Presaturation Of Solvent Resonance

Often the solvent gives rise to a very large peak in the spectrum that interferes with observation of the solute peaks. For example, many biological samples are run in 90% H20/10% D20 so that exchangeable peptide amide and nucleotide base N-H resonances can be observed (the 10% D20 is for locking). In this case the water protons are present at a concentration of about 100 M (55 M x 2 x 0.90) and the solute may be at 1 mM or less—a factor of 105 difference in concentration. Without some strategy to suppress the solvent signal, the solute peaks in the spectrum will be very difficult to observe. The receiver gain would have to be turned down dramatically so that the FID, which is dominated by the water signal, will not overflow the digitizer. At a low receiver gain setting (low amplification of the FID signal) the signal-to-noise ratio of the solute peaks is greatly reduced. Furthermore, the solute portion of the FID can be lost in the accuracy limits of the digitizer because the solvent signal is now filling the range of the digitizer. For example, with a 12-bit digitizer a signal that is 0.024% (100%/212) or less of the maximum signal cannot be represented in digital form because the entire signal is less than the least-significant digit (bit) of the digitizer. A 1 mM solution in 90% H20 corresponds to a 0.001% solute signal, or 24 times smaller than the smallest digital "currency" of the digitizer, if the water signal is just filling the digitizer. A 16-bit digitizer improves the situation, but still provides only two thirds of a bit to digitize the solute signal! Without active suppression of the water peak, you will not see the solute at all.

There are many methods for suppressing a strong solvent signal, but we will consider here only the simplest: presaturation. This technique involves irradiating at the precise solvent frequency with a long (~ 1 s) low-power signal to saturate (equalize populations of) the solvent protons. Then a normal high-power (nonselective or "hard") pulse is immediately delivered to excite the solute nuclei and obtain an FID. The solvent protons have no population difference (Mz = 0) and no coherence (Mx = My = 0) at the time of the high-power pulse and therefore produce no magnetization in the x-y plane and no signal in the FID.

Presaturation (as well as homonuclear decoupling and NOE difference) requires that you have two power levels of RF available at the proton frequency: high power (e.g., 30-50 W) for the 90° nonselective "hard" pulse, and low power (< 1 W) for continuous irradiation. In newer models of spectrometers, this is done simply by switching the power level (attenuation) of a single RF source from low power (during presaturation) to high power (during the "read" pulse). Older spectrometers cannot switch power levels rapidly, and some even use mechanical relays for power switching. The repeated switching of a relay with every scan would burn it out in the course of a few experiments. For this reason, the traditional way of presaturating the solvent resonance involves using the proton decoupler (which produces low power RF at the XH frequency during 13 C experiments) to deliver the low-power irradiation, and the broadband "transmitter" (the usual source of high-power pulses in both *H and 13 C experiments) to generate the "read" pulse. These two signals are electronically combined and delivered to the probehead. This allows the frequency for presaturation to be set independently from the observe frequency (the pulse and reference frequency—center of the spectral window). In a practical sense, this means that the solvent resonance does not have to be placed at the center of the spectral window. The

Decoupler frequency n

Figure 5.19

same hardware setup is used for NOE difference and homonuclear decoupling experiments. Current spectrometers can switch both power level and frequency of the channel in a few |xs or less without mechanical relays, so there really are no longer any hardware issues for CW irradiation experiments.

Presaturation is very precise and cuts a razor-sharp swath out of the spectrum, but there will be some attenuation of peaks near the solvent peak. For 90% H2O (10% D2O) samples typically used in biological NMR, the method is very demanding because if the shimming is not perfect, there will be broad components at the base of the water peak and these will not be removed by the narrow slice of the presaturation. This leads to broad "humps" at the water frequency that can still be very large compared to the solute peaks, usually caused by higher order shim problems (Z4, Z5, andZ6: see Chapter 3, Fig. 3.5). With very good shimming, the water signal appears as a sharp, negative dip between two humps (Fig. 5.19). Since even with the best shimming the FID is dominated by the water signal, the quality of water suppression can be measured by how high the receiver gain can be increased without overloading the digitizer. Better water suppression means a less intense FID signal in the probe coil, and more amplification is possible before reaching the digitizer. Good presaturation should allow a receiver gain of 64 (Bruker) or 22 (Varian). Solute signals, such as amide HN protons that exchange with water protons, can be strongly attenuated ("bleached") because these protons spend some of their time at the water resonance. Depending on the rate of exchange with water, protons bound to nitrogen may be slightly attenuated or completely wiped out. For this reason, solvent saturation methods involving gradients and selective pulses (e.g., Watergate, Chapter 8) have largely replaced "presat." It is still useful as a very demanding test of the experimental setup: if the presat spectrum looks good, the shimming must be excellent!

Figure 5.20 shows the 500 MHz presaturation 1H spectrum of a cyclic octapeptide in 90% H2O/10% D2O. H2O is used so that the exchangeable amide protons (CO-NH) can be observed; in D2O they would be replaced with deuterium. The tiny remaining H2O peak is seen as a sharp spike at 4.6 ppm with a broad signal around it. The tryptophane NH is seen as a singlet at 10 ppm, and the region from 7.5-9.0 ppm contains most of the amide NH resonances, as well as the histidine NH at 8.4 ppm. The aromatic protons fall in the region of 7.0-7.5 ppm, and the eight CaH protons and the threonine C^H (CH(OH)CH3) are near the water at 3.5-4.5 ppm. The 2.2-3.5 ppm region contains the other C^H protons and the lysine QH (CH2-NH3+), and the farthest upfield region at 0.5-1.5 ppm has the CYH protons and the methyl groups. The CaH protons near the water are somewhat attenuated, as are some of the amide NH protons that exchange more rapidly with water.

0 0

Post a comment