Summary of demo/setup/test sample survey
Gerald Pearson (gpearson@umaxc.weeg.uiowa.edu)
Tue, 27 Jul 93 19:26:04 -0500
Many thanks to everyone who contributed to the survey. I've grouped the
responses somewhat arbitrarily into 4 sections, and I've taken the liberty of
editing the notes as seemed appropriate. The dates and subjects of my
original postings were:
20 May 1993 What's a good demo/tune-up sample for NOESY?
7 Jun 1993 Survey of demo/setup/test samples
8 Jun 1993 Add SOLIDS demo/setup/test samples to the survey.
--- Gerry
-------------------- cut -------------------------------------------
SUMMARY OF AMMRL SURVEY OF NMR SAMPLES FOR
TUNE-UP, CALIBRATION, TESTING, AND DEMONSTRATIONS
Tuesday, 27 July 1993
Compiled by:
Gerald A. Pearson INTERNET: gpearson@umaxc.weeg.uiowa.edu
Chem. Dept., Univ. of Iowa VOICE: 319-335-1336
Iowa City, IA 52242-1219, USA FAX: 319-335-1270
Sections
------------------
NOESY TEST SAMPLES
GENERAL CALIBRATION/TEST/DEMO SAMPLES -- MAINLY SOLUTIONS NMR
SOLIDS NMR
COMMENTS
====================
NOESY TEST SAMPLES
====================
-------
From: William Stevens <wstevens@igate.c-mols.siu.EDU>
I'd say a dodecapeptide with a trace of bacteriacide (sodium azide?)
thrown in.
-------
From: macura@tesla.mayo.edu
PDMPO
The simplest is CDCl3 solution of PDMPO (poly(2,6-dimethyl-1,4-phenylene
{oxide). Proton spectrum has only TWO lines (ring CH and methyl). Effective
correlation time is 1-5 ns, depending on concentration. Very inexpensive
polymer (unfortunately I do not have it here). If you want J-coupling problems
then make a mixture with another compound.
Spectra and more details on PDMPO you may find in J.Mag. Reson:
63, 484-493 (1985) 70, 493-499 (1986) 80, 1-22 (1988).
PDMPO is like small protein with regard to NOESY. If you want a system
with NOE+chemical exchange, then mix (in CDCl3 or other solvent) PDMPO and
N,N- dimethyl acetamide. THe latter has three methyl lines, two of which are
in chem exchange (at 30-35 C the best).
-------
From: silber@chemvx.tamu.edu
COBESTER
Personally, I have a cobester sample (vitamin b-12 analog) that I have
used for noesy tests at 400 Mhz. This is only because I already had the
sample, not due to any inherent advantages spectroscopically.
-------
From: rmair@stan.xx.swin.OZ.AU
You might want to look at 1-3 heptanone..... Gives a superb COSY..... but
haven't run a NOESY of it.
-------
From: mmcgregor@chm.uri.edu
PAMOIC ACID
I've been using pamoic acid (1,1'-Methylene-bis[2-hydroxy-3-naphthoic
acid]), which has an easily observed NOE between two para protons on the two
fused aromatic rings of naphthalene. The compound is cheap and readily
available. An NOE can be observed with 20 mg of compound in only 8 scans of a
1D difference experiment. The large positive NOE at 300 MHz becomes zero at
400 MHz and negative at 500 MHz.
DCDMB
I would like to point out a note in TAMU (No. 355 April '88 page 14) in
which Page and Mazzola suggest DCDMB (1,5-Dichloro-2,4-dimethoxy benzene) as a
benzene) as an NOE standard, although it is not commercially available.
-------
From: steveC@al.nmr.varian.com
SUCROSE IN D2O.
There's a resonably intense noe between the anomeric and H2 proton. Use
500ms mix.
-------
From: rshoe@tinman.unl.edu
CYCLOSPORIN
I just recently started using cyclosporin as my NOESY standard for the
simple reason that "example" data came with my new version of Triad NMR
(formerly NMRI's software) software from Tripos, and the NOESY spectrum has
lots of cross-peaks, but it's not too busy (I used to use the old gramicydin-s
(spelling?) ... but WAY too much info in the cross peaks for a good standard).
The other "up" side is that the assignments are straightforward, and already
done for you if you have New Method's data!! <grin>.
