Fill volume for 1.7mm NMR tubes

The SSPPS NMR Facility routinely uses a 1.7mm micro-cryoprobe that takes capillary NMR tubes with a diameter of 1.7mm instead of the standard 5mm NMR tubes. These capillary tubes require much less sample volume than a standard tube, but exactly how much should be used? And what happens if there is too little? Or too much?

To determine the optimum volume for the 1.7mm NMR tubes a series of 1D ¹H NMR spectra were recorded with different volumes of methanol-d₄. Initially 20μl of methanol-d₄ was added and a spectrum recorded, then solvent was added in 5μl aliquots up to a total volume of 65μl. Spectra were recorded after each addition using a 90^o pulse, 2.0s relaxation delay and a 2.28s acquisition time.

The figure below shows expansions from a stackplot of the ten spectra. The hydroxyl resonance is on the left and the methyl resonance on the right. At the lower volumes both peaks are broad, misshapen and small. Gradient shimming was not possible for the 20μl and 25μl samples. As the volume was increased the peaks sharpened, lineshape improved, and the intensity increased until there was no further improvement with further additions.

At the lower volumes there is insufficient sample to fill the entire volume inside the detection coils in the probe. This results in poor shimming, and thus broad and distorted peaks. It also reduces the intensity of the signal because the number of atoms inside the detection volume is less than the maximum. Once the sample volume is large enough to fill the detection coils then the shimming improves and the signal intensity reaches a maximum. Increasing the sample volume beyond this simply adds solution above the coils, so the shimming does not improve and the signal no longer increases because the extra atoms are outside the detection volume.

One way to determine the minimum required volume is to identify when the shimming no longer improves. Linewidth, the width of a peak at half its maximum height, is a good measure of shimming. The narrower a peak and the smaller its linewidth, the better the shimming. Plotting the linewidth against volume in the figure below we can see that linewidth decreases until 35μl has been added, after which it remains fairly constant.

Another way to determine the minimum required volume is to identify when the signal reaches a maximum. When the signal is maximised the detection volume must be full. Adding further sample will not be useful as it will be outside the detection coils and not contribute to the signal. Peak integrals are a good measure of signal as, unlike peak intensity, the lineshape does not matter. Plotting integrals against volume in the figure below we find that the integrals increase up until the sample volume reaches 35ul, consistent with the value obtained from linewidths.

The Bruker marketing literature and manuals state that the 1.7mm micro cryoprobe needs a sample volume of 30ul which corresponds to a fill height of 23mm. I measured fill heights for all the volumes used and found 30μl corresponds to a fill height of 17mm and a fill height of 23mm corresponds to a volume of 42μl (figure below). My measurements may be a little inaccurate but I don't think they are that far off. It could also be that the Bruker numbers were obtained for aqueous buffer solutions, instead of an organic solvent, and surface tension differences could affect the volume delivered.

The measurements here suggest that 35μl is the minimum required volume for methanol-d₄. I would recommend adding a little more, 40μl, to compensate for sample loss during transfer. Anything more than 50μl is not necessary, and if material is limited, will only dilute your sample.