NMR workshop by 2IU1PX


									NMR workshop (Part II)

       USD, summer 2006
Available online at http://www.usd.edu/~gsereda

Literature: Timothy D. W. Claridge “High-Resolution NMR
   Techniques in Organic Chemistry”, 1999, Pergamon
   1. Summary of the RF-spin interaction
   2. Specifics of 13C spectroscopy. Broad band decoupling
   3. Nuclear Overhauser effect
   4. Different types of 13C spectra. Decoupling and NOE
   5. The spin echo pulse sequence
   6. The APT experiment
   7. Polarization transfer
   8. The DEPT experiment
   9. 2D-NMR. Source of the second dimension.
COSY experiment
   10. TOCSY and 1D-TOCSY experiments
   11. NOE and NOESY experiments
   12. Heteronuclear correlation and indirect detection.
HETCOR and HMQC experiments
   13. Solvent suppression
   14. Application of NMR in the student organic chemistry
                       1. Summary of the RF-spin interaction
    Z                  A magnetic nucleus (a spin) rotates around the magnetic field B0
                       with the Larmur frequency . We want to affect the spin with
                       the magnetic component B1 of the radiofrequency (rf).
        B1         X
                       The B1 is too weak to compete with B0.

                       How to eliminate the competition?

1. Apply very strong rf. (Not practical, not selective)
2. Stop rotation of the spin around B0 (How to achieve this?)

a. Remove the magnetic field B0 (Not practical).
b. Rotate with the spin at the   frequency
and apply B1 in the rotating system (rotating frame)
Consequence: The field B1 must oscillate at the  frequency.
Advantage: Selectivity toward spins, rotating at the frequencies, close to .
         B0            There are three possible situations:

              M        1. The rf frequency is exactly o .
                       There is no competition from B0. The spin is affected the most.
        B1         X
                       That is the resonance (NMR)
    What happens during the resonance?
    a. The oscillating rf pushes the rotating spin forward, trying to accelerate it,
    but it is impossible, because  is determined only by the magnetogyric ratio of the
    spin and magnitude of the magnetic field B0. The energy must drain somewhere
    and promotes the spin to a higher energy state (spin flip due to rotation around the
    magnetic field B1).
    b. The phases of the spins are affected. They adjust themselves to the phase of rf
    and become correlated.

    2. The rf frequency is far from o . The field B1 is not competitive with B0.
    The spin is not affected by rf.

    3. The rf frequency is close to o . The spin slowly rotates with respect to the
    oscillating rf and changes its phase. It allows us to play with the phases and acquire
    more information about the system, than just the set of resonance frequencies.
                              Stationary frame, single spin

                    B0                                                         B0
               Z                                                           Z

                         M               Nothing B1 oscillates at 
                                         B1 is too weak                                       X
                   B1          X
         Y                                                             Y


                              Rotating frame ( ), single spin
                    B0                                                         B0
               Z                                                           Z

                         M             B1 oscillates at 

                   B1          X                                                              X
                                                                                    M (Y=0)
         Y                                                             Y

Problem: A single spin does not emit rf (otherwise it would eventually lose energy
and stop rotating, which is impossible) and, consequently, is hard to be sensed and
monitored. Measuring absorbance of rf is hard technically.
                          Rotating frame ( ), many spins

                     B0           M>0 because the populations of the ground       B0
                Z                 and excited states are different            Z

                                            B1 oscillates at                         M

                    B1        X                                                            X

          Y                                                               Y

                              Affecting the total magnetization M:

                                        Spin-flip component

                                  or        Phase correlation component

Note: If Mz is changed after the sampling pulse, it does not affect the signal (too late).
          Different orientation of a spin in the magnetic field may create different
additional magnetic fields at other spins and, consequently change their
Larmur (resonance) frequencies.
          If the Larmur frequency of a spin of interest (A) depends on the
orientation of a neighboring spin (B), it will resonate at slightly different
frequencies (n1 and n2), when the spin B is in the spin state a or in the spin state
b. In the frame of reference, rotating at n0 (center of the doublet), the components
n1 and n2 will rotate (evolve) with the same speed in oposite directions.

                    Doublet (if there is one neighbor B):
          A resonates at n1                       B is in the spin state a
          A resonates at n2                       B is in the spin state b

                         Triplet (if there are two neighbors B):
          A resonates at n1                         B1 is in the spin state a
                                                    B2 is in the spin state a

          A resonates at n2                         B1 is in the spin state a or b
                                                    B2 is in the spin state b or a

          A resonates at n3                         B1 is in the spin state b
                                                    B2 is in the spin state b
     2. Specifics of 13C spectroscopy. Broad band decoupling
a. Low natural abundance of 13C (less than 1%)
Consequence: presence of more than one 13C per molecule is
very unlikely, so no 13C-13C coupling is normally observed
b. Low magnetogyric ratio for 13C, which makes the nucleus less
responsive to the magnetic field
Consequence: the resonance occurs at a lower frequency, than for 1H.
(13C would resonate at 50 MHz, if 1H resonates at 200 MHz in the
same magnetic field). It allows us to observe 1H and 13C spectra
c. Larger relaxation times (> 5 sec) than for 1H (less than 5 sec).
Consequence: Integration of signals requires larger relaxation delays,
which need to be optimized before the experiment
Consequence from a, b, and c: the experiment takes longer (normally
256 scans about 4 sec each vs 8 scans about 2 sec each for 1H) and
requires higher concentration of the material.
                Is it possible to suppress all couplings at once
                          and simplify the spectrum?

  Decoupling with a particular nucleus (intense irradiation at its resonance frequency
  during acquisition) rapidly interchanges its energy levels, averages frequencies of
  the components of the multiplet, which causes the loss of splitting in the multiplet.

  Another effect of coupling is saturation – loss of phase coherence while a- and b-
  populations are equal. Therefore, a saturated resonance produces no signal.

  Depending on the duration of the decoupling pulse, decopling can be selective
  (saturation of one particular resonance) or broad band (saturation of all
  resonances of a particular nuclide, for instance, 1H).

Excitation profiles of two rf pulses of different duration
Homonuclear decoupling – suppression of coupling between the same type of
nuclides (for instance 1H - 1H).
Heteronuclear decoupling – suppression of coupling between different types of
nuclides (for instance 1H – 13C).
Total (broad band) homonuclear decoupling is impossible without loss of the whole
While taking a 13C spectrum we do not care about 1H signals.
Consequence: We can suppress 13C – 1H couplings all at once.
It is called “broad band decoupling” and performed by making the decoupler to
cover the whole 1H frequencies range.
                      3. Nuclear Overhauser effect
Suppose, we have two coupled nuclei – I and S and want to decouple
them by irradiating the system at the resonance frequency of S.
It results in:
a)Disappearance of the signal S
b) Suppression of the coupling between I and S.
c) Change in the intensity of the signal I (“side effect” of decoupling).
The effect c) is called the “Nuclear Overhauser effect” or NOE.
If the intensity I increases, NOE is said to be positive.
If the intensity I decreases, NOE is said to be negative.
The source of NOE – correlated relaxation of both nuclei I and S,
so that the total spin of the system changes by 1 (transition W2, both I and S
increase or decrease their spins when the molecules are tumbling at the frequency
of I plus the frequency of S; characteristic for small, faster molecules),
or does not change at all (transition W0, either I or S increases its spins, and another
nucleus decreases its spin when the molecules are tumbling at the frequency of I
minus the frequency of S; characteristic for large, slower molecules).
Consequences of the relaxation transitions W2 and W0

                           1. a) to b) Populations of the ground and
                           excited states of S become equal.
                           The signal S disappears.

                           2. b) to c) The W2 relaxation increases
                           the population of a ground state of I and
                           decreases population of its excited state.
                           The positive NOE develops
                           (characteristic for small molecules).

                           3. b) to d) The W0 relaxation increases
                           the population of an excited state of I and
                           decreases population of its ground state.
                           The negative NOE develops
                           (characteristic for large molecules).
       4. Different types of 13C spectra. Decoupling and NOE
Two effects of decoupling (decoupling itself and NOE) can be separated,
resulting in different types of 13C spectra.
1. Gated decoupling    The NOE develops before and during the acquisition. The
                       coupling is retained, because the nucleus S does not jump
                       back and forth during the acquisition, so the nucleus I
                       precesses at two frequencies, depending on the spin of S.
                       It produces a coupled spectrum with NOE.

2. Inverse-gated decoupling
                       The NOE develops and affects the longitudinal
                       magnetization, but since acquisition is already in progress,
                       the registered transversal magnetization is not affected, so
                       the NOE does not show up in the spectrum. The coupling
                       is suppressed, because the nucleus S jumps back and forth
                       during the acquisition. The nucleus I precesses at two
                       frequencies, depending on the spin of S, but it is all
                       averaged in the NMR time scale.
                       It produces a decoupled spectrum without NOE.
3. Power-gated decoupling
                     The decoupler is “on” throughout the experiment, so
                     both decoupling and NOE take place.
                     During the lengthy relaxation delay, the decoupler power
                     is lower to avoid overheating the sample and to increase
                     the lifetime of the probe.
                     It produces a decoupled spectrum with NOE
                     (the most popular routine 13C spectrum).

4. No decoupling
                     It produces a coupled 13C spectrum without NOE.
Different types of 13C spectra of isobutyl butyrate
Influence of relaxation delay on the integration of 13C signals
                  5. The spin echo pulse sequence
With respect to the rotating frame, spins with higher Larmor
frequencies rotate forward, and the spins with lower Larmor
frequencies rotate backwards. If we let them rotate for a certain time
t and flip them by 180o about the Y axis, they will refocus after
another period t.
Spins (nucleus A), which are coupled with another spin (X), or in other
words, are components of a multiplet, behave differently, if the spin X
also undergoes a 180o pulse (spin flip). Each component of a multiplet
may become coupled with another spin state of X, change its Larmor
frequency and drift the opposite way. It causes defocusing of
components of multiplets, rather than refocusing.

That is how we can sense if the spin is coupled with another spin.
                        6. The APT experiment
               Evolution of some multiplet components over the time
                   (differences in 13C chemical shifts are ignored)

The components of multiplets drift at different speeds and end up at opposite phases
after the time 1/J. Problem: It works only for carbons with same chemical shifts.
SEFT (Spin-Echo Fourier Transform) sequence
                   1. The decoupler-gated variant
                   Purpose: to refocus chemical shifts, but still distinguish
                   different multiplets (see the previous slide).
                   During the first 1/J period we let the multiplets arrive to
                   opposite phases.
                      During the second 1/J period we stop their evolution
                   by decoupling and refocus carbons with different
                   chemical shift by applying a 180o pulse (spin echo
                   2. The pulsed variant
                       At the end of the first 1/2J period we apply a 180 o
                   pulse on the X-channel to refocus carbons with
                   different chemical shifts.
                   (spin echo sequence)
                     At the same time we apply a 180o pulse on the 1H-
                   channel to keep components of carbon multiplets from
                   refocusing and allow them to arrive to opposite phases
                   at the end of the second 1/2J period.
                   It results in the same situation as on the previous slide,
                   but with opposite phases.
APT (Attached Proton Test) is the decoupler-gated variant of
The SEFT sequence, modified in two ways:

   1. The first pulse is less than 90o to optimize acquisition,
but it leaves –z components.
   2. One more 180o pulse inverts back these undesired components to +z.
                              7. Polarization transfer
Polarization transfer enhances signal intensity by transferring the greater population
differences of high-g spins onto their coupled low-g partners, leading to a signal
enhancement by a factor of ghigh/glow (For the 1H/13C pair – by the factor of about 4).
It can be achieved by selective inversion of population of a high-g component of a
             Before the inversion                       After the inversion

     Problem: How to simultaneously and selectively invert populations
                      of one half of each miltiplet?
                          8. The DEPT experiment
                               The refocused INEPT
        (Insensitive Nuclei Enhanced by Polarization Transfer) sequence
Purpose: generate a greater population difference for the less sensitive nucleus
along the z-axis.

On the first step, polarization is transferred to CH-carbons (doublets).
The spin-echo sequence ensures that the process is independent on
proton chemical shifts. Then a 90o pulse samples the enhanced
population difference of 13C and allows components of all CH-doublets
to refocus. The spin-echo sequence ensures that the process is
independent on 13C chemical shifts.
The farthest signals of the CH3 quadruplet also end up in antiphase, so
only CH-doublets show up in the spectrum.
To detect other multiplets, durations of both steps should be changed.
The DEPT (Distortion Enhancement by Polarization Transfer)
Sequence makes the same effect as the refocused INEPT sequence, but simplifies
editing of the resulting spectrum.
         1. The 90o pulse on the X channel starts together with the 180 o pulse on the
1H channel and triggers coherent evolution of both 1H and X spins when the

components of the multiplet are antiphase.
         2.The q-pulse on the 1H-chanel transfers polarization to the transverse plane
on the X-channel.
         3. Instead of varying the spin-echoes durations, the width of the third pulse
on the proton channel is adjusted to each separately recorded multiplet.
         4. Two 180o pulses serve in two spin-echo sequences, eliminating the
influence of 1H and X chemical shifts.

Due to the polarization transfer, the signal to noise ratio for DEPT-spectra are better than for
           9. 2D-NMR. Source of the second dimension.
                  COSY experiment

To add a second dimension to the frequency domain, we need to add a
variable to the time domain. It is accomplished by running a series of
experiments with a variable time somewhere in the pulse sequence.
           The COSY (Correlation Spectroscopy) sequence

It is a homonuclear analogue of the basic INEPT sequence.

A part of magnetization not transferred to the coupling partner,
precesses with the original resonance frequency n 1 during both t1 and t2
periods. This results in a diagonal peak in the frequency domain
(n 1-n1).

