Hello.
Some time back I asked a question about radiation damping on a cryoprobe.
Here is the summary of the feedback I got from you guys, as well as a
concise report of the results obtained here from the student who
originally asked the question. Thank you to you all, as usual, AMMRL is
an invaluable source of knowledge and kindness.
As you can see, we could not get rid of damping on a cryoprobe by any
of the means suggested (we had tried most of them already). The
measured relaxation time was OK from the beginning though, as
suggested.
Cheers, Stefano
>
> OUR CONCLUSIONS:
>
> Thanks for so many useful informations. We performed a lot of
> experiments in order to measure the relaxation delay on a cryoprobe.
> We used both a 500 MHz equipped with a cryoprobe and a 200 MHz with a
> regular probe in order to compare our results. As a conclusion, a
> consistent value of 1s was obtained for the T1 of water in DMSO.
>
> -1st: We tried the inversion recovery sequence on the cryoprobe, as
> well by using gradients, capillaries or with the probe detuned. It is
> still impossible to invert the water signal while an inversion is
> observed on a 200MHz and a reasonable T1 value was obtained.
> - 2nd: We get more interesting results with the saturation recovery. A
> T1 value of 1s was obtained, even with a regular tube of 5mm, which
> correspond to the value found on the 200MHz by inversion recovery and
> saturation recovery.
>
>
> -----------------------------------------------------------------------
> ---------------------------------------------------------------
> OUR QUESTION:
>
> Spinlanders,
>
> another point on cryoprobes that seems not to be in the list kindly
> posted by Josh.
> We have a student trying to get reasonable T1 measurements for water
> in DMSO in a Bruker cryoprobe (500MHz).
>
> Apparently, it is very hard to get around the radiation damping
> effects, even trying capillaries, gradients along the recovery time.
> He always find about one sec relaxation time, way too short ?
> Could you please provide some advice ? I'll post a summary.
>
> -----------------------------------------------------------------------
> -------------------------------------------------------------
> THE ANSWERS FROM AMMRL:
>
> Stefano,
>
> When radiation damping is a problem, T1's must be measured with the
> saturation recovery technique rather than the inversion recovery
> technique. Also the fit to the data must be done using the integrals
> rather than the intensities.
>
> see Chem Phys Lett vol 222, pp 417-421 (1994)
>
> Good luck.
>
>
> -----------------------------------------------------------------------
> --------------------------------
> Stefano,
>
> What size capillaries are you using? Doing T1 measurements on a
> conventional probe, I didn't see any problems with as large as 50 uL
> capillaries. I generally use 10 uL capillaries centered in a 5 mm
> tube; I still see a surfeit of signal when I use 5 uL capillaries, but
> I began to be concerned about surface effects. T1 of distilled water
> was still 3 sec as in larger capillaries. There was a fast-relaxing
> component of the echo train in T2 measurements, which in 5 uL
> capillaries was a greater percentage of the initial magnetization, but
> the T2's otherwise agreed measurements made in larger capillaries. The
> practical limit in our case was that the dextran contrast agents
> seemed to find the glass within a few minutes. If your application
> permits you might try fused silica capillary tubing, which is
> available in 100 um, 50 um, or even 20 um i.d. Coating the interior
> with e.g. DCDMS could easily abrogate surface effects, if you see
> them. If you still need to go smaller, there's a cool trick of
> floating a bubble of sample in an immiscible solvent that would let
> you put a tiny perfect sphere into the sweet spot.
>
> How far have you been able to detune the probe? I'm curious to know
> how much this actually helps.
>
>
>
>
>
> -----------------------------------------------------------------------
> --------------------------------------------------------------
>
>
> This is a serious problem on cryoprobes (because of the very high Q
> factor).
>
> You may decrease this by de-tuning the probe, i.e. turn the match
> knob far (!) from usual fitting.
> Of cause, you have to determine the pulses after this.
>
> Besides this you may try to determine T1 not with "inversion
> recovery" , but with "saturation recovery" pulse sequence.
