March 29, 2015
Europe's NMR spectroscopy above 1 GHz
About a month ago I have started a discussion entitled "Who owns, or ordered, a >= 1GHz NMR spectrometer?" on the LinkedIn Group Magnetic Resonance Research Network. The Group has over 4800 members, but it is nevertheless classified as "private". I therefore want to copy my own conclusions from that discussion here, with only minor edits:
It looks like this thread is coming to an end. We (mostly thanks to the links collected by Bert Heise of Spin-Doc) have established several facts, I think:
(1) Apart from Bruker, there does not seem to be anybody else developing a mere 1 GHz magnet. The fact that for Bruker 1 GHz is a 'catalog product' probably makes a non-Bruker, independent development less attractive.
(2) 1.3 and 1.5 GHz magnets are under development in a couple of public Institutions in the US, but appear to be still far from usable.
(3) Since 600 MHz is the current upper limit for Jeol/Jastec, Bruker is presently the undisputed world-wide commercial monopolist above that field strength.
(4) Bruker is accepting orders for 1.2 GHz systems. They probably have in-house an energized prototype, or even prototypes; in fact, I heard rumors to that effect already 2 years ago. Nothing much is known about the specs of such magnets (persistency, MTBQ - the mean time before quench, helium consumption, etc), neither the currently verified ones, nor those expected after delivery.
(5) World-wide, at least 7 (maybe more) 1.2 GHz systems were actually ordered and are slated for delivery and installation between 2015 and 2018, all of them in Europe (links: Frankfurt, Göttingen, Munich, and Jülich, all in Germany, Utrecht in Netherlands, Birmingham in UK, and Florence in Italy).
(6) The heavily, and curiously, unbalanced World distribution of the ordered systems looks 'suspicious', but one must take into account that most of the 1.2 GHz orders are for the super-national EU Large Scale Facilities and similar Centres, basically controlled by the same Commission in Brussels, so they can not be counted as truly independet orders (they appear as such at first sight, but I do not believe they are).
(7) The decision to endow a significant portion of pan-European facilities by 1.2 GHz NMR's might be correct, I do not know; I just wish that there were more public discussions of such EU decisions.
(8) Though the standing orders imply a total cost around $150M, it is actually trifle compared with total LSF's running costs. It is also not too much for the Bruker Group whose yearly revenue is 12 times bigger. If, on the technical side, everything goes well with the installations and the cycle will be completed in 3 years, it will boost Bruker sales by about 2.5% per year - a short-term blessing, but not a big one.
- Installed and operative >=1.0 GHz spectrometers: To my knowledge, there are two successfully operating 1.0 GHz spectrometers: one in Lyon, France, and one in Bayreuth, Germany. The purchase of a 1.1 GHz system at ETH in Zürich, Switzerland, was announced, but the current status of this project is not clear.
- Pricewise, a 1.2 GHz NMR instrument weighs as about ten 800 MHz instruments.
- In the US, quite a few people are bewildered and getting worried, asking whether they are not missing something. See this Community letter to NRC.
- Above, I use [Hz] as magnetic field units (1H Larmor frequencey). As everybody knows, the conversion factor is 42.576388 MHz per Tesla, so that 1.2 GHz corresponds to 28.184636 T. My apologies to the magnet guys.
29 March 2015, Tweets between Christian Fernandez and Olivier Lafon, Uni Lille:
Christian: I think Lille, France, is also ordering 1.2 MHz, right?
Olivier: Yes, CNRS has also ordered a 1.2 GHz NMR. It will be installed in Lille and will be a national facility.
30 March 2015, Stan
There are rumors that Jeol is also working on a really high-field magnet. I believe (though there may be more going on) that they stem from the recent JMR article signed by 13 authors from 6 different Japanese institutions (including RIKEN) and Companies (including Jeol). It is a kind of high-quality work-in-progress report, nothing imminent, but very serious. It is certainly pertinent to this discussion, since the second part of its long title says: "Towards an ultra-compact super-high field NMR spectrometer operated beyond 1 GHz". Judge for yourself; this is the DOI: 10.1016/j.jmr.2014.10.006.
One never knows; it might yet happen that in the next 10 years the price of a 1.2 GHz instrument might drop by an order of magnitude. Which means that the coming European installations should better start producing results very soon, before becoming obsolete.
March 23, 2015
Premature departure of Alessandro Bagno
A word has reached me that Alessandro Bagno, professor at University of Padova and Head of an NMR Group), has died prematurely (at 57, of infarct). It left me truly saddened. I got to know him only relatively late, but our contacts were intensifying during the last years.
Alessandro was one of the few Italian experts on ab-initio DFT calculations of NMR parameters (chemical shifts and J-coupling constants). For me, contacts with him started when I got interested in "predictions" of NMR parameters as inputs for what I am doing lately with Mestrelab. I had many questions and he was very good at answering them. We have first met socially as lecturers at the GIDRM Advanced NMR Course in 2007 in Torino, where he presented a very clear review of the current status of ab-initio NMR predictions in "normal" diamagnetic molecules, based on his seeding 2001 paper (Chem. Eur. J. 7, 1652-1661). I can also hardly count the number of times I have studied the paper he and his colleagues Federico Rastrelli and Giacomo Saielli published in 2006 and, in particular, pored over the predicted values in strychnine (shown below, just to give you an idea of how far these computational chemistry techniques got). In the last few years, Alessandro moved into ab-initio computations of the apparently totally irregular chemical shifts in paramagnetic molecules. Those can be really huge (up to thousands of ppm for protons) and are therefore of interest, among others, to people who wonder about NMR applications in really low magnetic fields (see the second Figure). In this area Alessandro had ongoing international collaborations, most notably with Alex Pines at Berkeley (yes, he was a "pinut").
