An hour long FID from an 11 Hertz NMR spectrometer

The two FIDs and the spectrum on the right were recorded on an 1 micro-Tesla NMR spectrometer (!) which is probably the current low field NMR spectroscopy record. The Figure is borrowed (with permission) from the site of the group of Michael Romalis (see the previous entry) where it appears as a kind of fun. It shows the spectrum (top) of a liquid mixture of cyclopentane and pre-polarized 129Xe with a proton signal at 42.6 Hz and 129Xe signal at 11.9 Hz (cyclopentane protons get polarized by the xenon through dipolar spin-spin interactions). When only the protons are excited, the resulting FID (middle) decays with a T2 of about 6 s, corresponding to the linewidth of approximately 0.05 Hz. This is the real T2 (presumably equal also to T1) since magnetic field inhomogeneities are orders of magnitude smaller, as witnessed by the xenon FID (bottom) of which only the first 9 seconds are shown. Complete decay of the xenon signal below noise level would require more than one hour! Needles to say, the signals were detected using a SQUID (Superconducting Quantum Interference Device).

In a sense, this Figure makes a pair with the opposite record: the highest-field NMR spectrum ever measured (2D at 58 T) mentioned in the June 12 entry to this blog.

Will NMR jolt the Foundations of Physics?

What caught my attention recently is an NMR experiment which is being set-up at the Princeton University by the Atomic Group headed by Michael Romalis. The goal of the experiment is nothing less than imposing limits on the CP asymmetry of physical laws [Note 1] which has been experimentally confirmed in 1980 by Nobel Prize winners J.Cronin and V.Fitch, but which does not appear to be large enough to explain, for example, the matter/antimatter imbalance of our Universe. The experiments which the Princeton scientists plan to carry out might either demonstrate the existence of a permanent electric dipole moment (EDM) in an atomic nuclide [Note 2] or, more likely, reduce the experimental upper limits on CP asymmetry and thus exacerbate the matter/antimatter imbalance puzzle.

For that purpose, they have selected 129Xe. The reason is not that they suspect the EDM of 129Xe to be large (for a given CP asymmetry, it is expected to be about 10 times smaller than in 199Hg), but rather the fact that 129Xe has unique experimental advantages which might permit to push the CP asymmetry limit several orders of magnitude below the values obtained by earlier studies on 199Hg.

Since the EDM is expected to be extremely tiny (if any), interactions of 129Xe nuclides with anything else should be kept as small as possible. Consequently, the scientists at Princeton work in a very low magnetic fields of about 1 μT (50 times below Earth field), with 129Xe Larmor frequencies of 11-12 Hz! To compensate for the lack of sensitivity at such low fields, they use a sample of several ml of liquid xenon (boiling point 165 °K) which has been previously pre-polarized and detect its nuclear magnetization by a pair of SQUIDs. From the figures they have made public I estimate that the resulting FID S/N ratios are of the order of 100:1 and those of their spectral lines are in thousands to one. They are also aided by the fact that interactions of xenon nuclides with the outside 'lattice' are extremely weak and, consequently, its relaxation times are very long. In liquid, its T2 is about 1300 seconds, while its T1 is approximately 1800 s (the difference is probably due to the formation of Van der Waals molecules). Another feature of xenon which is of importance in this context is its high dielectric strength: it can stand electric fields of about 400 kV/cm before sparking.

The principle of the planned experiment is simple (see the original diagram). The sample is placed into small, uniform magnetic field and divided into two halves, each of which is subject to a large electric field aligned parallel to the magnetic field in one half and anti-parallel to it in the other half. The two squid detectors are arranged so that each perceives prevalently the magnetization in only one of the two halves. Should 129Xe possess an EDM, the energy due to its interaction with the applied electric fields would shift the Larmor frequency up in one of the two halves and down in the other. Romalis and his group intend to detect the frequency splitting by monitoring the evolution of the phase difference between the two signals picked up by the respective detectors after a 90° pulse. With their T2* of about 800 seconds, the response (an FID, only there is no induction in this case) is detectable for up to one hour, so that the sensitivity of the phase-shift method for detecting a Larmor frequency splitting might be as high as 1 μHz!

Reading the papers of Romalis, Ledbetter, Savukov et al, I could not but appreciate the fact that even before they started, they have already discovered a new NMR phenomenon, confirming that whatever you do in the lab, you always discover something, even though it is often not what you had in mind.

It turns out that at such low static fields and using pre-polarized 129Xe, the nuclear magnetization of the sample is comparable to the main field and gives rise to an interesting auto-feedback phenomenon (nuclides precessing around their own magnetization). The evolution of the rather complex system depends critically upon long-range dipole-dipole interactions between the 129Xe nuclides. The main conclusions of a theoretical analysis, confirmed by experiment, are that:
(i) For excitation pulse nutation angles greater than 35°, the phenomenon enhances the effect of any Larmor frequency differences on phase-shifts of the response signals, making the phase difference evolve non-linearly with time. For 90° excitation pulses, this enhances the sensitivity of the phase-shift method by almost two orders of magnitude (unfortunately, it also enhances Larmor frequency difference's due to unwanted static field inhomogeneity).
(ii) For excitation pulse nutation angles smaller than 35°, the feedback effect is no longer positive and response signals behave normally.

Though it is impossible to guess the final outcome of the planned experiments, it is clear that, as usual, the final experimental limit on a hypothetical EDM is going to be determined by the static field homogeneity. In other words, one never stops shimming - not even at 1μT !

Note 1: Potentially, laws of physics could - but need not - be symmetric under any of the following operations: reflection in a mirror plane (parity, P symmetry), simultaneous change of the signs of all electric charges (C symmetry) and time reversal (T symmetry). Experimentally, examples of a violation of each of these symmetries (as well as of the CT and PT combinations) are known since a long time. Until 1980 it was believed that the combined CP symmetry might hold, but it turned out that it does not. At present, only the CPT symmetry (invariance under a simultaneous combination of all three symmetry operations) is believed to be obeyed by Nature. According to the so-called CPT theorem, if CPT-symmetry did not hold, neither would the invariance of physical laws under Lorentz transformation which is so far too much to swallow even for physics cranks.

