Monday, April 18, 2011

FT NMR Spectra Without Pulses

An FT NMR spectrum is obtained by applying a pulse at the Larmor frequency to a sample in a magnetic field. The precession of the spins induces a voltage in the receiver coil which is recorded as a function of time. The Fourier transform of the time dependent signal is the NMR spectrum. What happens if you do not provide any pulses? You might think that you would not observe a signal - but this is not the case. Even without any pulses there is sufficient noise present to allow incoherent precession of the nuclear spins. This precession can be measured and indeed produce an NMR spectrum. This is demonstrated in the figure below. The bottom trace shows a conventional 300 MHz 1H NMR spectrum of ethyl acetate collected with one scan using a 30° pulse. The top trace was collected on the same sample by adding 256 single scan magnitude spectra using no pulses whatsoever. Although very weak, one can clearly see the NMR spectrum of ethyl acetate.

12 comments:

Unknown said...

thanks, could I say that this 'sufficient noise' come from other spectrometer pulse or outside, because the noise has not define frequency?

Glenn Facey said...

Mustafa,

The noise is not coming from the spectrometer. There is no pulse and no power. The sample is sitting in the tuned circuit of the probe and the spin noise emitted from the sample is monitored. I should emphasise that the NMR signals are of random phase. When all 256 spectra are added together, there is no resultant signal. The spectrum plotted is the result of adding 256 magnitude spectra.

Stan has a nice post about this subject in his BLOG with references. You can find it here:

http://www.ebyte.it/stan/blog08a.html#08May27

Glenn

Anonymous said...

I don't get it.
So what is the source of the magnetic wave, that flips the magnetization vector randomly?
The spins of neighbouring nuclei?

Glenn Facey said...

Anonymous,
Thank you for the question. It is not an extenal source of radiation that "flips the magnetization vector" but rather a random speontaneous emission from the nuclei.

Stan has a nice post about this subject in his BLOG with references. You can find it here:

http://www.ebyte.it/stan/blog08a.html#08May27

Glenn

Anonymous said...

Random emissions would not be in phase so no FID should be seen even for one scan.
Or is the claim that you are seeing a series of simulated emission caused by a random emission event?

Glenn Facey said...

Anonymous,
I do not know the answer to your question.
Glenn

Anonymous said...

Would turning on the receiver applies a pulse in the coil? Thanks.

Glenn Facey said...

Anonymous,

Thank you for your question. Turning on the receiver does not put any power into the sample coil.

Glenn

Unknown said...

So, here you collect the total spectrum as the result of 256 individual scans and then sum the magnitude spectra (i.e. |x|), because while for random data the sum of x = 0, the sum of |x| is nonzero, in fact, and that is what you see.

Glenn Facey said...

Ricardo,

Correct. The NMR signals are of random phase. When all 256 spectra are added together, there is no resultant signal. The spectrum plotted is the result of adding 256 magnitude spectra.

Glenn

David Rovnyak said...

Great spectra- do you know if you or others have looked at this with (i) cryoprobe vs RT or (ii) amplifier blanked vs unblanked? -David

Glenn Facey said...

Hi David,
I have observed spin noise spectra on a cryoprobe. See this link:
http://u-of-o-nmr-facility.blogspot.com/2017/09/optimizing-signal-to-noise-ratio-with.html
I have not compared the spectra to that of a RT probe. I have not compared spectra with the amplifier blanked vs unblanked but I have observed spin noise spectra using 0 Watts of power with the output cable disconnected from the amplifier entirely.

Glenn