3 C N CH3
5
J = 2.5 Hz
JACS 1968, 4666

Reich, U.Wisc. Chem. 605 5-HMR-6.8 J-long range


Copyright Hans J. Reich 2017
All Rights Reserved
5.7 Pople Nomenclature for Coupled Spin Systems University of Wisconsin

The analysis of complex NMR patterns is assisted by a general labelling method for spin systems introduced by
Pople. Each set of chemically equivalent protons (or other nuclei) is designated by a letter of the alphabet. Nuclei are
labeled AX or AMX if their chemical shift differences are large compared to the coupling between them ( > 5J).
Nuclei are labeled with adjacent letters of the alphabet (AB, ABC, MN or XYZ) if they are close in chemical shift
compared to the coupling between them (i.e. if they are strongly coupled).
If groups of nuclei are magnetically equivalent, they are labeled AnBn, etc. Thus CH3 groups are A3, or X3. A group
of magnetically equivalent nuclei must have identical chemical shifts, and all members of the group must be coupled
equally to nuclei outside the group. If nuclei are chemical shift equivalent but not magnetically equivalent, then they
are labeled AA', BB'B'' or XX'. Thus in an A2X2 system the A nucleus must have identical couplings to the two X
nuclei. In an AA'XX' system, on the other hand, JAX JAX'. There are usually profound differences in the appearance
of A2X2 compared to AA'XX' patterns.
JAX JAX
A X Magnetically equivalent: A X Magnetically nonequivalent:
JAX = JAX JAX JAX'
A JAX X thus A X A' JAX' X' thus AA'XX'
2 2

Two-Spin Systems
AX First order. Significant parameters: JAX. A and X are each doublets.

AB J is directly measurable, A and B must be calculated. Intensities are distorted: the doublets are not 1:1;
the inner lines are larger, the outer lines smaller.

AB
AX

B
X A B A
Three-Spin Systems
AX2 First order. Significant parameters: JAX. A is a triplet, X is a doublet.
AB2 Second order. Both JAB and AB must be calculated - neither can be directly measured from the spectrum.
Significant parameters: JAB, AB.
R1

R2 R2 R2 AB2
1 AX2
R
R2
X X A B
AMX First order. Significant parameters: JAM, JAX, JMX. A, M and X are each doublet of doublets (assuming all
three couplings are large enough to detect). Typical systems are trisubstituted benzenes, vinyl groups, and
monosubstituted furans and thiophenes.
ABX Second order. This is a very common pattern. JAB is directly measurable. The parameters JAX, JBX, A and
B can be calculated from the line positions of the spectrum once it has been properly analyzed.

ABC Second order. This pattern can only be accurately solved using computer simulation methods. Manual
analysis as a distorted ABX or even AMX pattern will lead to approximate values of coupling constants, which
in severe cases can be drastically wrong.
AMX
ABX ABX ABC

X B A C B A

Reich, U.Wisc. Chem. 605 5-HMR-7.1 Pople


Four-Spin Systems

AX3 First order. Significant parameters: JAX. Commonly seen in CH3CHXY groups.
AB3 Second order. Computer simulation required to solve.

AX3 AX3 AB3 AB3

A2X2 First order. This is a very rare pattern. A and X are each triplets.
A2B2 Second Order. Rare.
AA'XX' Second order. Common pattern. Can be solved by hand, but there are several ambiguities. For example,
one cannot distinguish JAA' from JXX'. Significant parameters: JAA', JXX', JAX, JAX'. The AA' and XX' patterns
are each centrosymmetric, and they are identical to each other, hence the inability to distinguish A from X
parameters. A common type is X-CH2-CH2-Y, which is often just two triplets. Also common are
p-disubstituted benzenes: and dioxolanes:

AA'BB' Second order. This is a common pattern. Requires computer simulation to solve. Seen in X-CH2CH2-Y
groups where X and Y have similar shift effects. The entire multiplet is centrosymmetric (i.e., the AA' part is a
mirror image of the BB' part).

A2X2 A2B2 AA'XX' AA'BB'

AMPX First order. Commonly seen in ortho and meta disubstituted benzenes, 2- or 3-substituted pyridines and
related aromatics.

SCl

NO2

8.3 8.2 8.1 8.0 7.9 7.8 7.7 7.6 7.5 7.4
Cl

NO2

8.2 8.1 8.0 7.9 7.8 7.7 7.6 7.5

Reich, U.Wisc. Chem. 605 5-HMR-7.2 Pople


Five-Spin Systems
A2X3 First order. Very common pattern: ethyl groups: CH3CH2-R where R is an achiral electron withdrawing
group (if R is chiral then we get an ABX3 pattern)

A2B3 Second order. Seen in ethyl groups CH3CH2-R where R is a metal: e.g. CH3CH2-SiR3.

R A2X3 A2X3
A2B3

B A

ABX3 Second order, but some types are soluble by hand. Commonly seen in ethyl groups in chiral
molecules where the CH2 protons are diastereotopic. For these the X part is always a triplet.
Methyl-substituted alkenes can also give this pattern. Shown below are the A2 / AB parts.

HA HB HB
R
CH3 R* CH3 ABX3 ABX3 ABX3
X3 X3 HA
A2X3 ABX3

B A

AAMM'X Second order. Part structures like -CH2-CH2-CH- can appear to be first order (t, q, t).

AA'BB'C Always second order. Commonly seen in monosubstituted phenyl groups.

R
HA' HA

HB' HB
HC AA' C BB' BB' C AA'
R = electron withdrawing R = electron donating

ABMXY Second order. Part structures like -CH2-CH-CH2- are actually two ABX patterns which share a
common X.

R2 HM
R1 R3

HB HA HY HX

Reich, U.Wisc. Chem. 605 5-HMR-7.3 Pople


Six-Spin Systems
1 2
AA'MM'XX' Not actually first order, but a common type, R -CH2-CH2-CH2-R often looks nearly first-order, especially if
1 2
R and R are relatively small groups. Usually two triplets and a pentet, (or a pentet (2H) and a triplet (4H) if
R1 and R2 are identical).

AA' XX'
1
R R2
MM'

ABMX3 Second order. There are two types here (at least): those with R1 and R2 on the same carbon, and those
with the R groups vicinal. For the former, if R1 and R2 are different, the AB part is an AB quartet, each half of
which is split by three protons, thus an AB quartet of pentets, or an AB quartet of doublets of doublets. If R1
and R2 are the same, or the chemical shift between A and B is very small, then the AB part might be a pentet
or a doublet of quartets.

AB
R1 ABMX3
X3 M
R2

If R1 and R2 are vicinal then the AB part always has the appearance of the AB part of an ABX system (an
ab quartet of doublets), the M part is ddq, and the X3 part a doublet.

R2
R1
M
X3 AB

Reich, U.Wisc. Chem. 605 5-HMR-7.4 Pople


Seven-Spin Systems
AX6 First order. Common pattern: isopropyl groups: (CH3)2CH-R where R is a heteroatom or a carbon bearing
no protons. Usually a septet and a doublet. When R is a metal like Si or Sn the CH3 and CH protons can be
close in chemical shift, and give a complex pattern (AB6) or even a singlet.

i
Pr3Si-Cl
(AB6)3

1.2 ppm

A3MM'XX' Not actually first order. However, a common type, n-propyl groups CH3-CH2-CH2-R, are usually nearly first
oder if chemical shifts are large enough. Usually approximately two triplets and a sextet.

Reich, U.Wisc. Chem. 605 5-HMR-7.5 Pople


Copyright Hans J. Reich 2017
All Rights Reserved
University of Wisconsin
5.8 Symmetry in NMR Spectra
Protons and other nuclei in NMR spectra can be classified as heterotopic, diastereotopic, enantiotopic and
homotopic. Heterotopic and diastereotopic protons will have different chemical shifts and couplings to neighboring
magnetic nuclei, enantiotopic and homotopic protons will have identical chemical shifts. They may or may not have
identical couplings to other nuclei. Distinction can be made by the substitution test.

The Substitution Test for Equivalance of Protons


For a pair of protons to be tested, replace one and then the other with another group (one not present in the
molecule). Compare the two structures formed. If they are identical, the protons are homotopic, if they are
enantiomers, the protons are enantiotopic, if they are diastereomers then the protons are diastereotopic, if they are
structural isomers, the protons are constitutionally heterotopic.

Homotopic Protons:

H H H Ph Ph H The structures A and B are


identical, the two protons
CH3 CH3 CH3 CH3 CH3 CH3
are homotopic.
A B

Enantiotopic Protons:

Enantiotopic protons normally have identical chemical shifts. However, when the molecule is placed in a chiral
environment (say with an optically active solvent, cosolvent or Lewis acid) then the protons can become
diastereotopic. This is in contrast to homotopic protons, which are always identical.

H H H Ph Ph H
The structures C and D are
OH enantiomers, the two protons
CH3 OH CH3 OH CH3
are enantiotopic.
C D
Diastereotopic Protons:

The concept of diastereotopicity was first introduced during the early days of NMR spectroscopy, when certain
kinds of molecules gave unexpectedly complex NMR spectra, leading to some confusion about the orgins of this
hitherto undetected phenomenon (Nair, P. M.; Roberts, J. D. J. Am. Chem. Soc., 1957, 79, 4565). A typical situation
where diastereotopic protons are seen is a CH2 group in a chiral molecule (one with an asymmetric center, or other
types of asymmetry).

H H H Ph Cl H H Ph The structures E and F are


H Cl H Ph
diastereomers, the two
CH3 O CH3 CH3 O CH3 CH3 O CH3 protons are diastereotopic.
E F
A more subtle form of diastereotopism is demonstrated in the classical example of diethyl acetal below. Even
though diethyl acetal has no asymmetric centers, the CH2 group is diastereotopic. This can be shown by applying the
substitution test, which creates a pair of diastereomers G and H. Thus the ethyl group forms an ABX3 pattern (see
Section 5-HMR-13). The key to understanding this type of diastereotopicity is that the molecule has a plane of
symmetry (hence is achiral). However, there is no plane of symmetry that bisects the CH2 protons, so they are
nonequivalent.
H H H H H D H H H H H D H H H

CH3 O O CH3 CH3 * O * O CH3 CH3 O O CH3


G H

Reich, U.Wisc. Chem. 605 5-HMR-8.1 Symmetry


The dibromocyclopropane spectrum illustrates this effect in a different context - the protons of the CH2Cl group
are diastereotopic. However, the protons of the cyclopropane CH2 group are not, since they are related by a plane
of symmetry.

300 MHz 1H NMR spectrum


Source: Aldrich Spectra Viewer
H H
H Br
Cl

Cl H
H Br
H C5H6Br2Cl2

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5

Not all CH2 groups in chiral molecules are diastereotopic - in the chiral molecules below the CH2 is on a C2
axis of symmetry, and the protons are homotopic. In general, CH2 groups (or other similar groups like CHMe2,
CHF2, etc) will be diastereotopic when part of a chiral molecules unless the CH2 group is on a C2 rotation axis.

OH
Ph Ph
H
H
H H
OH
Exercise: Why are three of the aromatic carbon 13C NMR signals in this compound doubled?

75 MHz 13C NMR spectrum in CDCl3 O Click for full Spectrum


Source: Bill Sikorski/Reich

SPh

SPh

134 133 132 131 130 129 128


ppm

150 140 130 120 110 100 90 80 70 60 50 40 30 20


ppm

Constitutionally Heterotopic Protons:

NO2 NO2 NO2


The structures I and J are
structural isomers, the two
H F H protons are heterotopic.
H H F
I J

Reich, U.Wisc. Chem. 605 5-HMR-8.2 Symmetry


Magnetic Equivalence
There is an additional element of symmetry which is important for NMR spectroscopy, the magnetic equivalence
or inequivalence of nuclei. Protons that are enantiotopic or homotopic will have the same chemical shift, but they
will not necessarily be magnetically equivalent. For two protons to be magnetically equivalent they not only have to
have the same chemical shift, but they must also each have the same J coupling to other magnetic nuclei in the
molecule. This is easiest to see from some specific examples.
The two vinyl and two allylic protons in cyclopropene are each magnetically equivalent because each of the A
protons is equally coupled to the two X protons. The spectrum consists of two identical triplets (A2X2 system).

EtO2C HA
HX HA HX

HA'
HX
HX HA EtO2C JAX JAX'
Magnetically equivalent Magnetically inequivalent
A and X protons (A2X2) A and X protons (AA'XX') 2.20 2.15 2.10 1.45 1.40
On the other hand, the two pairs of equivalent protons in trans-bis(carbomethoxy)cyclopropane are NOT
magnetically equivalent, because each of the A protons is coupled differently to the two X protons (one is a trans
coupling, the other a cis). In the Pople nomenclature, such magnetically inequivalent nuclei are given an AA'
designation. Thus the bis(carbomethoxy)cyclopropane is referred to as an AA'XX' system, where A and A' refer to
protons that are symmetry equivalent but not magnetically equivalent. The spectrum will be much more complicated
than two triplets, and both sets of protons will be identical.
Two more examples are 1,1-difluoroallene, which is an A2X2 system, and 1,1-difluoroethylene, which is an AA'XX'
system (see 5-HMR-14.2 for a spectrum).

FA FA HX
HX
JAX JAX'
FA HX FA' HX'
A2X2 AA'XX'

In general any system which contains chemical shift equivalent but magnetically inequivalent nuclei of the AA'
type will not give first order splitting patterns, although sometimes the spectra may appear to be first order
("deceptively simple" spectra). For example, X-CH2-CH2-Y systems are of the AA'XX' type, but the coupling
constants JAX and JAX' are often close enough in size that apparent triplets are seen for each CH2 group. See
Section 5-HMR-14 for examples.
Two important generalizations:
Coupling between symmetry equivalent but magnetically inequivalent nuclei typically will affect the
appearance of the NMR spectrum. In fact, it is the coupling between the equivalent nuclei that is responsible
for the complexity of spectra of the AA'BB'.. type.
Coupling between magnetically equivalent nuclei does not affect NMR spectra, cannot be detected, and thus
can be ignored.
The NMR Time Scale
It is important to recognize that diastereotopic and magnetic equivalence effects are subject to the time scale of
the NMR experiment, which is on the order of tenths of a second (see Sect 8-TECH-3). Flexible molecules will often
have several conformations, some of which may have lower symmetry than others. However, since these
conformations are typically interconverting rapidly on the NMR time scale, the observed symmetry in the NMR
spectrum will be that of the most symmetric conformation reachable. Thus cyclohexane is a sharp singlet at room
temperature, whereas at -100 C the ring inversion is slow on the NMR time scale, and a much more complex
spectrum results (see Sect 5-HMR-5.3).

Reich, U.Wisc. Chem. 605 5-HMR-8.3 Symmetry


5.9 Second Order Effects in Coupled Systems
For first order systems J and values are directly measurable from line positions. However protons or groups
of protons form first order multiplets only if the chemical shift differences between the protons () are large
compared to the coupling constants between them (J), i.e. if /J (all in Hz) is < 5 then second order effects
appear. When /J < 1 then second order effects become very pronounced, often preventing detailed manual
interpretation of multiplets, or giving incorrect coupling constants if first order behavior is assumed.
There are a number of changes that occur in NMR spectra which are the result of degenerate or near-degenerate
energy levels in strongly coupled systems i.e. when /J becomes small (in English: whenever the coupling
constant between two nuclei is similar in magnitude to the chemical shift between them, the spectra get
complicated). If the spectrum is measured at higher spectrometer frequency the chemical shifts (in Hz) become
larger, whereas the coupling constants stay the same, so the spectrum usually gets simpler. Exceptions are the
AA'XX' type of systems, which are field independent, and usually cannot be completely solved from line positions
and intensities alone.

