# Why is no MS-gas phase NMR coupling possible?

With ever stronger magnets coming out every year, LC-NMR coupling becomes more attractive and is now used frequently in the pharmaceutical industry as well. Considering the rapid progress that is made in that direction I wondered why no MS-NMR coupling has been reported yet, especially as gas phase NMR is possible already. So I wondered: When will it be possible to do MS-NMR coupling? What are the biggest obstacles for achieving MS-NMR coupling?

• Perhaps because the concentration of molecules is very much smaller than in LC. MS is very sensitive and works with very small numbers of molecules. The sensitivity would have to be incredibly high. – matt_black Oct 22 '16 at 12:34
• And why isn't it possible to just do mass spectrometry with a higher number of molecules, so adapt it to higher concentrations? – logical x 2 Oct 22 '16 at 15:16
• How do you keep the charged molecules in your NMR probe for hundreds of millicesconds, so you can record an FID? And why couple the two methods at all? Just run an NMR of the same sample! For separation according to size, use SEC/GPC, those have been coupled with NMR. – Karl Oct 22 '16 at 19:20
• Keeping charged molecules in NMR probe for hundreds of milliseconds $\Rightarrow$ could be possible with quadrupole ion trap (linear ion trap) Why couple them? $\Rightarrow$ because MS is very powerful for large molecules and complex mixtures while NMR is better for small molecules, combination of those would be incredible. – logical x 2 Oct 23 '16 at 8:30
• An ion trap can hardly keep a macroscopic amount of sample, and in MS, there is no macroscopic amount of ions flying. Even with hyperpolarisation (which is likely not possible in an ion trap), NMR does not work with a few 10000 of particles. And "incredible" is a weak point. What actual problem could this solve? You can do HPLC-NMR, SEC-NMR, DOSY, etc. pp. today! – Karl Oct 23 '16 at 13:46

(The magnets are not really getting stronger. The > 20 T magnets are still too big and expensive to be widely used. First 900 MHz was from 2001, the 1GHz Bruker magnet came out five years ago. Since then seemingly nothing... except this 1020 MHz hybrid prototype by JEOL et al. Looks a bit unwieldy. There is a reason it is a hybrid (inner coils made of Bi-2223, i think): The classic superconducting materials (Nb3Sn, NbTi) have an absolute upper limit of ~25 T. HTSC can go higher, but they are very difficult in manufacturing. )

Fundamental problem is T1 relaxation in the gas phase. It is much too slow. You need hyperpolarisation techniques, that you have to interface with the GC. Difficult. (Impossible with mass spectrometry, because there the sample is in vacuum.)

The amount of analyte would be tinytiny. Likely impossibly small for NMR, even with hyperpolarisation. Certainly too small with mass spectrometry.

Other point: Why? GC is not useful for complex molecules, because the have very little vapour pressure. For less complex molecules, GC-MS is fine. There are a number of working NMR techniques to separate different analytes. HPCL-NMR, GPC-NMR, DOSY, etc.

I guess nobody has seen a problem that needs GC-NMR to solve it. Or at least he has not been able to convince anyone it would be worth financing it.

(I initially misread the original question to be about a GC-NMR combination. Hence all the GC in my answer.)

• I see that GC-NMR would not be that much better than LC-NMR. But what about MS-NMR? I mean it would be a very powerful technique for e.g. proteomics. – logical x 2 Oct 21 '16 at 19:34
• MS-NMR? Not sure i understand that. NMR on single molecules? – Karl Oct 21 '16 at 19:44
• MS doesn't consist of single molecules more like $10^6$ molecules something in that range. – logical x 2 Oct 21 '16 at 19:44
• But there will never be more than a few molecules in your probe at a time. Plus, they will never polarise, because they are in vaccuo. – Karl Oct 21 '16 at 19:48