Dear members,
Thank you all for the responses, especially to Mr. Keifer for your
prescription on this issue.
I sum up here the main points of the replies and also enclose portions of
original messages.
The major impurity in normal grade (HPLC grade) ACN is propionitrile,
CH3CH2CN (Reidel-de Haen's is <0.0002%).
It is not quite visible by UV thus not a big issue for HPLC. It's probably
not quite visible in NMR with few scans, thus maybe ok for flow-NMR. But
'real' LC-NMR experiment does need long hour acquisition. It will gives a
triplet at around 1.3 and one quadruple at 2.36. Mr. Keifer offers a
detailed do-it-yourself method to get ride of CH3CH2CN.
H2O contian in ACN (0.01% for LC-NMR grade, much better grade can be found
from those used for oligo synthesis) is not that an issue, simply because
we are using low quality D2O (99%-99.5% -d) anyway.
Cheers,
Xiaogang
************************
Paul Keifer:
To get any good grade of acetonitrile suitable for LC-NMR, all you
need to do is distill it. The main contaminant in acetonitrile that hurts
1H LC-NMR is propionitrile. The boiling point of propionitrile is so
different from acetonitrile (20-40 degrees C?) that it is a trivial
distillation.
The following is what I passed out to all of the users that I trained
during my LC-NMR years at Varian:
CH3CN often is contaminated with visible amounts (0.1%?) of CH3CH2CN. This
annoying impurity can be eliminated by using CH3CN which has been
"glass-distilled". (Grades of solvent labeled "UV Grade" or "HPLC Grade"
or "Pesticide Analysis Grade" are by themselves insufficient; it needs to
be glass distilled for 1H NMR applications.) We are most satisfied with
glass-distilled Omnisolve CH3CN from EM Science (#AX0142-1).
************************
Clemens Anklin:
the major problem with HPLC grade acetonitrile was contamination with
propionitrile. This gives unwanted signals in the NMR spectrum but is of
no concern to HPLC alone. The higher water content is acceptable since
in most experiments H2O/ACN solvant gradients are run anyway.
If your Fisher brand ACN is consistently clean enough, then go ahead and
use it. I would run more than 16 scans to test. Often 1D experiments
with more than 128 scans are performed to observe minor components in a
separation.
************************
Andrew J. Jensen:
The problem commonly encountered with non LC grade acetonitrile is the
presence of a propionitrile impurity in overnight/weekend runs, which is
normally the case. If you can obtain a decent LC-NMR spectrum in 16 scans,
that is not what LC-NMR is intended for, atleast not in industrial
applications.
************************
Ian Marshall:
The usual contaminant in Acetonitrile is proprionitrile (which is not
detected by the routine UV purity tests). Reidel-de Haen LC-NMR grade is
'low in proprionitrile' and as LC-NMR work is done with D2O as well then a
small increase in H2O is acceptable
Having leapt to Reidel-de Haen's defence, I will now add that I don't use
it
because it cost more than the reduction in proprionitrile is worth. I've
found that Fluka Acetonitrile for Residue Analysis is almost as good as
costs about the same for 2.5l as 500mls of Reidel-de Haen LC-NMR grade.
************************
Roger Kautz:
We use normal HPLC grade ACN. When forced to do trace analysis I find
it's important to do a solvent blank as closely matched to the sample as
possible; for LC-NMR that may mean a blank injection.
************************
Wei Li:
We have been using Fisher HPLC for our flow NMR on our Varian VAST
system without any problem. I don't see much reason for an expensive grade.
This archive was generated by hypermail 2b29 : Tue Dec 17 2002 - 01:04:24 PST