The effects of spectrometer field strength on the


ability to resolve NMR coupling information is
illustrated in this set of spectra of ethylbenzene, plotted
at a constant Hz scale. The aromatic signals go from
nearly a singlet at 60 MHz to a reasonably resolved 60 MHz
set of peaks at 600 MHz (spectra courtesy of Kris
Kolonko). This increase in information content and 440430
greater ease of interpretation of NMR spectra at higher
magentic field strength is the main justification for the
additional expense of more powerful magnets. 90 MHz

660650

X
Ho Ho
300 MHz

Hm Hm
2200 2180 2160 2140
Hp
X

400 MHz

2960 2940 2920 2900 2880 2860

500 MHz

3680 3660 3640 3620 3600 3580 3560


X Ho
Hm
Hp
600 MHz

4420 4400 4380 4360 4340 4320 4300 4280 4260


Hz

Reich, U.Wisc. Chem. 605 5-HMR-9.1 Second Order


We can define a hierarchy of coupling patterns which show increasingly larger number of second-order effects:

AX and all other first order systems (AX2, AMX, A3X2, etc.)
AB (line intensities start to lean, J can be measured, has to be calculated)
AB2 (extra lines, both J and have to be calculated)
ABX, ABX2, ABX3, JAB can be measured, others require a simple calculation
ABC (both J and can only be obtained by computer simulation)
AA'XX' (these do not become first order even at higher fields)
AA'BB'
AA'BB'X (etc)

1. A universally observed effect is that as chemical shifts become comparable to couplings, line intensities are no
longer integral ratios (AB and higher). The lines away from the chemical shift of the other proton (outer lines)
become smaller and lines closer (inner lines) become larger (see the triplets below) - the multiplets "lean" towards
each other (some call this a "roof" effect). The leaning becomes more pronounced as the chemical shift difference
between the coupled multiplets becomes smaller.

C3H7BrO
300 MHz 1H NMR spectrum in CDCl3
Source: Aldrich Spectra Viewer/Reich

Br
"leaning" O
"leaning"

3.8 3.7 3.6 3.5 3.4

8 7 6 5 4 3 2 1 0
ppm

2. Line positions are no longer symmetrically related to chemical shift positions (AB), eventually calculations may
have to be carried out to obtain and J (ABX and higher).

3. Some or all of the coupling constants can no longer be obtained from line separations (AB2 and higher).
4. The signs of coupling constants affect line positions and intensities (ABX and higher).

Reich, U.Wisc. Chem. 605 5-HMR-9.2 Second Order


5. Additional lines over that predicted by simple coupling rules appear. First, lines which formally have intensities
of 2 or more split into the component lines. Eventually combination lines, which no longer can be assigned to any
one nucleus appear (AB2 and higher). A nice example is provided by the compound below. For the BrCH2CH2O
group the two methylenes at 3.48 and 3.81 have a relatively large chemical shift separation, and they form
recognizable triplets, although with a little leaning. For the MeOCH2CH2O group the chemical shift between the CH2
groups is small, and the signals are a complicated multiplet with only a vague resemblance to a triplet. There is
likely an additional complication from variability in the size of the two different vicinal couplings in the two patterns
(see Section 5-HMR-15 for more on this). The additional lines can lead to "Virtual coupling" effects: apparent
coupling to protons that are actually not coupled. See Section 5-HMR-16)

C5H11BrO2
300 MHz 1H NMR spectrum in CDCl3
Source: Aldrich Spectra Viewer/Reich
c
O a d Br
b O
c
d

b a

3.8 3.7 3.6 3.5 3.4

8 7 6 5 4 3 2 1 0
ppm

6. Coupling between equivalent nuclei (e.g., JAA' or JXX') affects line count and positions. Second order effects
will appear even if /J is large when groups of magnetically non-equivalent protons with identical chemical
shifts are coupled to each other (see Section 5.8). Thus Me3Si-CH2-CH2-OH is not just two triplets, since the
CH2CH2 is an AA'XX' system (see Section 5-HMR-14) These patterns do not get simpler at higher field strengths.

C5H14OSi
300 MHz 1H NMR spectrum in CDCl3
Source: Aldrich Spectra Viewer/Reich
Me3Si
OH

3.8 3.7 1.0 0.9

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0


ppm

Reich, U.Wisc. Chem. 605 5-HMR-9.3 Second Order


7. Computer analysis becomes mandatory to extract accurate J and values (ABC and higher). A typical
example is the spectrum below, where a near coincidence of H2 and H3 leads to a complex spectrum. Even here
one can make some sense of the multiplets - the one at 5.55 is H3, essentially a doublet of multiplets (J 10), at
5.72 another dm, with J 17 corresponding to H4. The peaks between 5.8 and 6.0 are H1 and H2, one can guess at
some of the couplings from the downfield half of the H3 ddd, but this becomes quite risky.

C6H7NO2
300 MHz 1H NMR spectrum in CDCl3
Source: Aldrich Spectra Viewer/Reich H2
H1
J12 = 5.4
H3 O Hz
30 20 10 0
J13 = -1.0
J14 = -1.5 H4 CN O
J23 = 10.1
Simulation J24 = 16.9

2 1 4 3

Spectrum

6.0 5.9 5.8 ppm 5.7 5.6 5.5

A computer simulation of this spectrum was performed with WINDNMR, and gave the simulated spectrum shown,
using the J values indicated. To give some idea of what this involves, it took about 60 minutes of manual fiddling
with J and values in WINDNMR to arrive at this simulation (which is not completly optimized). Second order
spectra like this are extremely sensitive to the paramater values - even a 0.1 Hz offset in several of the parameters
significantly reduced the quality of the fit. Note that the chemical shift difference between H1 and H2 is 13.2 Hz, more
that twice the value of the coupling between the two protons. The complexity is the result of the overlap between the
upfield part of the H2 multiplet (which is over 30 Hz wide) and the downfield part of the H1 multiplet, which makes
several of the energy levels degenerate, or nearly so.

8. Ultimately spectra become so complex that the only useful information is integration, chemical shift and
general appearance, as in the spectrum of cyclohexyl chloride, where only the proton to the chlorine gives an
interpretable multiplet (a tt), and even here the coupling constants obtained could be in error because of second
order effects (possible virtual coupling).

C6H11Cl
300 MHz 1H NMR spectrum in CDCl3
Source: Aldrich Spectra Viewer/Reich Cl

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0


ppm

Reich, U.Wisc. Chem. 605 5-HMR-9.4 Second Order


Copyright Hans J. Reich 2017
All Rights Reserved
University of Wisconsin
5.10 AX and AB Spectra
The simplest molecules that show J coupling contain two spin 1/2 nuclei separated by 1, 2, 3 (occasionally 4 and
5) bonds from each other. If the chemical shift between the protons is large compared to the coupling between them
(AX >> JAX), we label them as HA and HX. If the chemical shift is comparable to the coupling between the protons
(AB < 5 JAX),we have an AB system. Some molecules that give AB/AX patterns are shown below (spectra are all at
300 MHz):

Disubstituted alkenes
O
I OEt

H H
7.5 7.4 7.3 7.2 7.1 7.0 6.9
C5H7IO2
ppm

H CN

H Cl
6.3 6.2 6.1

1,2,3,4- and 1,2,3,5-tetrasubstituted benzenes; polysubstituted furans, pyridines, and other aromatic systems

H O
H

MeO OMe
OMe 7.6 7.4 7.2 7.0 6.8 6.6
Benzyl, methoxymethyl and related protecting groups in chiral molecules, and other isolated diastereotopic CH2
groups.

300 MHz 1H NMR Spectrum CDCl3


CH3 H H

N
OH
H
Ph O Ph
C15H17N
O H H
C16H16O3

5.1 5.0 3.7 3.6 3.5

Reich, U.Wisc. Chem. 605 5-HMR-10.1 AB - AX


Exercise: Assign the protons in this partial 1H NMR spectrum.

OH
N
H
Ph
H
C12H17NO 4.0 3.8 ppm 3.6 3.4

Exercise: Assign the protons in this spectrum.

O
O
O
O

O O
C12H18O6
4.8 4.6 4.4 4.2 4.0

Reich, U.Wisc. Chem. 605 5-HMR-10.2 AB - AX


Energy Levels of AX and AB Spectra

The four energy levels for an AX system are given in a very straightforward way by the equation below, by
substituting the four possible spin combinations of mA and mX (++, +-, -+, --):
E = -(mAA + mXX) + mAmXJAX
There are four states: , , , . We will use the convention: is the lowest energy state ( is aligned
with the field, m = +) and is the highest energy state ( is aligned against the field, m = -). The first term in
the equation is the chemical shift part, the second term the coupling part. If the coupling is a small perturbation,
then the energy is simply the sum of the two parts. In energy level terms, this means that the energy separation
of the and states is large compared to J.

State mA mB mT Energy Only transitions with mT


= 1are allowed. The
- - -1 -[(-)A + (-)X] + (-)(-)J = (A + X) + J/4 to (double quantum)
- + 0 -[(-)A + (+)X] + (-)(+)J = (A - X) - J/4 and to (zero
quantum) transitions are
+ - 0 -[(+)A + (-)X] + (+)(-)J = -(A - X) - J/4
forbidden, since they
+ + +1 -[(+)A + (+)X] + (+)(+)J = -(A + X) + J/4 involve simultaneous
changes in both spins.
If both A AX AB A2
and X are
protons J=0 J>0 J>0 J>0
>> J >> J J = 0
+(A + B) + J/4
+(A + B) J/4
E = A - B
A1
MHz
A1 B1 A1 B1 A1 B1 (zero B1
intensity)
2D = 2 + J 2

+(A - B) +Q J
Q =
D - J/4 + 2D
1 + Q2
Q 0 >> J
J/4 Q 1 0
Hz 0 0

-3J/4
-(A - B) +Q
-D - J/4 B2
1 + Q2 (zero
intensity)
MHz A2 B2 A2 B2 A2 B2 A2

-(A + B) + J/4
-(A + B)

Reich, U.Wisc. Chem. 605 5-HMR-10.3 AB - AX


AB Spectra

When the energies of the and states approach each other, they begin to mix, the state develops some
character and vice versa (the mixing parameter Q specifies the degree of mixing). The energy of the state,
instead of (A-B), then becomes [(A-B)2 + J2] (here defined as D)

State mA mB mT Energy
D = [(A-B)2 + J2]
- - -1 +(A + X) + J/4 J
Q =
- + 0 +D - J/4 + 2D
+ - 0 -D - J/4 Q 0 >> J
+ + +1 -(A + X) + J/4 Q 1 0

In addition to these perturbations in energy levels, the probability of the transitions (i.e. line intensities) also
varies - the A1 and B2 transitions become weaker and eventually disappear (i.e. they become forbidden), leaving
only the A2 and B1 lines, which appear exactly at the chemical shifts of A and B when becomes 0.

Intensities:
A1 B1 (1 + Q)2
A2 B2 A2 B1 iA2 = iB1 =
(1 + Q2)
A1 A2 B1 B2 (1 - Q)2
iA1 = iB2 =
A1 B2 (1 + Q2)
JAB JBA
iA2 iB1 (A1 - B2)
A B A B A B A = =
B iA1 iB2 (A2 - B1)
J=0 J0 J0

Four lines are present, as for an AX spectrum,


and J is the same:
AB
|JAB| = (1 - 2) = (3 - 4)
2D
2D + J
The line intensities, i, are no longer 2 2D - J 3
1 : 1 : 1 : 1, but given by the ratios:
i2 i3 (1 - 4) 1 4
= = J
i1 i4 (2 - 3) J
A and B are not exactly halfway C
between line 1 and line 2, or between line
3 and line 4.
AB = A - B
C = center of AB pattern A B
= (2 + 3) = (1 + 4) 2D = (A - B)2 + J 2
AB = (2D)2 - J 2
If A - B were calculated as if the pattern were AX instead of AB,
one would get 2D instead of the correct value.
= (1 - 3)2 - J 2

= (1 - 4) (2 - 3)

Reich, U.Wisc. Chem. 605 5-HMR-10.4 AB - AX


AB Quartet with increasing AB AB Quartet with JAB = 10 Hz and AB = 0.1 ppm
2D/ is the error in
Constant JAB = 10 Hz at various spectrometer frequencies
the measured
chemical shift AB. all printed at same Hz scale all printed at same ppm scale

/J 2D/
AB = 50 Hz
600 MHz
AB = 60 Hz
5 1.02

A B
AB = 40 Hz 500 MHz
AB = 50 Hz
4 1.03

A B
300 MHz
AB = 30 Hz
AB = 30 Hz
3 1.05

A B
AB = 20 Hz 200 MHz
AB = 20 Hz
2 1.12

A B
AB = 10 Hz
100 MHz
1 1.41
AB = 10 Hz

A B

50 40 30 20 10 0 -10 -20 -30 -40 0.1


A 0 B -0.1
Hz ppm
The distinction between an AB q and a regular q is not always trivial. In fact, if an AB quartet has the same Hz
separation between the center two lines as the coupling constant J, then the intensities of the four lines are
1:3:3:1, exactly the same as for a regular q. Of course, an ABq must always integrate to at least 2 protons, and
that may help with a distinction in this peculiar case.
Consider the muliplet below, which at first glance might be identified as a ddq. However, a proper analysis,
which first removes the two smaller couplings, the dd, gives a child multiplet that does NOT have the correct line
positions for a q (separation A is NOT the same as J, as required for a ddq). Rather, the intensities are those of an
AB q, each line of which is split into a dd.

Hz
30 20 10 0 This multiplet is centered at 2.64 and appears in
the spectrum of 3-(methylthio)butyraldehyde. Which
protons are these?

SMe O

H
J
A
JA

Reich, U.Wisc. Chem. 605 5-HMR-10.5 AB - AX


Solving an AB pattern:

2036.2

2026.1

2022.0

2012.0
1. Determine the four line positions in Hz, and measure JAB
|JAB| = (1 - 2) = (3 - 4) = 10 Hz
2 3
2. Calculate the center position (in Hz):

center = (2 + 3) = 2024.1

3. Calculate AB.
AB = (1 - 4) (2 - 3) = 9.94 Hz

4. Calculate A and B (spectrometer frequency: 300 MHz). 1 4


A = center + AB = 2029.1 Hz
B = center - AB = 2019.1 Hz
A B
A = A / MHz = 6.76
B = B / MHz = 6.73 2050 2040 2030
2020 2010 2000
Hz
Graphical method for determining the position of a leaning coupled partner. The point Q is the horizontal
projection of the tip of line 2 on the position of line 1, and point P is the projection of the line 1 on the position of
line 2. The line through P and Q intersects the baseline at the midpoint between the chemical shifts of A and B
(point C) (http://www.ebyte.it/library/docs/kts/KTS isoAB Geometry.html). You can use this method to quickly
estimate where a leaning doublet's coupling partner should be, if other peaks obscure the region of interest, or to
determine whether you are looking at a leaning doublet, or two unrelated peaks.

2 3
Q

1 P 4

AB = A - B

How to report an AB quartet.