Last year in May we met at NMR Valtice in Moravia (in Czech) where he presented his recent work, and where we had an opportunity to talk at length during a 10 km cross-country hike. We talked a bit about Italian politics - he was keenly interested in helping to halt the Italian decline. But mostly we talked about Science. He got me interested in simulating nuclear spectra of strongly coupled paramagnetic systems in low fields, where the high-field approximation can no longer be applied and the computations become exceedingly hard. We decided to do something together along this line - and now he departed for good, leaving me with a bad conscience because I have not finished anything yet ...
It is so bad when these things halt in mid-stride a rich life full of plans and interests!
February 19, 2015
GE ordered by FDA to recall nearly 13'000 MRI scanners!
Yesterday the U.S. Food and Drug Administration (FDA) issued a Class 1 Recall order on almost 13000 General Electric (GE) MRI scanners of a staggering number of types installed in the USA. Class 1 recall is the most severe one, indicating a lethal danger.
The order apparently regards units which have a separate, self-standing Magnetic Rundown Unit (MRU) which can be switched OFF. It is not yet clear how the actual "recall" will be managed, nor what impact might the order have on yet unlisted scanner types and/or those produced by other manufacturers. Nor is it clear what will be its impact on MR sites outside of the US.
The story behind the recall seem to be this: there was a mortal accident in which a large magnetic object was drawn into a magnet, seriously hurting the patient, and obstructing access. Since the MRU unit was disconnected, it could not be used to run down the magnetic field in order to clear the magnet entrance. Prevented from receiving first aid, the patient died.
This raises several questions:
- Do vendors inform local technicians stringently enough that the MRU must stay always connected?
- Why the managers of so many MR sites so often disconnect the MRU's anyway?
- Why, if the MRU must stay connected, are the scanners not designed so as to prevent normal clinical use unless the MRU is attached, operational, and tested.
Technically, that would seem to be the most obvious precaution.
It is not likely that the installed scanners will be really "recalled". The solution might consist in just issuing a strict order to all MR site managers to keep the MRU's operative all the time, and test them on some regular basis. The best solution, it seems to me, might be a relatively minor, but important, in-place upgrade of all the scanners indicated by the FDA.
The total number of installed GE MRI scanners in the U.S. is over 19600, and there are over 9500 more outside the US. At this moment it is not clear (to me, at least) how are other manufacturers/vendors handling this safety aspect. I will investigate.
The photo below gives a clear idea about what we are talking about. Imagine a patient trapped inside the magnet and nobody knowing what to do because the "panic" button ... OOPS! ... does not work.
This photo was borrowed from MRI Metal Detector blog. Among other places concerned with MRI safety, see Frank Shellock's website.
February 18, 2015
Spin Systems, Math Theorems, Sister Celine, and Machine Proofs
In 2007, I wrote an article about a combinatorial identity, which kind of popped-out from my NMR musings, and which I considered novel. Naturally, nobody cared much about it which, to be sincere, I have expected - and reciprocated, kind of. As often happens, I have published it in Stan's LIBRARY, but also subjected it preventively to an informal peer review. As a result, I got a dense flak from mathematicians, basically pointing out that the proof could have been done automatically by of those symbolic math software packages, avoiding my dirty application-oriented approaches based on counting p-quantum transitions in NMR systems of spin-1/2 transitions or, alternatively, on strings of symbols "0" and "1". They did not show any love of NMR at all!
In this case, the suggested computer proofs are based on the time-honored Sister Celine's method. They certainly work; there is a nice book about it, written by Petkovsek, Wilf and Zeilberger and titled just A = B (it is also available online). The automated-proof techniques embrace a whole math jungle inhabited by hundreds of combinatorial theorems.
As I wrote in the revised version of the article, however, they leave you with a bitter after-taste: you know that an expression holds, but you do not really know why (behold the brave new math World!), nor do you gain any convincing feeling about why it should work. Once you have a probable identity, the machine-proof methods will tell you whether it holds or not - but you must have it first, and that does not happen spontaneously. You cannot just recline on a coach and wait until an identity materializes in your head! For these reasons, I prefer what mathematicians call a "model" for any particular identity, and if the model can be drawn from an NMR context, that makes me even happier.
A reason why I am digging this up, however, is a recent blog entry of mine, entitled Magnetic Resonance on OEIS. There, I have promised some explanations and, not having enough time to do so right now, the old article can help to, at least, introduce the context.
BTW, this is the identity, with C(n,k) denoting the binomial coefficients:
It expresses the total number of p-quantum transitions N(n,p) in a spin system composed of n coupled spin-1/2 nuclei.
January 31, 2015
MR fast sepsis
Table-top NMR diagnoses sepsis fast enough to save lives
It is quite rare that IEEE Spectrum - the prestigious news magazine of the Institute of Electrical and Electronics Engineers, would write about magnetic resonance (except, occasionaly, about MRI). Yet in the last issue it did (January 2015, page 19), though somewhat indirectly.
The article regards a Massachussets company T2Biosystems which IEEE defines as a start-up. They produce a table top instrument not larger than a medium-size home printer, T2Dx, which can diagnose various types of bacterial sepsis in the blood of a hospitalized patient in 3-5 hours instead of 1-2 days, a time difference which can often literally mean death or life. It may sound surprizing, but even in the USA, sepsis is among the major causes of death among hospitalized post-traumatic and post-surgery patients. It is still a major clinical issue.
