Almost all introductory chapters to correlation NMR spectroscopy, such as HSQC, TOCSY, COSY, ROESY, NOESY suggest that data collections should be done without spinning the tube (except perhaps DOSY at room temperature). Doing VT (variable temperature, i.e. with cooling or heating) 2D NMR, I get always better (up to superior) spectra (less noise and artifacts) in a 400 MHz Bruker probe head with spinning than without. Is there a clear explanation for this? Like bad heat control without spinning or something else?
As far as I know, spinning of NMR sample tubes was used as a way to improve the homogeneity of magnetic field around the sample, which increased the spectral resolution. However, modern NMR magnets very often have sufficiently good characteristics, and therefore spinning may not be necessary (considering also the improvements made in computer-assisted gradient shimming).
Speaking of 2D NMR, early implementations had a tendency to use relatively low resolution compared to 1D spectra, so spinning did not bring any noticeable improvements. Therefore, it was recommended not to spin the sample.
It is interesting, though, to see examples when spinning might indeed help to improve the resolution in 2D NMR. To me it seems to be the issue of magnetic field homogeneity; does anyone have any other explanations ?
For biomolecular NMR, we do not spin samples whether it is 2D or 3D NMR experiments. I agree what is written by Vladmir. New spectrometers do not need spinning to improve sample homogeneity. The shims do not change much over time. We check shims in between long experiments but they are very stable.
Spinning is generally a bad thing for 2D data on modern spectrometers. You can introduce modulation artefacts which can make a bit of a mess of your data.
As Vladimir Khlebnikov mentioned, spinning was originally used to help with poor magnetic field homogeneity but modern systems don't need this. However, it can also be good for poor *sample* homogeneity so we still use it for 1D spectra on open access systems (where samples may be less well-tended than those in the analytical lab)
Dear Imre,
Three remarks:
- I agree with Vladimir: if any spectra are substantially better with spin than without I'd guess there is something wrong with the off-axis shims of your NMR machine.
- We use spinning only for applications where we really have to quantify small J-couplings.
- Whenever you run a sequence where huge signals have to be subtracted by phase-cycling spinning (1D or e.g. HSQC 2D without gradients) is a problem, because it introduces imperfect subtraction. But actually for most state-of-the-art sequences/spectrometers this is no problem anymore since the introduction of gradients.
I hope this helps - kind regards
Alfred
Thank you all for your answers. As I wrote earlier my VT 2D spectra are substantially better with spin than without, despite the books. And just the contrary, I get less artefacts with spin than without. Shimming and resolution is always better with spin than without in 1D, especially in VT, but this is probably not the issue as you don't need 0.1Hz (or higher) resolution in 2D. I should emphasize that I am talking about VT, i.e.cooling or heating and thus also about controlling the temperature. Thus I tend to believe that there is something is wrong with the temperature control in the sample when not spinning causing homogeneity problems as Vladimir suggested. What do you think?
If you are heating or cooling the sample, you may have temperature gradients in your sample and spinning could help minimise them. I would have thought that a long experiment would eliminate these gradients in the steady state.
If spinning works for you, use it. Otherwise, you might want to leave your sample longer in the magnet to allow it to fully equilibrate at the non-ambient temperature. Don't forget that the thermocouple in the probe is only measuring the temperature at a particular point in the probe, (normally close to the bottom of your nmr tube) and not that of your sample. This means that the temperature make look stable on the thermocouple but may not be homogeneous in your sample.
Dear Imre,
all colleagues that point out the existence of a temperature gradients in NMR tubes are of course right. Homogeneous temperaturization of NMR tubes is a art of its own.
What we are doing if we have a suspicion for T-gradients is chemical shift imaging of the NMR sample. Here you see the chemical shift of NMR signals resolved along the z-axis of the NMR tube...results can be, to say the least...surprising.
For high or low temperatures one really has to play with the settings of the temperature unit (which is air-flow, heater current, temperature of the air-flow - our Bruker pre-cooling unit has three levels of cooling). On top of this all details depend strongly on the solvent in use (viscosity, heat-capacity).
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication