Highly Recommended NMR Book!
Dr. Tim Claridge
Director of NMR Spectroscopy & University Research Lecturer
Organic Chemistry NMR Facility
at the University of Oxford
http://www.chem.ox.ac.uk/spectroscopy/nmr/
THE COSY (COrrelation SpectroscopY) EXPERIMENT
Jean Jeener. 1971. Two-dimensional Fourier Transform NMR, presented at an Ampère International
Summer School, Basko Polje, unpublished.
A verbatim quote follows from Richard R. Ernst's Nobel Laureate Lecture delivered on December
2nd, 1992, ``A new approach to measure two-dimensional (2D) spectra has been proposed by Jean
Jeener at an Ampere Summer School in Basko Polje, Yugoslavia, 1971 ([6]). He suggested a 2D
Fourier transform experiment consisting of two 2 pulses with a variable time t1 between the pulses
and the time variable t2 measuring the time elapsed after the second pulse as shown in Fig. 6 that
expands the principles of Fig. 1. Measuring the response s(t1 t2) of the two-pulse sequence and
Fourier-transformation with respect to both time variables produces a two-dimensional spectrum
S(O1 O2) of the desired form (62,63). This two-pulse experiment by Jean Jeener is the forefather
of a whole class of 2D experiments (8,63) that can also easily be expanded to multidimensional
spectroscopy.''
THE BASIC IDEA OF nD NMR
For a single uncoupled proton resonance
P: Preparation
E: Evolution
M: Mixing
D: Detection
Amplitude modulation of a singlet resonance as a function of t 1
(signal decays due to spin relaxation)
FT t2
FT t1
The concept can be expanded to 3 and n dimensions
two-dimensional (2D) experiment
threedimensional (3D) experiment
In the previous slide two-dimensional experiment, a sample containing two uncoupled
spins A and X, of offsets nA and nX, will produce 2D peaks at their corresponding
chemical shift offsets in both dimensions (2D spectra are in fact 3D images).
Contour Plot
If spins A and X are J coupled, during the evolution time t1 the magnetization will
evolve into anti-phase components. A necessary condition for the magnetization
magnetization transfer to takes place during the mixing pulse (M).
As a result, each magnetization (A and X) will not only modulate with its own
resonance frequency but they will also modulate with the resonance frequency of
the coupled spin, giving origin to a cross-correlation peak between them.
Diagonal Peak
Cross-Correlation Peak
Jean Jeener’s brilliant idea has opened an endless avenue for what today is
Multinuclear and Multidimensional NMR Spectroscopy.
“Extensions of the standard COSY experiment. Relayed correlation, total correlation spectroscopy (TOCSY), and multiple
quantum spectroscopy (MQS) increase the information content, while exclusive correlation (E.COSY), multiple quantum
filtering (MQF), and spin topology filtration reduce the complexity. Both avenues can lead to three-dimensional
spectroscopy.” Richard R. Ernst Nobel Lecture*
*Richard R. Ernst. Nobel Lecture. http://nobelprize.org/nobel_prizes/chemistry/laureates/1991/ernst-lecture.pdf
Angewandte Chemie Int. Ed. 1992, 31(7), 805-930.
ADVANCE METHODS FOR STRUCTURE ELUCIDATION
OF
SMALL MOLECULES
2D NMR Spectroscopy
We will focus mainly in 1H and 13C nuclei at natural abundance,
although 2D NMR is not limited to these nuclei only.
Correlations Through Chemical Bonds:
-Homonuclear
-Heteronuclear
Correlations Through Space:
-Homonuclear
Through Chemical Bond 2D Homonuclear Chemical Shift Correlations Experiments
COSY-90
Correlating coupled homonuclear spins. Typically used for correlating protons coupled over two or three bonds but may be used
for any high-abundance nuclide. The basic COSY experiment.
