I do not have a procedure written up for obtaining _accurate_
integrations using DISNMR. Part of the reason is that through
trial and error, I decided there is no way to use any automation
sequence to get even reasonably good integrations (at least not
with baseline correction). These comments apply when small and/or
broad lines exist in the spectra; automatic procedures seem to
work well for "routine" spectra where only semi-quantitative
integrations are sufficient. Typically I advise the following
procedure for performing integrations (I guess now I'll have a
written procedure :-)
1. Obtain the data with a large sweep width (e.g. 20ppm) to
help reduce baseline rolling in data region. Make absolutely
certain that no signal clipping is occurring (reduce RG by
a factor of 2 if uncertain!). Take one data set with a
relaxation delay 5 times longer to make certain T1 effects
are not affecting results.
2. Take all data if possible at one time using identical (RG, RD,
etc.) settings. If this can be done, then PK will work well
and phasing should not be adjusted differently for different
spectra. Use an automation routine to EM;FT;PK and write out
spectra. If data is taken at different times, phasing should
be performed manually, with care taken to try to get baseline
to have identical slope over region of interest (use some
defined criteria to keep from becoming overly subjective).
3. You should use line broadening (I think) 5 times smaller than
than the line width of the _narrow_ line to preserve quantitative
results; a larger LB will reduce the narrow line integral more
than for the broader line.
4. Baseline corrections always have to be performed, and I
have never gotten consistant autobaseline routines to work
sufficiently well. For a lot of spectra, performing the
manual baseline corrections is the most tedious and subjective
part of the process. Do _not_ rely on slope adjustments to
correct a poor baseline!!
5. I have found that with good baseline correction, the slope
adjustment is straightforward and only has to be done once.
The integrals are not nearly as sensitive to slope as they
are to phasing and baseline correction (at least in my
experience).
I'd very much like to hear if others have had better success with
auto-baseline corrections. We often go to other software for
this type of work, but for small numbers of spectra, DISNMR works
quite well, albeit a bit tediously.
-------------------------------------------------------------
Charlie Fry TEL: 608-262-3182
Director, MR Facility FAX: 608-262-0381
Chem. Dept., Univ. Wisconsin
Madison, WI 53706 USA e-mail: fry@chem.wisc.edu
-------------------------------------------------------------
>Dear Members:
> This inquirey is directed toward members with Bruker AC console systems.
>Is there a standard procedure for obtaining _reproducible_ proton
>integration vaules. A post-doc here has pointed out that in cases where he
>is looking at small broad peaks versus relatively large, narrow peaks he can
>variations of up to 200% by simply re-processing/re-integrating the file
>(not re-acquiring!). In particular, I am looking for "hints" concerning:
>
>a) Using automatic baseline correction/phase corrections;
>b) Proper line-broadening values;
>c) Slope/Bias corrections.
>
>If anyone has a written procedure out there I would really appreciate
>getting an Email copy.
>
>Thanks,
>
>Dr. W. L. Jarrett
>University of Southern Mississippi
>Department of Polymer Science
>E-mail:jarrett@whale.st.usm.edu
>Phone:(601)266-4551
>FAX: (601)266-5504