Note 2: Elementary particles and atomic nuclides may possess a half-integer spin S with its associated permanent angular momentum, a magnetic dipole moment (MDM) when S > 0, and an electric quadrupole moment (EQM) when S ≥ 1. In NMR and ESR we are well accustomed to the observation of the effects associated with each of these quantities. There is a question, however, whether atomic nuclides might possess an electric dipole moment (EDM, S > 0) or a magnetic quadrupole moment (MQM, S ≥ 1). Since the existence of these properties stands or falls with the CP symmetry, and since CP symmetry is known to be imperfect, the answer is "in principle, yes". Nevertheless, neither of these two properties has been ever experimentally confirmed. Consequently, showing that an atomic nuclide actually possesses an EDM would be a revolutionary achievement.

Closing down 2006 MR meetings and opening the 2007 master list

Yesterday finished what appears to be the last MR event of 2006!

According to my NmrEvents list, this year's crop included 58 conference's, seminars, schools and open days directly pertinent to NMR, MRI, ESR and NQR - plus an unknown number of events which involved a magnetic resonance discipline as a minority topic. Even if the list were complete, which is not necessarily so, this would still mean that the average exceeds one per week - much more than what any single person can attend (and any Sponsor sponsor).

It seems that the last event of 2006 was the Wollongong NMR course in Australia (December 4-8), while the first one of 2007 will be the Keystone Symposium Frontiers of NMR in Molecular Biology in Utah (January 6-11).

The fast growing 2007 list already includes 28 entries. Among these I would like to point out the 2007 MRI Courses organized worldwide by the European Society for Magnetic Resonance in Medicine and Biology (ESMRMB). The series includes 14 highly qualified two-days events, starting June 7 in Vilnius (Lutuania) and ending November 10 in Cape Town (South Africa).

I have lumped each such series of events under a single entry in a new section of my list (Seminars and Courses); naturally with a link to the series' home page.

I would like to remind everybody that the list has now become a heavily visited announcement board. Let me know about any meeting(s), seminar(s) or course(s) with NMR, MRI, NQR or ESR content which you might be organizing and I will be glad to publish it free of charge. All I need to know is the link to your event's home page. Optionally, you can also include a brief descriptive text explaining the content and purpose of the event.

If there is enough interest from OEM producers, I would also like to start a Section listing Manufacturer-User Meetings. This, however, will be strictly a service in the sense that it is the manufacturers who must let me know of their planned User meetings. I will publish such announcements for free but I am not going to try and ferret them out myself.

NMR, MRI and ESR books published in 2006

Year 2006 is near to its end and the editorial programs of Publishing Houses are all but closed. Which means that one can set up nearly complete lists of all NMR and MRI books published this year. Almost all the books are available from Amazon: if you are interested, consult the NMR 2006 and MRI 2006 book previews on the right. A similar preview column is available also in the master collection of ESR books and, moreover, NMR 2005 and MRI 2005 previews of books available from Amazon are now available on the 2005 archive page.

NMR / MRI installations and earthquakes

On two earlier occasions (Sept.27 and October 8) I was wondering about the resistance of superconductive magnets to mechanical shocks and earthquakes and I promised to query manufacturers about the topic. Which I did, but with meager results.

It is very difficult to establish an e-mail contact with Jeol since their sites generally do not contain e-mail addresses, or else e-mails sent to them bounce back (I will have to use snail mail). E-mails to Bruker and Varian went through but I got no answer. I have got a reply from a man at Oxford Instruments, apologizing for the delay and for being busy that week, but those were the last words I heard from him.

It is a pity, since I am convinced that (i) it should be possible to build statistically significant indications about the likelyhood of a damage to a supercon due to an earthquake of a certain magnitude and (ii) such data would be most interesting to supercon Users in seismic areas.

There are hundreds of seismic areas around the globe, some in such densely populated states as Japan, Italy, and California. There must be thousands of supercon magnets installed and operating in those areas, hundreds of which have already undergone an earthquake. Yet it is apparently not known how often this resulted in a permanent damage, or a temporary damage (e.g., a quench), a drop in specs (e.g., loss of homogeneity), or no effect at all.

Does anybody have a first-hand experience with an NMR spectrometer or an MRI installation going through an earthquake? If so, let us hear about it (we might eventually build a database from such reports).

Collection of ESR monographs

The collection of references to NMR monographs has reached an extraordinary popularity. It is now linked-to even by the prestigious DMOZ Open Directory Project and visited / downloaded hundreds of times a month. It is evident that there is an acute need for such Collections, provided that they are nearly complete and well maintained.

It is with this goal in mind that I have started another such Collection, this time concerned with books about Electron Spin Resonance, ESR, also known as Electron Paramagnetic Resonance, EPR, or Electron Magnetic Resonance, EMR. The new Collection is organized along the same lines adopted for the NMR and MRI lists (sorting by year and then by the first author).

It starts with over 100 entries which, I hope, places it at about three-quarters mark on a hypothetical completeness scale. I will keep adding more entries to it whenever they come to my attention. Needless to say, I will be grateful for any help, such letting me know about any errors, new books, omitted books, etc.

Why is it important for NMR's to have stable field and good homogeneity

Colleen, a biology teacher from New York State send me this query:

... Presently, I am taking a Molecular Spectrometry online course offered through the University of Maryland. In our present module of study we are discussing NMR's and the science behind them. My web-based research brought me to your blog. As part of my assignment, I am trying to determine some more details of how a NMR functions. Could you tell me or direct me to websites that offer information about why it is important for NMR's to have stable field and homogeneity.

Dear Colleen, to a specialist your question is quite disarming, yet it is fully legitimate. I have been asked qustions like that by electronics engineers, patent lawyers and planning officers. Since I could not find anything among my web links that would be simple, brief, to the point, level with your question, and yet correct, I have decided to write it myself. Please, see whether my new One-Page MR Primer can settle your doubts.

Re: Superconducting magnets: earthquakes, quenches and other mishaps

Vanni Piccinotti, a valiant NMR technician operating from his premises in Tuscany, send me several amusing and instructive remarks which complement the September 27 entry. Since the missive is too long for the blog, I have given it the form of a separate, bi-lingual page (italian and english) entitled Storie di Supeconduttori (Supercon stories). On the basis of his first-hand experience, Vanni concludes that superconductive magnets are sentient quantum beings. Which may be exaggerated, but they sure can stand a lot - and then quench for no apparent reason at all!

Thanks, Vanni, for sharing with us these glimpses!

By the way, Vanni often keeps around a refurbished supercon system, available at a really convenient price. If you are looking for one, send an e-mail to his Magnetic Resonances Technical Services, Via delle Gore 39A, Firenze, I-50141 Italy.

Superconducting magnets: earthquakes, quenches and other mishaps

A friend asked me whether he could transport an energized superconducting magnet to a new lab two floors above its current location. He would have liked to save the time and expenses associated with de-energizing and subsequent re-energizing of the device.

