I would recommend using hmbcgpl2ndqf for "routine" 1H-13C HMBC use, provided you have gradients. This is a simple and robust sequence and contains a 2-step low pass filter for removal of one-bond responses (as opposed to hmbcgplpndqf which contains only a single step filter). Alternatively, hmbcetgpl2nd can lead to phase-sensitive data in F1 (13C) which has the potential for better resolution, but the processing is a little more complex and for most instances of small-molecule work is probably of minimal advantage.

From the above list I would use hmbcetgpl2nd if you have gradients. This is a HMBC with double low pass filter for better suppression of 1J(C,H) and it is a semi phase sensitive experiment resulting in better resolution in the carbon dimension. The processing is done by doing the 2D ft with xfb and then a magnitude calculation along the observed dimension with xf2m. In newer software there is also a version with triple low pass for even better suppression of direct correlations. Write me directly to get a copy.

I usually use hmbcgplpndqf. Has been satisfactory for me. I'll try the IMPACT-HMBC.

A couple of months back, I helped a user here to run hmbc using "hmbcgplpndqf". He is using that since then. No complaints. I guess that works fine for him.

-Ilango.

I routinely use the hmbcgplpndqf. On rare occasions I get better results with the simplified hmbcgpndqf. It seems the low pass filter can now and then decrease the sensitivity and supress some multiple-bounds correlations of interest. I wonder if this has been optimized in the 2-filtering step version mentioned by Tim?

I would be interested also to hear some user feedbacks about the STAR-HMBC (hmbcacbigpl2ndqf). Looks like a useful experiment for many cases of structural elucidation.

I use hmbcgplpndqf as routine for structural elucidation. However, some artifacts have been found in spectra.

Lately I was trying to use hmbcgplpndqf for routine HMBC measurments of metal-organic compounds. In all spectrum I have obtained doublets in F2 dimension. (for common 145 Hz coupling constant) Incresing level of coupling constant erased some of the doublets but a few new ones shows up. Whan I decreased constant a few of doublets was simplified but also a group of higher constant coupled doublet shows up. I think using sequnce with double filter of single bond coupling should be enough. In case of my compounds this doube filter sequence results with enourmous t1 noise, but after removal of it, spectra was quite easy to solution :)