This is my first time taking a $\ce{{}^{13}C-NMR}$ spectrum and while I'm aware the signal to noise ratio is much lower compared to $\ce{{}^{1}H-NMR}$, I didn't realize it was quite this bad. I took a spectrum and I only see the $\ce{CDCl3}$ triplet. What concentration range is generally appropriate given say 256 scans (10 minutes on our machines?), or is it rare to see any good results after without drastically increasing the scan time?

Spectrum for reference (taken on a 300 MHz Varian Mercury):
13C NMR Spectrum

  • $\begingroup$ What pulse sequence are you using? Are you sure your delays between scans is sufficient for relaxation? Have you calibrated your pulse power/length to ensure you get 90degree pulses? As its your first 13c spectrum, have you tried using a huge concentration of something else (Nujol, ethanol etc) to check its working? $\endgroup$
    – Nick
    Jun 11, 2012 at 15:06

3 Answers 3


There are several reasons why $\mathrm{^{13}C}$ experiments are far less sensitive than proton experiments.

The natural abundance for $\mathrm{^{13}C}$ is only 1.1%, while $\mathrm{^{1}H}$ has a natural abundance of 99,9%. This difference alone means that if you don't label your sample with $\mathrm{^{13}C}$, the spectra will be around 90 times less sensitive than proton spectra due to the lower natural abundance alone.

$\mathrm{^{13}C}$ also has a four times lower gyromagnetic ration than $\mathrm{^{1}H}$, which also lowers the sensitivity. But the way the experiment is usually measured, you get magnetization transfer from directly-attached protons to the carbons, which increases the sensitivity.

$\mathrm{^{13}C}$ also relaxes more slowly, which means that the time between experiments is longer and you can measure fewer scans than for proton experiments in the same time.

Another factor that is important is which kind of probe your spectrometer is using. For an inverse probe like e.g. a Bruker TXI, the inner coil is the proton coil, which means it's especially suited for proton experiments. For $\mathrm{^{13}C}$ experiments an observe probe like a BBO is more sensitive, as the carbon is measured on the inner coil. If you have multiple spectrometers, you should ask which one is better suited for $\mathrm{^{13}C}$ experiments.

If you have a very small amount of sample, a Shigemi tube allows you to measure in half the sample volume (200-250 $\mu$l) compared to standard tubes, you can achieve a higher concentration that way. They're pretty expensive though and they have to be matched to the solvent, so you would need a CDCl3 tube.

And there are various problems that could drastically reduce your sensitivity for $\mathrm{^{13}C}$, the most important ones if we disregard hardware problems, are proper tuning and matching for $\mathrm{^{13}C}$, and the correct pulse length for $\mathrm{^{13}C}$.

I've avoided your actual question, as the required concentration depends on quite a few factors. I'd recommend you to ask the person responsible for the spectrometers, as they should be able to give you a better estimate based on the actual hardware you have.


At 256 scans you should see something if your compound has reasonable concentration. The rule of thumb I was taught as an undergrad was ~10 mg of compound for a $^{13}$C scan in under a mL of solvent (400-500 uL if you are measuring precisely, modified by probe size and whatnot of course)

As for an exact concentration based on MW? I'm not sure, I've never really worked it out, since the rule of thumb works well enough.

Next question: Does your compound have protons on some of its carbons? Quaternary carbons will be much harder to see, so it is possible that you won't see them after 256 scans. However yes, concentration is probably the problem. I made this mistake yesterday and found that after 256 scans I could see a tiny blot on the baseline.

  • $\begingroup$ Thanks, I tried again with 1000 scans and still nothing. I suppose the concentration is far too low for my compound which is very heavy. $\endgroup$
    – Radu
    Jun 1, 2012 at 22:01
  • 1
    $\begingroup$ With the same sample? Remember that throwing more scans at it increases the resolution on $\sqrt{ns}$ where $ns$ is the number of scans. So you are almost always better off increasing the concentration. So going from 256 to 1024 is only doubling your S:N ratio. More concentration is a far better option. The other thing, is are you sure you are setting things up right? Good shimming, etc? Can you see the 1H spectrum decently? $\endgroup$
    – Canageek
    Jun 1, 2012 at 22:47

You might start seeing signal if you set lb=1 (or, drastically, lb=2) and then type "wft" It won't give you a decent spectrum but you may at least see the beginnings of peaks. Also, are you confident that this is a 1-H decoupled spectrum? If not, then your S/N is going to be very poor.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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