The end of an era at NIST

In preparation for the NIST (then NBS) Reactor producing neutron beams starting in 1969, the science leader, Jack Rush, assembled a top-notch team of scientists that included crystallographers Ted Prince, Antonio (Tony) Santoro and, just a bit later, Judy Stalick. With the passing of Ted, Tony and Judy, I’d like to put a few words in remembrance of what was likely at one time the most respected neutron crystallography team in the USA. I was fortunate that after Ted officially retired (but still came to work each day), I was able to work with them.

It may have been chance that the reactor started just as Hugo Rietveld’s work was published. The team was a very early adopter of his technique. I can recall Judy more than once telling me how gracious Hugo was on the occasions they met. I had the good fortune to experience this myself. 

Ted died a few years ago. Tony died about a week ago. Judy died this morning, after much more than a decade with severe pain and disability. I’ll write a bit about each of them individually.

Ted was a crystallographer’s crystallographer. He was an expert on so many different aspects of crystallography, but perhaps above all he was a leading authority on the math and statistics we use. In fact, it was Ted who proved the correctness of the Rietveld technique. Crystallographers prior to Rietveld would chain a fit for integrated intensities of individual powder diffraction peaks together with a structural fit, by using those output intensities as input to single-crystal software. With that approach, they would see very different standard uncertainties (then called e.s.d.’s) in comparison with the values obtained from Rietveld fits. This from their view indicated that something was wrong with Rietveld's technique. The “Rietvelders" argued that chaining least-squares fits was invalid. Ted demonstrated that chaining least-squares fits together serially is perfectly valid statistically, provided that the covariance matrix from the first fit is used as weighting in the second fit — something not possible in standard single-crystal software (as weights are input as a vector). When Ted wrote software that allowed this, he showed the s.u. values then agreed with the Rietveld values. QED, the Rietveld technique gives correct answers.

Ted was very proud to have gotten his crystallographic education in the UK; not a common thing for a US undergrad to pursue in the 1950s, at a time when the country was rebuilding from WWII. My father (who grew up in London during the WWII bombings) finished his Queen Mary College undergrad degree around then and came to Canada for his graduate work. Ted would sometimes mention different happy experiences he had in those years, but the one story I remember best was when he told us how impressed he was at the time to be introduced to Rosalind Franklin (1920–1958) when she came to see the lab where he worked. Back then crystallographers knew who she was and what she had done, even if the rest of the world was yet to learn it.

Ted was the editor for the initial edition of Volume C of International Tables of Crystallography and its first revision. He asked me if I would then take over for the second revision. I could not imagine ever finding the time for that, but now wonder how he managed. Ted’s book, Mathematical Techniques in Crystallography and Materials Science, is considered a classic in the field and nearly 20 years after its 3rd edition was released is still available on the Kindle. I will warn the reader that while I go back to this text often, I do budget at least a few hours per page. Ted assumed his readers know a lot. Larry Finger told me that when he retired and gave away his library of crystallography texts so that he and his wife could fit into a mobile home, Ted’s book was the only one he retained. Ted and Larry were also well known for their fine crystallography package, which included a Rietveld code customized to the 5-bank and later 32-bank BT-1 instrument at the NIST reactor.

Tony came to the US from Italy after his education. Tony was loud, opinionated, blunt and so, so much fun to spend time with. On one occasion, when I was discussing with him the classic work of Cagliotti from the 1950s, he corrected my pronunciation of the name and then added "and he was quite an SOB” but in slightly more colorful language. When I was asked to be the crystallography team leader, while still in my early 30s, I found it very daunting to supervise someone who was a world-level practitioner and also about the same age as my parents. Tony was never anything but gracious and kind towards me. In fact, at one point I came to him saying that the end of the fiscal year was close and this was a good time to put in a request for purchases, as the beloved benevolent despots (and I mean each of those three words for Mike and Jack) who ran the reactor always kept some money in the budget in case of unseen problems, but needed that to be spent before the books closed. So, smaller quick purchases might fly through if put in at the right time. Tony asked me if he could get his first PC and a Fortran compiler and then told me that I was the first person in his many years at NBS/NIST who had ever asked him if he wanted to order something. He got the PC and g77 before the end of the day. 