>
>
>
>
> -----------------------------------------------------------------------
> ------------------------------------------------------------
>
>
> Hi,
>
> There are some papers in the literature describing an electronic
> feed-back loop to counteract the current induced in the RF coil by a
> large macroscopic magnetization. How well this will work in an
> inversion recovery experiment, I do not know. See: SUPPRESSION OF
> RADIATION DAMPING IN NMR IN LIQUIDS BY ACTIVE ELECTRONIC FEEDBACK
>
> Author(s): BROEKAERT P, JEENER J
>
> JOURNAL OF MAGNETIC RESONANCE SERIES A 113 (1): 60-64 MAR 1995
>
>
>
> What I have used successfully myself in an experiment at 600 MHz
> where the solvent signal had to be either along +z or along -z for
> about 100 ms. was to give a z -gradient every 5 to 10 ms. For a
> cryogenic probe at 500 MHz where the Radiation Damping rate R is about
> 300 per second, z-gradients might have to applied every one to two ms
> during the inversion-recovery delay. There should be a dependence of
> the measured water "T1" time as a function of the spacing of the delay
> between the gradient pulses. I strongly recommend the use of a
> Shigemi tube, best a 3 mm one with a liquid column no longer than 16
> mm. B1 field inhomogeneity (less for a shorter sample) will create
> some transverse magnetization. This will start the radiation damping
> process immediately while the perfectly inverted magnetization is in
> fact a meta-stable state as far as radiation damping is concerned.
>
> Finally, although the T1 of pure water at 25 ºC is 3.3 seconds, I am
> not altogether certain that the rotational correlation time of H2O
> molecules in mixtures with DMSO remains that long. There is
> literature on the Dielectric relaxation of DMSO Water mixtures. See :
>
> KAATZE U, POTTEL R, SCHAFER M
>
> DIELECTRIC SPECTRUM OF DIMETHYL-SULFOXIDE WATER MIXTURES AS A
> FUNCTION OF COMPOSITION
>
> JOURNAL OF PHYSICAL CHEMISTRY 93 (14): 5623-5627 JUL 13 1989
>
>
>
> Dear Stefano
>
> the best way to get the good numbers for t1 is to use bipolar
> gradients in the TAU delay. The pulse sequence would then look
> something like this.
> ( I have not tested it, but it should work)
>
> The gradient strength only needs to be 1 or 2 %
>
> The other issue you have to worry about is radiation damping during
> the acquisition. For very short and very long delays you get a very
> strong signal with lots of radiation damping. Near the zero crossing
> the signal is weak with less radiation damping. This affects line
> width and using intensity for analysis is not working anymore. You
> MUST use integrals for the calculation of the relaxation time.
>
>
> ;t1ir
> ;avance-version (04/09/06)
> ;T1 measurement using inversion recovery
>
>
> #include <Avance.incl>
>
>
> ;;"p2=p1*2"
> ;;"d11=30m"
>
>
> 1 ze
> 2 d1
> p2 ph1
> 5u gron1
> vd*0.5
> 5u gron1*-1
> vd*0.5
> 5u groff
> p1 ph2
> go=2 ph31
> d11 wr #0 if #0 ivd
> lo to 1 times td1
> exit
>
>
> ph1=0 2
> ph2=0 0 2 2 1 1 3 3
> ph31=0 0 2 2 1 1 3 3
>
>
> ;pl1 : f1 channel - power level for pulse (default)
> ;p1 : f1 channel - 90 degree high power pulse
> ;p2 : f1 channel - 180 degree high power pulse
> ;d1 : relaxation delay; 1-5 * T1
> ;d11: delay for disk I/O [30 msec]
> ;vd : variable delay, taken from vd-list
> ;NS: 8 * n
> ;DS: 4
> ;td1: number of experiments = number of delays in vd-list
> ;gpz1: 1 or 2%
>
>
> ;FnMODE: undefined
>
> ;define VDLIST
>
> ;this pulse program produces a ser-file (PARMOD = 2D)
>
>
>
> ;$Id: t1ir,v 1.9 2004/11/23 15:22:18 ber Exp $
>
>
> --
> -----------------------------------------------------------------------
> ----------------------------------------------------------------
>
>
> I have encountered the problem of radiation damping in measuring 1H
> T1s in aqueous samples at 600 MHz using a normal 1H probe. An
> expensive
> solution is to use a probe with a Q switch (C. Anklin, et al, J. Magn.