-------
===============================================================
GENERAL CALIBRATION/TEST/DEMO SAMPLES -- MAINLY SOLUTIONS NMR
===============================================================
-------
SUCROSE
At the last ENC, at the ammrl discussion session on instrument performance
checks, (quote from Vera's summary):
> Dee [Huang] suggested sucrose/D2O in various concentrations as a very
> good standard for a variety of situations, including resolution for both 1H
> and 13C, sensitivity, lineshape, and VT stability.
-------
From: jbv@iris1.sb.fsu.edu (Joseph Vaughn)
SUCROSE
"REALISTIC" SENSITIVITY TEST FOR HIGH-FIELD SPECTROMETERS
... performance which Ad Bax described in TAMU News Letter ~1989 when they
got their 600 MHz. They used sucrose in NaCl/D2O.
-------
From: gpearson@umaxc.weeg.uiowa.edu
7% 1-PROPANOL + ^0.13% TMS in CDCl3
demo: routine proton & C-13 data acquisition, homonuclear decoupling
test/tuneup: proton pulse calibration, shimming, COSY
comments: very versatile sample; useful when training spectrometer
users; For shimming and pulse calibration on the TMS proton signal,
use a narrow (200 Hz) spectrum width to reject the other signals.
Cleanly-resolved Si-29 satellites of the TMS give a quick check of
high-order shims. Not optimum for a fast COSY, because spectral width
is a bit wide. Concentrations are ideal for 300 & 360 MHz
spectrometers. Seal the sample tube.
neat ETHYLENE GLYCOL-d2
demo: effect of different decoupling schemes on multiplicity and
intensity (NOE) of C-13 lines -- BB decoupling, gated, inverse gated,
and no decoupling.
test/tuneup: C-13 pulse calibration, decoupler pulse calibration,
optimize probe tuning by maximizing C-13 FID while using a small flip
angle (roughly 20 deg)
comments: very intense C-13 FID, particularly using a spectrum width
of only 500 Hz. Demos and calibrations are fast, because the C-13 T1
is only about 5 sec. Shimming should be done on the C-13 FID, because
the proton and deuterium signals are too wide. Seal the sample tube.
recipe: Mix ethylene glycol and some D2O. The hydroxyls exchange very
quickly. Strip off the water on a Rotovap. Repeat once or twice with
fresh D2O so you can get a high atom % D without using much D2O.
-------
From: rnunlist@bloch.cchem.berkeley.edu
60 dioxane, 1% trimethylphosPHATE, 0.2 % trifluorotoluene in C6D6
Test/setup: QNP probes (1H-13C-31P-19F).
Comments: One sample is used for the 3 X-nuclei for routione S/N checks.
Using EXE or xau programs the acquisition/processing and plotting can be
automated. Visual comparison of the plot is quite accurate, differences =>
10% are noticeable. Using only one sample considerably shortens the time
needed to test a probe after cleaning/tuning/repair. The DEPT expt is used to
calibrate 1H decoupler pulses (null @ 90); with the decoupler on-resonance,
the 1H pulse can be calibrated at different power levels (CPD decoupling etc).
Not suiteable for 1H shimming (radiation damping). TrimethylphosPHATE is less
objectionable then trimethylphosPHITE if it breaks.
-------
From: rnunlist@bloch.cchem.berkeley.edu
~ 5 mg Rotenone in CDCl3.
J. Heterocyclic Chem, 25, 51 (1988) [note: There is a typo, proton 1 is at
6.72, NOT 1.72, ppm]
Demo/testing: 2D: COSY, NOESY, TOCSY, Forward/Inverse 13C-1H, Long range
correlations
1D: DEPT, INEPT, etc.(use 50 mg?) T1 (0.4-3.6 sec)
Comments: Roteneone is a nice sample to show the results of different
techniques (and failures of the latest new pulse sequence). It covers ~1.7
to 7.7 ppm. The spectrum is mostly 1st order. Soluable in CDCl3. I have
also used rotenone to check shaped pulse sequences.
-------
From: wstevens@igate.c-mols.siu.EDU
D2O with some MnCl2 to give a residual water line about 4 Hz wide.
This relaxes quickly enough that you can do PW calibrations, amplitude and
phase stability tests, and whatnot without worrying about relaxation delays.
If it contains some DSS or TSP, it can be used to monitor temperature
stability, too.
-------
From: wstevens@igate.c-mols.siu.EDU
WATER SUPPRESSION CRITERION
SUCROSE
It is, I think, Ad Bax's "spec test" for adequacy of water suppression.