Tony loved to walk, commuting by Metro and bus from DC to the front gate of NIST, where he then walked a mile or so to the Reactor and did the reverse in the evening. Just once, on a cold rainy day, he accepted a ride from me. Another favorite memory was while at the 2005 IUCr Congress (Florence), Tony invited me to his apartment in Lido di Ostia. However, he had a bad experience in the Rome train station, and he insisted on meeting my train and escorting me to Ostia lest the same happen to me. He also insisted I should sleep in his bedroom while he slept on the couch, as it would be too noisy for me to sleep otherwise. Alas, with the non-stop partying all around us all night, even with his room and earplugs, the sound levels were still more than I could manage.

Ted, Tony and to a lesser extent Judy, were crippled in their work by a decision at the NCNR in the very early 1990s to move all computation and data analysis from a VAX computer to Unix machines. It was true at the time that Unix machines offered much better performance for the buck, but almost all crystallographers were using VAXes in those days and very little software was available for any other platform. Each was very comfortable with the VAX/VMS OS and found switching to Unix bewildering. Once they no longer had VAX access, Ted and Tony gave up doing Rietveld refinements themselves. Ted moved to more theoretical work and Tony switched to problems where he could use a pocket calculator, until about a decade later when he got that PC. Judy struggled along with Ultrix but was never comfortable with it. The NCNR continued to use VAXes to run instruments for at least another five years at that point, but the loss of effectiveness for crystallography vastly outweighed the high cost of a small microVAX. What a waste.

Ted Prince and Judy Stalick at the 2nd generation five-detector BT-1 instrument.

Tony hired an assistant scientist, Qingzhen Huang, more-or-less as a postdoc, despite Qing not having the Chinese equivalent of a PhD. Qing was a terrifically productive scientist and became a world-class expert in magnetic crystallography. At one point I was told by someone outside NIST that Qing was also the most prolific publisher of Rietveld results in the world. This was doubly impressive to me, as magnetic crystallography is not easy and even for non-magnetic materials, neutron data collection might require days/sample. One would expect a synchrotron or even lab scientist to have access to more data, but the truth is that Qing did excellent work and a number of notable synthesis experts knew that and sent their samples to Qing. With time, Qing’s expertise and productivity was recognized by the NCNR and he became an independent full-fledged member of the team. Qing was of great help to me as I started to learn about magnetic scattering and was always ready to step in to help users run samples at BT-1.

Judy was a superb but a more practical crystallographer, in comparison with Ted and Tony. Ted Prince is credited for the design of the 3rd generation (32-detector) BT-1 instrument, which still offers the highest resolution neutron diffraction in the US ~30 years after it was initially deployed. (Sad that it has seen no major updates since!) However, it was Judy who made that instrument happen, working with engineers and technicians to get it done and commissioned. She also never stopped tinkering with the instrument to improve performance. For example, by convincing a bunch of us to work on pressing and aligning Ge(311) and Ge(733) crystals, two seldom-used monochromators were replaced with capabilities that greatly enhanced the range of the instrument by providing significant wavelength shifts from the initial design. Later she figured out that use of a "zero-blade Soller collimator" would provide ~60' collimation, doubling the speed of measurements at the cost of some resolution, and that mod would be pretty simple to install — by replacing one of the two “off” positions in the neutron shutter with that open collimator to offer a third collimation option. Cheap and highly effective! (See for yourself here.)

Judy had been a grad student of Jim Ibers and had been back then “very married” (in the words of someone who overlapped with her) to another chemistry grad student. I can recall Judy saying that she would take breaks at her desk in Jim’s lab and read a novel or newspaper. Feeling very confident in her work, she shocked the other students that she felt no need to hide what she was doing when Jim came into the lab. Jim clearly respected that too. Following her husband’s hire at a faculty position in Virginia, she wanted to find a position nearby that would also allow her to start a family and took a job I feel was below her level, working on the NBS*CrystalData database. Despite an arduous commute, she felt NBS would be good for her overall goals, albeit perhaps not very professionally ambitious. As congestion, construction and for that matter her health issues got worse, she dropped to working a 30-hour week sometime later. In her CrystalData job, Judy was pivotal in the design of the first computerized data storage format (then called AIDS*83, later renamed as PDF-2 for some reason) for storage of both the JCPDS (now ICDD-JCPDS) and the NBS database, along with data entry software for the folks who abstracted from the literature. She would tell me how the ICDD had to be talked into using the computerized format. By that time, the ICDD was marketing a PDF-2 product on the recently-invented CDROM platform; it was already clear that computerization had saved the ICDD from obsolescence. After just a short period with CrystalData, Jack noticed Judy’s talents and recruited her to join the reactor staff. It was at the NCNR where her talents could be more fully recognized.

