Hi all,
a few days I made a short report concerning unexpected
frequency shifts of about 60 Hz after upgrading to XWIN-NMR
for Linux.
Meanwhile I have to correct and to extend this report.
First the correction: with the exception of a tiny problem with
DQD detection, XWIN-NMR for Linux has nothing to do
with the problem described. The real reason seems to be a new
cryo probe. We made two changes simultaneously and weren't
able to address the problem to the correct origin.
Now the extensions. Meanwhile we made a few experiments
and all we observed is very crazy and we are not able to
explain any observation.
The experimental contants
- a 2 mMol sucrose sample in 90% H2O/10%D2O
- a sample of 50% H2O and 50%D2O
- a sample of D2O with traces of HOD
- a 600 MHz spectrometer with a conventional probe
- a 600 MHz spectrometer with a cryo probe
- nuclei table as delivered by Bruker, SR set to 0
- cable connection to the transmitter for shielded gradients
opened (with the exception of watergate experiments)
Now the first experiments
We took all three samples mentioned above and put them
into the conventional probe. We irradiated at O1=4.7 ppm
and observed the water resonance at 4.7 ppm for all three
samples. Water suppression experiments worked well, if we
assumed the centre of the spectrum at 4.7 ppm.
We took the same three sample and put them into the
cryo probe. We irradiated at 4.7 ppm and observed the
water signal in the D2O sample at 4.7 ppm for the 50/50
sample at 4.75 ppm and for the sucrose sample at 4.8 ppm.
Water suppression experiments succeeded if we assumed
the water resonance at 4.7 ppm for all three samples and
failed, if we took the above mentioned frequencies of 4.75
and 4.8 ppm. This is valid both for presat and watergate.
Now some more details using the sucrose sample only.
We changed pulse length and power we mismatched the
probe. Always the observed signal appeared at 4.8 ppm
of course with different intensity.
We made some presat experiments using different presaturation
power levels (at 4.7 ppm). With a presat power of about 90 dB
a first slight change of the 4.8 ppm signal towards 4.7 ppm
was observable. Using 80 dB the resonance was located at
4.78 ppm, using 70 dB the resonance moved to 4.75 ppm
and with 50 dB we had prefect water suppression. During all of
these experiments the positions of the sucrose signals remained
unchanged.
That's all we know so far. Apparently the signal shift is a function
of the concentration. Furthermore the sum of z and x,y magnetization
seems to play a role (independence from the pulse length).
If one destroys z-magnetization (presat) or x,y magnetization
(watergate) the frequency shift vanishes.
All experiments work well, if we set the frequency acording the
assumption where the water signal should be. The experiments
fail if we adjust the water suppression according to the
water signal observation. With other words, all of the old data
sets continue to work with slight pulse length changes.
As already said, we have no explanation. The only weak idea:
B1 and B0 field are not strictly perpendicular by one another.
Could this cause the described effects? And comment would
be helpful.
Rainer
(TU Munich)
Received on Wed Jul 24 2002 - 11:12:55 MST