One must read the fine print of the IEEE article to realize that T2Dx is an NMR time-domain instrument. It contains a very small but very performant magnet developed years ago by Eiichi Fukushima and his collegues at a small New Mexico company ABQMR (they initially called the magnet and the NMR system build around it NanoMR). To detect the sepsis bacteria and discriminate between their different strains, T2Biosystems eploits relaxation effects due to magnetic nano-particles coated by a biochemically active layer.
What got me excited when I spotted the IEEE article was that I wrote about the development already over 6 years ago, under the heading The smallest NMR ever built. You might therefore wonder, perhaps, why IEEE still defines T2Biosystems as a start-up, and what has taken them so long to get on the market. One reason is, I think, that the application requires a combination of several different technologies, of which the NMR part was, possibly, the one resolved early. Another aspect are field tests and FDA approvals. T2Biosystems tested their method on 1500 patients at 7 hospitals! After that, they got the FDA rubber stamp in a record time, IEEE says, because with T2Dx the mortality rate of septic infections dropped from 40 to 11 percent. A study like that cannot be done overnight! In any case, they are now ready to go, with the 2015 revenue target sat at 35 M$ - and thousands of saved lives. Thumbs up!
9 February 2015, Eiichi Fukushima
Just saw your blog about T2Biosystems. I tried the link to the article and it did not work so I went to the Spectrum site and found the article by Prachi Patel, posted 23 Dec 2014. You might want to change the link.
Thank you for giving us credit for the magnet but I do not believe they use the same magnet we used. I'm sure they need more sample than our magnet did but can check with people who are more familiar with what they are doing -- I am not. Also, nanoMR was a company that was formed to license our technology -- and wrote about it in a couple of articles like in the 2008 Appl. Magn. Reson. article (Vol.35, pp.185-195, DOI 10.1007/s00723-008-0151-3) -- but they did not follow up the NMR development. They elected to pursue the immunochemistry of labeling, instead, and just (2 weeks ago) got bought out by a British company called DNAElectronics [...]
By the way, Andrew McDowell who invented the technology that was licensed to nanoMR left us a few years ago and is working on his own, still in Albuquerque, as NuevoMR, LLC. He is on LinkedIn; I need to ask him if he has anything else out in cyberspace.
Anyway, it is good to see that you and the blog are still active. Our activities have turned in a new direction; I have spent much of my energy recently on a large-scale Earth's field NMR project for detecting oil spills in the Arctic, something I never thought I'd do.
More on that later. Best wishes, Eiichi
17 February 2015, Stan
Eiichi, I have corrected the link (thanks), but you left me with my mouth wide open.
First, told by you, the T2Biosystems story sounds much different, and more real. Let me point out that nanoMR was sold for 24 M$; it could stimulate a potential start-up hesitating whether to move into some "funny" NMR. It also illustrates what I am preaching since some time now: what really counts are the applications, while NMR as such is just a tool.
Second, though I am well aware of your pioneering mountain climbing in the Antarctica, your new Arctic NMR interests really beat me. You are an amazing guy!
I am looking forward to any update.
January 25, 2015
the great guy
who started it
Detection of a single nuclear-spin announced ... and retracted
A micro-story of MRFM nano-imaging, including a recent hiccup.
From the very beginning of Magnetic Resonance Imaging , there emerged an interest in pushing its limits towards the highest possible resolution. It became soon clear that the "classical" MRI set-up had severe resolution limits which would be hard to beat due to limited sensitivity (typically, the smallest number of nuclei an inductive sensor can detect is about 10^12). A different technology was needed.
Starting from a seeding 1991 paper  by John Sidles from Univerity of Washington, Seattle, a group of researchers and engineers at IBM, led by Daniel Rugar, tried to apply to spins a detection method known as MFM (Magnetic Force Microscopy). MFM uses the mechanical resonance of nano-sized, vibrating cantilevers, a nano-technology capable to detect extremely tiny forces (there is nothing that beats it). In their nice 1995 review , Sidles has named the new MFM-based MR technique MRFM (Magnetic Resonance Force Microscopy).
With this new technology, the Sidles - Rugar group started pushing down the minimum number of detectable spins. There was a rapid series of successes, with microscopic images of biological cell structures, viruses, and the like. Finally, in 2004, there came an exciting announcement  of the first detection of a single electron spin. The clever experiment combined cantilever detection with a magnetic field, a pulsed ESR micro-spectrometer, and an ingenious pulse sequence, all designed to make sure that what was detected was indeed a single electron spin. A true jewel of science and engineering!
That achievement started a "race" towards the detection of a single nuclear (not electron) spin. That, of course, is a much bigger challenge. Since electron magnetic moment is over 658 times larger than that or proton, and since the sensitivity of this type of sensors increases roughly with the square of the magnetic moment, a single electron beats in terms of sensitivity more than 400'000 protons! But the fata morgana of nuclear spin imaging at nanometer scale is a strong attractor. If it could be done, structural analysis of molecules would change dramatically: in principle, we could produce 3D images of individual molecules in all their complexity!
Naturally, the experiments of the Sidles - Rugar group were repeated by others. This brings us to the 2005 Thesis of Christian Degen . Though it is oriented towards nano-scale MRMF spectroscopy (implying many spins), it is pertinent also to our topic. It gave rise to a group led by Christian Degen and Kai Ebenhardt at the ETH in Zurich. They did fine job on improving the sensitivity in 2007  and applied their MRFM setup to 1D and 2D spectroscopy in 2008 . In the 2007 experiment they were able to detect a very sharp "ridge" of 19F nuclear spin magnetization, produced by MRI techniques, in a single crystal of CaF2 and "image" it with a resolution better than 10 nm. Not only that: they were actually able to monitor the spreading of the edge due to the spin-diffusion phenomenon and estimate the spin-diffusion rate, which turned out to be in full agreement with previous, indirect methods. A lovely paper which is today carved in stone in the MR Valhalla.
