Friday, June 13, 2008

The Effect of Sample Volume on Deuterium Gradient Shimming

Yesterday's post emphasized that the sample depth with respect to the probe coil has a huge effect on the success of deuterium gradient shimming. For similar reasons, the volume of the sample in the NMR tube also has a huge effect. The shimming routine will work best when the volume of the sample is similar to the volume of the sample used to create the shim map. If the volume deviates too much from that, then the shimming will be unsatisfactory. This is illustrated in the figure below where different volumes of the same sample were run. Each sample was gradient shimmed prior to acquisition.

6 comments:

Anonymous said...

Glenn, this post is quite interesting. I was particularly interested in the result from the last tube in the figure, where the volume used was excessive. I was always under the impression that over-filling the NMR tube would not cause problems with shimming, although I never tried it myself. My thinking was that as long as the sample volume "appeared" as an infinite cylinder to the coil, no problems would occur. As a rough rule of thumb, an infinite cylinder to me was at least twice the coil height.

Your post shows that this thinking is not correct...presumably a function of the gradient shimming experiment. Do you have a feel for why this reasoning is flawed?

Mike

Glenn Facey said...

nmrfreak,

First of all, I should say that these spectra were recorded with the lineshape sample using an 8second acquisition time and that the error in shimming for the overfilled sample would probably not show up as well if "routine" parameters (eg. a 4 second acquisition time) were used. You will also notice that the error in shimming for the overfilled sample is the smallest in the series. In fact, for a routine sample, I don't think that overfilling the tube would cause enugh shimming error to bother the average organic chemist. It could very well be that the sample does not appear as an "infinite cylinder" unless the sample column is 3 or 4 times the length of the coil.

Glenn

Unknown said...

Hello Glenn,

I have a question about under-filled samples. I am to believe the distortion is due to the miniscus at the air-water interface. So, if I were to use slice-selective nmr* to to probe deeper into the sample, would I observe the same distortion?

* z-gradient active during hard pulse excitation to select depth.

Glenn Facey said...

Unknown,
The main magnetic field experienced by an under-filled sample is indeed distorted due to the magnetic susceptibility difference between your sample and the air above it. This field distortion makes it very difficult to shim magnet such that the under-filled sample is in a homogeneous field. The inhomogeneous field leads to broad distorted lines. In principle if the distortion is only along the Z axis, one would be able to obtain sharper lines by looking at only a very thin slice of the sample as you describe. The slice would have to be thin enough so that the field is essentially homogeneous within the slice. This would not be practical though as the SNR would be very low and there would also be some broadening effects due to things like gradient recovery issues.
Glenn

Unknown said...

Ah, I don't mean to be unknown, that will be corrected shortly.

Yes, the magnetic field is diverted by the susceptibility difference in the same way as light is effected by a refractive index. This would mean that a flat surface should have least distortion, correct?

-David

Glenn Facey said...

David,
I'm not sure about the analogy with light and refractive index so cannot comment on whether or not there would be more distortion for a curved meniscus vs a flat surface.
Glenn