I have of course told him not to do it, but ... it actually is possible to move an energized supercon magnet, provided one avoids mechanical shocks. I know about at least two occasions when an energized supercon was [slowly] moved by a pick-up truck to a location distant many kilometers!

The risk involved in such an operation is a mechanical damage to the suspensions which hold in place the superconducting coil and the inner part or the liquid helium dewar in which it is housed. All the parts which sustain the innermost section of a supercon are extremely thin in order to limit inward heat conduction. Should the suspensions break, the liquid helium would first boil off, possibly quite fast, and a short time afterwards the magnet would quench. One would then have to warm-up the magnet and, depending upon design, either dismantle or cut the outer shell and the dewar vessels to recover the coil.

During normal transport, the coil and the innermost dewar are kept in place by robust transport restraints, but those can't be inserted while the magnet is energized.

While no manufacturer would ever condone transporting an energized magnet (you have to do so at your own risk), the same kind of mechanical damage might occur during an earthquake. For this reason manufacturers should specify the degree of resistance of their magnets to mechanical shock. I could not find any info of this kind on any of their web sites. Bruker has a page on safety regulations but nothing about magnet shock-resistance, while Oxford Instruments, Varian and Jeol have nothing relevant at all, nor did I find something on the sites of major producers of MRI scanners. Should any manufacturer like to comment on the matter, I would be glad to publish the contribution.

Contrary to common belief, neither a spontaneous quench (whose probability is never zero) nor a stimulated one (such as when you forget to refill the magnet) are particularly dangerous. A supercon produced during the last 20 years should withstand any quench without any permanent damage, though no producer is likely to put it in writing as a warranty clause.

Apart from unauthorized transport or an earthquake, there are other ways how an energized magnet could get harmed.

A funny incident happened many years ago at Bruker Spectrospin premises in Fällanden, Switzerland. A worker was asked to move a standard-size gas cylinder. He put it on a small manual cart and, unaware of any possible problem, drove it in a nearly vertical position to its new destination. Unfortunately, he has chosen to pass between two energized 200 MHz magnets standing in a test area about 1 m from each other. The result was that the two magnets literally jumped towards the gas cylinder and stuck to it on each side as though it were an old man and they were propping him up. They did not quench and the only way to disassemble the conglomerate was to de-energize them. Amazingly, neither magnet was damaged and, once re-energized, they worked as well as ever.

Non-magnetic tools for MRI and NMR

The Section "MR compatible tools and devices" in my directory of NMR/MRI Companies keeps growing, indicating that I am hardly aware of all relevant producers. My latest addition is AADCO Medical, a Vermont Company producing, among other things, MRI-safe overhead suspensions, an item which does not seem to be covered by anybody else.

This is good news since overhead suspensions with their jointed arms are great when it comes to freeing floor space and positioning displays and accessory devices exactly where you need them. In MRI the need for overhead suspensions is obvious, but they can be really handy in NMR Spectroscopy as well, especially when the access to your probe becomes crowded (such as doing MRS of perfused organs or living cell cultures, or measuring something at 10 degrees Kelvin).

In general, the importance of non-magnetic tools for NMR and MRI installations can hardly be exaggerated. The number of accidents which occur because an electrician or a bricklayer carrying one of their tools walked by a magnet is staggering. We just do not hear much about such events (unless the consequences are really serious, of course) since those responsible for them tend to maintain a distinguished silence ...

A graph of NMR-books publishing activity

Since the Collection of NMR monographs is now nearly complete, it makes sense to use it as a basis for statistical studies such as the one summed-up by the following graph.

It shows the number of NMR books published every year (the dots) and five-year averages (the green line) for the periods (2000-2005], (1995-2000], ..., (1950-1955].
One notices several features which, at least to me, appear a bit surprising since they contradict the widespread impression of NMR as a rapidly evolving evergreen:
*** There is a sharp, statistically quite improbable peak in 1987. Why ???
*** There were five good years between 1994 and 1998, followed by a rather sharp recession between 1999 and 2003 when NMR books production barely matched that of the seventies (we are now snapping out of it).
*** During the last 30 years (excluding 1987), NMR books were published at a nearly constant rate of 12 ±2.5 books/year, with a long-term growth trend of just 0.2 books/year.

These trends would be even more pronounced were it not for the fact that in the case of multiple editions, my list includes only the latest one. This increases the numbers for more recent years at the expense of earlier ones and thus partially masks the broad maximum around late-eighties and early nineties.

By the way, the book detaining the record number of Editions is Fukushima and Roeder's "Experimental Pulse NMR. A Nuts and Bolts Approach" (Addison-Wesley) which was reprinted 10 times over the relatively short period of 1981 - 1986. Evidently, there is a demand for introductory technical NMR texts frustrated by an inadequate supply (though Nuts and Bolts are today considerably obsolete, they still sell a lot).

The graph does not include medically-oriented MRI books. Once I complete the Collection of those (now at 170 and growing), I will amend the graph with separate MRI data. It is already evident that the MRI books curve which started in early eighties has currently reached the 'classical' NMR (in the last 2 years, it might be even taking over).

Functional MRI (fMRI)

When I first saw this on the site of Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB), I was for a moment convinced that it was a pair of brain images of a volunteer who (i) was asked to think about scissors and (ii) got hurt with them on the head |:-))

Jokes apart, the left image is just a photo of coils for TMS or transcranial magnetic stimulation (an auxiliary tool of fMRI) placed over a head phantom. The picture gallery of this site is a nice gateway to understanding recent progresses of fMRI.

The thing which puzzles me is the name of fMRI. The best definition of fMRI I can think up for the discipline is "nonivasive, fast MRI detection of biochemical changes in specific areas of the brain due to its various activities". If I did not know that already, however, I would have never guessed it when hearing about 'functional MRI'.

Collection of NMR monographs tops 500

Many of you (I hope) noticed my collection of references to NMR books. When I started it nine months ago with the first 100 items, I thought at that time that it was quite close to completion. Now that the canonical gestation period is over, I see that I was very, very wrong. The number just reached the 500 items mark (including some yet to-be-published titles) and is still growing, though at a snail's pace. I believe that now it is really quite close to the final goal of a complete list of all NMR books ever published. So far, not all the books have their ISBN number listed (I am working on it). There are of course many books which just mention NMR in a few paragraphs and which I will never be able to list in this Collection - nor do I believe that they should be there.

The NMR books Collection does not include medical MRI monographs which have a separate, less complete listing (currently containing 165 titles). However, books on MR imaging principles and non-medical MRI are included.

The references are sorted primarily by year and, secondarily, by the first Author. Books which reached three or more editions are prominently marked. For searches, I recommend that you use the Find command of your browser's menu. Unlike fiction, the titles of scientific books are usually closely related to their content (try it with protein or logging).

Naturally, if you let me know about an NMR or MRI book which I overlooked, I will be most grateful. I will be equally grateful if you press an Amazon GO button on one of my pages and purchase a book or anything else (like NMR tubes) from them. This will result in a small commission to me at NO extra cost to you and thus help me maintain this site.
Thanks in advance.

MR jobs and salaries

In the USA they have all kinds of amazing things so it is not really surprising that they also have an Association of Managers of Magnetic Resonance Laboratories (AMMRL) which has published a most interesting survey regarding the salaries of academic MR lab managers in 2004. It turns out that MR managers range from poor things earning barely 35K US$/year and living in welfare housing projects, to lucky barons making 105K a year and living in vast mansions on private asteroids. The average is 66K±20 (gross, pre-tax) which looks respectable, especially if you live in sub-Saharan Africa. If you live in New York, though, and yearn to become an NMR lab technician, think twice since that number is the salary of your future boss, not yours. You will be lucky to get half of that!

Not that your chances to get hired are very bright, anyway. The average number of subordinates an NMR manager has is 1±1.5! Unless you are lucky, chances are that the guy to whom you have sent your CV is in the -0.5 subordinates category.

I also enjoyed some of the correlation graphs which show that correllations of any kind are merely a popular fiction. The NMR job market is a virgin jungle! The graph which really shattered my fragile psyche shows salary versus years of experience: there is no correlation at all !!! A guy with 50 years of experience and a meter long beard is paid only 42K while the clean-shaven top shot with his 105K has just 27 years of practice. There are even babies with mere 8-9 years of toddling practice pocketing 70K a year! Having over 40 years of credit myself, I find that totally outrageous. Other correlations are just as depressive. The only one I can agree with is salary versus the total GHz you manage (from 0.3 to 4). Again, the correlation is almost nill which is good news for the FFC NMR relaxometry guys fooling below 0.01 GHz.

By the way, I was wondering whether I should add a kind of job announcements service to my NMR pages but, in the end, I have refrained from doing so since the "Jobs" page on the Spincore NMR Information Server appears to be quite active and up to the task. If you have an opening, you can announce it there (the process is simple and completely automated), while if you are looking for a job, there are recent entries you should consider.

News from the industrial front

Today I have added two more entries to my NMR/MRI Companies directory.

One is Progression a privately owned Company based in Massachussets, USA, which produces 1H-, 19F- and 31P-LR-NMR process analyzers (Magneflow and iPulse series) built for a number of industrial applications involving polymers, coal, foodstuffs, etc. Some of their products are table-top units for industrial labs, while others are designed for production lines. Though they are in operation since 1989, I was not aware of them.

Which, of course, is my fault. I was convinced that my directory was quite complete but, evidently, there are still gaps. Please, if you have a venture which is not listed, do not hesitate to point it out to me.

The other entry is Invento, a new Italian Company based on a partnership between Stelar (a producer of LR-NMR instruments) and a University of Torino initiative known as the Incubator. The scientific board of Invento includes senior researchers from several European countries whose skills cover a wide area of NMR relaxometry.

I was somewhat at a loss about how to classify this venture. In the end, considering their home-page exhortation 'we would be delighted to discuss your idea', I have decided to list Invento among NMR Services, even though they clearly plan to service ideas rather than folks and carry out a lot of R&D on their own. There is also little doubt that the Stelar's recently announced bench-top fast-field-cycling relaxometer will play a major role in their operations.

The highest man-made magnetic fields

The January 1 entry to this blog contains a table of magnetic fields one can encounter in different places. I have committed an error there, claiming that the highest field ever generated in a laboratory was about 50 T (Tesla). This is not quite correct since values more than twice as high can be achieved for brief periods (a few milliseconds). So what is really the current status of man-made magnetic fields ?

We can skip all the magnets based on iron and rare-earths because natural ferromagnetism can only bring us up to about 2.3 T (the iron group) or 5 T (the rare-earths group) - far too low for what are the 'routine' fields in present-day use at almost every chemical institute.

Next come the superconductive magnets (supercons) produced commercially by Companies like Oxford Instruments and Bruker for NMR in-vitro spectroscopy (in MRI, the requested field strengths are more limited).

Officially, the magnetic fields of current persistent supercons go up to 22.31 T (950 MHz proton Larmor frequency). At slightly lower values (19.97 T, or 850 MHz) they are available even with complete shields against their stray field which otherwise complicates life in surrounding laboratories and provides a nasty link between the field in the center of the magnet and objects located on the outside, such as bunches of keys in the pockets of unaware visitors a floor below or above.

Unofficially, maximum supercon fields are a bit higher and I am sure that both Companies have already breached the 1 GHz tag (23.49 T). For good reasons, however, they do not yet consider 1 GHz systems reliable enough as commercial products.

A persistent supercon, once energized, looses energy and field so slowly that it takes the precision of NMR to notice the drift. Such a device needs a tiny 'nudge' to bring its field back to its starting value maybe once every 10 years. During all that time the current in the main coil keeps circulating in a zero-resistance loop without any need of an external power supply. Magnets of this type are probably the largest and most striking quantum devices which, according to classical physics, should not exist at all !

Then, there are the non-persistent supercons which extend the achievable field values a few percent above the persistent levels sustainable by the same technology. That, however, comes at a price. Trying to feed more current into a persistent supercon, one hits an upper limit above which the current loop becomes unstable - it keeps 'cracking' and 'healing' all the time in a myriad of places. Though the average electric resistance of the coil is still extremely small, it is not zero and, on top of it, it is unstable. Consequently, the magnet loses energy and the only way to keep it on the desired field is to have it permanently connected to a stabilized current supply. In addition, the energy losses dissipate as heat in the liquid helium used to cool the coil and this implies the employment of an on-line helium liquefier. The running costs of non-persistent supercons are therefore very high, while the gains are very small.

At present, supercon technology is stuck at the 21-23 T value. This is reminiscent of the times before superconductivity was practically viable (over 35 years ago) when the upper limit was about 2.3 T, limited by saturation values in ferro-magnetic materials. In that case, however, there were solid theoretical grounds for asserting that the limit was unsurmountable. In the case of superconducting materials, no such theoretical upper bound exists so that somebody might one day come up with a new superconductor and push the limit another order of magnitude higher. After decades of intensive research, however, that looks a bit unlikely.

