You can buy NMR tubes in a huge variety of qualities, with an equally huge difference in price between the cheapest and the most expensive NMR tubes. They are usually rated for a specific spectrometer frequency, e.g. 300 MHz+ or 600 MHz+.

What is the difference between these different grades of NMR tubes? And how large is the effect on the quality of the spectra if a lower-grade tube is used?

  • $\begingroup$ Bad NMR tubes can also have curves in them, which can bring them into contact with the probe, damaging them. This is usually a result of heating the tube, rather then it being purchased through; I can give details on how/why this happens and how to avoid it if people want. $\endgroup$
    – Canageek
    Commented May 25, 2012 at 21:31

3 Answers 3


A general rule is: crap goes in, crap comes out. A large-sample low-field 1D NMR at room temperature is usually only minimally affected by using a cheap NMR tube. There are important differences though and I’ll highlight a few.

The first distinction between prices is what the tube is constructed from: quartz obviously costs more than borosilicate. Why would a chemist ever use the more expensive quartz? You can heat/cool quartz faster (nice for thermal studies), the UV cutoff is lower (think 190 nm opposed to 320 nm) which is important for photolysis, you can work with quartz at higher temperatures (around 1300 °C instead of 250 °C), and the purity of quartz is better controlled than your typical Pyrex. There are different grades of quartz, fused and synthetic, and there are different grades of borosilicate, such as the high-quality Pyrex or the lower-quality Class B, each comes with its own limitations as far as purities are concerned and so forth.

Three more important parameters have to do with the manufacturing of your tube are: concentricity, camber and wall thickness. Lower quality tubes will tend to have less precision and accuracy over each of these parameters and as a result your sample may wobble while spinning (introducing problems such as modulation sidebands). A particularly bad tube can hit your RF coils and cause damage to your probe over time slowly or quickly if it is ignoring any reasonable standard – even more apparent for a tube at this level of “quality” is that it may be easier to break while acquiring your sample and we all should be aware of how much fun that is for everybody involved.

Shimming can deal with impurities present in the glass (such as ferric oxide) and increased impurities in the glass/inhomogeneities will result in taking longer to get a good shim. Time is money.

A lot of these things have lower tolerances in more complex experiments and at higher fields. It really does depend on your particular experiment and what you’re hoping to get out of it.

  • 1
    $\begingroup$ Really nice answer. I was wondering if you could comment on whether or not it is possible for adsorbates to quasi-permanently contaminate tubes and whether tubes exist with surface coatings for reactive species? (just casual interest) $\endgroup$ Commented May 2, 2012 at 10:35
  • 1
    $\begingroup$ It's quite possible and is something to seriously consider when doing trace analysis, for example iron, silver, indium, uranium, cobalt, rubidium will all adsorb into pyrex in varying quantities. These are usually ppm changes though and I'd be surprised if NMR were sensitive enough. For cases where glass is incompatible, go to fluoropolymers. $\endgroup$
    – Chris
    Commented May 4, 2012 at 17:55
  • $\begingroup$ I do not agree with your statement that shimming can deal with paramagnetic impurities in the glass. These impurities would create strong local gradients, which will be impossible to shim out. Furthermore using a modern gradient shimming system the length of the shimming routine is not as severely affected. $\endgroup$ Commented May 21, 2012 at 20:56
  • 1
    $\begingroup$ the overall shimming quality is certainly reduced with an increasing impurity of paramagnetic species, but I expect that the impurities of interest are not aggregated and largely homogeneously distributed throughout the NMR tube. I'm not an expert in NMR though. I've used these newer techniques for acquiring 1D 1H NMR spectra before, but the synthetic chemists I know (many are inorganic) stay away from these auto-shimming techniques. $\endgroup$
    – Chris
    Commented May 24, 2012 at 7:46

There are additional aspects concerning NMR tubes. Besides their precision of wall thickness and concentricity the material itself can be of great importance.

1H and 13C-NMR

Borsilicate (pyrex) glass will be fine for standard applications. For high throughput measurements these NMR-Tubes rated for 400 MHz can be disposed after usage avoiding any cross contamination. These tubes have de facto no background signal for these nuclei. Good quality tubes rated at 400 MHz can be used at higher field machines, too, however one has to be aware that some degradation in signal quality can occur. This has to be cross checked with higher rated tubes.


In this case both borsilicate and quartz tubes will produce a glass hump, a broad signal resonating at the position where typically Q-groups are found. A possible solution is the usage of PTFE (teflon) liners, which are tubes made of plastic. However one has to be aware that the probe head itself contains quartz tubes which are built in to support the RF coils. Only changing the tube material could still lead to a significant background signal.

In addition special care has to be taken concerning gradient-equipped probes as the gradient assembly can contain silicon as well.

Changing the NMR-tube can help, however it is worthwhile to measure just the empty probe to get an impression how big the background of the tube is.

There are spectroscopic means to at least partially suppress these background signals like relaxation filters or using a DEPTH sequence.

  • 1
    $\begingroup$ And for some reason if you need to do boron NMR, then using the quartz tubes is a good idea... chem.ch.huji.ac.il/nmr/techniques/1d/row2/b.html $\endgroup$
    – Curt F.
    Commented May 6, 2016 at 15:04
  • $\begingroup$ The coil on the probe is usually fixed on a tube of borosilicate glass (duran/pyrex), which will show up in your spectrum as well. Bruker will happily sell you a boron-free probe, however. ;) $\endgroup$
    – Karl
    Commented May 6, 2016 at 16:04
  • $\begingroup$ According to Bruker the coil support is made of Quartz. At least at the recent probes. $\endgroup$ Commented May 6, 2016 at 16:59

High-priced tubes are guaranteed to be cylindrically symmetric. This is very important for the shimming, because otherwise the susceptibility differences between air, glass and solvent will make local gradients* that are hard to shim out (i.e. you need a thorough re-shim for every sample).

This used to be partially counteracted by spinning, but modern probes don't spin the sample any more, plus that gave you spinning sidebands.

The high-priced "high frequency" tubes are also supposed to be made of higher rated glass, which is more homogeneous and has low and constant susceptibility. Not sure if that's true or if it really makes a difference.

Don't dry your expensive tubes in the oven standing tilted in a beaker at 200°C! The thin glass bends very quickly.

(*If your sample tube is not overfilled and located at the correct height in the probe, you will get very bad suceptibility effects from the upper and lower end of the sample. See e.g. Olivier Buu, J Mag Res 257, 2015)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.