In 90% H2O/10% D2O with a trace of sodium azide as bacteriocide, make
solution 2 mM sucrose and 0.5 mM DSS. Using a 50 Hz presaturation field ...
you examine the water peak at half the height of the DSS peak. It should be
less than 100 Hz wide. Also, the splitting of the anomeric doublet of sucrose
at 5.4 ppm should be less than 40% of its intensity (that is, the doublet
should be split to lower down than 40%).
This is supposedly the Ad Bax test (I think) relayed to me by Dennis
Sandoz.
-------
From: "Brian G. Sayer" <sayerbg@mcmail.cis.mcmaster.ca>
STANDARD TEST SAMPLES
LINESHAPE, CALIBRATION CONSISTENCY, INTEGRATION ACCURACY, CP/MAS
... We operate as a "contract lab" for several organisations and have been
asked to comply with the International Standards outlined in ISO 9004. Section
13 is the main concern for NMR Facilities and, in brief, states that "control
should be exercised over instruments and related computer software"....
"Procedures should be established to monitor and maintain the measurement
process" ..."Documentary evidence covering identification of instruments,
frequency of re-calibration" etc.must be maintained,...."Traceability to
reference standards of known accuracy and stability, preferably to national or
international standards". ... For anyone who missed it, there was an article
in C&E News, March 1, 1993, on ISO 9000 series standards.
In order to comply with these requirements, a monthly record is kept for 4
NMR instruments and includes a lineshape measurement using 10% chloroform in
acetone-d6 and a check on calibration consistency and integration accuracy
using 5% ethylbenzene in deuterochloroform. Calibration is checked for both
protons and carbon-13.
Adamantane is used to check CPMAS performance.
In terms of other "demo" samples, sucrose (from the coffee lounge) is
usually used for CPMAS and gives a good spectrum in a few minutes. For
liquids, the "Standard Test Procedures" from Bruker are usually followed.
-------
=============
SOLIDS NMR
=============
-------
From: pratum@uwchem.chem.washington.edu (Tom Pratum)
TETRAMETHYLAMMONIUM IODIDE
C-13 CPMAS solid test/tuneup sample
I feel that this salt is far superior to any other common compound for
C-13 CPMAS because: (1) it has only one kind of carbon, which leads to a high
spectral intensity (2) it has a very short C-13 T-1 (50 msec at 7 T, 120 msec
at 11.7 T) which means that you can see repetitive single scan shots when
attempting determine the C-13 90 directly (3) it has a negligible csa due to
the isotropic tumbling of the tetramethylammonium ion, which means it is no
good for setting the magic angle (which should probably be done with KBr
anyway, but adds to its high spectral intensity (4) it is not noticably
hygroscopic, as all other common tetramethylammonium salts are, and (5) it has
a very sharp CP matching condition (like adamantane, but has very little
enhancement from CP unlike adamantane. This is not a problem since you can
still see it on one scan), so it is good for setting the CP power level.
-------
From: gpearson@umaxc.weeg.uiowa.edu
MIXTURE OF KBR & ADAMANTANE
A few years ago, ... Martine Zelios at Bruker suggested trying a 50/50
mixture of KBr & adamantane. This packed just great, and we've been using the
sample ever since, both to tune the magic angle *and* to set power levels
without changing samples or re-tuning the probe.
A mixture of tetramethylammonium iodide and KBr might be even better.
-------
From: pratum@uwchem.chem.washington.edu (Tom Pratum)
TETRAMETHYLAMMONIUM IODIDE WITH KBr IN IT
I have a sample of TMA iodide with KBr in it myself (used it mostly to
check the spinning rate when I didn't have a spin rate meter). There seems to
be no problem with the mixture. I made it by putting some Kbr and TMA iodide
in a mortar and grinding it up prior to placing it in the rotor. The only
thing which I notice about TMA iodide samples is that they tend to "solidify"
if you keep them around long enough. I think that this must have something to
do with the absorption of water, but this doesn't seem to keep them from
spinning well.
DEUTERIUM SOLID WORK
I have another suggestion for a sample for Deuterium solid work:
deuterated ammonium chloride as a narrow line solid sample. It has a line
width of a few hundred Hz, and a fairly short T1 (much less than 1 second),
and a strong signal. It is easily made by exchange of NH4Cl with D2O. For wide
line work I have used a less common material which I just happen to have
around: choline bromide C4-d2 (doubly deuterated on the methylene carbon next
to the nitrogen). It has an apparent eqQ of 140 kHz, eta of zero, and T1 of
about 100 msec at 6 T. Its main attraction is its short T1 and reasonably
large coupling constant. Its main drawback is its very hygroscopic nature. For
synthesis and other info: J.Magn.Reson. 81,350(1989).