Beyond the realm of supercons-as-you-know-them, the generation of magnetic fields becomes much more arduous. One generally uses Bitter solenoids made from special materials and immersed in cryogens such as liquid nitrogen to keep their resistance low (even though reaching superconductivity is ruled out). The magnets are operated either in a continuous (DC) mode or, for the highest fields, in brief pulses. The DC systems are often built as hybrids composed of a resistive Bitter coil surrounded by a superconductive shell.

The pulsed-magnets category splits further into two sub-classes: (i) reusable pulsed magnets which can stand repeated bursts and (ii) one-shot pulsed magnets which get destroyed during a single burst. The highest fields ever generated (over 100 T) were obtained by self-destructive devices of the latter type. In both cases, a typical pulse lasts a few milliseconds during which the field describes an approximate sine-lobe arc.

There is a lot of research going on in this area and there exist several large centers which offer public access to the highest achievable fields. Let us see a few of them:

National High Magnetic Field Laboratory (NHMFL) at Tallahassee, Florida, is

the world's largest laboratory of this type (they also have a facility at Los Alamos, New Mexico, but the Tallahassee center is the one more involved in what interests us here). They have spent over 12 M$ to build the world's highest-field NMR spectrometer operating at 45 Tesla (1.92 GHz for protons). It consists of a 35 tons hybrid using a 31 T Bitter coil placed inside a 14 T superconducting shell and, when the field is on, requires 30 MW of electric power and 400 l/s of cooling water. In addition, they have a number of traditional supercons up to about 21 T, several resistive Bitter magnets which can produce up to 35 T in DC mode, and reusable Bitter coils which can be driven by massive capacitive-discharges beyond a 70 Tesla mark.

EuroMagNet is a European consortium concerned with research in high magnetic fields.

It has several centers, complemented by a loose Theory Group. There is a call for proposals posted by EuroMagNet with a deadline of Sept.15, 2006, so if you have any ideas to test, don't hesitate to apply. Nearly all the highest-field magnets currently operated at the European laboratories come from the Tallahassee NHMFL facility so, in a sense, the US and UE installations form a single network. The six European centers are listed below, but only the first two have NMR capabilities:

- HFML at Nijmegen, Netherlands, where you can access an NMR spectrometer

build around a hybrid magnet operating at 33 T (1.4 GHz) with a resolution of 3 ppm. Just keep in mind that NMR is not their top priority. They have a nice page showing all the stuff one can do in high magnetic fields, but NMR is not even mentioned! There is also an often-quoted page on magnetic levitation of objects such as living frogs.

- IFW at Leibniz Institute in Dresden, Germany, with its

Laboratory for pulsed high magnetic fields. Here you can book for NMR experiments with transient field bursts up to 60 T. The IFW group holds the claim to the highest-field NMR spectrum ever measured (2D at 58 T, consisting of a single spectral line and, unfortunately, undated. Note added on December 25,2006: the link now appears to be broken or inaccessible).

- LNCMP in Toulouse, France. Long-duration field pulses up to 70 T.

- LVSM at the Catholic University in Leuven, Belgium. Field pulses up to 70 T.

- LCMIZ at University of Zaragoza, Spain. Long-duration field pulses up to 35 T.

- MegaGauss (or MegaTransport) at Humboldt University in Berlin, Germany.

There are several other places which are - or had been - involved with generating high magnetic fields: the US Lawrence Livermore National Laboratory (a split-solenoid project) and Francis Bitter Magnet Laboratory (35 T hybrid), the EU Grenoble High Magnetic Field Laboratory (30 T hybrid) and Nijmegen HMFL (30 T hybrid), the Japanese National Research Institute for Metals (40 T hybrid and smaller Bitter coils), and probably quite a few others (see list1 and list2). It seems, however, that most of the relevant worldwide R&D is presently going on at NHMFL in Tallahassee with other facilities concentrating more on scientific applications.

The real limits to further progress are given by (i) the high operating costs and (ii) the mechanical stresses within the devices. Close to or beyond 100 T there is no earthly material strong enough to handle the stress which tries to make the magnets collapse unto themselves. Even if somebody came up with a novel superconductor sustaining exceptionally high current densities in high fields, that would only solve the first problem but not the second one. It takes the condensed nuclear material of collapsed stars to withstand such forces.

The conclusion is that while it took 35 years or so to cover the decade of magnetic field strengths from 2.3 to 23 T, it is likely to take a much longer period and many major discoveries in quite disparate areas to conquer the next one.

On the other hand, my prophecies are known to be wrong quite often ...

[it] Quattro K bastano ed avanzano

Questo è un mio ricordo della situazione della RMN in Italia agli inizi della diffusione commerciale della FT-NMR, vista da un venditore (allora, dal autunno 1975 alla fine del 1980, ero alle dipendenze della Bruker Spectrospin Italiana). Credo che vi troverete diversi punti divertenti. A differenza delle storie più recenti, questa è abbastanza vecchia per poterne tirare fuori persino qualche insegnamento positivo per il futuro.

Dovrei scriverne altri. A pensarci, di ricordi in chiave amarcord ne avrei tanti.

Magnetic-field noise and NMR signals

Almost a month passed since my last entry into my own blog !!! Shame on me ! I feel justified, though, since (a) I am a bit under pressure work-wise these days and (b) I have finally managed to organize two major chunks of my NMR field-noise studies into a presentable format and publish them as two articles in Stan's Library:
Field noise effects on FID's and spectra and
Field noise effects on Hahn echoes and CPMG's.

This is something I was working on in my spare time in an on-and-off manner since about 2002 and, though the matter fascinates me, my inability to wrap it up in a presentable form was becoming a major psychological problem. Not that the two articles exhaust the topic. On the opposite, they barely scratch the surface and I also have mountains of experimental data which could/should be used to support the math. However, an exhaustive treatment might take ten more years and require a dedicated monograph! In addition, I must earn my living and (so far) nobody is willing to pay me for analysing noise, be it magnetic or else.

What I find incomprehensible is why almost nothing has been published so far on the topic of field noise effects on NMR signals. In my opinio, the topic is central and absolutely crucial to all branches of NMR, including spectroscopy, MRI, FFC, fixed-field relaxometry, ex-situ NMR, well logging, etc... (you name it).

Since everything in the real world is subject to instabilities and fluctuations, nobody in his/her sane mind can think that there might exist a perfectly stable magnetic field. Even if, through some sorcery, one could produce a perfectly stable magnet, that would still not eliminate the environmental magnetic noise present in any laboratory and, indeed, anywhere on this planet. In a typical lab such a noise is of the order of 1 µT, corresponding to proton Larmor frequency of over 40 Hz! And yet we pretend to measure our spectra in such environments and resolve spectral line splittings of maybe 0.01 Hz. Well, if I did not know better, I would be inclined to say that we were crazy!

Though magnetic resonance obviously needs magnetic field, have you seen many serious studies about exactly how - and how much - do magnetic field instabilities affect NMR signals? I have looked hard for them and I have seen close to none. Apparently, for the producer the rule is: if anybody complains about the stability of the magnet, do whatever it takes to make him stop and then forget the episode. For the academic user, the equipment is a sacred cow which produces the standard textbook milk - it hardly ever crosses their minds that the milk might be adulterated right at the source.

I am sure that the situation will eventually change and the topic will receive the top attention it merits. I hope that the two articles might contribute to make a crack in the present conspiracy of silence. Modesty apart, I honestly believe that my approach to the problem and the respective math are sound and can be safely used for further explorations.

NMR is teaching tricks to IR and UV-VIS

On May 27-30, at a pleasant Swiss mountain hotel, will take place the 3rd International Conference on Coherent Multidimensional Spectroscopy. Now, since the meeting regards essentially IR and UV-VIS spectroscopy, you might wonder what does it have in common with NMR and why am I mentioning it. The answer is contained in the following excerpt from the conference home page:

Vibrational and electronic 2D spectroscopy is a novel spectroscopic tool to investigate the structure and dynamics of solution phase systems with unprecedented detail, in a way very similar to 2D-NMR spectroscopy. Optical 2D spectroscopy measures couplings between certain vibrational or electronic transitions, which are related to the relative distance and orientation of the corresponding chromophores. In this way, structural information can be extracted. Essentially all elementary pulse sequences known from 2D-NMR spectroscopy (NOESY, COSY, EXSY) have been applied to vibrational and electronic transitions. The potential advantage of vibrational and electronic 2D spectroscopy is the inherent time resolution which is orders of magnitudes faster than what NMR can achieve. Potential applications are in protein structure determination, protein folding, DNA structure and dynamics, proton transfer, vibrational dynamics, dephasing processes, liquid dynamics, solvation dynamics, and in new analytical approaches.

For one like me, who has had exposition to all these spectroscopies but spent most of his life in NMR, these are very refreshing news. Of course, there are lots of technical problems to tackle, such as the generation of excitation pulses in the IR to UV region (they use synchrotrons, plasma produced by laser-pulses, and the like). However, the field is on the move and might well lead to new and revolutionary activity in those classical spectroscopies which we have so far considered, scientifically speaking, as exploited as a squeezed, dry lemon.

Most interesting is the fact that we have in this case NMR teaching tricks to other spectroscopies. It is no accident that the opening and closing talks of the Conference will be given by our colleague Goeffrey Bodenhausen. The titles of the two talks are "Two-dimensional NMR: A tool for mapping the transfer of populations, or coherent superpositions of states" and "Could non-NMR 2D benefit from a description in terms of coherence transfer pathways?"

New book: Nuclear Spin Relaxation in Liquids

I can't review/promote all new NMR books and selecting just a few may be unfair to the other Authors. Occasionally, however, I can't resist the temptation either because I like the book too much, or because it has a direct relevance to my recent NMR work or, as in this case, both. I am talking about a new book,

Nuclear Spin Relaxation in Liquids: Theory, Experiments, and Applications,

written by Josef Kowalewski and Lena Maler. The book is just rolling off the CRC presses and it is not yet easy to find on the internet (it took me some time to locate the link associated with the title). The book's trailer says that it

   * Provides a detailed survey of spin relaxation theory, experimental techniques, and illustrative applications
   * Focuses on the physical nature of spin relaxation phenomena rather than on their intricacies
   * Avoids rigorous mathematics where possible and requires only an understanding of the basics of NMR
   * Contains many original illustrations along with references to further sources of information

It then continues saying that

   Collecting relaxation theory, experimental techniques, and illustrative applications into a single volume, this book clarifies the nature of the phenomenon, shows how to study it, and explains why such studies are worthwhile ...

and, knowing Josef, I have no doubt that it indeed does just that. Congratulations, Josef and Lena, this is going to be a very helpful starting point for all those who intend to study or use or even just understand NMR relaxation phenomena.

A text of this kind was particularly badly missing in the area of variable-field NMR relaxometry (VF-NMR). Novel users of fast-field-cycling relaxometers (FFC NMR) often caught me off-guard by the innocent question Ok, now that I have measured an NMRD profile, what do I do with it?. From now on, I will have something solid to refer them to!

Forsight of a genius

The following sentence appears to be due to Charles Babagge (1791-1871), the inventor of computing machines:

Errors using INADEQUATE data are much less than those using no data at all. 

Might it be that the genius' forsight went as far as NMR spectroscopy ???
[Quote borrowed from a V.P.Ananikov's page]

COST meeting on Lanthanide Chemistry for Diagnosis and Therapy

This two-days meeting, to be held starting March 31, is to a considerable part concerned with the use of lanthanides as MRI contrast agents and MR Spectroscopy shift agents, including novel applications as constituents of chemically specific spectroscopic and relaxation probes (a still tentative program is available from the above link).

As such, it is certainly of great interest to all of us, especially since it is the final meeting which will close a six-years long COST action project (D18).

Yet I became aware of the meeting only two days ago because - like most public COST meetings - it was inadequately advertized. COST (European Cooperation in the field of Scientific and Technical Research) is one of those Brussels mega-organizations which handle huge money in an excessively introspective way. Personally, I view it essentially as a set of interlocking old-boys rings concerned mostly with their intricate intra-mural relations, with little time left for public visibility and accounting. The thing I wrote ten years ago about Meetings was partly stimulated by my participation in a couple of planning sessions on EU projects, so if you read it, you will get a better idea of what I mean. Visit the COST site and see for yourself how many links are either missing or broken. Or click on their Events & Calendar and see (a) whether the Orleans meeting is listed and (b) whether ANY of the listed meetings has a link (the answer is NO on both counts).

I am very much tempted to start a new blog functioning as a [critical] observatory on EU Science organization, epistemology and deontology. If you want to contribute - no matter whether supporting, opposing or ignoring my own views - let me know. It could help to kick off the initiative.

