AMMRLers,
Thanks to everybody that responded to my question on 59Co chemical shift
broadening in the standard for more than one scan. All I have to say is
enough, enough, I give, I give.
I received 20 responses so far, all saying that it is the temperature
dependence of the chemical shifts.
Probably the most succinct response is the following (from an applications
person for one of the NMR vendors):
59Co has a temperature dependent shift of 1.4 ppm/degree C. This can lead to
line broadening due to temperature gradients in the sample and due to
temperature changes when running longer experiments. Decoupling would
further introduce temperature changes.
Make sure you have very stable temperature conditions and keep the airflow
relatively high to minimize temperature gradients in the probe.
A sample tube that is filled to high can also cause problems, because the
top part of the sample might be reaching into the top of the probe where it
can be cooler. This can lead to convection, another reason for bad
lineshapes. Try a sample with ca 40mm filling height.
Based on what was in Brevard and Granger's "Handbook of High Resolution
Multinuclear NMR" by Wiley (out of print, but used copies are out there:
this book belongs on everybody's bookshelf), the sample was saturated (in
d2o) and the researcher who made up the sample probably used close to a mL
of solution. Using a smaller volume of solution would have helped, as would
have regulating the temperature (VT was off when we ran).
59-Co relaxes quickly and so 1H decoupling is superfluous (not to mention
that the K3Co(CN)6 has no protons - again, I had just started with 13C
parameters changed the observe nucleus. The standard showed a line that was
135 wide at half height, the actual samples had line widths of 250 Hz and
larger.
But considering that the chemical shift range for 59Co has a chemical shift
range of about 20000 ppm, the line broadening observed here was not that
bad. The actual samples had 59Co(III) shifts of about 10000 ppm (~950 kHz)
relative to the standard, so having the standard's resonance go from 135 Hz
to perhaps 200 Hz was the equivalent of having a standard proton line that
is 0.56 Hz wide go to 0.83 Hz wide - noticeable, but not really something
that calls into question the measured chemical shift of the sample.
On reflection, what I find most interesting is how the standard line shows a
distinct, lumpy pattern that was likely due to somewhat uneven power
dissipation by the 1H decoupler into my salty solution. That is, it appears
that one can gather an image of the power dissipation in the sample (or at
least a histogram of it). Somebody could turn this into nice project to
examine decoupler power distribution in various types of probes.
A practical note: When moving the transmitter (o1/tof), be sure to retune
the probe every couple hundred kHz.
Jeff
Jeff Simpson, Director
Department of Chemistry Instrumentation Facility
Massachusetts Institute of Technology
77 Massachusetts Avenue, 18-0090
Cambridge, MA, 02139
617-253-2016 (1806 lab)
_____
From: Jeff Simpson [mailto:jsimpson_at_MIT.EDU]
Sent: Thursday, September 03, 2009 4:09 PM
To: 'AMMRL'
Subject: AMMRL: solution state Cobalt-59
AMMRLers,
I have a researcher that got me to do some 59Co NMR recently, and we
observed a strange line shape for the standard when summing multiple scans.
I believe that the strange broadening I get of the signal from the K3Co(CN)6
is due to a quadrupolar effect, but a quick search of the web only generated
hits for solid state 59Co NMR work, which would obviously give spectra with
none or at least very little motional averaging of the quadrupolar, dipolar
and chemical shift tensors.
So my question is, why would I get a broader line (even with the 2H lock on)
when I scan more than once on my standard in d2o. I am running at about 95
MHz on a Bruker AVANCE (DPX) 400 with a broadband probe, proton decoupling
(I modified the 13C parameters and just used eda to change the transmitter
nucleus).
Even with a single scan the line appears as if there are horrendous z4
shimming problems, but I suspect this is not a shimming problem (beating on
z4 didn't make a hump on the right more anywhere).
Any comments are much appreciated. A summary will be sent out if there is
sufficient interest.
Thanks for your time,
Jeff
Jeff Simpson, Director
Department of Chemistry Instrumentation Facility
Massachusetts Institute of Technology
77 Massachusetts Avenue, 18-0090
Cambridge, MA, 02139
617-253-2016 (1806 lab)
Received on Fri Sep 04 2009 - 04:00:20 MST