Thank you AMMRLers,
I have received following replies to my following mail, which i would like to share with you:
i have a question for you all. is the intensity/area of NMR signal (in quantitative mode) dependent only on the number of observed nuclei and independent of the kind of molecule that one is looking into? i think logically this should be the case provided- Inter pulse delay > 5*T1 for the slowest relaxing nucleus observed under the same set of instrumental conditions.
kindly respond.
Avik Mazumder
Gwalior (India)
………………………………………………….Responses………………………………………………………………..
Your recipe is correct for protons, although _only_ for areas. It is in general _not_ true for intensities; intensities only work if all the protons have the same T2 relaxation, and thus all have the same line width.
None of this will be true, however, for a standard 13C setup where NOEs are involved. An INVGATE type of experiment will of course solve that problem. For carbon, people try to use intensities, but again this is
often problematic. Areas (integrals or line shape fitting) are the only accurate way to do quantitation with any nucleus. And for 13C, finding out the "T1 of slowest relaxing carbon" can be difficult.
Note that even for integration of protons, your phrase "independent of the kind of molecule" assumes you have accurate knowledge of the "T1 of slowest relaxing proton". These relaxation values vary considerable,
depending on MW, shape, solvent, temp, etc.
==================================================================================
Even if you assume that the spectrum is a fully relaxed one, the safe bet will be to use the integrated area rather than intensity since we don't know if there is a difference in the T2* for different nuclei (for whatever reason, say, exchange broadening and the like). Only other comment I would like to add is, one has to wait the integrated area with the concentration if you are comparing NMR signals from molecule A and B which are not in 1:1 ratio.
==========================================================================
I think you are correct but I am checking for some authoritative basis for the assumption. I think that NMR is quantitative if all nuclei relax at the same (or almost the same rate). This means that 13C is generally not quantitative but a colleague is doing quantitative 13C by using a relaxation agent which presumably forces all nuclei to relax at a rate mainly determined by the agent. Also nuclei with spin quantum numbers other
than 1/2 are probably not quantitative, but, again, perhaps a relaxation agent helps. All I can say tonight is that there is a lot of literature out there and the consensus seems to be that NMr is a quantitative technique
except for various factors which make it not so ... which is hardly helpful. I will try and help more tomorrow.
================================================================================
That's correct. So it's great for quantitation. One more thing: if the nucleus is C13 (or any other nucleus which is observed while decoupling), you must also ensure that you are acquiring with no NOE.
==========================================================================
Hi Avik,
I am still thinking but there is one clear caveat. It is fine setting the pre-scan (relaxation) delay to 5*T1 but then one must
1. Know the value of T1. One can use the inversion-recovery experiment to discover this.
2. All the observed nuclei in the molecule must have the same T1 or the recovery delay must be set to 5* the largest T1 value. Again the inversion-recovery experiment will quickly show you the longest T1.
=================================================================================
That is a good question. I believe you are right provided the T2 of the sites is not too short. For example, one of the advantages in the solid-state TRAPDOR experiment is the ability to indirectly detect 27Al sites
that cannot be observed due to broad lines induced by its quadrupolar moment.
I would be interested in hearing other responses on this question.
=================================================================================
Yes! We quantify pharmaceuticals against primary standards al the time.
==================================================================================
For liquids what you said is true, provided the T1 and T2 relaxation time is not too fast. For example there is some exchange broadening going on. If T1 or T2 is too fast, a faster digitizing rate etc.. are required. Sometimes spin echo experiments are required to recover the true lineshape/intensity.
==================================================================================
I would agree with your statements. If you have measured T1 for every nucleus in your sample at a particular temperature, and your delay is >5*T1 of the slowest relaxing nucleus in your sample, and you are measuring signal quantitatively (no NOE effects) at that temperature and there are no paramagnetic impurities, then the results should be quantitative (keeping in mind that there is error associated with any measurement).
The issue is that T1 is strongly dependent on "the kind of molecule" one is observing and the experimental conditions including sample preparation, etc. Other issues that can affect signal include spectrometer performance, which in turn can be affected by the temperature of the room, fluctuations in the electrical supply, etc.
================================================================
I guess that if different samples are involved, probe tuning could affect the absolute area. Probably necessary to have an internal standard in those cases. The "same set of experimental conditions" also has to include receiver gain.
Received on Thu Aug 09 2007 - 19:37:29 MST