Back to the Orléans meeting, though. As it often happens, when one gets to the bottom line, i.e., the scientific program, things look much better than in Brussels. I really like the program and I think that the meeting will be very good and successful. Unfortunately, due to the lack of prior notice, I will not be able to attend.

Multi-sample probes

Almost by chance, I have stumbled on an idea which, in its absolute simplicity, made me exclaim "why have I not thought about this myself some 30 years ago!". It is embodied in a relatively recent US patent (US2004222796 of November 11, 2004) by Munson E.J., Offerdahl T.J. and Schieber L.J., inocuously titled Solid-state nuclear magnetic resonance probe. The patent is owned by the University of Kansas which apparently selected Revolution NMR as its exclusive licensee.

The basic idea of the patent consists in lodging within the main magnetic field several NMR samples and measuring them cyclically one-sample-one-scan at a time. While one sample is subject to a pulse sequence and its response is being collected by the receiver gear, the others are peacefully relaxing in (approximately) the same B0 and thus getting ready for their next turn at a scan. The result is a reduction of the incidence of relaxation delay times by a factor as high as the number of samples lodged inside the high B0 area.

Dr.Munson and his colleagues propose to fit several (their drawing shows 6) solid-state NMR rotors, all complete with RF coils and loaded with a sample, into the bore of a supercon magnet. The vertical array is then shuffled up and down in order to keep the currently measured sample in the spot where B0 has the best homogeneity.

When doing HR-NMR in solids, cutting down the idle times due to relaxation delays is particularly important because the T1's of solids can be particularly long. To think about it, though, we are all aware of the fact that in all kinds of NMR spectroscopy our costly instruments do most of the time absolutely nothing, except waiting for the sample magnetization to lazily plod back to its equilibrium value. Multi-sample probes might be a neat response to this problem and, in my opinion, to design such devices for liquid NMR spectroscopy should be even simpler than doing it for solid-state MAS NMR.

[it] MestRe Day al CISI (Univerità di Milano)

La nostra collega Silvia Mari mi ha chiesto di dare notizia dell'evento Mestre Day al CISI (Centro Interdisciplinare di Studi bio-molecolari e applicazioni Industriali) dell'Università di Milano il 9 Marzo 2006. Si tratta della presentazione da parte della Mestrelab Research di alcuni pacchetti software low-cost (MestReC, i-NMR, NMR Predict, NMR Dev) per applicazioni spettroscopiche della RMN, seguita da dimostrazioni pratiche.

Esaote SpA changed hands

There are only two Italian Companies producing NMR and MRI equipment. One is Stelar srl producing NMR relaxometers and the other, considerably larger, is Esaote SpA, an established producer of intermediate-scale MRI scanners.

Esaote started from a small initiative within Ansaldo SpA, a huge but troubled industrial conglomerate. It then became part of the Bracco SpA group (Bracco is a major producer of MRI contrast agents). This year, on January 20th, Bracco sold Esaote to a group of Italian investors (a mix of banks and individuals) led by Banca Intesa.

The Italian NMR and MRI communities naturally hope that, after the ownership change, Esaote will continue to expand its successful operations with the same vigor as before.

PERFIDI !?! a new general-purpose NMR sequence

PERFIDI is the acronym of a new NMR general-purpose pulse sequence (or sequence preamble) which stands for Parametrically Enabled NMR Relaxation Filters with Double & multiple Inversion. It also indicates an Italian patent of which I am one of the Authors (together with Paola Fantazzini). The patent was registered in July 2005 by the patent office of the University of Bologna and its full text is still confidential. However, a click on the title of this blog entry will lead you to an english-language abstract and a brief description of the invention.

The University and the Authors are available for in-depth discussions with prospective licensees and with Companies and/or Institutions interested in the exploitation and further development of this promising new NMR technique.

Don't be surprised by the fact that the long official title of the patent does not mention PERFIDI and that its meaning appears a bit impenetrable. Patent attorneys are strange people and encountering them is a tough and unforgettable experience. Once they are done with an inventor's text there is no chance that, reading it, the said inventor could ever recognize his invention.

In order to alleviate this problem, the Authors of PERFIDI have started new web pages dedicated to the various aspects, applications and bifurcations of the technique.

[it] Prendere fischi per fiaschi

Questo divertente mini-racconto dell'amico Ing. Vanni Piccinotti (forse dovrei aggiungere mitico o nazionale) non ha certo bisogno di commenti ... Cliccate sul titolo e godetevelo.

CHMOOGLE - a new on-line chemical search engine/database

Though I am not a chemist, many of those concerned with NMR are. Besides, an on-line chemical search engine and de-facto database with access to about 13 million chemical structures is extremely useful to every scientist. I am sure that this relatively new (November 2005) google-like chemistry site will become extremely popular with all of us.

Origin of the APSR sequence

A friend of mine (Villiam Bortolotti, MRPM) noticed that I have implemented the APSR (APeriodic Saturation Recovery) sequence in the pulse-sequence library of Stelar NMR relaxometers and asked me whether I could give him some references on the topic. Trying to oblige, I have realized that, though I know the sequence since my Bruker days back in the 70's and have always given it for granted, there are few pertinent literature studies.

The sequence consists of a pre-saturation train of 90° RF pulses, followed by a relaxation period τ and a 90° readout pulse. The goal is to measure longitudinal relaxation rates as fast as possible. Its basic principle is the same as in the Saturation Recovery sequence (SR) - the pre-saturation train of pulses in APSR simply replaces the first 90° pulse of the SR sequence. In order to optimize the capability of the saturation preamble to null sample magnetization regardless of its initial state, the intervals between the pulses are made to decrease progressively. This is because a train of pulses with even spacings (like in the DANTE sequence) suppresses signals at some offsets and leaves them intact at others. Such offset sensitivity can be exploited in high-resolution spectroscopy, but it is highly undesirable in relaxometry.

The scheme of the complete APSR sequence is

90 - (n-1)δ - 90 - (n-2)δ - ... - 90 - 2δ - 90 - δ - 90 - τ - 90 - Acquisition,

with the bold section being the saturation preamble. In practice, the delay δ is set to about 1% of the typical T1 in the sample and n to 10 or 15, making the preamble much shorter than the canonical 4*T1 'relaxation delay' interval and thus speeding up the measurement since, obviously, no relaxation delay is required between consecutive APSR scans.

