Pure shift NMR - Broad band homodecoupled HSQC

In most 2D NMR spectra the position of the peaks, the chemical shift, is the most important piece of information. If splitting of the peaks by scalar coupling is observed, it is usually ignored. In fact, in many cases scalar coupling in 2D experiments can be detrimental. The splitting increases the chance of overlap and makes it harder to determine accurately the chemical shift. Pure shift experiments, in which the scalar coupling is removed, do not have these problems and are also more sensitive because the peaks are narrower and taller. Heteronuclear 2D experiments, such as HSQC, can be converted to pure shift experiments without loss of sensitivity by applying broad band homo decoupling during acquisition.

The figure below shows an expansion of the aromatic resonances of ethyl benzene. In red is a standard HSQC and in blue is a broad band homo decoupled (BBHD) HSQC plotted with the same contour levels. Both experiments were recorded with the same number of scans and t₁ increments and processed in the same way. The inset shows ¹H projections of the aromatic region with the standard experiment in red and the pure shift experiment in blue.

The overlaid 2D spectra show how the BBHD has reduced the multiplets to single peaks without affecting their chemical shift. Clearly, BBHD improves resolution. In overcrowded spectra a BBHD HSQC makes it much easier to count the number of resonances and to determine their chemical shifts. The projections in the inset also demonstrate how collapsing the multiplets increases the signal to noise.

To implement BBHD a technique similar to one used in the 1D pure shift PSYCHE experiment is used. In the PSYCHE experiment a series of FIDs, each starting at a different time, is collected and a portion from each is used to stitch together a complete FID in which the scalar coupling has not been able to evolve. To apply BBHD, recording of the FID is repeatedly interrupted to apply a BIRD filter (a combination of ¹H and ¹³C pulses) which removes the ¹H-¹H coupling. During the periods that the BIRD filter is applied the receiver is switched off and the FID is not recorded. This produces "chunks" of the FID in which there is no scalar coupling. Typically, somewhere between 8 and 32 chunks are recorded. Concatenating the chunks gives an FID, albeit a slightly shorter one, which can be processed in the normal way.

 

A BIRD filter selects protons bound to ¹²C or ¹³C. In this experiment it is used to select ¹³C-bound protons. At natural isotopic abundance it is unlikely that adjacent carbons are both ¹³C nuclei and so the protons are effectively decoupled. It should be noted that the BIRD filter cannot decouple geminal protons since they are attached to the same carbon.

Using the BIRD filter imposes a sensitivity loss equal to the natural abundance of the heteroatom, but in a HSQC selection for ¹³C already takes place and so there is no further sensitivity loss. BBHD also has the advantage that all the information is recorded in one FID, so a BBHD HSQC takes no longer than a standard HSQC. One disadvantage of the BBHD HSQC is that the FID becomes shorter, since parts of it are not recorded, and this leads to slightly broader resonances. Another disadvantage is that stitching the chunks together leads to discontinuities or "steps" in the FID that produce artifacts after fourier transformation to give the spectrum. These artifacts appear as noise spikes at regular intervals around the true peaks. The spacing of the spikes is a function of the size of the chunks. Methods to minimize the noise spikes by randomly varying the size of the chunks have been reported, but I have not had much success with those. Nevertheless, the BBHD HSQC sequence is available on both of the Facility spectrometers and should prove useful for simplifying crowded spectra.

References
1. K. Zangger
"Pure Shift NMR"
Prog. NMR. Spec. 2015 86-87:1-20.

2. L. Castanar and T. Parella
"Broadband ¹H homodecoupled NMR experiments: recent developments, methods and applications"
Magn. Reson. Chem. 2015 53(6):399-426.

3. J. Mauhart, S. Glanzer, P. Sakhaii, W. Bermel, K. Zangger
"Faster and cleaner real-time pure shift NMR experiments"
J. Magn. Reson. 2015 259:207-215.