Journals require that NMR spectra be reported in text format. There are several ways an AB quartet could be
reported:
1. Treat the pattern as first order (i.e., as two doublets). This is OK for AB quartets with a large AB / JAB ratio,
say > 4, where the error in chemical shifts caused by simply taking the middle of each doublet is small:
3.68 (d, 1H, J = 10.3 Hz), 3.79 (d, 1H, J = 10.3 Hz)

2. For closely spaced AB quartets (AB / JAB < 4) the AB character should be explicitly shown, to indicate that
the pattern was recognized, and the shifts were calculated correctly. One way is to report the chemical shift of the
center of the AB quartet, and AB and JAB.
2.66 (ABq, 2H, AB = 0.05, JAB = 12.2 Hz)
3. A third way is to report the two chemical shifts, and the coupling.
2.63, 2.69 (ABq, 2H, JAB = 12.2 Hz)
Note that the latter two formats not only use less journal space but also contain more information than the "first
order" format (1). There is nothing in the first description that specifies that the two doublets are coupled to each
other, yet that would be obvious from observing the spectrum.

Reich, U.Wisc. Chem. 605 5-HMR-10.6 AB - AX


Shown above is the 60 MHz spectrum of Abel's ketone in CDCl3 solution. There are three sets of protons that one
would expect to form AB quartets. Exercise: Identify them on the structure.

0.82

2.08

O O
1.35
2.20

1.62
0.80

372.0

362.1

247.5

231.5

211.5

195.7
9 8 7 6 5 4 3 2 1 0
ppm
The AB quartet at 3.7 can be analyzed as follows:

JAB = 247.5 - 231.5 = 16.0 i. e. JAB = 15.9 Hz


JAB = 211.5 - 195.7 = 15.8

= [(247.5 - 195.7)(231.5 - 211.5)] = 32.19 i. e. = 32.25 Hz


= [(247.5 - 211.5)2 - (15.9)2] = 32.30
Center = 221.60 Hz or 221.50 Hz so:
A = 237.6 Hz, 3.96
B = 205.4 Hz, 3.42
Intensity Ratios:
Observed: i2/i1 = 2.75; i3/i4 = 2.54 2 3
Calculated: (1 - 4)/(2 - 3) = 2.59 4
1
The doublet at 6.1 is the B part of an AB quartet. Where is A?

JAB =372.0 - 362.1 = 9.9


i3/i4 = 1.2 = [(2 + 9.9) - 362.1] / (2 - 372.0)

Solve for 2: 471.0


2
2+J

372.0
362.1

1 = 2 + JAB = 480.9
Solve the AB quartet: A = 475.9 Hz or 7.93
This is a little off because the intensity ratio is not very accurate, but allows proper assignment. You could also use
the graphical method illustrated on the previous page.
Exercise: Why are the peak heights of the downfield doublet (ca 7.7) lower than those of the upfield one at
6.1? Hint: Section 5-HMR-06

Reich, U.Wisc. Chem. 605 5-HMR-10.7 AB - AX


Exercise: The dibenzyl ester of aspartic acid salt has two different diastereotopic Ph-CH2 groups. Identify the peaks
and calculate AB and AB for each.

1512.4

1500.2

1494.0

1481.9
1480.4

1467.5
1466.4

1453.6
300 MHz 1H NMR spectrum
Source: Aldrich Spectra Viewer

+ TsO
O NH3
O Ph
Ph O
O
C18H19NO4 TsOH

5.0 4.9
ppm

Reich, U.Wisc. Chem. 605 5-HMR-10.8 AB - AX


Copyright Hans J. Reich 2017
All Rights Reserved
5.11 The AX2 and AB2 Patterns University of Wisconsin

Some examples of molecules containing three spin


systems in which two of the nuclei are magnetically
equivalent are shown below. Perhaps the most common
type is 1,2,3-trisubstituted benzenes in which the 1 and

3 substituents are identical. = 40
J
Y
FX Cl Cl X X
56
FA Cl Cl A B
HB
FX HB HA HB HB
= 50 Hz 78
HA 4
J = 10 Hz
23
AX2 and AB2 patterns are not as common as AMX,
ABX and ABC, but there is an important reason for 1 = 5
J
examining them in some detail. The transition of an AX2
to an AB2 spin system provides additional insight into the
appearance of second order effects. In the AB pattern
there are two effects of this type: A B
The line intensities no longer follow simple rules
The arithmetic average of line positions no longer
gives true chemical shifts, although the coupling
constant JAB can still be directly measured from the
= 2
spectrum. J
In AB2 spectra a third and fourth effects appear:
none of the line separations correspond to JAB, and
additional lines appear which are not predicted by
simple multiplet rules. The additional lines arise from
splitting of double-intensity lines, as well as from the A B
appearance of new transitions.

The AB2 spectra illustrate this process. In the top


spectrum we have /J >> 5, and the system is

effectively AX2, it consists of an A triplet and a B = 1
J
doublet. As /J becomes smaller, the double-intensity
middle line of A and both B lines split into two lines. An
additional line which is not a direct descendant of any of
the AX2 lines appears (a combination line: ).
Since it is essentially a forbidden transition it is usually A B
quite weak, but can sometimes be observed. In the
bottom spectrum (/J = 0.7) the intensity of line 9 is 0.4%
of the most intense line (line 5).
When /J < 1 the spectrum takes on the
appearance of a triplet, with a very intense and = 0.7
J
broadened central line. Finally, as /J approached 0 (A
= B) the outer lines disappear completely, and we are line 9
A B
left with a singlet.

A B

Reich, U.Wisc. Chem. 605 5-HMR-11.1 AX2


Solving an AB2 Pattern
To analyze an AB2 pattern, we number the lines as shown, the four A lines 1- 4, the four B lines 5- 8, and the
very weak combination line 9. The arithmetic is simple:

Combination

= 1
J

A B2 A = 3
B = (5 + 7)/2
JAB = (1 - 4 + 6 - 8 ) / 3

1 2 3 45 6 7 8 9
A B

Some points to remember about AB2 patterns:


1. The spectrum depends only on the ratio /J.
2. Note that lines 1-4 must correspond to the one-proton part, lines 5-8 to the two-proton part. Thus, if the
pattern is A2B then the numbering proceeds in the reverse direction. Distinguish the one and two proton parts by
integration.
3. Line 5 is the most intense line. The lines 5 and 6 often do not split up.
4. When /J is much less than 1 the spectra appear nearly symmetrical since 1, 2 and 8 become very weak.
The spectrum then has the appearance of a distorted triplet with a 1:10:1 area ratio (the peak heights will not be in
this ratio since the center line consists of several closely spaced ones.
5. Neither JBB nor the sign of JAB affect the appearance of the spectrum.
40.7
39.1

32.5
31.9

15.6

9.0

6.7

0.0

OH
Cl Cl

= 3.35 6 5 The origin is 400
J
Hz from TMS
60 MHz 4
7
8 2.06 CDCl3
3
2 1
1.00

40 30 20 10 0
Hz
The "1H" part is upfield of the "2H" part here so numbers run from right to left: 1 = 0.0 . . . 8 = 40.7
B = 3 = 9.0 Hz (B = (9 + 400)/60 = 6.82)
A = (5 + 7)/2 = (31.9 + 39.1)/2 = 35.5 Hz (A = (35.5 + 400)/60 = 7.26)
|JAB| = |(1 - 4 + 6 - 8)/3| = |(0 - 15.6 + 32.5 - 40.7)| / 3 = 7.9 Hz

Reich, U.Wisc. Chem. 605 5-HMR-11.2 AX2 - AB2


Copyright Hans J. Reich 2017
All Rights Reserved
5.12 ABX Pattern University of Wisconsin

AMX, ABX and ABC patterns, and various related spin systems are very common in organic molecules. Below
some of the structural types which give ABX patterns.
HX
HX HA HX HB
R' R
R C
' R'' C C HX
R C C HA
HA HB R'' HB
HB R
HA

AMX Patterns. Three nuclei coupled to each other and separated by a large chemical shifts compared to the
coupling between them can be analyzed in first order fashion (Sect. 5-HMR-3): the A, M and X signals are each a
doublet of doublets, and the couplings can be extracted by inspection.

Exercise: How can the assignments for the A and M protons be done in the example below (see Sect. 5.5)?

An AMX Pattern
HX O Hz
HM 30 20 10 0

Cl HA

C8H7ClO

X M A

3.8 3.7 3.6 3.5 3.4 3.3 3.2 3.1 3.0 2.9 2.8 2.7 2.6
ppm

ABX Patterns. When two of the protons of an AMX pattern approach each other to form an ABX pattern, the
characteristic changes in intensities of a strongly coupled system (leaning) are seen, and, as the size of J
approaches the value of AB more complicated changes arise, so that the pattern can no longer be analyzed
correctly by first order methods. A typical ABX spectrum is shown below:

300 MHz 1H NMR spectrum in DMSO-d6


Source: Aldrich Specral Viewer/Reich
Hz
30 20 10 0
Cl
NH3+ Cl-
OMe

X B

4.3 4.2 A
3.3 ppm 3.2 3.1
Figure 5-12.1. Sample ABX pattern

For this spectrum AB is less than twice J, and a first order (AMX-type) interpretation starts to become imprecise,
although, in this particular case, it is unlikely to lead to a substantial misinterpretation. On the other hand, for the
spectrum below (which is actually an ABMX3, where X = 19F), the second order effects are so large than a first order
interpretation may lead to grossly inaccurate couplings, both in magnitude and sign, and a possible misassignment
of the structure.

Reich, U.Wisc. Chem. 605 5-HMR-12.1 ABX


(C6H9F3O3)
300 MHz 1H NMR spectrum in CDCl3
Source: Aldrich Specral Viewer/Reich
Hz
30 20 10 0
OH O
M
CF3 OEt
AB
X3

M A/B

4.5 4.4 4.3 4.2 4.1 4.0 2.7 2.6


ppm ppm

Figure 5-12.2. A borderline ABX pattern (actually an ABM3X pattern, since the X proton is coupled to the three
fluorines. First order analysis of this one is problematic.

Even more likely to mislead is the ABX pattern below, for which any form of first order analysis could lead to
wildly incorrect structure interpretations, or even a "false negative" during synthesis of a molecule (i.e., your reaction
was actually successful, but you conclude that if failed because the NMR spectrum does not appear to fit the
expected structure).

300 MHz 1H NMR Spectrum in CDCl3


Source: Aldrich Spectral Viewer/Reich

OH H
Hz
N 30 20 10 0

C9H13NO

X A/B
4.8 ppm 4.7 4.6 2.7 ppm 2.6

Figure 5-12.3. A deceptive ABX pattern, in which one of the ab sub-quartets has collapsed to a singlet. No
first-order analysis possible.
For these reasons, we will examine ABX patterns in some detail. In the progression from first-order NMR patterns
to incomprehensible jungles of peaks, they represent the last stopping point where a complete analysis (by hand or
hand calculator) is still possible, and where insights into the problems that arise in the analysis of more complex
systems can be achieved. Specifically, ABX patterns are the simplest systems which show the phenomenon
sometimes referred to as "virtual coupling" (see Sect. 5-HMR-16) and they are the simplest systems in which both
the magnitude and the sign of J coupling constants is significant. Furthermore, as illustrated above, there are
several pathological forms of ABX patterns which are sufficiently nonintuitive that the unwary spectroscopist can
mis-assign coupling constants and even structures.

Reich, U.Wisc. Chem. 605 5-HMR-12.2 ABX


Development of an ABX Pattern. Consider the stick diagram below which represents an ABX pattern in which
we sequentially turn on first the A-X and then the B-X coupling:

A B X

Turn on JAX
JAX > 0

X-spin
Turn on JBX
JBX > 0

X-spin
Turn on JBX
JBX < 0

X-spin
One of the two lines in the A-pattern arises from those molecules with the spin of the X-nucleus aligned against
the field () and the other from those which have the X-spin aligned with the field (). Similarly for the B-pattern.
Note, however, that the line assignments of the pattern with both JAX and JBX nonzero will be different depending on
the relative sign of JAX and JBX, as illustrated in the Figure. Up to this point the line positions are identical.
The key to understanding ABX patterns is to realize that the A and B nuclei with spin of X = and those with
spin of X = are actually on different molecules, and cannot interact with each other. Thus, when we finally turn
on JAB, it will be the X = line of A and the X = line of B that will couple to form an AB-quartet. Similarly, the two C
X = lines will form a second AB-quartet. Since the line intensities and line positions of an AB quartet depend on
the "chemical shift" between the nuclei, it is clear that the different relative signs of JAX and JBX will result in different
spectra. The ABX pattern is thus the simplest spin system for which the discerning spectroscopist can identify the
relative signs of coupling constants by analysis of the pattern. The figure below shows the final AB part of the ABX
pattern for the two cases.

Effect of Relative Sign of JAX and JBX on an ABX pattern


Same sign JAB = 13 Hz Different signs
A B A B
JAX = 5 Hz
JBX = 10 Hz JBX = -10 Hz
JAX JBX

B = 90 Hz
A = 110 Hz

1.0 1.0

Reich, U.Wisc. Chem. 605 5-HMR-12.3 ABX


Solving ABX Patterns
Recognizing an ABX Pattern. A typical ABX spectrum consists of an unsymmetrical 8-line pattern integrating to
two protons which has 4 doublets with the same separation JAB (each doublet shows strong "leaning"). This is the
AB part. The X part is a symmetric 6-line pattern, integrating to one proton, with four lines dominant (often looking like
a dd). The 5th and 6th lines (marked with red arrows) are usually small, and not often seen. JAB and X are directly
measureable, the other parameters (JAX, JBX, A, B) must be calculated.

JAB

X
The AB part consists of two superimposed ab quartets (8 lines) which have normal intensities and line
HT
separations, both of which have identical JAB values, but can have very different ab values. We will use "a" and "b"
pa
for the AB-subquartets of the AB part of an ABX pattern. Occasionally one of the ab quartets has ab = 0, and
appears as a singlet. Such systems appear as a five line pattern, with one ab quartet and a singlet (see Fig. 5-12.3
for an example). There are also several other deceptive forms with one or more lines superimposed.
The X part usually consists of an apparent doublet of doublets, although apparent triplets are not uncommon.
There are two other lines which are often too weak to be detected (total of 6 lines). They become large when JAB >
AB.
First Order "AMX" Type Solution. Many ABX patterns are sufficiently close to AMX (i.e., AB>>JAB) that a
first-order solution has an excellent chance of being correct. We identify the distorted doublet of doublets (JAB, JAX)
which make up the A portion, as well as the dd (JAB, JBX) for B, and begin the analysis by first removing the JAX and
JBX couplings, respectively. This leaves us with an AB pattern, which we can solve in the usual way. Since this is a
first-order analysis there is no information about the relative signs of JAB and JBX.
A B
Exact solution to an ABX Pattern
2. Then remove JAX and JBX (solve two d)
JAB = 13 Hz 1. Solve the two AB patterns
JAX = 5 Hz
JBX = 10 Hz
Note that the approximate analysis at the
bottom proceeds in the reverse order
A = 110 Hz
as the exact one at the top.
B = 90 Hz
Approximate "AMX" Solution
JAX JBX 1. Remove JAX and JBX (solve two dd)
2. Then solve AB pattern

A B

Reich, U.Wisc. Chem. 605 5-HMR-12.4 ABX


For ABX patterns which are of the "Solution 1" type (see below) this analysis will lead to J and values that are
quite close to correct. The errors become larger when JAX and JBX are very different in size (especially if they are
different in sign) and, of course, when AB is small compared to JAB. However, such an analysis, carelessly applied,
can be completely wrong if the system is of the "Solution 2" type.
Correct Analysis of ABX Patterns. In order to correctly analyze an ABX pattern of arbitrary complexity we have
to reverse the order of extraction of coupling constants compared to the AMX solution above. We have to first solve
for JAB, and then for JAX and JBX. Proceed in the following order:
1. Identify the two ab quartets. These can usually be recognized by the characteristic line separations and
"leaning." We will use the notation ab+ and ab- for the two quartets (+ identifies the one with the larger ab). Number
the lines of one ab quartet 2,4,6,8 and the other 1,3,5,7 (NOTE: these line numbers will not typically be in sequence
in the spectrum). Check to make sure that Jab+ = Jab-, and that the ab quartet with the taller middle lines has the
shorter outer lines. Note that ABX patterns are not affected by the sign of JAB.
If the ABX pattern verges on AMX (AB/JAB >> 2), then line intensity patterns will not allow unambiguous choice of
ab subquartets. Such systems can normally be analyzed as an AMX pattern, but with the limitation that the relative
sign of JAX and JBX is indeterminate. If you complete the full ABX treatment with the wrong assignment of quartets,
the signs of JAX and JBX will be wrong, and there will be small errors in their magnitude. This could ultimately lead to
a wrong Solution 1/2 assignment (see below) if you use the signs of couplings to make the distinction.
Another situation in which the choice of ab subquartets can be difficult is in systems verging on ABC, where all
of the line intensities are distorted. This is where computer simulations might become necessary.
C

Correct choice of ab quartets Incorrect ab quartets:


Jab+ = Jab-, and intensities are OK Jab is OK, but intensities are wrong (i.e. the
(i.e. the ab quartet with more closely spaced thin ab quartet should have much taller
inner lines has the smaller outer lines. central lines)

Reich, U.Wisc. Chem. 605 5-HMR-12.5 ABX


2. Solve the two ab quartets. Treat the ab subquartets as normal AB patterns, and obtain the four "chemical
shifts," a+, b+ and a-, b-.