It is not easy to estimate how many 19F spins were involved in the experiment; the linear resolution being about 10 nm, the count boils down to something around 1000, give or take a factor of 5. Some optimistic estimates of detectability I have heard around 2010 [verbal communications] were as low as 50 - 200 spins.
Again, all these achievements were reproduced elsewhere and a lot of impressive results were obtained. See, for example, the top-quality Review of reference .
Then, in October 2014, just 3 months ago, came an announcement  from the ETH group that a single proton spin has been detected. In a very special system (diamond crystal with a nitrogen defect buried just under the surface, close to the proton to be detected) but, nevertheless, a single nuclide! Wow!!!
Unfortunately, something went wrong: the authors themselves, very honestly, pointed out a potential problem  and, a few days ago, retracted  the Science paper. The retraction says that the observed proton signal might have been confused with the 4th harmonic of 13C Larmor frequency (13C nuclides are of course abundant in diamond). The retraction leaves open a chance that, in one out of the three cases described in the paper, a single proton might have been indeed observed.
What can one say? An error, of course, can happen. It may be that the present-day academic pressure to publish fast is just too high. Especially when there are groups tackling the same targets by competing detection technologies such as, in this case, optically detected magnetic resonance (ODMR). A November 2014 ODMR paper , in fact, also claims to have detected a single proton. They use the same diamond surface with nitrogen defects, and similar pulsed RF excitation schemes; just the detector is different. Which makes me wonder: might not the reasons behind the retraction of the Degen's group be pertinent also to the ODMR paper?
Anyway, the single-proton MRMF detection milestone is presently lost in a fog, kind of.
Hence, rejoice, because the race is still on! But better do not rejoice for too long, I think.
- Lauterbur P.C., Image Formation by Induced Local Interactions: Examples Employing Nuclear Magnetic Resonance, Nature 242, 190 (1973). DOI 10.1038/242190a0.
- Sidles J.A., Noninductive detection of single-proton magnetic resonance, Appl.Phys.Lett., 58, 2854 (1991). DOI 10.1063/1.104757. The title of this seeding paper expresses the author's dreams, not an actual achievement.
- Sidles J.A., Garbini J.A., Bruland K.J., Rugar D., Züger O., Hoen S., Yannoni C.S., Magnetic Resonance Force Microscopy, Reviews of Modern Physics, 67, p.249-265 (1995). DOI 10.1103/RevModPhys.67.249.
- Rugar D., Budakian R., Mamin H.J., Chui B.W., Single Spin Detection by Magnetic Resonance Force Microscopy, Nature 430, 329-331 (2004). DOI 10.1038/nature02658.
- Degen C.L., Magnetic Resonance Force Microscopy: NMR Spectroscopy at the Micro- and Nano-Scale, Diss. ETH No. 16304 (2005). DOI 10.3929/ethz-a-005128707.
- Eberhardt K.W., Mouaziz S., Boero G., Brugger J., Meier B.H., Direct Observation of Nuclear Spin Diffusion in Real Space, Phys.Rev.Lett. 99, 227603 (2007). DOI 10.1103/PhysRevLett.99.227603.
- Eberhardt K.W, Degen C.L., Hunkeler A., Meier B.H, One- and Two-Dimensional NMR Spectroscopy with a Magnetic-Resonance Force Microscope, Angew. Chem. Int. Ed. 47, 8961 (2008). DOI 10.1002/anie.200802978
- Seppe Kuehn S., Hickman S.A., Marohn J.A., Advances in mechanical detection of magnetic resonance, J.Chem.Phys. 128, 052208 (2008). DOI 10.1063/1.2834737.
- Loretz M., Rosskopf T., Boss J.M., Pezzagna S., Meijer J., Degen C.L., Single-proton spin detection by diamond magnetometry, Science Express, October 16 (2014). DOI 10.1126/science.1259464.
- Loretz M., Boss J.M., Rosskopf T., Mamin H.J., Rugar D., Degen C.L., Spurious harmonic response of multipulse quantum sensing sequences, arXiv:1412.5768 (December 2014).
- Loretz M., Rosskopf T., Boss J.M., Pezzagna S., Meijer J., Degen C.L., Retraction, DOI 10.1126/science.347.6218.139-a.
- Sushkov A.O., Lovchinsky I., Chisholm N., Walsworth R.L., Park H., Lukin M.D., Magnetic Resonance Detection of Individual Proton Spins Using Quantum Reporters, Phys.Rev.Lett. 113, 197601 (November 2014). DOI 10.1103/PhysRevLett.113.197601.
25 January 2015, Stan. Sorry, I forgot a very useful MRFM review:
Suter A., The magnetic resonance force microscope, in Progress in Nuclear Magnetic Resonance Spectroscopy 45, 239-274 (2004). DOI 10.1016/j.pnmrs.2004.06.001.
3 February 2015, Stan
Errata: Christian Degen pointed out to me that their experiments, like the ODMR ones, are done at room temperature. In the original version of my article I have got it wrong, thinking that they were cry-cooling the detection assembly. To minimize any spread of inaccuracies, I have corrected the article text.