-------
From: charlie@telemann.pse.umass.edu (leonard charles dickinson)
PARA-DITERTIARYBUTYLBENZENE
CP/MAS TUNE UP STANDARD
(Adamantane is is pretty inadequate CP standard--it does most of the
averaging for you at room temperature and doesn't have a large CSA.) Paul
Murphy of the defunct IBM Instruments used para-ditertiarybutylbenzene. This
has many virtues: strong CH3 which you see even if you aren't spinning,
aromatics and aliphatics. The aromatics test the magic angle because the
protonated CH's are split by crystal packing to the baseline if you are right
on. The anisotropy of the unprotonated aromatic C line can be used to tune
the magic angle as it gives directional information with the sign of the error
in the magic angle. The relative heights of the CH3 and aromatics are a good
test of the Hartmann-Hahn match. The aromatic linewidths also test the extent
of decoupling--power and frequency. The samples are pretty stable
morpholigically though after 2 or 3 years we find they don't spin well and
have to be reground and repacked.
-------
From: colorado!cindy@csn.org (Cynthia_Ridenour, Chemagnetics)
DEUTERIUM QUAD ECHO
A good standard for deuterium quad echo is perdeuterated PMMA. This is a
good standard because it has both a narrow and a broad component. The narrow
component can be observed in a single transient, and has a fast T1. The broad
component has a long T1, and requires more time to observe, but is a good
final check once the spectrometer is optimized.
I have favorite standards and protocols for a number of solid-state NMR
experiments, including CP/MAS of various nuclei, and CRAMPS. ... If anyone
requests a standard for an "odd" nuclei, let me know. Chances are, we've run
it.
-------
==========
COMMENTS
==========
-------
From: William Stevens <wstevens@igate.c-mols.siu.EDU>
COMMENT ON NOESY SAMPLES
Don't use too simple a molecule - insufficient molecular weight and your
nOe's are likely to disappear. (Small molecules need the ROESY experiment.)
-------
From: silber@chemvx.tamu.edu
COMMENT ON NOESY SAMPLES
One obvious concern is the field strength you are using. Not all samples
will respond well at all field strengths, especially smaller molecules.
-------
From: rshoe@tinman.unl.edu
COMMENT ON NOESY SAMPLES
... one quick note that several of the molecules mentioned are GREAT
standards for NOE difference, but make lousy standards for NOESY (ok for
ROESY, though) due to the extremely small correlation time.
-------
From: gpearson@umaxc.weeg.uiowa.edu
COMMENT ON CP/MAS OF SUCROSE
Proton T1 is very long. Use a relaxation delay of at least 20 seconds.
-------
From: slp@al.nmr.varian.com (Steve Patt @ Varian NMR Instruments)
COMMENT ON INTEGRATION ACCURACY
... a pet peeve of mine, which is the idea that multiple repetitions of a
small pulse width is somehow better for quantitation than a smaller number of
repetition of a 90 degree pulse. The simple truth is, that unless you are
waiting 5*the longest T1, you can never get quantitative results (unless all
T1's are exactly equal, of course, then it doesn't matter how long you wait;
in that case you should pulse at the Ernst angle for best S/N).
Here are some simple calculations, assuming three signals with T1's of 30,
15, and 5 seconds:
Case 1: 15 degree pulse, 30 second delay, 8 scans
T1 Total Signal S/N
30 2.038 0.72
15 1.998 0.71
5 1.923 0.68
Case 2: 90 degree pulse, 120 second delay (actually not quite long
enough), NT=2
30 1.98 1.40
15 2.00 1.41
5 2.00 1.41
Case 3: 90 degree pulse, 240 second delay, NT=1
30 1.00 1.00
15 1.00 1.00
5 1.00 1.00
As you can see, Case 1 suffers BOTH from worse quantitation (relative
ratios of the areas of the peaks with different T1s) AS WELL AS worse S/N!
Case 2, where the delay time is almost long enough, has quantitation to 1%,
and the best signal/noise; Case 3 actually had a delay that was too long, but
perfect (of course) quantitation.
You can (and I did once, a long, long time ago in a Varian workshop in the
'70's) prove this analytically, instead of numerically as above, but the
simple answer is always - 90 degree pulses for quantitation!
-------
-------------------- end cut ---------------------------------------