121.2
115.7

108.1
102.7
99.3

89.8
86.3

76.7
c- = (5+3)/2 = 103.7
ab- = - = (7-1)(5-3) = 17.5
c- -/2 = 103.7 8.76 = 112.5, 94.9

c+ = (6+4)/2 = 96.25
JAB
JAB ab+ = + = (8-2)(6-4) = 22.43
8 6 4 2 c+ +/2 = 96.25 11.21 = 107.5, 85.0
7 5 3 1

At this stage, we know one of the bold lines is a,


and the other b, but we do not know which is which.
112.5

107.5

Similarly for the thin lines.


95.0

84.9

a- a+ b- b+ Sol. 1
or
b- a+ a- b+ Sol. 2

Reich, U.Wisc. Chem. 605 5-HMR-12.6 ABX


3. Identify the correct solution. At this point in the analysis we encounter an ambiguity. We know that each of the
ab quartets consists of two a and two b lines, but we do not know which half is a and which is b. There are thus two
solutions to all ABX patterns which have two ab quartets. (The only exceptions are those ABX patterns in which one
of the ab quartets has collapsed to a singlet. For these there is only one solution.) The two solutions are obtained
by pairing up one each of a + and a - line (i.e. in the stick spectra shown, pair up one bold and one light line -
solution 1 corresponds to pairing up the nearest neighbors, solution 2 to the remote ones, where we have swapped
the a-/b- assignments of the bold lines). The analysis is completed as below:

Solution 1 Solution 2
121.2
115.7

108.1
102.7
99.3

89.8
86.3

76.7

121.2
115.7

108.1
102.7
99.3

89.8
86.3

76.7
JAB JAB
JAB JAB
8 6 4 2 8 6 4 2
7 5 3 1 7 5 3 1
112.5

107.5

112.5

107.5
95.0

84.9

95.0

84.9
JAX = 112.5-107.5 JAX = 95.0-107.5
= 5.0 Hz = -12.5 Hz
a- a+ b- b+ b- a+ a- b+
JBX = 95.0-84.9 JBX = 112.5-84.9
110.0

90.0

101.3

98.7

= 10.1 Hz = 27.6 Hz
A B or AB
A = (112.5+107.5)/2 = 110.0 A = (107.5+95.0)/2 = 101.3
B = (95.0+84.9)/2 = 90.0 B = (112.5+84.9)/2 = 98.7

The relative sign of JAX and JBX is given by the direction of vectors from bold to thin lines (i.e., whether
the X = lines are upfield or downfield of the X = lines). Note that in this specific case, for Sol. 1 the two
red arrows (each going from a "-" to a "+" line) are in the same direction, meaning that JAX and JBX have
the same sign, whereas for Sol. 2 they are in opposing directions, and thus JAX and JBX have different
signs.

As part of the solution we obtain the relative signs of JAX and JBX. In the example above, this means that for
Solution 1 the couplings are either both positive or both negative, and for Solution 2 one is positive and one
negative. Note that there are also ABX patterns where the signs of JAX and JBX are the same in both solutions, and
where they are different in both solutions.

Reich, U.Wisc. Chem. 605 5-HMR-12.7 ABX


The relative signs of JAX and JBX are determined by the way in which the ab quartets overlap. For the statements
below, "lines" refers to the a and b line positions obtained by solving the ab- and ab+ quartets. Solution 1 is defined
as the one with the larger difference between A and B. Thus Solution 1 always has the least distorted X-part. The
vectors are drawn from a- to a+ and from b- to b+ (bold to thin) in each case.

(1) If the lines of ab- and ab+ (2) If the lines of ab+ do not (3) If the lines of ab- are inside
alternate, then JAX and JBX have the overlap those of ab-, then JAX those of ab+, then JAX and JBX
same signs in one solution, and and JBX have the same sign in have opposite signs in both
opposite signs in the other. This is the Solutions 1 and 2. Solutions 1 and 2.
case for the current example.
a- a+ b- b+ a- b- a+ b+ a+ a- b- b+

Solution 1

A B Swap these A B A B
assignments
for Sol. 2
b- a+ a- b+ b- a- a+ b+ a+ b- a- b+
Solution 2

AB AB A B
Distinguishing Between Solutions 1 and 2. Which solution is the correct one? Several criteria can be used to
make the assignment:
1. Magnitude of the couplings. Sometimes one of the solutions gives unreasonable couplings. In the example
above, if we are dealing with proton-proton couplings, Solution 2 looks dubious because one of the couplings, JBX at
27.6 Hz, is larger than usually observed for JHH. A coupling this large is not impossible for a proton spectrum, but
rather unlikely.
2. Signs of coupling constants. Sometimes the sign of the coupling constants is definitive. If the structure
fragment is known, the signs can sometimes be predicted, and may rule out one solution. For example, all vicinal
couplings (3JHCCH) are positive, geminal couplings (2JHCH) at sp3 carbons are usually negative. A common
structure fragment which gives ABX patterns is CHX-CHAHB. Here both JAX and JBX must have the same sign. On
the other hand, if the pattern is CHA-CHBHX (a much less common situation) then the signs must be different. Note,
however, that if you misidentified the ab subquartets, then the signs of the coupling constants you calculated may be
wrong.
3. Analysis of the X-Part. It is important to note that all lines have identical positions in both Solutions 1 and 2.
The intensities of the AB part are also identical for both solutions. However, the intensities of the lines in the X-part
are always different, and this is the most reliable and general way to identify the correct solution.
For the vast majority of ABX patterns encountered in organic molecules, Solution 1 is correct. Solution 2
spectra are found when A and B are close in chemical shift and the size of JAX and JBX are very different, and
especially when they have different signs.

Checking your solution arithmetic. There are a couple of checks you can run to make sure that there has not
been a calculation error:
1. The difference in the centers of the two ab quartets should be half the average of JAX and JBX:

| c+ - c- | = 1/2[JAX + JBX]
2. The average of the two centers should be equal to the average of the two chemical shifts:
1/2 | c+ + c- | = 1/2 | A + B |

Reich, U.Wisc. Chem. 605 5-HMR-12.8 ABX


Analysis of the X-Part of ABX Patterns. The X part of an ABX pattern is maximally a centrosymmetric 6-line <a
pattern. However, in many cases it closely resembles a doublet of doublets, and it is often treated as such.
However, the couplings obtained are only approximate. The errors become larger when JAX and JBX differ greatly in
size, and especially if they have different signs. The sum of JAX + JBX will be correct, but the individual values will be
incorrect, with the errors increasingl as AB becomes smaller. The values of JAX and JBX will be completely wrong if
we are dealing with a Solution 2 pattern.
The X-part consists of 6 lines, of which only four are usually visible. The two additional lines are often weak, but
can be seen in Solution 2 patterns for which A and B are close together, and the X-part is consequently
significantly distorted. The lines are numbered as follows: the two most intense are 9 and 12, they are separated by
JAX + JBX. The inner pair of the remaining lines are 10 and 11, their separation is 2D+ - 2D-. The outer lines (often
invisible) are 14 and 15, separated by 2D+ + 2D-. Line 13 has intensity of zero. These line assignments are not
always straightforward: sometimes lines 10 and 11 are on top of each other, resulting in a triplet-like pattern,
sometimes 10 and 11 are very close to 9 and 12, leaving just a doublet.

Solution 1 Solution 2
121.2
115.7

108.1
102.7
99.3

89.8
86.3

76.7

2D+ JAX+JBX 2D+ + 2D-


2D- 2D+ - 2D-
-7.50
-2.16
2.16
7.50

JAX+JBX
8 6 4 2
7 5 3 1 14
15
12 9
12 11 10 9
11 10
24.0

-24.0

Definition of 2D+ and 2D- 15 14


AB part
X X

To carry out an intensity calculation we define lines 9 and 12 to have intensity 1 (i9 = i12 = 1.0), and proceed as
outlined below:
Solution 1 Solution 2
1+ = 0.5arcsin(JAB/2D+) 2+ = 1+
= 0.5arcsin(13.0/26.0) = 15.0 = 15.0
1- = 0.5arcsin(JAB/2D-) 2- = 90-1-
= 0.5arcsin(13.0/21.9) = 18.2 = 90-18.2 = 71.8
2
i10 = i11 = cos (1+ - 1-) i10 = i11 = cos2(2+ - 2-)
= cos2(15.0 - 18.2) = 0.997 = cos2(15.0 - 71.8) = 0.30
i14 = i15 = sin2(1+ - 1-) i14 = i15 = sin2(2+ - 2-)
= sin2(15.0 - 18.2) = 0.003 = sin2(15.0 - 71.8) = 0.70
arcsin = sin-1
Below is another complete worked example of an ABX pattern solution. The "eyeball method" is the one
described in the previous pages, the "formula method" is the one commonly presented in NMR books. We
recommend the "eyeball" method" because it follows the actual coupling tree in a systematic manner, whereas the
"formula method" extracts the information in a mathematically correct but non-intuitive fashion. Both will give
identical answers.

Reich, U.Wisc. Chem. 605 5-HMR-12.9 ABX


Solving an ABX pattern - Summary
1. Pick two quartets in the AB part. The quartet with the largest effective chemical shift is the + quartet, the other the
- quartet, i.e. + > -; D+ > D-; ab+: [2], [4], [6], [8]; ab-: [1], [3], [5], [7].
2. Solve the AB quartets: JAB = [8] - [6] = [4] - [2] = [7] - [5] = [3] - [1] = 11.2 Hz
ab+ = + = ([8]-[2])([6]-[4]) = 12.2 Hz; ab- = + = [7]-[1])([5]-[3]) = 10.4 Hz
c+ = ([6]+[4])/2 = 221.5 c- = ([5]+[3])/2 = 215.2
2D+ = [8]-[4] = 16.6 2D- = [5]-[1] = 15.2
Choose one method of doing the calculations. The "eyeball" method on the left, which was described on the
previous pages, or the "formula" method on the right, in which no attempt is made to follow the development of the
pattern but the equations governing the line positions are solved directly.
3f. Calculate

221.5

215.2
3e. Calculate the line positions (solve ab+, ab-)
Solve ab+ + + - = 12.2 + 10.4 = 22.6

213.1
217.2
c+ +/2 = 221.5 12.2/2
224.2

218.8
+ - - = 12.2 - 10.4 = 1.8
a+, b+ = 227.5, 215.4
c+ + c- = 436.7
Solve ab- c+ - c- = 6.3
2D+ 2D-
c- -/2 = 215.2 10.4/2 4f. Solution 1
235.3

228.5

202.0
207.6
a-, b- = 220.4, 210.0 A + B = c+ + c- = 436.7
[6] [4] [5] [3]
We don't know at this [8] [7] [2] [1] A - B = 1/2(+ + -) = 11.3
point which line is a and |JAX + JBX| = 2(c+ - c-) = 12.6
which line is b
|JAX - JBX| = + - - = 1.8
4e. The two solutions for the AB part
Add and subtract each pair of equations:
are obtained by pairing up one
each of a + and a - line, i.e. each 436.7 + 11.3 436.7 - 11.3
210.0
227.6

220.4

A =
215.4

B =
ab+, ab- is half A and half B, but 2 2
don't know which. Thus: = 224.0 Hz = 212.7 Hz
Solution 1
12.6 + 1.8
227.6 + 220.4 215.4 + 210.0 JAX = JBX = 12.6-1.8
A = B = 2 2
2 2
= 7.2 Hz = 5.4 Hz
= 224.0 Hz = 212.7 Hz Solution 2
JAX = 7.2 Hz JBX = 5.4 Hz A + B = c+ + c- = 436.7
A - B = 1/2(+ - -) = 0.9
224.0

212.7

|JAX + JBX| = 2(c+ - c-) = 12.6


A B |JAX - JBX| = + + - = 22.6
Solution 2
Add and subtract each pair of equations:
227.6 + 210.0 220.4 + 215.4
A = B = 436.7 + 0.9 436.7 - 0.9
2 2 A = B =
2 2
= 218.8 Hz = 217.9 Hz
= 218.8 Hz = 217.9 Hz
JAX = 17.6 Hz JBX = -5.0 Hz
12.6 + 22.6
JAX = JBX = 12.6-22.6
2 2
217.9

= 17.6 Hz = -5.0 Hz
218.8

B
A Same for both solutions

Reich, U.Wisc. Chem. 605 5-HMR-12.10 ABX


5. Analyzing the X Part

Solution 1 and Solution 2 are defined such that Solution 1 has the larger A - B value (i.e. the larger chemical
shift difference between the A and B nuclei). Hence Solution 1 always corresponds to the one with the least
distorted X part. To properly identify the correct solution in ambiguous cases it is necessary to do an intensity
calculation. The six X lines are numbered as follows: the two most intense are 9 and 12, they are separated by JAX
+ JBX. The inner pair of lines are 10 and 11, their separation is 2D+ - 2D-. The outer pair of lines (often invisible) are
14 and 15, separated by 2D+ + 2D-. Line 13 has intensity 0.

Define the intensity of lines 9 and 12 = 1.0, and calculate the relative intensity of lines 10 and 14. For this
example Solution 1 has a fairly normal appearance close to a dd, whereas Solution 2 has all 6 lines clearly visible.

