What are the best practices for handling NMR spectroscopy samples?

The solvent needs to dissolve the sample. No overlap with needed peak from the solvent. Sample is stable in solvent. Options are D2O, DCCl3, DMSO-d6. Usually DCCl3 works or DMSO-d6 are what I like.

Choosing right solvent is very important. Solubility of the compound in the solvent is the first criteria for preparing the NMR sample. It is also important to note that the solvent peak doesn’t interfere with the compounds peak i.e., If the solvents peak comes in the same region as that of the compound then it can give misleading information while integrating.

Correct solvent choice when preparing NMR samples is of absolute importance. The solution of the compound is also important, however, solvent peak should not overlap with product peak as such it is also important to make sure your sample is dry

I would say use solvent that your compound is readily soluble in. Because higher the concentration better the signal even when you take lower number of scans. You can always use your solvent as a reference standard. Sometimes you can use a drop of deuterated water in your sample to get rid of exchangeable protons like basic or acidic protons.

This as a chemist should be a rule of thumb. Poor quality tubes can lead to instrument damage at the worst and in the least poor resolution. Thus, purchasing of high quality NMR tubes should be done well.

Using a high quality tube will be important. On a 600 MHz I have had issue with low quality tubes. You can have difficulty with shimming.

Tubes should be chosen for the type of analysis. Routine NMR analyses only need screening type tubes, no need to get high concentricity tubes to run an HSQC on a concentrated sample.

Use better quality tubes. It will be a mess if your tube breaks inside the instrument. Clean the tube properly and keep them in oven for long enough to remove the solvents used to clean the tube. For variable temperature NMR measurement, make sure your NMR tube can withstand the changes in temperature, there are specialized NMR tubes for that purpose.

A simple trick that generally works is that the solvent needs to be between 3 and 4 fingers high on the tube. That generally works without measurement.

Quite straightforward. 500 uL for 5mm tubes, 160 uL for 3mm tubes, and 45 uL for 1.7 mm tubes. It is better to have a bit less volume and center the tube in the active coil volume than too much!

Not much to say about this topic. You need 1 to 1.5 inches of solvent. High concentration is useful for carbon NMR to have high signal.

Sample should be prepared with 0.5mL to 0.8mL of solvent for good NMR data. Adding more solvent will dilute the compounds concentration which might result in small peaks and large residual solvent peak or water peak. For carbon NMR high concentration of compound in the sample is necessary. Otherwise it would be difficult to see quarternary peaks for eaxample carbonyl carbon peaks. On the other hand adding too little sample will give improper data. Sample may not run properly as well.

I would say use concentrated sample of proper quantity. Using more solvent would dilute the sample but you have to make sure the sample is enough to cover the measurement area. You have to use depth gauge that helps to determine the volume of sample to be added.

I generally disagree with spinning the sample. As long as the shims are reasonable, there is little improvement in line width but there is a significant risk of spinning side bands that detract from the quality of the spectrum.

Depends on the probe. Newer probes feature longer and more efficient RF coils thus making any benefits of spinning hardly noticeable on a well shimmed sample. Elimination of convection is a primary benefit of spinning but BRUKER’s counter flow VT probe design eliminates convection .

Fully agree with Dimitris and Sean, don't spin sample in modern probes (if you use nitrogen gas for cooling for example). I have never spun a tube in more than 10 years.

I personally have not had issues with different speeds. Some of the experiments require no speed. The default of 20 rpm was used and I did not find issue.

As soon as the data is received, solvent should be removed using rota to prevent decomposition of compounds. It might happen some sample are sensitive to acidic condition, in that case keeping the compound in solvent like CDCl3 might ruin the compound as it is slightly acidic.

I would not store the sample that is needed for very long. I would roto-vap the solvent off and dry under vacuum as soon as possible.

It is important to ensure that tubes are "de-metalized". Metal contamination in tubes can lead to line broadening effects and reduce relaxation times. This is particularly a problem for paramagnetic transition metal ion contaminants. Another consideration is Lanthanide contamination, although rare, could distort any signals arising in coordination spheres which bind calcium. This precaution requires extensive treatment with EDTA, thorough rinsing, careful handling, and that all buffers and solvents are also treated with chelex.