Christian also pointed out to me this pertinent link to the ETHzürich News & Events page:
Extremely high-resolution magnetic resonance imaging.
January 23, 2015
New NMR/MRI books
The recent crops of NMR and MRI books are abundant and all the books would no doubt merit a few words. Let me point out at least two of them, for two different reasons:
The Introduction to Spectroscopy by Donald Pavia et al, could easily escape attention of NMR spectroscopists because it does not have "NMR" in its title. However, it dedicates full five Chapters (over 360 pages) to NMR spectroscopy, in addition to IR (infrared), MS (mass), and UV (ultraviolet) spectroscopies. It would be an excellent buy, I think, were it not so expensive, despite being presented as a textbook ($284 to buy, $66 to rent, and there is no Kindle edition). Anyway, if you can afford it, go ahead and give it a try. Especially if you need also an introduction to one of the other spectroscopies. The accent is on organic analytical chemistry aspects, though a fast introduction to physical principles is also given.
The 2014 edition of the popular Reference Manual for Magnetic Resonance Safety, Implants, and Devices, edited by Frank Shellock, is intended for MRI developers, manufacturers, site managers, as well as plain Users. Managers of in-vitro NMR facilities, however, can also learn a lot from it. I mention it here not because it is an unusual novelty (the Shellock safety manual has now become an almost periodical publication), but because safety is of paramaunt importance. A single lapse in safety can give a bad name to a whole facility or even to the whole trade. Price: $70. Again, an updated Kindle version would be appreciated, but it does not exist (yet).
January 11, 2015
A personal view of the state of NMR at the start of 2015
NMR is presently going through an unusual phase. For a number of reasons, and for the first time in history, traditional NMR spectroscopy is commercially unprofitable. It was totally abandoned by one major manufacturer, while others are anything but enthusiastic about it. At the same time, there are many new commercial ventures aimed at considerably different types of NMR instruments (compact, sub-compact, table-top, cryogen-free, portable, ex situ, ffc, ...).
The worrisome situation is limited to high-field NMR spectroscopy (HFNMRS) which was for many decades the staple of just three manufacturers: Varian (later Agilent), Bruker, and Jeol. It does not regard MRI, whose market keeps growing at a steep yearly rate of 11-12%. Sincerely, even in MRI, I do not think that the growth is a merit of innovation. My feeling is that, compared to three years ago, there is less talk about fNMR, spectroscopic MRI, molecular MRI, interventional MRI, diffusion and susceptibility imaging, or even just novel pulse sequences. Rather, the steady growth of MRI is driven by the fact that the majority of people on this planet still have only a very limited access to this important diagnostic tool. Even with limited innovation efforts, MRI is therefore enjoying a vast potential market. With current production capacity, any saturation of that market is out of question for 100 more years.
HFNMRS, on the opposite, faces a saturated market, mountains of used but recyclable hardware, and disenchanted manufacturers who have nearly forgotten the word "innovation" (indeed, I suspect that they might be dreading it).
Recent market reports show that between 2012 and 2014 the market for HFNMRS has slumped. The drop does not correlate much with the overall economy (it is true that Europe is in an economic quagmire, but USA are running strong, and some parts of the World are doing even better). Some explanations, apart from the European weakness, might be the following:
- In 2010 - 2011, many big Pharmas reduced their research facilities. Since, excepting academic NMR Centres, Pharmas represent the principle HFNMRS users, their crisis has dealt a sizable blow to that market.
- Many Pharmas pushed their downsizing as far as actually dismantling existing facilities (some with tens of NMR spectrometers per facility), sending lots of NMR hardware into the second-hand instrumentation market to compete with newly manufactured units.
- NMR as a structural-analysis chemical tool is no longer as prominent and unique as it used to be. Even though my friends among the chemists assure me that NMR is as highly considered as ever, it is evident that other analytical methods have made impressive advances. HFNMRS looks today as a mature technique, developing at a substantially slower rate than in the 80's when its innovation rate was peaking.
While the NMR market was slowing down, the principal manufacturers of HFNMRS equipment were (and some still are) facing other, more specific, challenges:
On October 14, 2014, Agilent Technologies have abruptly terminated all their NMR activities. The brusque, almost offensive, announcement came as a surprise even to most Agilent employees. Well, not all, perhaps. The eminent scientist Erics Kupce, for example, must have figured it out already by the end of 2012 because he resigned from a nominally excellent position at Agilent, effective January 1, 2013. But Agilent NMR guys in Italy, for example, were in the dark until the announcement.
Agilent entered NMR on July 27, 2009, by acquiring Varian Inc, the company born in 1999 from a split of Varian Associates. The latter, in turn, used to be a much studied textbook example of Silicon Valley industrial innovation capabilities in the post-WWII period (roughly up to 1975 - then it became more complicated). The Agilent's decision to drop NMR de-facto guillotined the last vestiges of a once glorious Chapter in NMR history.
It is ironic to think that in 2009 Agilent entered NMR at the sound of fanfares: they spent 1500 million dollars to acquire Varian, including its magnets production facility in Yarnton, UK, (ex Magnex), only to drop it all at a huge loss mere five years later! I was at the public grand open-house kick-off of new Agilent NMR premises (Santa Clara, April 8, 2011, prior to ENC). At the meeting, where historic personalities like Weston Anderson were present, American Chemical Society conferred to the would-be continuators of Varian traditions a plaque stating that the mythical A60 NMR spectrometer was declared a National Historic Chemical Landmark. There were bright talks, great buffet, and a lot of hype. However, the two questions that I fielded (how would Agilent handle the maintenance of the installed base of legacy Varian instruments, and how would Agilent react to the forthcoming helium crisis) were answered in a way that did not convince me.
My guess about why Agilent totally dropped NMR only 5.5 years after the merger:
- They declared to SEC, and therefore to their investors, that the 'merger' would cost 1500 M$, and that they 'expected a 20% return on it in 4-5 years'. After 5 years, instead, there was a large loss. Hence, since the takeover was no bleeding NMR heart operation, it was scrapped and written off as a deductible loss.
- The problem emerged just when Agilent was going through another internal crisis. The following announcement appeared on all their websites, starting at the same time as the NMR dismissal (it is still on most Agilent pages):
"As of November 1, 2014, Keysight Technologies, the division of Agilent Technologies that develops and sells electronic measurement (EM) products, will de-merge from the Agilent Technologies group into a separate publicly traded company, ..."
Now, Agilent's EM products were highly valued by electronic engineers worldwide. To thousands of electronics professionals, the term Agilent was simply equivalent to top-quality electronics-lab-instruments. I have realized myself only after I have seen the above announcements that this regarded just an acquired subdivision.
And Agilent was about to lose it!
- In those 5 years Agilent did not contribute any innovation worth mentioning to the ex-Varian products. They redesigned the closet on the DDR console, so that they could put the new logo on it and call it DDR2. Internally, they changed only one board, as far as I know (the filters). They did development on probes, but nothing revolutionary. And that's it. Everybody expected that the masters of RF and MW would redesign it all (the DDR console was an ageing heredity of the ex Chemagnetics takeover by Varian Inc). But those masters (Keysight) were in another department and, as it turned out, ready for a divorce.
- To top the mess, Agilent decided to relocate the production of the DDR2 console from New Mexico to Malaysia, while, at the same time, software development was being moved from India back to California! Can you imagine!
- World-wide maintenance of the ex-Varian legacy base must have been a nightmare. At the grand opening, they said they would rely on a software package to handle it! And already then they, themselves, anticipated problems due to lack of RoHS compliance of the old stuff. It is illegal nowadays to sell or even just to ship around old tin-and-lead containing electronic boards. So what you do if that is the only kind of spare parts you will ever have? With no plans for spare-parts production you are stuck to what you can cannibalize from discarded units, and you can't ship it anywhere without breaking the law. It did not make much sense to me.
Anyway, all this is History today.
And Agilent should officially return the A-60 landmark plaque to ACS, I think.
Bruker Corporation, of Swiss-German origin, is now US-based, with an impressive (excessive?) range of activities. NMR represented most of their business from 70's to mid 90's, but it is nowadays in minority, though not yet marginal. In the last decade, investments into innovation targeted mostly branches other than NMR.
My very personal opinions:
- The overall quality of Bruker NMR is high, but it is based largely on the heredity of the past (except for high-field magnets - they did a good job on them).
- Bruker business structure is geared towards commercially heavy items. In the past they went into cryoprobes. More recently, they went into DNP, adopting its most costly variant. To address the helium supply problem, they have chosen online helium re-liquefaction - again the most costly option. Their 'normal' probes are good, but outrageously costly. Etc. It is not a technical problem, I think. They probably cannot make profit on anything costing less than ~70 K$.
- On the applications side, it is 7 years I hear the same talk about orange juice ...
- They remind me of IBM when the PC appeared on the horizon. There are impending NMR revolutions, but I suspect that Bruker is dreading any of them that might mean lower prices!
Bruker hard facts gleaned from public documents (study the link, particularly the 'COMPANY' section on the left, including the SEC filings):
Bruker is today a public Company with 'insiders' still detaining a sizable part of its shares, but not a majority (about 35%). Presently it declares 6200 employees, compared with 6700 in January 2014. Among major shareholders there are banks of the like of J. P. Morgan. In the third quarter of 2014 they declared a loss. Commercially speaking, they estimate that NMR will perform poorly in 2015 - 2016 (at best in the 'lower single digits'). A year ago, to tranquillize their shareholders, they announced ongoing drastic measures: reduction of operating divisions (from 10 to 3 just in USA), reorganization of all businesses, continuing 'headcounts' (HC) reduction, continued outsourcing, relocation, and dismissals ... (source: Bruker JPMorgan 2014 Healthcare Conference presentation, pages 14,15,18,20).
As I wrote on a LinkedIn forum, I do not enjoy all that at all. I feel sympathy for Bruker and I would much prefer to see them snap back into shape. But right now, they are going through serious troubles and they view NMR more as a culprit than as a solution.
I cannot find out much about Jeol (the Japanese 'SEC' is not as open-access as its US counterpart). They were keeping low profile recently, mostly focusing on probes (MAS microrotors, cryo-cooled MAS), in collaboration with Universities. Somewhere I read that they had kept NMR on a stand-by for three years. On the other hand, they have recently installed a 600 MHz instrument in my birth town, Olomouc (Moravia, Czech Republic). I was amazed because Olomouc is a small place, and Czech Republic is in the heart of Bruker territory, jealously guarded by Scientific Instruments Brno. As always, Jeol's presence is a patchwork: significant in some Countries (UK, Scandinavia) and weak or null in others (I don't know about USA). Overall, they do not appear to be a formidable Bruker competitor, though on the technical side they do have all the prerogatives. Their magnets supplier is Jastec Superconductor, a very respectable Company indeed.
Wrapping it up:
Something worrisome is happening to HFNMRS instruments manufacturing and R&D. The market is depressed and there are dangerously few manufacturers. One manufacturer has closed shop, and the remaining two are anything but enthusiastic about NMR. At the same time, there are several new players, but they are still far behind in both volume and technology.
