Hi Ben,
To get meaningful integrals (maybe an oxymoron... all NMR integrals should
regarded with suspicion) you need:
--excellent signal-to-noise
--full relaxation ( at least 5 x T1 of slowest relaxing species)
--no NOE... collect 13C or 31P with inverse-gated decoupling where
the acquistion time (decoupler-on) is very short compared to relaxation delay
(decoupler-off) .. or better yet, run coupled (no decoupling at all)
--properly phased
--adequate digital resolution
--good baseline... this is non-trivial
19F typically has wild baselines to contend with, but even
"flat" baselines can be deceptive.
--all baseline correction methods are imperfect and will almost
certainly impart their own distortions to the "true" baseline.
--keep in mind that Lorentzian lines have intensity extending
to infinity.
--any distortion of early part of FID (due to receiver ring-down,
left-shifting, noise, etc etc) will modify integrals.
--IMHO, the best integrations can be obtained by fitting the resonance
lineshape (assuming the lineshape model, e.g. Lorentzian, is valid) with
amplitude, frequency, half-height-linewidth, AND baseline offset as
parameters to be explicitly fitted. This requires non-linear regression
or comparable methods that are not usually present in NMR processing
software. One can get this to work for overlapping resonances, but may
require indidualized optimized treatment for each resonance.... not very
general.
An interesting new variation that I have yet to explore is the new
CRAFT feature in Agilent VnmrJ 4.0 that deconvolutes resonances in the FID
and is claimed to avoid baseline and phase errors.
--However you obtain integrals, I would recommend comparing real and
calculated spectra... and analyzing the residual differences.
--Test your integration protocol on an appropriate pure model compound
with known stoichiometry.
--You will be amazed at how hard it is to get the right answer!
--Did I mention you should never trust any integrals?
Good luck!
--Mike
-----------------------------
Michael Strain
Director, CAMCOR NMR Facility
University of Oregon
Eugene, OR 97403-1253
mstrain_at_uoregon.edu
541-346-4605 office/lab
541-556-4077 mobile
http://nmr.uoregon.edu
On Tue, 3 Dec 2013, bbanks_at_pharmacore.com wrote:
> Folks,
>
> Is it possible to integrate resonances in spectra of nuclei such as 19F,
> 31P and 13C?
>
> Thanks
>
> Ben
>
>
> Benjamin Banks, Ph.D.
> Associate Director, Analytical R&D
> Pharmacore
> 4180 Mendenhall Oaks Pkwy
> High Point, NC 27265
> (336) 841-5250 ext 127 phone
> (336) 841-5251 FAX
> bbanks_at_pharmacore.com
> www.pharmacore.com
>
> This email message (including any attachments) is for the sole use
> of the intended recipient and may contain confidential and privileged
> information. Any unauthorized view, use, disclosure or distribution
> is prohibited. If you are not the intended recipient, please contact the
> sender by reply email and destroy all copies of the original message.
>
Received on Thu Dec 05 2013 - 12:11:16 MST