Solution 1 Solution 2

1 JAB 2+ = 1+ = 21.2
1+ = arcsin ( )
2 2D+
2- = 90-1- = 90 - 23.7 = 66.3
1 11.2
= arcsin ( ) = 21.2
2 16.6 i10 = i11 = cos2 (2+ - 2-)

= cos2 (21.2 - 66.3) = 0.498


1 JAB
1- = arcsin ( )
2 2D- i14 = i15 = sin2 (2+ - 2-)

1 11.2 = sin2 (21.2 - 66.3) = 0.502


= arcsin ( ) = 23.7
2 15.2

i10 = i11 = cos2 (1+ - 1-)

= cos2 (21.2 - 23.7) = 0.995

i14 = i15 = sin2 (1+ - 1-)

= sin2 (21.2 - 23.7) = 0.0019


-0.7

-6.3
6.3

0.7

JAX + JBX
-15.9
15.9

[12] [11] [10] [9] 2D+ - 2D-


2D+ + 2D-
[15] [14]

Reich, U.Wisc. Chem. 605 5-HMR-12.11 ABX


A Simple ABX Pattern as AB is Changed

JBX JAB = -11.20 JAX


AB = 52.00 Hz
JAX = 7.20
JAB JAB
JBX = 5.40
Vary AB

B A

AB = 32.00

AB = 22.00

AB = 12.00

AB = 5.00

AB = 3.00

40 30 20 10 0 -10 -20 -30 -40


Hz

Reich, U.Wisc. Chem. 605 5-HMR-12.12 ABX


Effect of Relative Sign of JAX and JBX on ABX pattern

JAB = 13 Hz
JAX = 10 Hz
JBX = 5 Hz JBX = -5 Hz

A = 70 Hz
B = 130 Hz
AB = 60 Hz

1.0 1.0

A = 85 Hz
B = 115 Hz
AB = 30 Hz

1.0 1.0

A = 90 Hz
B = 110 Hz
AB = 20 Hz

1.0 1.0

Here one of the ab


quartets has collapsed
to a singlet
A = 95 Hz
B = 105 Hz
AB = 10 Hz

1.0 1.0

Reich, U.Wisc. Chem. 605 5-HMR-12.13 ABX


ABX with Accidental Coincidences

AB = 18
JAB = 11.20
JAX = 7.20 Normal
JBX = 5.40 AB
X

X A B

120 100 80 Hz 60 40 20 0

AB = 13.2
JAB = 11.20
JAX = 7.20 Middle two AB lines accidentally
JBX = 5.40 superimposed

X A B

120 100 80 Hz 60 40 20 0

AB = 12
JAB = 11.20
JAX = 7.20
JBX = 5.40 Here the middle two lines have
moved past each other

X A B

120 100 80 Hz 60 40 20 0

AB = 3
JAB = 11.20
JAX = 7.20
JBX = 5.40 Both ab quartets are collapsing
to singlets, one a little ahead of
the other

X A B

120 100 80 Hz 60 40 20 0

Reich, U.Wisc. Chem. 605 5-HMR-12.14 ABX


ABX with Accidental Coincidences

AB = 3
JAB = 10.2
JAX = 5.0 One of the ab quartets has
JBX = -2.0 collapsed to a singlet
X AB
Solution 1 = Solution 2
AB = 0.5 (JAX - JBX)
(ab quartet + singlet)

X A B

120 100 80 60 40 20 0
Hz
AB = 4
Solution 1 = Solution 2
JAB = 2.0
JAX = 17.0 AB = 0.5 (JAX - JBX)
JBX = 10.0 (ab quartet + singlet)

X A B

120 100 80 60 40 20 0
Hz

AB = 14
Here is an interesting "pathological" ABX
JAB = 8.0
pattern. If JAB = JAX = JBX and AB is just
JAX = 8.0 X
right, the AB part completely mimics a
JBX = 8.0
pentet
6.08
JAB = JAX = JBX 4.03
4.11
(AB part is fake pentet)
1.00 1.01

X A B
120 100 80 60 40 20 0
Hz

AB = 0
JAB = 12.0 AB = 0
JAX = 2.0 X (X part is more or less a triplet, even
JBX = 12.0 though JAX and JBX are very different)

X B
A

120 100 80 60 40 20 0
Hz

Reich, U.Wisc. Chem. 605 5-HMR-12.15 ABX


ABX of the Vinyl Type
B A
JAB = 2.50 HX HB
JAX = 16.80 Mimic:
JBX = 10.00 t
R HA
AB = 25 Hz
300 MHz NMR

AB = 10 Hz

The chemical shifts of A and B are often close


together if R is a tert-alkyl group. Note the
increased size of the extra lines in the X part

AB = 5 Hz

(AB part is an ab quartet + singlet)


Solution 1 = Solution 2

AB = 3 Hz

A
Here we have a Solution 2 situation

AB = 1 Hz

5.9 5.8 5.7 5.6 5.5 5.4 5.3 5.2 5.1 5.0 4.9 4.8
ppm

Reich, U.Wisc. Chem. 605 5-HMR-12.16 ABX


ABX Going to ABC

C = 270.00 HC
JAB = -13 A = 60
JAC = 9 B = 70
JBC = 2 C varies
HB HA

C B A
C = 210.00

C B A

C = 170.00

C B A
C = 120.00

C B A
C = 100.00

C B A
C = 90.00

C B A
How badly off are we, if we treat this as an AMX
C = 80.00 pattern. taking the peaks at face value?
JAB = 12.6, 11.3 (actual: 13)
JAC = 8.4, 7.1, 8.7 (actual: 9)
JBC = 1.8, 3.2, 3.2 (actual 2)

C B A

2.6 2.4 2.2 2.0 1.8


ppm

Reich, U.Wisc. Chem. 605 5-HMR-12.17 ABX


ABX Going to A2X

A = 10 HA
JAB = 8
JAC = 0.8 B varies
JBC = 2.3 C = 180 HC HB
AB = 15

C B A

AB = 8

C B A

AB = 6

C B A

AB = 4

C B A

AB = 2

C BA

Note the "fake" triplet for C AB = 1

C BA

180 160 140 120 100 80 60 40 20 0


Hz

Reich, U.Wisc. Chem. 605 5-HMR-12.18 ABX


Sample ABX Spectra

300 MHz 1H NMR Spectrum CDCl3


Source: Aldrich Spectra/Reich
OH
Ph O Ph
10.39
H H O
C16H16O3
4.454.404.35

Hz
30 20 10 0

3.2 3.1 3.0 2.9 2.8


3.12

1.97

1.00

7.3 7.2 7.1 5.155.105.05

10 9 8 7 6 ppm 5 4 3 2 1 0

300 MHz 1H NMR spectrum in CDCl3


Source: Aldrich Specral Viewer/Reich Hz
30 20 10 0

OH O

CF3 OEt
C6H9F3O3

2.7 2.6
ppm

4.5 4.4 4.3 4.2 4.1 4.0


ppm

10 9 8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-12.19 ABX


Sample ABX Spectra

300 MHz 1H NMR spectrum in CDCl3


Source: Aldrich Specral Viewer/Reich

30 20 10
Hz
0

O OH

C4H8O2 3.0 2.9 2.8 2.7 2.6

3.00

2.14 2.11

1.00

3.8 3.7 3.6 3.5

9 8 7 6 5 ppm 4 3 2 1 0

300 MHz 1H NMR spectrum in methanol-d4


Source: Aldrich Specral Viewer/Reich

O
Hz
30 20 10 0
O (CH2)14CH3
+
Me3N

Cl
O OH
C23H48ClNO2
5.7 5.6

2.8 2.7 2.4

3.9 3.8 3.7

10 9 8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-12.20 ABX


ABXmYnZo Patterns

In real molecules AB patterns that are coupled to more than just one X proton appear frequently. There may be
several X protons (ABX2, ABX3) or there may be two or more different protons coupled to the AB part (ABXY,
ABXYZ). Some part structures illustrate these common types:
H H HB HA HB HA HB HA HB HA
H HB A B A
HX

HY HX HY HX HZ HY HX HX HX
HX
AB ABX ABXY ABXY ABXYZ ABX3

Most such spectra can be at least approximately analyzed in a straightforward fashion using an AMX-type
approach, by treating the two A lines of the parent AB quartet as each being split into a dd, ddd, dddd, dt, etc by the
X, Y and Z protons. Similarly for the two lines of the B part. Fortunately, except in very unusual cases, there is no
Solution 1/Solution 2 ambiguity, since the A lines and B lines can be distinguished because each shares common
couplings.

1. Here is an AB pattern, and a simple ABX 4. Here an aditional coupling has been added
pattern derived from it to HB, note that each of the original B lines is
JAB = -12.00 now a ddd (ABXYZ)
AB = 41 JBX = 6.0
JBY = 2.0
JAX = 14.0
JBZ = 1.0
JAY = 4.0
A2 1
B
JAX = 14.0
JBX = 6.0

2
A 1
B 2
A 1B

2. Here one additional coupling has been added 5. As we move A and B closer together,
to both A and B. Note how the original A and B the leaning increases, and some small
lines are now each dd (ABXY) second order effect are seen

JAX = 14.0 JBX = 6.0


JAY = 4.0 JBY = 2.0 AB = 30

2 1B
2
A 1 A
B

3. Here JAX JAB, giving a triplet of doublets 6. When A approaches JAB the spectrum becomes
on the downfield side complicated, no meaningful analysis possible by hand

JAX = 12.0 JBX = 6.0


JBY = 2.0
AB = 11
JAB = 12.0

2A 1
B
2A 1
B

Reich, U.Wisc. Chem. 605 5-HMR-12.21 ABX


Examples of ABXmYnZo Patterns

Identify the protons that are shown


300 MHz 300 MHz
AcO

AcO
O

AcO SH
AcO
C- Fry

5.2 5.1 4.3 4.2 4.1

Ph2P SiMe3 O
O
OMe
O. Daugulis/Vedejs
H CO2CH3
C19H25OP
H

C18H32O4Si
M. Jones/Burke

2.3 ppm 2.2 2.1 2.70 ppm 2.60 2.50

270 MHz 270 MHz


H O

H
O
M. Jones/Burke

5.6 5.5 ppm 5.4 2.00 1.90 1.80 1.70


ppm

250 MHz
OH

N
H

2.4 2.3 2.2 2.1 2.0 1.9


ppm

Reich, U.Wisc. Chem. 605 5-HMR-12.22 ABX


Copyright Hans J. Reich 2017
5.13 ABX3 Spectra All Rights Reserved
University of Wisconsin
ABX3 patterns are very common in organic molecules. 1-Substituted-1-alkenes show this pattern. As can be seen
in the example of trans-1-propene below, the AB part of the pattern is essentially an AB quartet, each line of which
is split into a quartet by the 2J and 3J coupling to the methyl groups. As is common for most 4-spin systems, when
the chemical shift between the two of the protons becomes small (as in cis-1-bromo-1-propene) the spectrum
becomes very complicated, and can no longer be analyzed so simply.

HB Hz HB
30 20 10 0
Br
HA
CH3 CH3
X3 HA X3
HB Br
HA

X3
JAB
JAB

JAX HA/HB X3

JBX

6.2 6.1 6.0 1.7 6.2 6.1 1.8 1.7


ppm

Another sample ABX3 spectrum of methyl crotonate is shown below (this one is really AMX3).

Problem R-27L C5H8O2


250 MHz 1H NMR spectrum in CDCl3
Source: Adam Fiedler/Reich

O
A
X3
O
B Hz
30 20 10 0

3.17
3.04

7.05 7.00 6.95 6.90

1.00 1.01

5.90 5.85 5.80 3.75 3.70 1.90 1.85

8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.1 ABX3


Most ethyl groups in chiral molecule will have diastereotopic CH2 protons, and thus form an ABX3 system. The
partial 1H NMR spectrum of 2-ethoxycyclohexanone below illustrates a typical pattern. The A and B signals are well
separated, and can be readily understood and solved as a first order "AMX3" pattern.

A B
1
H NMR spectrum in CDCl3
Source: Aldrich NMR Library

O X3
O CH3

HB HA
C8H14O2

4.0 3.9 3.8 3.7 3.6 3.5 3.4 3.3


ppm

Just as for ABX systems, there is an exact solution, in which one first solves the four AB quartets, which are
present in a 1:3:3:1 ratio (i.e., they represent the subspectra resulting from the four combinations of X spins: ;
//; //; ). The solutions to these AB quartets give a 1:3:3:1 quartet for the A proton, and
another for the B. These can then be solved as first order patterns.

Exact Solution of ABX3 Approximate Solution: AMX3


A B A B
Creating

Turn on Turn on JAB

Creating
JAX, JBX
Take out JAB
//
//

//
//

X-spins Turn on
JAX, JBX
Take out
JAX, JBX
Turn on JAB
Solving
Solving

Fortunately, it is rarely necessary to do an exact solution. If JAX = JBX (or very nearly so), which is usually the
case, then a first order treatment of the pattern as an "AMX3" type is quite accurate. What is done here is to treat the
pattern as an AB quartet of 1:3:3:1 quartets. In other words, we view the pattern as an AB quartet, each line of
which is split by the X3 protons into a 1:3:3:1 quartet. The four 1:3:3:1 quartets will have the normal intensity ratios
of an AB quartet. To solve, identify the AB-quartet of q and then remove the X coupling. What remains is an AB
quartet which can be solved in the usual way. Note that this corresponds exactly to the "AMX" solution for ABX
patterns (see 5-HMR-12.3), in which we treat the pattern as an AB quartet, each half of which is split into a doublet
by the X nucleus.

Reich, U.Wisc. Chem. 605 5-HMR-13.1 ABX3


The simulated spectra shown mimic ABX3 patterns (AB part)
of OCH2CH3 groups in chiral molecules. In these spectra all 16
of the lines are resolved, and recognition of the pattern is 200 MHz
relatively easy. In real molecules it is common for several of the JAB = -11 Hz
lines to be superimposed (especially since JAB is often nearly JAX = JBX = 7 Hz
twice as large as JAX), making recognition of this pattern more
difficult. In situations where the distereotopic shift is small, the
pattern can be mistaken for a quartets of doublets (see the AB
= 6 Hz spectrum).
AB = 20 Hz
It is not necessary for a molecule to have a center of chirality
to show diastereoptopic CH2 groups. Molecules with two ethyl
groups attached to a prochiral center can also have ABX3
patterns, as illustrated in the spectra of the diethoxysilanes
below. The left structure has enantiotopic CH2 protons, the
right has diastereotopic ones (see Section 5-HMR-8 for the
substitution test).
AB = 12 Hz

Ph Ph

EtO Si OEt EtO Si OEt

Ph Me

AB = 6 Hz
3.90 3.85 3.80 3.85 3.80 3.75
ppm ppm

The spectrum of diethyl sulfite below is of historical interest -


this type of diastereotopicity was first recognized for this
molecule (Finegold, H. Proc. Chem. Soc., 1960, 283), with the
correct explanation and analysis described in a classic paper
AB = 0 Hz
(Kaplan, F.; Roberts, J. D. J. Am. Chem. Soc. 1961, 83, 4666),
which also reported the first recognition that 2J and 3J at sp3
carbons have different signs. Diethyl acetals of aldehydes or
diethyl ketals of unsymmetrical ketones also form ABX3
patterns.
4.0 3.9 3.8
ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.2 ABX3


The spectra of dichloroacetaldehyde diethyl acetal illustrate that a molecule which contains diastereotopic protons
can give decidedly simpler NMR spectra at low field than at high field. The CH2 group appears as a simple quartet
at 90 MHz: the outer lines of the ABX3 pattern are too small to see, and the splitting of the inner lines is too small to
resolve. At 300 MHz, on the other hand, the CH2 group is much more complex since HA and HB of the ethoxy are
now far enough apart to give a well-developed ABX3 pattern,

R-18AG - 90 MHz

Cl O

Cl O

C6H12Cl2O2

5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5


ppm

300 MHz
Aldrich Spectra Viewer

Hz
30 20 10 0

5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0
ppm

Exercize: Identify the peaks in the CH2 pattern at 300 MHz.