There is a lot of used hardware, so the NMR community can go on for 2-3 years regardless of manufacturers. Competent tinkerers and maintenance engineers will be much needed, of course, but there are lots of them. They are being made 'redundant' even while you read these lines. Btw, the capability of the community to go on is the only thing that can keep Bruker from exaggerating with prices. The vicious price wars of the last decade are over, of course, and the net tabs for new equipment will go up, no doubt about that. But, maybe, at least the list prices will not change.
Might it be that something big is looming on the horizon?
For would-be players of tomorrow, this is probably the best time to step in! It could be one of the smaller ventures (there are several candidates), or it could be Jeol. Or, more likely, one of the 'official' competitors of Bruker, which are Thermo Fischer and Danaher, both about ten times bigger. Though, have a look at this.
Oh well, so many things could happen ...
COMMENTS are welcome (especially from the Companies in question):
12 January 2015, Carlos Cobas, Mestrelab Research
Hi Stan. Just to let you know: Ray Freeman and Garret A. Morris have recently published an article entitled The Varian Story which you will no doubt enjoy.
12 January 2015, Stan
Thanks, Carlos. I must also acknowledge that you kept me posted with Agilent news during last three months. I liked and commented the article Agilent Draws Academics Ire on C&EN, and I have printed and saved your email containing a copy of the Agilent Open Letter which appeared on Agilent's Spinsights Blog. Btw, I can no longer access Spinsights - I guess it is dead now.
13 January 2015, Craigh Butts, Bristol University
Hi Stan, Just a couple of comments on your 'State of NMR' blog:
Firstly, I think that is a nice overview, balanced and pretty fair given the scope of the situation!
Secondly there are a couple of relatively minor (but perhaps important) points I'd like to add:
(1) I agree that from 2009-2013, Agilent had very little impact on the hardware but they did have a massive impact on the software - the latest versions of VNMRJ are much more powerful than anything I've seen in the past (I just installed a new probe, and the software shimmed it from zero shims and had it at lineshape spec, with full multinuclear calibration and signal:noise spec within 13 hours all under full automation (I did nothing after I had loaded the calibration samples onto the robot and pushed the 'Go' button). My research chemists can set-up and run a dozen different selective multi-dimensional NMR experiments in less than a minute (if they are being slow) and the gradient shimming on our routine CDCl3 samples takes less than 12 seconds. That just wasn't possible 5 years ago.....
(2) In 2014 Agilent released some absolutely gorgeous new hardware - in particular their new coldprobe which whipped the performance of the existing cryo/cold probe technologies in terms of sensitivity (the 600MHz coldprobe had better sensitivity than 800MHz cryo/coldprobe). And they had a new console and probe ready to release (it would be out by now if they had not closed up shop) which had a pretty large number of modifications and improvements.
(3) JEOL have actually had a number of hardware developments in the past 12-24 months that are directly competitive with Bruker in the HFNMRS field e.g. 13C-cryogenic (helium) probe, a liquid-N2 cryogenic probe (both of those cryoprobes are at 500-600MHz, so limited field availabilities at the moment) and ultra compact (400MHz) spectrometers (and a few less interesting ones like He-recycling magnets and new sample-changer). Their software needs a lot of work, but their hardware is actually pretty competitive.
13 January 2015, Stan
On your first 2 points: You are right that I left out the period of Agilent NMR enthusiasm following the Varian takeover. Agilent had retained almost all Varian STEM employees, and since they were all unhappy with the old Varian management, they had been truly delighted. They would say "finally things will start moving" and in fact they did, including R&D support. This lasted 4 years. The first layoffs started in Autumn 2013 when probably the nasty decision was taken (at the investors level, I would guess). So, there were four years of motivated and competent work! That must leave a trace, and your comment illustrates that it did. Were the results insufficient, or were they too late? Or were the initial expectations too high? Or did the market really drop so hopelessly that developing for one more year was out of question? The 2014 period is the most puzzling one: rumors about great new hardware were persistent, but I am 99% sure that the decision to stop was already taken. I can only speculate on what was going on there between October 2013 and October 2014. One thing is sure: the destinies of HFNMRS are no longer in the hands of people who know what a spin is.
14 January 2015, Vanni Piccinotti, ex VPNMR (Italy), ex Varian
Ciao Stan. From this situation there MUST emerge in Europe at least 3-4 persons willing to start an independent technical service for the Varian/Agilent installed instrumentation, as well as sales of used, reconditioned units. Should this not happen, it would be - in my opinion - a terrible signal of total lack of entrepreneurship and of willingness to accept any risk. As you know, I have done it many years ago, it worked very well for me, and if I just had 10 years less, I would go into it again.
15 January 2015, Stan
Hi, Vanni, what a pleasure to hear from you. You are totally right, and I am afraid that EU entrepreneurship is, indeed, in a woeful shape. But the situation is not totally bleak. The French company RS2D is well positioned for refurbished Varian (and Bruker) instruments market (look for RMN/Reconditionné on their website). The German Spin-Doc, run by Bert Heise (an ex Varian/Agilent guy whom you certainly know well), is available for maintenance, as well as whole units recycling. Another existing German company, NMR Service is well positioned. They are, among other things, a Tecmag console dealer, a feasible option for ageing ex Varian systems, at some point.
But you are right, given the volume, and the geographic spread of the installed ex Varian/Agilent 'park', much more is needed. For example, what are the Italian Agilent NMR engineers waiting for? Or, with your competence as a resource, you could start a youngster on this road; an independent one, or maybe someone hired for the specific purpose by an Italian NMR company such as Stelar and/or SpinLab.
