Introducing Three New Pulse Sequence Tools

In many nuclear magnetic resonance (NMR) experiments, we wish to excite only one resonance or peak in the spectrum, corresponding to a specific position within the molecule. Once this resonance has been "selected," we can transfer magnetization via J couplings (through bond) or via NOE (nuclear Overhauser effect) transfer (through space) to other positions nearby in the molecule. The new, transferred magnetization can be observed in the free induction decay (FID) and identified by its chemical shift, allowing us to "connect" or "correlate" two resonances in the spectrum (two different chemical shifts). This process of establishing a relationship (through bond or through space) between two spins is central to all advanced NMR experiments.

So far, the only way we know to select a resonance is through saturation: a long, very low power radio frequency (RF) pulse set to the exact frequency of that resonance (Section 5.9), equalizing the populations and destroying any net magnetization on that proton. This technique is used by the 1D NOE difference experiment (Section 5.12), which allows us to select a single resonance in the spectrum and observe an enhancement of the peak intensities of any resonance corresponding to a proton close in space to the selected proton. New technology in NMR hardware began to be commonly available in the 1990s allowing not just saturation but specific pulse excitation (e.g., 90° pulse or 180° pulse) with specific phase (e.g., x, y', —x', or —y') of any resonance in the spectrum, without affecting the other spins in the molecule in any way. This is much more powerful than saturation because now we can create coherence (90° pulse) on a single spin (a single position within the molecule) or invert (180° pulse) specifically just one resonance in the spectrum.

Another new tool in our arsenal comes from the technology of NMR imaging (MRI). In MRI, there is only one chemical shift (that of H2O) and the chemical shift scale is used

NMR Spectroscopy Explained: Simplified Theory, Applications and Examples for Organic Chemistry and Structural Biology, by Neil E Jacobsen Copyright © 2007 John Wiley & Sons, Inc.

instead for indicating the physical location of a spin within the sample volume. By "bending" the homogeneity of the magnetic field during the acquisition of the FID, the Bo field becomes dependent on the position within the sample. This intentionally nonhomogeneous field is called a "field gradient." Because the NMR resonance frequency is directly proportional to the field strength (vo = yBo/2n), this means that the resonant frequency of each spin now depends only on its position within the sample. The 'HNMR spectrum becomes a physical "map" (or image) of where the spins are located.

In NMR spectroscopy (as opposed to imaging), we do not use field gradients during the acquisition of the FID, but the gradient technology can be used for another purpose: for destroying coherences that we do not want to see. Gradients are applied for a brief period of time (typically 1-2 ms) and then removed, returning the magnetic field to its very high degree of homogeneity. The gradient affects coherence because the precession frequency changes during the gradient (again, vo = yBo/2n), and this makes all the "identical" spins have different precession frequencies depending on their position within the sample tube. The end result after a millisecond or two of this chaos is that the phase of the coherence is now scrambled throughout the sample and no longer adds together to make measurable net magnetization. The technique of "pulsing" the gradient on and then off again rapidly is called "pulsed field gradients" (PFGs), and it has become an integral part of all modern NMR experiments. We can think of PFGs as the janitorial service of NMR, sweeping up and discarding all of the signals we do not want to see (the "artifacts") and leaving the clean spectrum we are interested in.

A third new technique or pulse sequence tool will be introduced in this chapter: the spin lock. A spin lock is just a long RF pulse, applied to give a very large number of rotations of the net magnetization: hundreds or thousands rather than the one-fourth (90° pulse) or half (180° pulse) we usually think of for pulses. Because we can not shim it, the B1 (RF) field is notoriously inhomogeneous compared to the Bo field. This means that depending on where you are in the sample, a well-calibrated 90° pulse might give a rotation of 88° or 91° rather than 90°. This does not create a big problem for most pulse sequences, but imagine what happens if you rotate 100 times (400 times the duration of a 90° pulse): the 88° pulse rotates 97.777 cycles and the 91° pulse rotates 101.111 cycles. The integer number of cycles does not matter much, but the fraction is 280° (0.777 times 360°) for the 88° pulse and 40° for the 91° pulse. Depending on where you are in the sample, you will see every possible rotation of the sample magnetization, and the total net magnetization throughout the sample will be zero. Again, as with the pulsed field gradient, we have destroyed the magnetization by scrambling it as a function of location in space.

But we have been assuming that the sample net magnetization is forming a 90° angle to the Bi field. What if the magnetization is colinear with the Bi field? We have seen with simple pulse rotations that the B1 field has no effect on magnetization that is aligned with it; for example, a 90°y pulse does not change Iy. The same is true for a spin lock: All magnetization perpendicular to the B1 field is scrambled and all magnetization aligned with it is preserved. Not only is this magnetization preserved, but it is also locked to the axis of the B1 field for the duration of this long pulse, preventing it from moving. Some very interesting things happen to this magnetization while it is locked on the B1 axis: Magnetization can transfer from one spin to a nearby spin either by NOE (through space) or by J coupling (through bond). These processes are complex and even with the powerful theoretical tools we have developed, we will only get a glimpse of how they work. But they are the basis of two extremely useful NMR experiments: TOCSY (total correlation spectroscopy) and ROESY (rotating-frame Overhauser effect spectroscopy). We will attempt to at least get a feel for what is going on in the spin lock and how we can get transfer of magnetization.

To understand selective (shaped) pulses and the spin lock, we need to look in detail at the effect of pulses on spins as a function of their resonant frequency, vo, that is to say the position of a resonance within the spectral window.

0 0

Post a comment