Reich, U.Wisc. Chem. 605 5-HMR-13.3 ABX3


Exercize: Analyze the spectrum below, determine and J values

300 MHz 1H NMR spectrum inCDCl3


Source: Olafs Daugulis/Vedejs (od2177001)

Et3N
Hz
OH 30 20 10 0

C10H23NO

2.70 2.65 2.60 2.55 2.50 2.45 2.40 2.35 2.30 2.25 2.20

9.40

3.30 3.25 3.20 1.10 1.05 1.00 0.95 0.90 0.85


6.31

5.74

0.88 1.00

5 4 3 ppm 2 1 0
812.3

805.0

799.2
797.7

792.1
790.5

784.9

777.6

757.4

750.6

744.5
743.5

737.4
736.5

730.6

725.3
723.4
721.3

712.9

709.0

705.0

694.5
692.6

682.1

2.70 2.65 2.60 2.55 2.50 2.45 2.40 2.35 2.30 2.25
ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.4 ABX3


Sample ABX3 Spectra

300 MHz 1H NMR Spectrum in CDCl3


Aldrich Spectra Viewer/Reich
Hz
O 30 20 10 0

S
O O
C4H10O3S

4.1 4.0 1.35 1.30

9 8 7 6 5 4 3 2 1 0
ppm

300 MHz 1H NMR Spectrum in CDCl3


Aldrich Spectra Viewer/Reich
This CH2 (at 4.2)
O O is diastereotopic Hz
30 20 10 0

O O
Ph C15H20O4

This CH2 (at 2.35)


1.2 1.1 1.0 0.9
is not diastereotopic

2.4 2.3
4.3 4.2 4.1

10 9 8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.5 ABX3


More ABX3 patterns:

300 MHz 1H NMR Spectrum


Aldrich Spectra Viewer
Hz
80 60 40 20 0

O
OH
HO
O

C7H12O4

2.6 2.4 2.2 2.0 1.8 1.6 1.4 0.8

12 11 10 9 8 7 6 5 4 3 2 1 0
ppm

300 MHz 1H NMR Spectrum


Aldrich Spectra Viewer

HO CO2H

Hz
C6H12O3 30 20 10 0

0.8
1.8 1.7 1.6 1.5 1.4

12 11 10 9 8 7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.6 ABX3


The spectra below provide some details of a typical fully developed ABX3 system. If the CH3 group is decoupled,
then we are left with a simple AB quartet, exactly analogous to AMX3 solution described above.

Sample ABX3 Pattern


200 MHz 1H NMR
Source: Nelsen

decouple 1.3 N decouple 5.1


OEt

OEt
840.9

833.8
831.5

826.9
824.7

819.7
817.6

814.3

810.4
807.5
804.9

800.5
798.0

793.3
790.9

783.7
3.00 2.95 2.90 2.85

4.20 4.15 4.10 4.05 4.00 3.95 3.90

5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.7 ABX3


ABMX3 Patterns
Spin system where the CH2 protons of a diastereotopic ethyl group are coupled to another proton (ABMX3) can
also be readily understood and analyzed by a simple extension of the ABX3 analysis. The CH2 of the ethyl group is
an AB quartet, each line of which is split into quartets from coupling to the CH3 protons. The quartets are then split
again by coupling to M, giving a ddq for each of the CH2 protons. Some structure fragments which give such
patterns are shown below. For 2 the M nucleus is the spin 1/2 phosphorus. In each case R1 and R2 must be different
to make the CH2 group diastereotopic.
HA HB
R1 HA HB O
M
CH3 R2 P
CH3 O R1
X R2
HM X
1 2
The spectrum of 3 below is of this type. One of the two dq of each proton is shown schematically above the
simulation, which is plotted with a narrow line width so all of the lines can be resolved. Note that in this case the
coupling of A and B to the X3 protons are identical, but A and B are coupled differently to the M proton. The M proton
in addition also coupled to two others labeled P and Q.

300 MHz 1H NMR Spectrum in CDCl3


Aldrich Spectra Viewer
AB = 86.0 Hz
JAX = JBX = 7.3
I O
PQ
JAB = -13.7 Simulation
M JAM = 9.9
OH
JBM = 4.6
AB W 1/2 = 0.30
I I X3

OH 3

Spectrum

A B

1.9 1.8 1.7 1.6 1.5

X3
P Q M
1.000.95
3.5 3.4 3.3 3.2 3.1 3.0 2.9 2.8 2.7

PQ M A B X

8 7 6 5 ppm 4 3 2 1 0

Reich, U.Wisc. Chem. 605 5-HMR-13.8 ABX3


It is fairly common for ABMX3 patterns of the CH-CHAHB-CH3 type (1) to show nearly equal JAX, JBX, JAM and JBM.
In this case the CHAHB group appears as an AB quartet of pentets. The partial NMR spectrum below shows the
CH2 signal of isopropyl 2-methylbutyrate (4). The downfield signal is a clean doublet of pentets, the upfield one
closer to a ddq.

300 MHz 1H NMR Spectrum in CDCl3 (C8H16O2)


Aldrich Spectra Viewer

HA HB O
4 1.7 1.6 1.5 1.4

The phosphonate 5 has the diastereotopic OCH2 protons coupled equally to the CH3 group and the phosphorus,
so here also a dp is seen for each proton (AB quartet of pentets).

300 MHz 1H NMR Spectrum in CDCl3 (C19H23O3PSi)


Olafs Daugulis/Vedejs JOC 1998, 63, 2338

SiMe3

O-CH2-CH3
P O-CH2-CH3
O

5 3.95 3.90 3.85 3.80 3.75 3.70 3.65


ppm

Exercise: Assign the protons and completely solve the pattern below,

300 MHz 1H NMR Spectrum in CDCl3


Olafs Daugulis/Vedejs (OD.070) JOC 1998, 63, 2338

Si
Hz
P 30 20 10 0

O OCH2CH3

3.85 3.80 3.75 3.70 3.65 3.60 3.55 3.50


ppm

Reich, U.Wisc. Chem. 605 5-HMR-13.9 ABX3


C
Copyright Hans J. Reich 2017
F
All Rights Reserved
5.14 A2X2 and AA'XX' Spectra University of Wisconsin

In A2X2 and A2B2 patterns the two A nuclei and the two X (B) nuclei are magnetically equivalent: they have the
same chemical shift by symmetry, and each A proton is coupled equally to the two X (or B) protons. True A2X2
patterns are quite rare. Both the A and X protons are identical triplets. More complicated patterns are seen when
the chemical shift difference approaches or is smaller than the JAB coupling. However, both A2B2 and AA'BB' always
give centrosymmetric patters (A2 part mirror image of the B2 part).

Some molecules with A2B2 / A2X2 patterns:


HB HB
HA HA
HB HA FA H HA
FB FB
S
FB FB F
HB HA FA F H
HA
Change from A2X2 to A2B2

AB = 200 Hz
A2X2
JAB = 10 Hz

B A

AB = 70 Hz

B A
AB = 25 Hz
A2B2

B A

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

AA'XX' and AA'BB' spectra are much more common. Here each A proton is coupled differently to the B and B'
protons (or X, X' nuclei). Some molecules with such patterns are:

H F Cl Cl

H F O Cl O2N

H H
Cl H HO2C
H Cl H
Br CH2 CH2 Cl
Cl MeO O OMe
H
H Br H
H H HO2C

Such molecules give inherently second-order multiplets. Only if the JAB coupling is identical to the JAB' coupling by
accident does the system become A2B2 or A2X2, and a first order pattern is seen (if AB is large enough).

Reich, U.Wisc. Chem. 605 5-HMR-14.1 AA'XX'


AA'XX' Spectra

AA'XX' spectra consist of two identical half spectra, one for JAA' = -10
AA' and one for XX', each a maximum of 10 lines, each JXX' = -11
symmetrical about its midpoint, A and X, respectively. See JAX = 12
example B below. The appearance of the spectrum is defined by JAX' = 2
four coupling constants: JAA', JXX', JAX and JAX'. The spectrum is
sensitive to the relative signs of JAX and JAX', but not to the
relative signs of JAA' and JXX'. The relationship between these,
and the directly measurable values K, L, M, and N are given 1.2 1.1 1.0 0.9 0.8
below and in the graphic.
|JAX + JAX'|
|K| = |JAA' + JXX'| "J" of one ab quartet |N|
|L| = |JAX - JAX'| "" of both ab quartets
|M| = |JAA' - JXX'| "J" of other ab quartet
|N| = |JAX + JAX'| "doublet"

Each half-spectrum consists of a 1:1 doublet with a separation of A


N (intensity 50% of the half spectrum), and two ab quartets, each
with "normal" intensity ratios and ab = |L|. One has apparent 2P
couplings (Jab) of |K| and the other of |M|, as indicated. |JAA' + JXX'|
|K| K2+L2
Unfortunately, K and M cannot be distinguished, the relative signs of
JAA' and JXX' are not known, nor is it known which number obtained
is JAA' and which is JXX'. It is also not known which coupling is JAX |L|
and which is JAX', but the relative signs of JAX and JAX' can be
determined: if |N| is larger than |L|, signs are the same. Thus the 19F |JAA' - JXX'| 2R
and 1H spectra of 1,1-difluoroethylene (B) are identical, so it is not |M|
2R = M2+L2
possible to distinguish which coupling is 2JFF and which is 2JHH, nor
can one tell which is the cis JHF and which is trans JHF. This would |L|
have to be done using information about such couplings obtained
from compounds where the assignments are not ambiguous. |JAX-JAX'| = |L| = (2P)2 - K2
= (2R)2 - M2

H SeC(CH3)3
H F
OMs
A Ph B C MsO
Si(CH3)3 H F

JCP 1963-38-226
JACS-1972-(94)-34

200 MHz

60 MHz

140 120 100 4.5 4.0 3.5 3.0 30 20 10 0


ppm

Reich, U.Wisc. Chem. 605 5-HMR-14.2 AA'XX'


Solving an AA'XX' Pattern. If all 10 lines are visible, and can be assigned to the large doublet and the two ab
quartets, the process is straighforward, as shown for the solution of the 19F NMR spectrum of 1,1-difluoroethylene
below:
1. Determine N from the doublet separation (35.3 Hz).
2. Measure K (41.2 and 41.4 Hz) and M (31.7, 32.0 Hz) from the appropriate line separation ("J" of the two ab
quartets).
3. Calculate L - it is the "ab" of each of the ab quartets. For the K quartet we get:
SQRT[(276.2-181.3)(235.0-222.7)] = 33.8 Hz, for the M quartet: SQRT[(268.1-189.8)(236.4-221.8)] = 34.2 Hz
4. Calculate JAA' and JXX' by summing and subtracting K and M: JAA' = (K+M)/2 = (41.3+31.8)/2 = 36.5 Hz; JXX' =
(K-M)/2 = (41.3-31.8)/2 = 4.7 Hz. Because we do not know which ab quartet is K, and which M, we do not know the
relative signs of JAA' and JXX', nor do we know which coupling is which.
5. Calculate JAX and JAX' by summing and subtracting L and N: JAX = (N+L)/2 = (35.3+34.0)/2 = 34.7 Hz, JAX' =
(N-L)/2 = (35.3-34.0)/2 = 0.7 Hz. Again, we do not know which coupling is which, but the relative signs can be
determined: if |N| is larger than |L|, the signs are the same, as in this case.
276.2

268.1

247.0

236.4
235.0

222.7
221.8

211.7

189.8

181.3
|K| = |JAA' + JXX'| "J" of one ab quartet
|N| |L| = |JAX - JAX'| "" of both ab quartets
35.3
|M| = |JAA' - JXX'| "J" of other ab quartet
|K| |K| |N| = |JAX + JAX'| "doublet"
41.2 41.4
If we make the reasonable assumption that
|M| |M| JFF > JHH and JHF (trans) > JHF (cis) we get
31.7 32.0
the following values:
0.7

F H
4.5 4.0 ppm 3.5 3.0 36.5 4.7
F H
|L|
34.0 34.7
Special Cases of AA'XX' patterns: Unfortunately a large fraction of AA'XX' patterns are missing lines, which
means that some or all of the coupling constants may be indeterminate. Below are summarized several common
(and some less common) situations where a reduced number of lines is seen.
In the situation where JAX = JAX' (i.e. L = 0, A2X2) the spectrum collapses to a triplet. In other words, the
effective "chemical shift" of each of the ab quartets is zero, and thus each gives a single line at A. This is more or
less the situation with many acyclic compounds of the X-CH2-CH2-Y type, provided that X and Y are not too large,
but cause very different chemical shifts. See example C.
In the situation where JAA' JXX' (which is often approximately the case with X-CH2-CH2-Y and p-disubstituted
benzenes) the M ab quartet collapses to two lines since M = 0. See example A.
In cases where JXX' is zero, both ab quartets will have the same Jab (M = K) and will be identical, leaving only 6
lines. This is nearly the case for situations like symmetrical o-disubstituted benzenes or 1,4-disubstituted
butadienes, where JAA' is a 3-bond coupling, and JXX' a 5-bond coupling. In these situations L is small (i.e.JAX is close
to JAX') and the central lines of the K and M quartets will likely be superimposed, whereas the small outer lines may
be distinct -- the outer lines are separated by just under twice the value of JXX', the inner lines by just a fraction of
JXX'.
If the signs of JAX and JAX' are different the N lines will be relatively close together. This is the case for AA'XX'
patterns of the AA' vicinal type, where JAB is a geminal coupling, hence negative, and JAB' is vicinal, and hence
positive. In the limit, if JAX = -JAX' then the N lines can collapse to a singlet.