DQF-COSY
Correlating coupled homonuclear spins. Typically used for correlating protons coupled over two or three bonds. Higher-resolution
display than basic COSY. Additional information on magnitudes of coupling constants may be extracted from 2D peak fine
structure. Singlets suppressed.
COSY-b
Correlating coupled homonuclear spins. Typically used for correlating protons coupled over two or three bonds but may be used
for any high-abundance nuclide. Reduced 2D peak structure over basic COSY. Vicinal and geminal coupling relationships can be
differentiated in some cases.
Delayed COSY
Correlating coupled homonuclear spins through small couplings. Often used to identify proton correlations over many bonds (>3),
hence also known as long-range COSY.
TOCSY
Correlating coupled homonuclear spins and those that reside within the same spin system but which may not share mutual
couplings. Employs the propagation of magnetization along a continuous chain of spins. Powerful technique for analyzing complex
proton spectra.
INADEQUATE
Correlating coupled homonuclear spins of low natural abundance. Typically used for correlating adjacent carbon centers at natural
abundance but has extremely low sensitivity.
ADEQUATE
Correlating coupled homonuclear spins of low natural abundance, primarily 13C–13C, but employing 1H excitation and detection for
sensitivity improvement.
The choice of the COSY Experiment. Which COSY should we use?
Absolute-value (magnitude-mode) COSY-90
Pros: Simple and robust, magnitude processing well suited to automated operation
Cons: Phase-twisted lineshapes produce poor resolution, which require strong
resolution enhancement functions. Crosspeak fine structure not usually apparent
Phase-sensitive COSY-90
Pros: High-resolution display due to absorptive lineshapes. Crosspeak fine structure
apparent; J measurement possible
Cons: Diagonal peaks have dispersive lineshapes that may interfere with
neighboring crosspeaks. Requires high digital resolution to reveal multiplet structures
Phase-sensitive DQF-COSY
Pros: High-resolution display due to absorptive lineshapes. Crosspeak fine structure
apparent; J measurement possible. Diagonals also have absorptive lineshapes. Singlets
suppressed
Cons:Theoretical sensitivity loss by a factor of 2 relative to the COSY-90 variant.
Requires high digital resolution to reveal multiplet structure
COSY-b
Pros: Simple and robust. Magnitude processing well suited to automated operation.
Simplification of crosspeak structures reduces peak overlap. Vicinal and geminal
couplings can be distinguished in some cases from tilt of peaks
Cons: Usually requires magnitude-mode presentation as phase sensitive variant has
mixed-phase lineshapes
Delayed COSY
Pros: Enhances detection of small- and long-range couplings (< 2 Hz) such as
between protons in allylic systems or those in w-relationships
Cons: Requires magnitude-mode presentation. Crosspeaks due to larger couplings can
be significantly attenuated
Relayed COSY
Pros: Provides two (or more)-step transfers and can reduce ambiguities arising from
crosspeak overlap
Cons: Typically has low sensitivity and responses show mixture of lineshapes, so
magnitude-mode presentations may be required. TOCSY preferred
TOCSY
Pros: Provides multi-step (relayed) transfers to overcome ambiguities arising from
crosspeak overlap. High sensitivity. In-phase lineshapes can provide correlations even
in the presence of broad resonances
Cons: Number of transfer steps associated with each crosspeak not known, a priori.
In-phase lineshapes tend to mask crosspeak fine structure and may preclude J
measurement
COSY Magnitude Mode
Gradient-selected COSY
Best option
No phase cycling
One scan per increment is enough
The Multiple Quantum Filter (MQF)
AX Spin System
E
SQ
ZQ
Single Spin
(Singlet)
- 1/2
DQ
b
SQ
SQ
SQ
SQ
+ 1/2

A1, A2, X1 and X2 are Single Quantum Transitions (SQ). Total Spin Change is 1
DQ: Double Quantum. Total Spin Change is 2
ZQ: Zero Quantum. Total Spin Change is 0 (zero)
Phase Sensitive COSY
*DQF: Double Quantum Filtered
Phase Sensitive DQF-COSY*
Phase Sensitive DQF-COSY
Suppressed by DQF
HDO
MeOD
Antiphase Absortive Structure
1D double-quantum filtration of the spectrum of the peptide Leu-enkephalin 5.4 in
CD3OD. The singlet resonances of the solvent, truncated in the conventional 1D
spectrum (a), have been filtered out in (b). The remaining peaks in (b) display the
characteristic anti-phase multiplet structure (which may be masked by magnitude
calculation if desired
Multiplets Structure in Phase Sensitive DQF COSY
AX Spin System
AMX Spin System
Overlapped resonances in the 1D Spectrum of the peptide Leu-enkephalin
Resolved Multiplicity in the DQF COSY
F2 slice. J extraction possible
High Resolution DQF COSY
COSY-b
In the COSY-b experiment, the mixing
pulse is usually set to 45o or 60o
V: Vecinal J (Positive), G: Geminal J (Negative)
Delayed COSY: detecting small couplings
The delayed (or longrange) COSY sequence.
Additional fixed delays are inserted into the
basic COSY sequence to enhance the
appearance of correlations from small
couplings.
The small J’s (1 Hz),
COSY-90
Delayed COSY (D = 200ms)
TOtal COrrelation SpectroscopY – TOCSY (Former HOHAHA)
- Correlates all protons within the same spin system
- Number of transfer steps associated with each crosspeak not known, a priori
- A second significant feature of TOCSY that contrasts with COSY is that it utilizes
the net transfer of in-phase magnetization; so it does not suffer from cancellation
of anti-phase peaks under conditions of low digital resolution or large linewidths. In
these instances, this feature makes TOCSY the more sensitive of these two
methods.
The TOCSY sequence.
The spin-lock mixing time, m, replaces the
single mixing pulse of the basic COSY
experiment.
VIRTUAL COUPLING
Suppose that:
- (Ha) is coupled to another (Hb) which is far away from it in chemical shift
- Hence, we expect for Ha a simple doublet
- But suppose that Hb is coupled to a third spin (Hc) which is very close to Hb in
chemical shift.
-Hence, we say that Hb and Hc are strongly coupled, which will distort the
multiplet patterns of these two spins
-The odd thing, however, is that the Ha multiplet becomes more complex, as if it
were coupled to Hc as well as to Hb.
This is called "virtual coupling" because the Ha nucleus, which has no J coupling
to Hc, appears to be coupled to it because of the strong coupling between Hb
and Hc
A general way of stating this is that:
Any nucleus which is coupled to one member of a
strongly-coupled group of nuclei will behave as if it
is J coupled to all of the members of the group”
“
A classic example is the CH3 signal shape in fatty acids
d 2.7
d 1.2
d 0.8
HO-C(O)-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH3
Broad distorted triplet
VIRTUAL COUPLING CAN BE TEMPORARILY INDUCED BY SPIN-LOCK
The spin-lock in its simplest form is a
single, long, low-power pulse. This can be
viewed as a continuous sequence of
closely spaced 180 pulses bracketed by
infinitely small periods, .
Spin-Locked Magnetizations
Chemical shift evolution terms are cancelled
But not the J evolution terms
Oscillating nature of the in-phase magnetization transfer in the TOCSY
for protons Ha and Hb as a function of the mixing time t.
t=0
Mxb
Mx a
t = 1/2J
At 70ms full transfer for Jab= 7.14 Hz
For Jab = 2 Hz, 18% of M transfer
For Jab = 0.5 Hz, only 1.2% of M transfer
Sample: Chentobiose
Solvent: CDCl3
Spectrometer: AVANCE 400
TOCSY
B
A
1
Probehead: Inverse Broadband with z-Gradients
1
A Anomeric H
B Anomeric H
B 1H NMR
A 1H NMR
http://rmn.iqfr.csic.es/guide/tutorials/specdata/spectra/dis_tocsy.html
Through Chemical Bond 2D Heteronuclear Chemical Shift Correlations Experiments
HMQC (Heteronuclear Multiple Quantum Correlation)
Correlating coupled heteronuclear spins across a single bond and hence identifying directly connected nuclei, most often 1H–13C.
Employs detection of high-sensitivity nuclide, e.g. 1H, 19F, 31P (an ‘inverse technique’). Experimentally robust sequence, well
suited to routine structural characterization.
HSQC (Heteronuclear Single Quantum Correlation)
Correlating coupled heteronuclear spins across a single bond and hence identifying directly connected nuclei. Employs detection
of high-sensitivity nuclide, e.g. 1H, 19F, 31P (an ‘inverse technique’). Provides improved resolution over HMQC, so it is better suited
for crowded spectra but can be more sensitive to experimental imperfections.
HMBC (Heteronuclear Multiple Bonds Correlation)
Correlating coupled spins across multiple bonds. Employs detection of high-sensitivity nuclide, e.g. 1H, 19F, 31P (an ‘inverse
technique’). Essentially HMQC tuned for the detection of small couplings. Most valuable in correlating 1H–13C over two or
three bonds. Powerful tool for linking together structural fragments.
HETCOR (Heteronuclear Correlation)
Correlating coupled heteronuclear spins across a single bond. Employs detection of the lower- nuclide, typically 13C, so has
significantly lower sensitivity than inverse techniques. Benefits from high resolution in 13C dimension, so may find use when this is
critical, otherwise superseded by above methods.
H–X–Y
Triple-resonance methods for correlating protons and a heteratom (Y) whilst using a second heteroatom (X) to either relay the
correlations or edit the correlation spectrum.
13C
(CARBON) NMR SPECTROSCOPY
Some facts:
- NMR is not limited to study protons, you can also observe 13C, 31P, 15N, Si29, etc.
- 13C has nuclear spin I=1/2 (same as proton)
- 13C natural abundance is 1.109%. Hence, the probability of having a 13C in a
molecule is 1 (one) 13C in 100 (hundred) 12C.
-The probability of having 2 (two) carbons in a molecule is 0.01 x 0.01 = 0.0001.
So 1 in 10000 12C!!!... That is why you do not see 13C coupled to other 13C atoms.
A fact that makes the 13C NMR spectrum much more simpler.
- Just imagine a proton NMR spectrum if the natural abundance of 13C were 100%.
- The 1H g/2p is 42.5781 MHz/Tesla and the 13C g/2p is 10.71 MHz/Tesla
- Their relative individual receptivity at same abundance is:
1H
g3 / 13C g3 = (42.5781)3/(10.71)3 = 62.83
At natural abundance is 62.83 / 0.01109 = 5666
THE POWER OF
CHEMICAL SHIFT DISPERSION OF
13C NMR
(AS WELL AS OTHER X NUCLEI)
1H
NMR Spectrum of Cholesterol (300 MHz – CDCl3 – 300K)
C27H46O
H-6
5.5
5.0
H-3
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
ppm
1H
NMR Spectrum of Cholesterol (300 MHz – CDCl3 – 300K)
All we can identify at
first glance is:
5 CH3 and 2 CH groups.
Only 17 out of 46
total protons in the
molecule
1.05
1.00
0.95
0.90
0.85
0.80
0.75
0.70
ppm
13C
NMR Spectrum of Cholesterol (75 MHz – CDCl3 – 300K)
CDCl3
C-O
3
2
1
C=C
140
130
120
110
100
90
80
70
60
50
40
30
20
10
ppm
13C
NMR Spectrum of Cholesterol (75 MHz – CDCl3 – 300K)
10
6
4 5
7,8
11 13
9
14
12
42.4
59
58
57
56
55
54
53
52
51
50
49
48
42.3
47
46
ppm
45
44
43
42
41
40
39
38
37
ppm
13C
15,16
NMR Spectrum of Cholesterol (75 MHz – CDCl3 – 300K)
We can count all but two overlapped signals.
Note that the two overlapped signals have twice
the height of the rest.
18
17
32
31
30
29
22
20 21
23 24
19
28
27
26
25
24
23
22
21
26
27
25
20
19
18
17
16
15
14
13
ppm
DMSO-d6
1D 13C NMR, DMSO-d6, 125MHz
Soluble part of 1mg of dye 10 (lots of sample precipitate)
36 hours
S
O
+
N
N
O
O
dye 10
190
180
170
S
160
150
O
140
x64
x64
-
130
120
110
100
90
80
70
60
50
40
30
20
ppm
Dye 10
gp-COSY-90 Magnitude
500 MHz
2 scan per F1 increment
512 increments
20 minutes
Edited HSQC (1H:500MHz, 13C:125MHz)
16 scans per F1 increment
256 increments in F1 (Echo-Antiecho)
90 minutes
SENSITIVITY IN HETERONUCLEAR 2D NMR SPECTROSCOPY
-The major concern for chemist is the horrible low sensitivity of NMR compared to
other spectroscopies
For experiments involving spin ½ nuclei:
N: number of molecules in the observed sample volume
A: a term that represents the abundance of the NMR-active spins involved in
the experiment,
T: temperature
B0: the static magnetic field,
gexc and gobs represent the magnetogyric ratios of the initially excited and
observed spins, respectively
T2*: the effective transverse relaxation time
S: is the total number of accumulated scans
Relative Sensitivity of Different 2D Schemes
P: preparation
E: evolution
M: mixing
D: detection
1D Carbon/DEPT NMR or HSQC/HMQC/HMBC?
Directly observing carbon (13C) at natural abundance does not longer make any sense
if they can be observed through the much more sensitive 1H nuclide using inverse
detection NMR probes.
Just imagine experiments that can combine the sensitivity of 1H with the chemical
shift dispersion of 13C.
Those experiments are basically the HMQC (Heteronuclear Multiple Quantum
Correlation), HSQC (Heteronuclear Single Quantum Correlation) and the HMBC
(Heteronuclear Multiple Bond Correlations)
ISOTOPOMERS AND ISOTOPOLOGUES
According to IUPAC:
Isotopomer
Isomers having the same number of each isotopic atom but differing in their positions.
The term is a contraction of 'isotopic isomer'. Isotopomers can be either constitutional
isomers (e.g. CH2DCH=O and CH3CD=O)
or isotopic stereoisomers [e.g. (R)- and (S)-CH3CHDOH or (Z)- and (E)-CH3CH=CHD].
Isotopologue
A molecular entity that differs only in isotopic composition (number of isotopic
substitutions), e.g. CH4, CH3D, CH2D2.
IUPAC. Compendium of Chemical Terminology, 2nd ed. (the "Gold Book"). Compiled by A. D. McNaught and A. Wilkinson.
Blackwell Scientific Publications, Oxford (1997). XML on-line corrected version: http://goldbook.iupac.org (2006-) created by
M. Nic, J. Jirat, B. Kosata; updates compiled by A. Jenkins. ISBN 0-9678550-9-8. doi:10.1351/goldbook.
1
1
12
12
H H
Cl
C
C
Br
1H-NMR
Cl Br
A
1
1
13
12
98%
C
C
1
12
13
H H
H H
Cl
1
Br
13C
NMR
HSQC/HMQC/HMBC
Cl
C
C
Cl Br
Cl Br
B 1%
C 1%
1
H 1H
Cl
13
C
13
C
Cl Br
D 0.01%
Br
INADEQUATE
Br
HMQC or HSQC?
Lets See
Unwanted
1H-1H
Evolutions in F1
Better Resolution
in F1
At this point there are no questions that HSQC is the best option over HMQC!!!
Now, the next question is: What HSQC experiment should you use?
1) Use a gradient version of the experiment. Gradients can be used to destroy
magnetization that would otherwise lead to artifacts. Hence you will
obtain cleaner spectra.
2) Use a version that includes sensitivity enhancement in the pulse program.
Sensitivity matter, particularly when you have limited amount of sample, or
just want to do it faster.
3) Use the edited version of HSQC. By the same token you will obtain carbon
multiplicity information (CH, CH2, CH3).
4) Finally, use a version that includes adiabatic pulses for 13C inversion and
refocusing. You will remove those very nasty phase artifacts at the resonance
edges of the carbon chemical shifts.
Versions that accounts for all these points are now standard and are available
in the Bruker pulse program library. Setting up this experiments is no longer
problematic. They are routinely run in my laboratory.
Comparison of 12C-1H signal suppression methods used in proton detected
heteronuclear correlation experiments
Clean
12C-1H
(a) Conventional 1D proton spectrum without suppression of the parent
resonance and displays the required 13C satellites.
(b) Phase cycling
(c) Optimised BIRD presaturation
(d) Pulsed field gradients to remove the parent line (Notice loss of sensitivity)
Echo-Antiecho Gradient HSQC
Echo-Antiecho Gradient HSQC with Sensitivity Enhancement (PEP)
x2 Gain in
Senstitivity
Editing based on multiplicity
DEPT-HMQC
D = 1/2J
Edited HSQC
Edited HSQC
CH2 (Black)
CH and CH3 (Red)
Simulation with NMRSim
At 125 MHz in 13C, 200 ppm corresponds to 25kHz
HSQC: Sensitivity Enhancement with Square 13C 180 degree pulses
HSQC: Sensitivity Enhancement with Adiabatic 13C 180 degree pulses
Adiabatic Inversion
Adiabatic Refocusing
Hymenistatin
Heteronuclear Multiple-Bond Correlation Spectroscopy (HMBC)
HMBC is in principle an HMQC experiment tuned for small 1H-13C coupling constants
12a
Magnitude Mode HMBC
with low-pass filter
12b
14
C7
2
3
4
10 9
1
8
5
6
O
H
One-Bond
Cross-Peak
O
15
7
11
13
O
1
C11
Ludartin
C13
12
Through Space 2D Homonuclear Chemical Shift Correlations Experiments
NOE difference
Establishing NOEs and hence spatial proximity between protons. Suitable only for ‘small’ molecules (Mr << 1000), for which NOEs
are positive. Observes steady-state or equilibrium NOEs generated from the saturation of a target.
NOESY(2D or 1D)
Establishing NOEs and hence spatial proximity between protons. Suitable for ‘small’ (Mr << 1000) and large molecules (Mr >
2000) for which NOEs are positive and negative respectively, but may fail for mid-sized molecules (zero NOE). Observes
transient NOEs generated from the inversion of a target. Estimates of internuclear separations can be obtained in favourable
cases.
ROESY(2D or 1D)
Establishing NOEs and hence spatial proximity between protons. Suitable for any molecule but often essential for mid-sized
molecules; NOEs are positive for all molecular sizes. Observes transient NOEs in the rotating-frame, but is prone to
interference from other mechanisms so requires cautious interpretation. Estimates of internuclear separations can be
obtained in favourable cases.
HOESY
Establishing heteronuclear NOEs and hence spatial proximity between different nuclides, for example, 1H–13C. Can provide
useful stereochemical information when homonuclear NOEs are insufficient or inappropriate. Suffers from low sensitivity
but 1H detected variants can help.
EXSY (2D)
Qualitative mapping of exchange pathways in dynamic systems when exchange rates are slow on the NMR chemical shift
timescale, meaning separate resonances are observed for each exchanging species. Quantitative data on exchange kinetics
can be obtained in favorable cases.
NOESY OR ROESY?