A part of magnetization, transferred to the coupling partner,
precesses with the original resonance frequency n 1 during t1 and with
the partner’s resonance frequency n 2 during t2.
This results in an off-diagonal peak in the frequency domain
(n 1-n2).
            10. TOCSY and 1D-TOCSY experiments
   Spin-echo sequence                    Spin-lock sequence

Spin-lock sequence – is the Spin-echo sequence, applied continuously.
The simplest spin-lock sequence is just a continuous pulse.

  The Spin-lock sequence makes all spins strongly coupled
  (differences in chemical shifts are less than coupling constants)
        TOCSY (Total Correlation Spectroscopy) sequence

In contrast to COSY, the TOCSY sequence has a Spin-lock pulse,
instead of a 90o pulse.
All coupled spins (each spin and its partner) undergo a 180o pulse, so
all multiplet components are not locked and continue to evolve,
propagating further along the chain during the mixing time tm
(tens of milliseconds). All spins become strongly coupled and lose their
identity. It leads to sharing of coherence between all spins and produces
additional (comparing with COSY) off-diagonal peaks between signals,
not coupled directly, but belonging to the same spin system in the
                          1D-TOCSY sequence

This sequence is close to TOCSY sequence, but instead of a 90o pulse,
a particular spin is selectively excited by a field gradient or by one of
numerous sequences for selective excitation. Additionally, there is no
need to array the time between excitation and the spin-lock.

This experiment allows to see how coupling propagates along the chain
of spins until it embraces the whole spin-system. It allows observation
of spin-systems (fragments) of a molecule separately.
    1. Acidic and alcoholic part of esters
    2. A molecule in the presence of impurities
    3. Fragments of individual amino-acids in a molecule of peptide
                  11. NOE and NOESY experiments
The NOE effect, which is observed when saturation of 1H-resonances
changes intensity of coupled 13C signals, is not limited to coupled spins.

NOE may be observed between non-coupled spins, located within 5A
from one another. Hence, it provides significant stereochemical
information. For small molecules NOE is usually positive, for large
molecules – negative. For some molecules it is close to zero.
So, NOE results may tell us “yes” or “may be”, but never “no”.
                       NOE-difference sequence
The steady-state NOE is measured by continuous saturation of the spin of interest.
The transient NOE is initiated by a one-moment population disturbance and
measured by a 2D-experiment, called NOESY (Nuclear Overhauser Effect

The NOESY sequence is similar to COSY sequence, but after the second 90 o pulse,
instead of acquisition we wait for the mixing time tm to let the NOE develop. The
third 90o pulse samples the NOESY affected population by placing the NOESY
component of magnetisation to the transverse plane.

The NOESY component accumulates along the Z-axis, which makes it insensitive to
the rf phase. Therefore, the COSY components can be removed by phase cycling.
         12. Heteronuclear correlation and indirect detection.
                 HETCOR and HMQC experiments

The refocused INEPT sequence (a reminder)         The HETCOR
                                                  (Heteronuclear Correlation)
The INEPT sequence transfers polarization from 1H to 13C and acquires
information on which hydrogens and which carbon are connected.

To plot this information as a spectrum, we just need to add the second dimension
(left of the dotted line) to the refocused INEPT sequence (right of the dotted line)
and a 180o pulse to refocus the couplings of X (effective decoupling during t 1).
  The HMQC (Heteronuclear Multiquantum Coherence) sequence

The first 90o pulse excites protons and their magnetization is transferred
 to carbons by the 90o pulse on the carbon channel.
Then, during the variable time t1, proton and 13C spins evolve coherently
and sampled by the last 90o pulse on the carbon channel.
To remove the influence of proton chemical shifts during t1, a spin-echo
sequence is applied to the proton channel. The acquisition starts after
another period 1/2J to allow components of proton multiplets refocus
(spin-echo sequence, applied to the whole time range).
The extent of coherent evolution depends on the time t1, which brings
correllation between proton and 13C frequencies to the FID.
                     13. Solvent suppression

Before the acquisition the solvent resonance is saturated for the
period d2 (decoupler mode “nyn”).
The standard delay time d1 is not needed
Protons chemically exchangeable with the solvent, are also suppressed.
14. Application of NMR in the student organic chemistry laboratory
         Dehydration of 4-methylpentanol-2 (DEPT spectrum)
                     Selectivity of oxime formation

                                        The reaction is very selective

A pseudo-singlet in the aromatic area
Identification of unknown alcohols and acids by esterification
                  Iodochlorination of styrene

                                 An example of the “AB” system

Practicing the concept of
diastereotopic groups
                     Synthesis of a porphyrin.
Practicing the concept of aromaticity and setup of the tof parameter
Explicit identification of two singlets versus one doublet, using COSY

To top