> Reson., Series B, 106, 199-201 (1995)). However, there are other ways
> around the problem.
> For background on the problem, there is an excellent article on
> Radiation Damping Diagnostics ( C. Szantay and A. Demeter, Concepts in
> Mangetic Resonance, 11(3), 121-145 (1999)). An important point to
> remember
> is that the initial point of the proton FID is the initial
> magnetization,
> M(0), which is proportional to the number of nuclei. The decay gives
> the
> linewidth in the Fourier transform. Radiation damping can cause a
> regrowth
> of the magnetization. The only way the integral of the Fourier
> transform
> can keep the same value equivalent to M(0) is basically to "screw up"
> the
> phase of the spectrum such that it has both positive and negative
> contributions to correct for the magnetization regrowth. However, the
> integral of the Fourier transform of the entire resonance is still
> proportional to the initial point of the FID. In other words,
> establish
> the "correct" phase correction with a single pulse. A T1 experiment
> can be
> then be run, plotting the integrated intensity as a function of the
> delay.
> Be aware that at short times the spectrum will appear with both
> positive
> and negative components. The integral though is still proportional the
> initial value of the FID.
> Rather than using an inversion recovery technique, I have had success
> in measuring 1H T1s in aqueous solution (in the presence of radiation
> damping) using both a Pulsed Saturation Recovery T1
>
> (Pi/2 - 25 us) repeated 48 times - tau- (Pi/2) - acquire
>
> and also with a gradient version
>
> (Pi/2) - GzT - tau - (Pi/2) - acquire
>
> where GzT is a Z-axis gradient applied for T time to completely wipe
> out
> (saturate) any 1H signal.
> I checked the performance of these sequences with that of an inversion
> recovery experiment on the residual protons in D2O to make certain
> that
> they provided the same answers. This sample of D2O did not show
> problems
> of radiation damping. All three techniques basically gave the same
> value.
> Then I ran the sequences on 95% H2O in 5% D2O. The inversion recovery
> sequence gives a 1H T1 of 39 ms due to the radiation damping. The
> pulsed
> saturation recovery sequence yielded a 1H T1 of 3.2 s on this same
> sample
> (as measured on the same instrument using the same probehead).
> My apologies if my explantation is not clear. Some things are easier
> to explanation in direct conversation rather than typing in an email.
> Good
> luck with the project.
>
> Bob
>
> -----------------------------------------------------------------------
> ---------------------------------------------------------------
>
>
>
> Yeah the cryoprobe sensitivities are getting better, and radiation
> damping is quite efficient; there's no way to give a pi pulse
> on most H2O samples due to the RF inhomogeneity which contributes
> to the damping.
>
> Why don't you try saturation recovery (instead of inversion recovery)?
> Make sure that the read pulse is not a 90degree either but something
> like a 10degree.
>
> Good luck,
>
>
>
> -----------------------------------------------------------------------
> -----------------------------------------------------------------
>
>
> Hi Stefano,
>
> One way to reduce the effects of radiation damping is to heavily
> detune the probe. This will cause a serious loss in s/n and increase
> the pw90, but if detuned enough will eliminate the effects of
> radiation damping. You could determine the linewidth or T1 and then
> detune some more. If the linewidth or T1 does not change you should be
> ok.
>
>
> -----------------------------------------------------------------------
> -----------------------------------------------------------------
>
>
> We find that the key to measuring T1's under radiation damping
> conditions is to keep a relatively weak gradient on for the whole of
> the interpulse delay in the inversion recovery pulse sequence. If
> damping is severe you will also need a strong gradient pulse
> immediately after the 180 degree pulse.
>
> That said, I don't know what T1 should be here - I'd expect H2O in
> DMSO to have a relatively short T1.
>
> Why not measure the excess linewidth of the water signal, to check
> how strong the radiation damping is?
>
>
>
>
> -----------------------------------------------------------------------
> ------------------------------------------------------------
>
Stefano Caldarelli
JE 2421 TRACES, Directeur
Université de Provence et Paul Cezanne
Site de Saint Jérôme Service 511 13013 Marseille
tél 0491282895 fax 0491282897
Received on Mon Sep 26 2005 - 10:49:43 MST