Judy specialized in intermetallic systems, but like everyone else in the neutron field, also did plenty of oxide crystal structures. She never lost the NBS credo that anything published from there had to be 100% above any critical dissent — even to comma placement — but she also never gave up on any project. She was willing to keep picking up a project again and again to get it to “done." I can recall one intermetallic with a huge cell that she revisited perhaps a dozen times over 15+ years.

The scooter provided by NIST that allowed Judy to get around Building 235. She was provided with a special parking spot in front of the building (hers was the only non-government vehicle allowed to drive on that section of road). Photo thanks to Peter Gehring.

Alas, if being a woman scientist and the expected roles for women in the 1980s were not enough of a handicap, she also had to face several others. Judy lived with crippling back pain and operations and drugs to treat that, at one point getting methadone treatment for addiction to prescribed pain meds. Her professional visibility was undercut by her inability to travel to meetings, as family responsibilities and the long period of pain that followed a plane flight prohibited travel. Sadly, if her life was not already demanding enough, Judy’s then husband left her when her son Jon was a pre-teen. That shortened Judy's commute, but left her as a single parent. She was apparently very effective in that. Jon has grown into an impressive human being, who was a devoted caretaker over the past two decades as his mom disabilities only compounded. To the credit of the NIST Center for Neutron Research’s management, she was offered amazing levels of physical accommodation that allowed her to keep coming into work until just a few years ago. She loved her work and the NCNR community and gave it up only when there was no other possibility.

I came to NIST originally as a facility user when employed at Air Products, so I could use the newly upgraded BT-1 for zeolite studies. I felt very much to be Judy’s student in CW neutron diffraction; I had done TOF neutron PDF work with Takeshi Egami and Jim Jorgensen before, and some, but not a lot of Rietveld. I was shocked when Jack suggested that I should consider applying for the job Ted was leaving, since I could not envision anyone appropriate to fill his shoes and certainly not me. I don’t remember how I managed to put together a seminar, as I could not talk about any of the work I did for Air Products (or perhaps being vague was good), but I got the offer. Unlike my prior terrible industrial job at Union Carbide (Rest In Pieces), Air Products had been a great place to work and I felt a lot of loyalty until a massive layoff not long before the interview, so my wife Diane, then 13 month-old son Jason and I left Allentown for Potomac as quickly as we could, living in temporary housing while house hunting. For what it's worth, I don’t think a single one of my coworkers from back then remained at Air Products long enough to retire. Only a few left voluntarily. Employee loyalty is overrated unless reciprocated. 

I continued to learn from Judy once her coworker. I’d like to hope she learned from me too. She would have been the team leader if she wanted the role. Instead, she had to put up with me as her supervisor, but she also did not hesitate to tell me on occasion how to do my job. I listened and learned.

My final project at NIST came from my belief (still held) that the best thing one can do with a thermal beam from a reactor is to build a Laue diffractometer. (Some triple-axis folks might disagree.) I infected Tony with that belief and together we worked on Laue diffraction geometries together as I built a prototype instrument without any funding. It was a fun project. I wish it had gone farther. 

I spent a decade at NIST. It is now just a few years short of two decades since I left. The NCNR was an environment filled with top-level scientists who (perhaps unusually) were almost all gracious, caring and collaborative. In all the places I have worked, that environment was unique, and I still miss it. But even more, I miss the collaboration and learning with Ted, Tony, Judy and Qing. Perhaps my arrival was the beginning of the end there, but at least during my professional lifetime, I don’t think any other American institution can claim to have employed three crystallographers of that stature. I personally owe them a debt that I cannot repay; I will just try to pay it forward. Farewell and thanks, Ted, Tony and Judy. 

Postscript: In the week or so since I wrote this, I have found myself listening over and over to the tape that I made of a concert I organized for the late Stan Rogers (some Canadians will recognize the name) while a grad student. It is a comfort as much as it also reopens a deep feeling of pain. I suspect that as I process the loss of people who meant so much to me, and flirt with thinking about my own mortality, I want to remind myself of how many wonderful things I have had the chance to do. I wish the same to all. Brian

This article was first published in ACA Reflexions, the Newsletter of the American Crystallographic Association.