Reich, U.Wisc. Chem. 605 5-HMR-14.3 AA'XX'


5.15 The AA'BB' Pattern

As A and X of the AA'XX' spectrum move closer together, the lines of the 1:1 doublet (N) each split into two lines,
for a total of twelve in each half-spectrum. The AA' and BB' parts are no longer centrosymmetric, but develop a
mirror image relationship, so that the entire pattern is centrosymmetric. As is found for all AX to AB
transformations, "leaning" occurs, the inner lines increase and the outer lines decrease in intensity, culminating in
just a single line when AB = 0. AA'BB' patterns can be solved manually, but the process is difficult, and is better
done with computer simulations.
Typical molecular fragments which give AA'BB' patterns are:

HA X
HA'
X HA HB HA HA'
X HB R
HB'
HA HB' R
HA' Y Y HB' HB HB'
HB
HA' Y

(a) AA'-Gem (b) AA'-Vic (c) ODCB (d) p-Aromatic


o-Dichlorobenzene

(a) AA' Geminal (X-CH2CH2-Y). This is the most common type of AA'BB' pattern. The appearance can range
from essentially perfect triplets, to rather complicated patterns which cannot be easily analyzed. The two spectral
parameters which control appearance of the spectrum are chemical shift difference AB, and the difference between
the two vicinal coupling constants JAB and JAB'. If AB is small (AA'BB'), then the pattern will always be complicated,
no matter what the coupling constants are. If JAB JAB' then the pattern will mimic A2X2/A2B2 and triplets will be
seen if the chemical shift is large enough.
Changing chemical shifts while keeping coupling constants static for an AA'BB' system with JAB = JAB'

JAA' = 15.0
JBB' = 15.0
JAB = 7.0
AB = 210.0
JAB' = 7.0

AB = 110.0

AB = 50.0

AB = 27.0

250 200 150 100 50 0


Hz

Reich, U.Wisc. Chem. 605 5-HMR-15.1 AA'BB'


The key feature that determines the complexity of AA'BB' patterns of CH2-CH2 groups is the relative size of JAB
and JAB', which is determined largely by the conformational properties of the X-CH2CH2-Y fragment. For acyclic
systems, if the X and/or Y groups are small, then the populations of the anti and gauche conformations will be close
to statistical (1:2). As can be seen from the table and the simulated spectra below, the two averaged couplings
become approximately equal when there is 67% of gauche isomer, and the spectrum will mimic an A2B2 pattern --
triplets if AB is large enough (AB >> JAB). If X and Y are large, then the anti isomer will be favored and the pattern
will be more complex. In practice, adjacent CH2 groups often look like triplets, and thus the gauche conformation
must usually be favored. For cyclic systems (e.g., N-cyanomorpholine) the ring constrains the -CH2CH2- fragment to
mostly the gauche conformation, and clean triplets are not usually seen.
The coupling constants were calculated using the simple Karplus
equations below. The spectra shown were calculated using %anti/gauche
AA'XX' formulas (i.e. the chemical shift between A and B is large
compared to JAB). 100 / 0 JAB = 2.3

J = Jo cos2 JAB' = 12

Jo = 12 Hz for > 90
Jo = 9 Hz for < 90
JAA' = JBB' = -13 Hz
80 / 20 JAB = 3.2
Y HB' HA JAB' = 10.1
HA HA' Z HA Y HB'

HB HB' Y HB Z HA'
Z HA' HB
anti gauche gauche 60 / 40 JAB = 4.2
JAB = 2.3 JAB = 2.3 JAB = 12 JAB' = 8.1
JAB' = 12 JAB' = 2.3 JAB' = 2.3
JA'B' = 2.3 JA'B' = 12 JA'B' = 2.3
JA'B = 12 JA'B = 2.3 JA'B = 2.3

Thus for 60/40 anti/syn we have: 33 / 67 JAB = 5.5


JAB = (0.6)(2.3) + (0.2)(2.3) + (0.2)(12) = 4.24 JAB' = 5.5
JAB' = (0.6)(12) + (0.2)(2.3) + (0.2)(2.3) = 8.12

20 / 80 JAB = 6.2
CN
JAB' = 4.2
N

Line width
O = 1.5 Hz
300 MHz (CDCl3)
Source: Aldrich Spectra Viewer
0 / 100 JAB = 7.1
JAB' = 2.3
3.8 3.6 ppm 3.4 3.2

It is a common misconception that the equalization of coupling constants (and hence the appearance of triplets)
is a consequence of free rotation around the CH2-CH2 bond. In fact, there is free rotation around almost all such
bonds in acyclic molecules at accessible temperatures. The appearance of more complicated patterns is the result
of a preference for the anti conformation over the gauche (or vice versa), and has nothing to do with the rate of
rotation. See section 5-HMR-09 for some additional examples. New 35-01 Rev 42-11

Reich, U.Wisc. Chem. 605 5-HMR-15.2 AA'BB'


X
AA'BB' Spectra X-CH2-CH2-Y
A A' B'
B

AA'XX' H SeC(CH3)3
Br
Ph
Si(CH3)3

C8H9Br 200 MHz


Reich (Bowe)

Why is the triplet at 3.55


much taller than the one
at 3.15?

3.6 3.5 3.4 3.3 3.2 3.1


2.2 2.1 0.7 0.6 0.5

O O
AA'BB'
Br-CH2-CH2-Cl
H OMe
60 MHz
100MHz
Reich
C8H14O3

2.4 2.2 2.0 1.8 1.6

AA'BB'
Ph
PhS H
O
100 MHz
SeCH3
Reich (Olson)
C10H12O
200 MHz
Reich (Bowe)

2.7 2.6 2.5


3.5 3.0 2.5

Reich, U.Wisc. Chem. 605 5-HMR-15.3 AA'BB'


(b) AA' Vicinal. This type of AA'BB' pattern is much less common than type (a). It appears principally in <a
1,1-disubstituted cyclopropanes, 2,2-disubstituted 1,3-dioxolanes and other similar structures. The patterns are never
triplets because JAB is invariably quite different from JAB'.
The N doublet peaks are close together in these patterns, often inside the K and M ab quartets, because JAX is
negative, and JAX' is positive (N = JAX + JAX'). Since JAA' JXX' the K ab quartet has very small outer peaks, and the
inner peaks will then be close to the N doublet peaks. Similarly the "ab" M quartet will have a near zero "coupling" (M
= JAX + JAX') and so will be essentially be two singlets, or two very closely spaced doublets, not resolved the spectrum
below. These features can be seen in the dioxolane spectrum below, which has been simulated using the parameters
shown (red trace). Peak labels are for the AA'XX' assignments. Since this is actually an AA'BB' pattern, the N lines
each split into two closely spaced ones. Because the A/X shift difference is diastereotopic in nature, most of these
patterns tend toward AA'BB' except at very high field strength.
HA'
HA HA
X
X O HA'
O HB'
Y HBH Y
B'
HB

JAA' = 7.00 N N
O O
JBB' = 7.10 M Br
JAB = -7.70 M
M
JAB' = 6.20 MK K AA'BB' Simulation

C10H10BrO2, 200 MHz

K
K |K| = |JAA' + JXX'|
"J" of one ab quartet
|L| = |JAX - JAX'|
"" of both ab quartets
Spectrum
|M| = |JAA' - JXX'|
"J" of other ab quartet
|N| = |JAX + JAX'|
"doublet"
4.3 4.2 4.1 4.0 3.9 3.8

Here some AA'BB' spectra of dioxolanes.

O Br

O
200 MHz C5H9BrO2 270 MHz
Reich (Bowe) Reich (Bowe)

4.0 3.9 3.8 3.7 4.0 3.9 3.8

Reich, U.Wisc. Chem. 605 5-HMR-15.4 AA'BB'


Unsymmetrical 1,1-disubstituted cyclopropanes often have AA'XX' patterns that have a quartet-like appearance.
Because JAA' is close to JAX', the M quartet is essentially a doublet, and the L quartet is very strongly leaning (see the
spectra and simulation below). This means that the central two line clusters have essentially 3/4 of the intensity
(N+K), and the M lines 1/4, just as for a regular quartet
N
AA'BB' Simulation |K| = |JAA' + JXX'|
JAA' = 10.00 N HA' "J" of one ab quartet
K K HO2C HA
JBB' = 10.30 |L| = |JAX - JAX'|
JAB = -4.20 "" of both ab quartets
JAB' = 7.70 HB'
HO |M| = |JAA' - JXX'|
VAB = 59.45 MM HB "J" of other ab quartet
MM 300 MHz |N| = |JAX + JAX'|
CDCl3/DMSO-d6
"doublet"
K
K

1.2 1.1 1.0 0.9

The half-spectrum of 1-phenyl-1-cyanocyclopropane illustrates this effect. Here JAA' and JXX' are a little more
different than in the above example so the M quartet is visible as four lines. This 2-proton multiplet could be
mistaken for a doublet of quartets by the unwary spectroscopist.
NN
JAA' = 9.40 K K
JBB' = 10.20 300 MHz
JAB = -5.30
AA'BB' Simulation
JAB' = 7.60
VAB = 93.52 MM HA'
MM NC HA
VA
K K HB' JA
Ph W
HB

1.75 1.70 1.65

O O

100 MHz
100 MHz Reich (Renga)
Reich (Renga)

1.2 1.0 0.8 0.6 0.4 1.2 1.0 0.8

Reich, U.Wisc. Chem. 605 5-HMR-15.5 AA'BB'


(c) o-Dichlorobenzene (ODCB) Type. This kind of AA'BB' pattern never approaches A2X2 because JAB (ortho
coupling) is always much larger than JAB' (meta coupling). It is seen in symmetrically 1,4-disubstituted dienes and
ortho disubstituted benzenes. Note that for all AA'XX' / AA'BB' systems the A and X / B patterns are identical
(although they have a mirror-image relationship). This is in spite of the fact that the coupling relationships of A and
X / B are often quite different in molecules of this type. The spectra often have the appearance of a dd.
For most AA'XX' patterns of this type JXX' is a 3-bond coupling, and JAA' a 5-bond coupling, Thus the K and M
ab quartets will be nearly superimposed (the Jab values differ by 2JAA'), often leaving only 6 resolvable lines
(Simulation A, naphthalene, o-dichlorobenzene). Even with the small JAA' values that are usually seen, the central
lines of the K and M quartets will likely be superimposed, whereas the small outer lines may be distinct -- they will be
separated by just under twice the value of JAA', the inner lines by just a fraction of JXX' (Simulation B, spectrum of
biphenylene and the 1,4-diacetoxybutadienes).

Simulated Spectra (AA' part only):


N MM N N N |K| = |JAA' + JXX'|
KK "J" of one ab quartet
HA
A B MM
KK
JAA' = 0.00 JAA' = 0.40
|L| = |JAX - JAX'|
HX JXX' = 7.00 JXX' = 7.00 "" of both ab quartets
JAX = 7.00 JAX = 7.00 |M| = |JAA' - JXX'|
JAX' = 2.50 JAX' = 2.50 "J" of other ab quartet
HX'
M M
|N| = |JAX + JAX'|
HA' K K KM MK "doublet"

160 150 140 160 150 140

Spectra:
HA
HB

300 MHz, CDCl3


HB'
HA'

7.9 7.8 7.7 ppm 7.6 7.5 7.4 7.3

N N
Cl M M
KN KM
M
300 MHz, CDCl3 K N K

Cl

300 MHz, CDCl3


K K
M MK
M
KM

7.5 7.4 ppm 7.3 7.2 7.1 6.75 6.70 6.65 6.60 6.55

Reich, U.Wisc. Chem. 605 5-HMR-15.6 AA'BB'


Only X part shown
AcO HX 7.35
HA HX 2.10
HA'
OAc
HA 2.13
60 MHz AcO 60 MHz
5.95 HX' OAc 7.10
5.87 HX' HA'
JOC 1968, 2835 (HX shown)
(HX shown)
JOC 1968, 2835

JAX = 12.5 JAX = 6.7


JAX' = -0.7 JAX' = -1.3
JXX' = 11.5 JXX' = 11.0
JAA' = 0.8 JAA' = 1.6

6.3 6.1 5.9 5.7 6.0 5.8 5.6


ppm ppm

S
300 MHz, CDCl3
Reich (Whipple)

7.3 7.2 7.1


ppm

As with all such NMR spectra, the patterns get more complex when the chemical shift between the protons
becomes smaller (AA'BB'). Some examples (both A and B shown):

H H Ph Ph
Ph Ph H H

H H H H
100 MHz 100 MHz
I. L. Reich I. L. Reich

7.2 7.0 6.8 6.6 7.2 7.0 6.8

Rev 41-06

Reich, U.Wisc. Chem. 605 5-HMR-15.7 AA'BB'


(d) p-Disubstituted Benzene Type. These usually resemble a two AX doublets or an AB quartet, with some
small extra lines.
For AA'XX' patterns of this type, the two 3J meta couplings (JAA' = JXX') are likely to be very close in size so that
the M ab quartet collapses to two lines since M = 0 or very small (Simulation B). In addition, the para-coupling (JAX')
is nearly zero, which makes L N. Thus the two lines of the M quartet will appear at or close to the chemical shift of
the N doublet, leaving only a doublet and one ab quartet, whose doublets will be on both sides of the N+M lines
(simulation A). This is why p-substituted benzenes are often incorrectly described as doublets, since the extra lines
corresponding to the K quartet amount to only 1/4 of the total intensity, and appear fairly close to the dominant M+N
doublet.

|K| = |JAA' + JXX'| |M| = |JAA' - JXX'|


"J" of one ab quartet "J" of other ab quartet
|L| = |JAX - JAX'| |N| = |JAX + JAX'|
Simulated Spectra (AA' part only): "" of both ab quartets "doublet"

A B C
N+M
Equal Jmeta Equal Jmeta Different Jmeta
Jpara = 0 Jpara = 0.8 Hz Jpara = 0.8 Hz
Jo + Jp N N N N

JAA' = 2.00 JAA' = 2.00


JAA' = 1.80
JXX' = 2.00 JXX' = 2.00 JXX' = 2.20
JAX = 8.00 JAX = 8.00 M
M JAX = 8.00
JAX' = 0.00 JAX' = 0.60
K K K JAX' = 0.60 K K
K
M "ab quartet" M M
K K K
collapsed (J = 0) K K K

160 150 140 160 150 160 150 140

Spectra: HA' HA'


HB' S S HB'

MeO HA HA OMe
HB HB

300 MHz (CDCl3)


Reich (Sikorski)

7.4 7.3 7.2 ppm 7.1 7.0 6.9 6.8

Reich, U.Wisc. Chem. 605 5-HMR-15.8 AA'BB'


When the chemical shift between A and B is small (as in p-bromochlorobenzene below), the usual leaning effects
are seen, additional lines appear, and the extra lines become more pronounced. As 300 MHz a more "normal"
AA'BB' pattern is seen
Br
HA' HA 300 MHz
60 MHz
HB' HB
Cl
C6H4BrCl

7.4 7.2 7.4 7.3 7.2

(e) Miscellaneous. In the 1,3-substituted cyclobutane below the 4JHH (JAA', JBB' and JAB') are large enough that
the expected first-order AB quartet is not seen, but a much more complicated pattern. In this case, as in the
p-disubstituted benzene case above the "AB" character dominates the appearance of the multiplet, largely because
the coupling interactions between either of the AB protons and the A'B' protons are much weaker than the A to B
and the A' to B' couplings (i.e. the AB pair is nearly isolated from the A'B' pair).

HA HB
Ph
OH

Ph
HA' HB'
Reich (Wesley)
3.40 3.35 3.10 3.05

Trans-1,2-disubstituted cyclopropanes can form AA'BB' patterns. Some examples

HO2C
H HO2C
H
H H
HO2C
HO2C
In CDCl3 & DMSO-d6
CDCl3
300 MHz
300 MHz

5.65 5.60 5.55 2.70 2.65 2.60 2.2 2.0 1.8 1.6 1.4
ppm ppm ppm

Reich, U.Wisc. Chem. 605 5-HMR-15.9 AA'BB'


Origin of Complexity in Patterns of the AA'XX'... Type

There are two situations where spin systems containing AA'XX' type do not show unusual complexity. One is
where JAX = JAX', in which case the pattern becomes first order A2X2.
The second is systems in which there is no coupling between both of the chemical shift equivalent protons, i.e.,
JAA' = JXX' = 0. In such cases the degeneracy between spin states is no longer present, and first order systems
result. Consider two examples. A monosubstituted benzene is an AA'BB'C or AA'MM'X system. A simulated
spectrum is shown below
R
Simulation data: Simulated spectrum
HA HA' JAM = JA'M' = 8
JMX = JM'X = 8
JAA' = 2 2.03 2.03
HM HM' JMM' = 2
HX JAX = JA'X = 2.0
1.00 JAM' = JA'M = 0.5

9.5 9.0 8.5 8.0 ppm 7.5 7.0 6.5 6.0


If we recalculate the spectrum after setting JAA' = 0 and JMM' = 0 then the spectrum becomes essentially first order
(it would be completely first order if the chemical shifts between A, M and X were made larger).