Now, while I somehow know all this, I am not completely aware of where the knowledge comes from, nor who and when coined the APSR acronym. It seems that the sequence was introduced in 1971 by J.L.Markley, W.J.Horsley and M.P.Klein (Spin-Lattice Relaxation Measurements in Slowly Relaxing Complex Spectra, J.Chem.Phys. 55, 3604-3605, 1971) though they did not coin the acronym and used the sequence in combination with a field-gradient pulse of duration Δ (homo-spoil pulse or HS), according to the scheme

90 - (n-1)δ - 90 - (n-2)δ - ... - 90 - 2δ - 90 - δ - 90 - HS - (τ-Δ) - 90 - Acquisition

Paradoxically, homo-spoil pulses were applied in exactly the same way to the plain SR sequence only two years later (G.G.McDonald, J.S.Leigh,Jr., A New Method for measuring Longitudinal Relaxation Times, J.Magn.Reson. 9, 358-362, 1973). The chronological sequence for the group of sequencies we are discussing is SR, followed by APSR with HS, then SR with HS and finally by plain APSR.

I was certainly using the APSR sequence already in 1977 within Bruker environment, without the HS pulse and with the acronym fully established. In the literature one finds occasional mentions of APSR having been used for this or that purpose but such mentions are rarely accompanied by any reference. In those cases where there is a reference, it is always to a Bruker pulse sequences library (usually the Aspect 3000 NMR Software Manual) or to the already mentioned Stelar pulse sequence library set up by myself.

In conclusion, I believe that the APSR sequence was born out of the Markley-Horsley-Klein idea and nicknamed by an anonymous Bruker guy who coded it into a Bruker library. Thereafter it achieved some limited but persistent diffusion through Bruker company literature. Unless I am wrong (if so, please correct me!) there does not exist any in-depth study of its presumed capability to achieve a good saturation starting from arbitrary initial magnetization. Which is a pity since such a capability would be very useful in many applications and since there are indications that the sequence works reasonably well.

Units and distributions of relaxation times

Looking up Vol.27A of Concepts in Magnetic Resonance (November 2005, pp 122-123) I have noticed a brief Note on Units in distributions of relaxation times by Paola Fantazzini and Bob Brown, whose abstract says:

Distributions of relaxation times or rates may be misinterpreted or wrongly compared if units are not consistently used or not correctly identified.

This, I understand, is an echo of an old discussion in which I was myself partially involved and which prompted me to write a Note in Stan's Library on this site. The standpoint of the two Authors is of course a bit different from mine since they respond to a direct challenge to their work. What I can add is that (i) I could not agree more with what they say and that (ii) it is not a bad idea to visit Stan's Library on a regular basis.

MRA: addendum

I have asked Robert C. Duncan, an astrophysicist and an expert on X-ray sources, pulsars and magnetars (see the January 1 entry below) the following question:

Are there any known magnetic bodies in the Universe with field values intermediate between white dwarfs and neutron stars? There is a gap of 6 orders of magnitude between the two and I always thought that Nature hated gaps.

His answer:

Some old neutron stars have weak magnetic fields in the range between magnetic white dwarfs and 10^12 Gauss pulsars [= 1e8 T]. So there is no clear gap in magnetic field when you consider all objects in nature. There is, however, a big jump in matter density between white dwarfs and neutron stars, by about a factor of 10^7! I don't know of any intermediate density stars.

He also most kindly allowed me to made the pdf copy of his Note on Physics in Ultra-strong Magnetic Fields downloadable directly from this site. Thanks, Robert.

MRA: Magnetic Resonance Astronomy

New Year's entry in a blog should be a bit special, which is why I want to talk about the range of magnetic fields in the Universe. After all, magnetic fields are something we should know about quite a lot and, given a field somewhere (anywhere), we should be able to do some magnetic resonance with it !!!

There are many fascinating magnetic bodies in the Universe. In our own solar system, the strongest magnetic fields occurring naturally are of the order of 0.1 Tesla, found in the sunspots and at Jupiter. The two environments are quite different, one very hot and saturated with radiations of all kinds, the other freezing cold. In both cases, however, the field-of-view is as large as hundreds of Earth-size planets - and it is full of stuff which might give rise to NMR and ESR signals.

The Table below lists some of the magnetic fields one can encounter in various places and other magnetic field data. Before talking about the strongest fields which are hard to find even in the vast Universe, let me point out that it is actually difficult to find any place where the field would be zero. Even the trifle magnetic field value of about 1 nT characteristic of the empty inter-galactic spaces is large enough to make a lonely electron precess at quite respectable 28 Hz!

The Table lists magnetic flux densities in Tesla units (T) with various SI prefixes. I will never understand why astrophysicists, those who should be the diamond tip of modern physics, apparently cling to the utterly obsolete Gauss units (1 T = 10000 Gauss)!

You probably know about neutron stars which are closely related to pulsars, discovered in 1967. Magnetars, discovered in 1979, are even stranger objects. They are very rare due to their short lifetime. At present maybe just six of them are known within our Galaxy and with all the magnet-quakes going on on them, they will not last much longer than 10000 years. They are probably neutron stars undergoing yet another implosive collapse.

Pulsars emit flashes of visible light, while magnetars emit flashes of X-rays. The theories behind these processes are complex and not quite convincing. What I find fascinating is the fact that the frequencies of visible light (between 4e14 and 8e14 Hz) are just an order of magnitude smaller than the highest Larmor precession frequencies of protons and neutrons on neutron stars. On magnetars, those Larmor frequencies exceed 10000 times the frequency of visible light and, in addition, wherever there are free electrons (like in ionized plasma atmospheres), their Larmor frequencies are still higher by the factor of 658. In other words, the Larmor frequencies do fit the picture! Might it be that the astrophysicists simply need to start thinking along standard MR ways?

Before you start pursuing this line of thought, however, you should take a look at the article by R. C. Duncan (the PDF version). It explains about the very strange physics which starts in magnetic fields above 4.4 GT where electron's Landau excitation energy exceeds its rest-mass energy. Vacuum becomes birefringent, electrons change mass, photons become unstable, atoms become thin needles, molecules are reduced to rigid linear rods, etc., etc.

What about still higher fields? It seems that somewhere between 1e47 and 1e49 Tesla, even vacuum would break down and convert spontaneously to matter. It is likely, of course, that that simply means that such fields are physically impossible. Fortunately, considering that this theoretical limit is at least 36 orders of magnitude above the strongest known magnetar field, we probably do not need to worry too much about it. At least not in the foreseeable future.

For a complete list of all entries since 2005, see the running INDEX