19 January 2015, Vanni Piccinotti
** Point 1, existing activities: I knew of most of them. Very good, but they all work independently, when the most obvious thing is to co-operate with each other to supply good service and products over Europe. To compete with each other would be total madness, there is plenty of room and work for everybody.
** Point 2, ex-Agilent people: They are still recovering from the bombshell, I guess. Give them time, something will hopefully come out.
** Point 3, training a youngster: I tried years ago, but could not find anybody interested in such a job. Reasons: too much to study and too much work, travelling around and often far from home, too many working hours, etc. This is the situation, at least in Italy (no comment). For what I know, big Companies also have problems to find willing (and capable, on top of it) servicemen for complex instrumentation like NMR.
20 January 2015, Stan
Vanni, your comments could start many new discussions!
For the moment, I hope others (like ex-Agilent NMR guys) will contribute. I have checked on LinkedIn; in Europe, most of them are apparently still with Agilent which means that they might eventually keep their jobs, but drop out from NMR.
On another front: JEOL seems to be waking up and moving. At least in Italy where they have even translated into Italian (an overkill!) their promotional materials for the new line of JNM-ECZ instruments.
27 January 2015, William Reynolds, University of Toronto
Dear Stan, here is a belated response to your Jan. 11 post.
On Oct. 16, 2014, just after the closing announcement, I posted on the Spinsights blog how inappropriate it was to have the A-60 plaque at Agilent. I also mentioned this to Judith Shoolery, Jim's wife. She contacted Martin Packard and Wes Anderson who intervened, presumably with ACS support, and I understand that the plaque has now been returned to Palo Alto. With respect to timing, I agree that the decision to close the NMR division was really made in the autumn of 2013. The departure of Nick Roelofs, the VP in charge of NMR occurred then and was, in retrospect, a clear warning sign. What I don't understand is why Agilent then still kept investing a large amount of money in development of new and very impressive NMR products, as mentioned by Craig Butts, when there was little possibility that they would ever be sold. Certainly that left the NMR staff optimistic about the future, as was apparent to me from discussions with some of them at SMASH in Sept., 2014. I know that, except possibly at the very top level of the NMR division, there was no advance warning of the closing to the NMR staff. In a broader sense, pressure from stock holders and SEC regulations certainly makes it difficult to have scientific innovation in the U.S. It seems to be better for companies to get a tax break by closing a division and writing off losses than to keep it running at a low profit margin.
28 March 2015, Stan Sykora
I have mentioned above that it will be interesting to see the official numbers regarding Bruker 2014 financial performance. Those public statements were released on February 19 and they certainly are of interest to the NMR community worldwide (especially now). Since I have already commented on them in a LinkedIn discussion, however, I will limit myself to copy and paste what I wrote there:
Bruker 2014 global, as well as 4th-quarter (Q4), financial results were made public on February 20. You can view also Frank Laukien's (Bruker CEO) presentation slides. They use double accounting: GAAP and non-GAAP, which makes it a bit difficult to understand them (GAAP stands for Generally Accepted Accounting Principles). The main results pertinent to this discussion are:
* As a whole group, Bruker reports -2% drop in 2014 revenues (slide 5) and -8% drop in Q4 (slide 4). That translates as stagnation turning into recession. The drop is mostly in Asia; Europe is flat, and there is a very mild growth in the Americas.
* Bruker has 4 major divisions, BIOSPIN (which includes NMR and pre-clinical MRI), CALID, NANO and BEST. The worst hit is BIOSPIN, even though MRI is doing quite well! The main culprit is NMR spectroscopy. Frank Laukien says: "Revenues decline in NMR business, as demand for high-field systems remains muted"!
* BIOSPIN is subject to "restructuring, right-sizing and footprint consolidation" (meaning shrinkage, slides 6 and 8). Discussions with labor are in progress since a 9% drop in headcounts (employess) is being planned (slide 6).
* While NMR in BIOSPIN is the worst performer, the best one is Daltonics, part of CALID, with its marvel product MALDI Biotyper (nothing to do with NMR).
* Slide 8 (2015 Priorities) is most revealing and requires no comments!
* 2014 investments as a percentage of revenues (which themselves dropped) went down -8% with respect to 2013 (slide 14), and Q4 R&D dropped -15% (slide 11). That means an accelerating R&D reduction!
One can read many more things from the slides, but not a single one is pleasant. To me, the dropping R&D is the most worrisome symptom. Where will be the reduced innovation efforts directed: on the worst performer (NMR) or the best ones (MAHDI, pre-clinical MRI, ...)? And keep in mind that, today, Bruker is not crucially dependent on NMR. They have tens of scientific instrumentation branches, and should they drop the worst performer, from the business point of view, they, as an Enterprise, would be certainly better off.
For us, depending on a similar monopolist is EXTREMELY dangerous.
Have a look at the data and judge yourself.
For a complete list of all entries since 2005, see the running INDEX
ADVERTISE with us
NMR Lab Store
NMR | MRI | ESR | NQR
Centres & Groups
Journals & Blogs
MATH | SOFTWARE
PHYSICS | CHEMISTRY
ELECTRONICS | DSP
Patents & IP | WWW
MRI | NMR | ESR
Biography of F.C.Yu
MR Antenna Theorem
NMR Dead Time
One-Page MR Primer
K-space and MRI
Support this site!
Shop from here:
New NMR books:
New ESR books:
New MRI books: