Summary of answers to the problem of Broad base on Lineshape Problem
We have a Varian Inova 500MHz NMR spectometers with a pfg triple resonance probe.
Normally we have obtained lineshape well within specs, but in a 3 week
period we have noticed an increase in the base of the lineshape peak.
(specs non-spin 0.65/6.00/12.00, our current specs 0.50/29/98). It is
still nicely symetrical but a very broad base. Probe placement is fixed.
What we have tried:
Shimming on the X3,Y3,Z3X,Z3Y shims
Setting all x,y shims to 0 and shimming
Automatic FID shim for X,Y shims
Cleaning magnet bore in case of contamination
Answers - Thanks to all
1. Maybe the inner quartz dewar within your probe is broken.
To check this out, remove your probe, turn it by 180 degree
and shake it a little bit.
2. I had a curious homogeneity problem on a Varian system awhile back, and
found a rather embarrassing problem to be the cause. Varian is inclined to
mount their depth gauge on the side of the console with cable ties. A user
apparently pushed too hard on it and the bottom plunger was pushed up, so
that samples were improperly gauged. Simply resetting the depth gauge fixed
our problem.....
3. How about cleaning the probe too? I use 7" wooden q-tips with ethanol.
If it's a 5mm probe, you might have to roll the q-tip with your fingers
to make it small enough to fit inside. Just a small bit of invisible material in
there can affect lineshape. If the first q-tip comes out brownish, then
most likely dirt in the insert is causing this. (In the US a q-tip is
also called an applicator...a small cotton ball glued to a stick.) If you have put
in any aqueous samples, a cleaning with H2O on the q-tip followed by
ethanol could be necessary. I like to avoid using CHCl3 and acetone, if possible.
4. We have seen similar problems on occasion. The most common fix is
carefully cleaning the probe, especially the glass components in the
sample region. Both sample breaks with resulting sample spill (this
will show broad "stuff" in other areas of the spectrum), and dust (which
can simply degrade line shape) have caused problems for us with line
shape. A glass break will also cause similar, although typically more
substantial, problems.
It is possible you've had a lead on the RT stack break; check all the
shims and make sure they have an effect on the lock signal. The last,
although again usually more substantive, issues could be a cryoshim
quench. These can be had to determine.
Likely your problem is a dirty probe. I'd start there.
5. I had a similar problem with my wide bore (89mm) VXR-400S. We were
using narrow bore (55mm) probes. As it turned out, I found a paper clip
inside the magnet bore sticking to the shim stack. After I removed the
paper clip and reshimmed everthing was back to normal.
6. Some of the following thoughts are basically stream of consciousness
from the so please take no offense;
Bad shims are bad shims no matter how they got there and it's hard to
tell what you have because you are so far off spec'. Most NMR facility
personnel don't have the skills to shim from zero to lineshape spec's in
any time frame under days. That's why we pay installers who do it
everyday. As you know, these are high order skills that require a great
deal of knowledge, concentration and a willingness to sit in one place
for up to days (I've seen weeks) at a time.
Has anything changed? New patch, He boil off, Lockfreq(Z0), a He fill
incident. Believe it or not, conservative filling can quench cryoshims.
I'm afraid that the magnet might have quenched or is quenching a
shimcoil. If you do find lineshape you might have to push the RT shims
too far. One way to tell without an installer or trained magnet man is
to shim it to spec and check the values. If the shim values for any
shim are very different and can not be moved (z1 ->z1C, etc.) then you
might have lost a cryoshim.
Perform an extremely thorough inspection of everything. You'd find
other problems and want to fix those later.
Check User logs. Who was the last user with reasonable lineshape? Ask
them what happened that was different.
Check flow meter logs and cryogen charts. Is the magnet loosing He
faster lately?
Do you use gmapsys? There are many PFG tricks to make a map better.
Begin shimming from the last best shim set. Make PFG shim map through
z6. Check it visually, if it is symmetrical without too many aberrant
points, run a PFG autoshim with gzsize=2. It should easily converge to
solution. If not the map is bad or something has really pushed the
magnet field out of shape. Get it to converge even if it is an ugly 7
iterations.
Exit gmapsys, leave spin off, open ACQI and adjust Z1C, Z2C for best
lock level moving fine current to course so the fine values are close to
what they were in the last best (latest) shim set. If Z1C or Z2C are
quite different then something is wrong with them, their complements
(Z3,z4 or Z5,Z6), a cryoshim or the local magnetic environment(new
neighbor?, probe issue etc. Regardless svs, scan for lineshape. If
it's better even by just 5% then go back to gmapsys set gzsize=4 and
autoshim. Perform the same procedure as before in ACQI on 1 and 2 and
note the values for 3 and 4. Hopefully you can get another 5%. You can
repeat these steps dozens of times to a point where you have the shims
at least 2x better than starting saving as you go.
If you would like to continue to cultivate your automated shimming
skills then I would suggest creating a shim map on CDCl3/acetone. I
have had great success creating specific maps only for the lineshape
sample. By having it ready to go I can shim the z's to spec within a
couple minutes where the normal user maps would fail. You more than
likely will need to do this 3 or 4 times to get it back to what it was.
PFG-shim => manual shims =>make new map => PFG-shim .... Until the top
of the broad base is below the spinning sidebands, it makes no sense to
work on other than low order xy. Course SSB's can be done in H2O/D2O
fast using lock and FID shimmethods.
There are a few other events that can cause a change. Particulate metal
dust collects inside the probe if your local environment is constantly
polluted. All it takes is one particle at 500Mhz to move the field.
Also probe or shim position errors. If someone has been working on it
and not documenting before and after it can be quite laborious to get
back. www.acornnmr.com <
http://www.acornnmr.com> has some good tips on
checking center of the
magnet.
Do you have another probe to try?
7. What ever you do leave yourself a way to back out of any changes you
make. Lot's of notes, pictures etc. If it turns out you only quenched
a cryoshim then that might be better than a service call if it is
shimmable. I do believe that the shim controllers on Inovas are way
more than they used to be on other consoles so if it's that it might be
worth a bench repair and the down time at Palo Alto vs a part trade out.
One last thing, Our Inova500 has at times in front of me with absolutely
no change in the shim settings changed lineshape horribly. It has been
a occasional glitch that we have grown used to but requires a full
console reset to get rid of. Su acqproc ==> power console down wait 1
minute == power back up===>su acqproc. Without the "treatment" the next
user would have bad shims and a normal reset does not fix it although
sometimes an MSR board rest fixes.
8. Alas, it appears that your magnetic field has drifted a bit.
It could be a probe problem (cracked quartz insert...costs about $5,000
USD to fix). Do you have a different probe to try?
9. Alternately, you can try shimming more to see if you can get the line
shape to improve. For shimming at 0.11% and 0.55%, I have always had my
best luck looking right at the FID itself or the spectrum in 'real
time'. To do this, you start acqi, then click FID (instead of LOCK or
SHIM). You may want to tweak your acquisition parameters to reduce AQ -
try aq=0.5 d1=1 gf (gf sends the parameters to acqi). Also make really
big changes in the higher order shims at first just to look for an
effect. Use the x64 button...good luck.
Result
We have tried probe cleaning. Checked for change in conditions etc.
After more shimming, although not within spec.s, it is acceptable for
the present.
--
Sandra Chapman
Professional Officer
Department of Chemistry
University of Wollongong
Northfields Ave
Wollongong NSW 2522
Phone 61 2 42213473
Fax 61 2 42214287
Received on Wed Nov 10 2004 - 09:12:54 MST