Simulation JAA' = 0 Simulated spectrum


JMM' = 0

9.5 9.0 8.5 8.0 ppm 7.5 7.0 6.5 6.0

For this reason, some spin systems which are formally of the AA'XX' type, but in which there is no significant
spin-spin coupling between the equivalent protons show first order spectra. For example, the fairly common spin
system below of the AA'BB'X type shows no unusual complexity (beyond that of normal ABX patterns) because
there is no coupling between the A and A' protons, nor between the B and B' protons. Such systems are sometimes
defined as (AB)2X to indicate that magnetic inequivalence is not a factor. The spectrum of 1,2,3-trichloropropane is
an example.
300 MHz 1H NMR in CDCl3
Source: Aldrich Spectra Viewer/Reich

HA HB HA' HB' Cl
Cl Cl
R R
R' HX C3H5Cl3

4.4 4.3 4.2 4.1 4.0 3.9 3.8 3.7

Reich, U.Wisc. Chem. 605 5-HMR-15.10 AA'BB'


Contrast this with the NMR spectrum of dibromosulfolane below.

O O Actual Spectrum 250 MHz, CDCl3


S

Br Br

4.85 4.80 4.75 4.10 4.05 4.00 3.95 3.60 3.55 3.50
Hz Hz Hz

Why is this spectrum so complex? In this AA'MM'XX' system, although there is no significant coupling between A
and A' or M and M', there is coupling between X and X', making the whole system a highly second order one
(remember: anything coupled to a strongly coupled pair like the XX' protons becomes second order).

O O
A = 3.7 JAM = JA'M' = -13
HA S HM' Simulated spectrum
B = 4.03 JMX = JM'X' = 5
HM HA' X = 4.40 JAA' = 0
JMM' = 0
Br HX' JAX = JA'X' = 8.0
HX Br JAM' = JA'M = 0
HX JXX' = 6 HM HA

JXX' = 6 Hz

4.5 4.4 4.3 4.2 4.1 4.0 3.9 3.8 3.7 3.6

If we remove the XX' coupling, the system becomes essentially first order (two isolated AMX patterns). We could
better describe it as an (AMX)2 rather than an AA'MM'XX' spin system.

Simulated spectrum

JXX' = 0

4.5 4.4 4.3 4.2 4.1 4.0 3.9 3.8 3.7 3.6

Reich, U.Wisc. Chem. 605 5-HMR-15.11 AA'BB'


Another example of the same spin system is shown below.

Note the extra peaks - these are not impurities or bad


300 MHz 1H NMR spectrum in CDCl3.
tuning. The additional complexity arises because we have
Source:Charlie Fry/Reich
an AA'BB'XX' system here. Even though JAA' and JBB' are
essentially 0, the fact that X and X' are significantly coupled
O causes the AB signals to become complicated
A
OMe
MeO B
X'

X
B' OMe
MeO
A'
O

2.8 2.7 2.6 2.5


The aromatic protons are also ppm
an AA'BB' system x AA'BB'XX'

6.8 6.7
ppm

8.00 3.45 3.40 3.35


ppm
6.175.80 AA'BB'XX'

3.90

2.11

7 6 5 4 3 2 1 0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-15.12 AA'BB'


Copyright Hans J. Reich 2017
All Rights Reserved
5.16 Virtual Coupling University of Wisconsin

The term "virtual coupling" refers to an NMR phenomenon in which apparently first-order multiplets contain false
coupling information. In extreme cases, protons that are not actually coupled will show splitting. More commonly,
the magnitude of coupling constants obtained by first-order analysis is incorrect. All virtual coupling effects arise
when protons, well isolated from other protons in chemical shift, are coupled to a group of other protons which are
strongly coupled to each other. By strongly coupled we mean that these protons are both close in chemical shift
and coupled to each other with J > .
On the following pages are examples which illustrate the virtual coupling phenomenon in several different
systems. One simple way to summarize these effects is as follows: when two or more protons are strongly coupled,
then any protons coupled to them will give multiplets with false coupling information and/or unexpectedly complex
multiplet structure. Furthermore, the strongly coupled protons themselves will contain misleading multiplet
structure. Virtual coupling effects are also frequently encountered in heteronuclear magnetic equivalence NMR
situations, e.g. in transition metal phosphine complexes (see Sect. 7-MULTI-2)
The simplest systems to show virtual coupling effects are ABX patterns, where the X-multiplet will give incorrect
values for JAX and JBX when it is analyzed as a doublet of doublets if AB is similar to or smaller than JAB. What is
particularly insidious about this effect is that analysis of the AB part by a first-order method (treatment of each part as
a distorted doublet of doublets) will give the same incorrect values for JAX and JBX.

If HA and HB are strongly coupled, then HX will not be a simple doublet, even if JA-X = 0

HA HX
HA HB HX will
mimic: C C
C C C
HB

If A-B = 0 then HX will be a triplet, even if JA-X = 0


It will appear that HX is coupled equally to HA and HB

Strongly-Coupled: JA-B > A-B

A typical example of virtual coupling is provided by the epoxide below. HA and HB are more or less first order
when the spectrum is taken in CDCl3 because HA and HB have a significant chemical shift (although typical second
order effects are starting to appear). However, in acetrone-d6, HA and HB are essentially superimposed, and HC
appears as a triplet, as if HA and HB were equally coupled to HC, leading to a very different structure assignment.
The small coupling visible for HB and HC in CDCl3 is probably not entirely real - it is the beginning of the "virtual
coupling" effect which eventually leads to a triplet for HC

Cl
200 MHz NMR spectrum
Cl I. L. Reich "Virtually coupled" triplet
O

HA HC

HB Cl CDCl3 Acetone-d6
Cl

HC HA HB HC HA HB

Reich, U.Wisc. Chem. 605 5-HMR-16.1 Virtual Coupling


These effects can be understood by examining the behavior of the X part of an ABX system as AB becomes
small. In this simulation, JAX is set to 0; nevertheless when AB < 15 Hz the X part shows clear indications of
coupling to HA, as does the AB part, i.e. virtual coupling.

HX JAB = 10 Hz Mimic:
JBX = 7 Hz HA HB HX
JAX = 0 C C C

As soon as AB approaches JAB, It looks as if X


is coupled to both A and B (i.e. X is a dd)

A, B Part AB = 10 Hz
AB = 40 Hz
The AB part also
shows the "fake"
AB = 20 Hz coupling.

AB = 10 Hz

AB = 5 Hz 40 30 20 10 0
Hz

A, B Part AB = 1 Hz
AB = 0 Hz

Looks like:
AB = 0 Hz
HA HX
C C
HB
20 10 0 -10 -20
Hz

The only hint that the X part is not a simple


triplet is the appearance of the small outer
peaks (lines 15 and 16 of the X part) and the
40 30 20 10 0 extra lines in the AB part.
Hz

Reich, U.Wisc. Chem. 605 5-HMR-16.2 Virtual Coupling


Virtual Coupling in X of an ABX system

Q = "Mixing coefficient"
AMX ABX J
Q =
M B + 2D
AM AB
A A


(+Q)/ 1+Q2
X= X=


(Q+)/ 1+Q2


These are transitions


of the X nucleus





(Q+)/ 1+Q2
X= X=
(+Q)/ 1+Q2



JAX JMX JAX JBX As AB 0, Q 1

As the chemical shift between HA and HB becomes smaller, the state begins to mix with the state, and the
state mixes with the (the mixing coefficient is Q). This results in their energies becoming closer together, and
eventually, when AB = 0, Q becomes 1, and the two energy levels consist of equal parts and , and their
energies are identical (the X part becomes a triplet).

The Illustration is for both JAX and JBX > 0


2
2
1+Q2
1+Q2

(+)/
(+)/
(+Q)/


(Q+)/

AB spins:

X X X
3.1 3.0 2.9 2.8 2.7 3.1 3.0 2.9 2.8 2.7 3.1 3.0 2.9 2.8 2.7
ppm ppm ppm

Reich, U.Wisc. Chem. 605 5-HMR-16.3 Virtual Coupling


"Virtual Coupling" Effects

An important generalization in this area is that the severity of the "virtual coupling" effects is very strongly
dependent on JAB - if JAB is larger than JAX and JBX then the X part is profoundly changed (i.e. doublet to triplet). On
the other hand, when JAB is smaller than JAX or JBX then the perturbations are much less dramatic, leading to small
additional lines and minor errors in apparent coupling constants. A case of this type is provided by the three
compounds below. The aromatic protons are well separated in 1 and 3 and thus are pretty much first order, but in 2
HA appears nearly on top of HB. As a result, HX shows non-first order structure.
O O
Br Cl

Cl
HB Cl HB Cl HB Cl

Cl HX Cl HX Cl HX
HA HA HA
1 2 3

HX HB HA HX HA HB HX HA HB

One of the main reasons that complex NMR spectra are simpler to interpret at higher field strengths is that many
virtual coupling effects are ameliorated or disappear altogether as chemical shift differences (in Hz) become larger.
C
Exceptions are AA'BB', AA'XX' patterns and more complicated analogs (such as AA'BB'X). These are always
non-first order (if there is coupling between A and A' or B and B') because they always satisfy the criteria for virtual
coupling: A and A' have zero chemical shift at any field strength. If A and A' are also coupled to each other (JAA' >
0) then the A and B protons, or any other protons coupled to A or B can give complex or misleading multiplets.

Reich, U.Wisc. Chem. 605 5-HMR-16.4 Virtual Coupling


What are those Impurities? A student in the course brought me the NMR spectrum of acrolein shown below,
with the question: why can't I get rid of the impurities, I've distilled the compound twice?
Acrolein: C3H4O
H2
200 MHz 1H NMR spectrum in CDCl3
Source: R. R. Dykstra RRDI-257/Reich H3 H1

H4 O
CDCl3
H2
Hz
H4 30 20 10 0
H1

H3
3.12

1.00
9.6 9.5 6.6 6.5 6.4 6.3 6.2

9 8 7 6 5 4 3 2 1 0
ppm
The answer? The compound is perfectly pure, as shown by its NMR spectrum in benzene-d6, which is essentially
first order. The problem with the spectrum in CDCl3 is that two of the protons (H2 and H3) are nearly superimposed
and are strongly coupled, so that all the others which are coupled to them become very complicated. Benzene, like
other aromatic solvents, causes significant upfield shifts (especially in molecules with a high dipole moment, like
this one) which removes the degeneracy and leads to an essentially first-order spectrum.

Acrolein: C3H4O
200 MHz 1H NMR spectrum in benzene-d6
Source: R. R. Dykstra/Reich Hz
30 20 10 0
H2
H3 H1 H2 H3 H4

H4 O

Benzene-d6

1.95

1.00 1.02

9.2 9.1 6.0 5.9 5.6 5.5 5.4

9 8 7 6 5 4 3 2 1 0
ppm
9/28 12/36
Exercise: Use WINDNMR to simulate first the benzene-d6 spectrum, and then the CDCl3 spectrum (all you
should need to do is adjust the chemical shifts). Which two protons are superimposed in the CDCl3 spectrum?

Reich, U.Wisc. Chem. 605 5-HMR-16.5 Virtual Coupling


Is it the Right Stuff? Accidental coincidences of chemical shifts and the resulting virtual coupling effects can
lead to surprisingly deceptive NMR spectra, sufficiently so that the unwary researcher could conclude that a
synthesis had failed. An interesting case is the phenyl region of the NMR spectrum below:

C13H17NO
300 MHz 1H NMR spectrum in CDCl3.
H Source: Alison Wendlandt/Stahl
N
ortho
+
O Hz
meta 30 20 10 0

6.80 6.75 6.70 6.65


ppm
para
3.93

7.40 7.35 7.30 7.25 7.20 7.15


ppm

0.99 1.04
0.87
5.75 5.70 5.65
ppm

7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6
ppm

Taken at face value the aromatic region of the cis isomer looks like a 4H doublet and a 1H multiplet (pentet or
sextet) with J of 4.3 Hz, hardly compatible with a monosubstituted phenyl group. Yet this is simply a phenyl group in
which the ortho and meta protons accidentally have identical chemical shifts (or nearly so). Since they are strongly
coupled to each other, the four protons behave as a unit, and the para proton appears to be equally coupled to all
four, leading to the apparent doublet and pentet - a classical case of virtual coupling. The apparent coupling
constant of 4.3 Hz is the average of the ortho and meta couplings.
In the trans isomer below the meta and para protons are a little further apart, and easily recognized for what they
are (although still not even close to a first order pattern).
H ortho C13H17NO
N 300 MHz 1H NMR spectrum in CDCl3.
Source: Alison Wendlandt/Stahl
O meta
Hz
30 20 10 0
4.01
para

7.45 7.40 7.35 7.30 7.25 7.20 7.15 7.10 7.05


ppm

1.01 0.98
0.80

6.30 6.25
ppm

7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6
ppm

Reich, U.Wisc. Chem. 605 5-HMR-16.6 Virtual Coupling


Why is my Spectrum so Ugly? In this spectrum there is an accidental superposition of two protons, HM and
HN, which are coupled to each other. All of the protons coupled to these two, HA, HB, HX, and HY, show strongly
second order distorted multiplets, with extra lines and non-centrosymmetric structure. A simple method to address
this kind of problem is to take the spectrum in an aromatic solvent (or even just add a few drops of benzene-d6 or C
pyridine to the sample), which in many cases moves the protons around enough that the second-order effects are
reduced or eliminated. When the spectrum is taken in benzene, HM and HN are shifted away from each other, and
HA and HB now show more or less first order multiplets.

O
1
270 MHz H NMR spectrum in CDCl3 O
(Source: Margaret K. Jones/S. D. Burke)
H CO2CH3
Me3Si H

HA
O
HB
Spectrum in C6D6: O

CO2CH3
HM
HN
HX HY
HA, HB Because HM and HN are strongly coupled, both the
HA, HB and HX, HY pairs are non-first order, with
abnormal intensities and line positions.
3.8 3.7 3.6 3.5 3.4

Hz
30 20 10 0

5.94

4.35 4.25 4.0 3.9 2.95 2.85 2.8 2.7 2.6 2.5 2.10
HA, HB HM, HN HX, HY 3.04

2.09 1.98
1.95

1.07 1.00 1.00 1.00


0.98

6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0
ppm

Reich, U.Wisc. Chem. 605 5-HMR-16.7 Virtual Coupling


Methyl Region AMX3 ---> ABX3 ---> AA'X3

Methyl groups usually give reliably simple multiplets in NMR spectra. However, if a methyl group is coupled to a
proton which is part of a strongly-coupled system, then its NMR signal can give complex and misleading information

In these simulated spectra, the methyl group is coupled to HA, which


is in turn coupled to HB. When the chemical shift of HA and HB
become close, then all signals become second order.

C C CH3 JAB = 7.5


HB HA JA-CH3 = 7.5
JB-CH3 = 0
AB = 0 Hz
Vary

AB = 1 Hz

AB = 2 Hz

AB = 4 Hz

AB = 6 Hz

AB = 10 Hz

AB = 14 Hz

AB = 20 Hz

AB = 30 Hz

B A CH3
300 MHz 1H NMR spectrum In this compound there is a small chemical shift
Source: Aldrich NMR Library difference between H2 and H3t ( close to J) which
leads to complicated multiplets for H3c and the methyl
group (as well as for H2 and H3t).

O
H2 H3t O O
H3c
H3c CH3
H3t H2

3.3 3.2 3.1 3.0 2.9 2.8 2.7 2.6 1.55 1.40
ppm ppm

Reich, U.Wisc. Chem. 605 5-HMR-16.8 Virtual Coupling