The Making of an Experimental Physicist

by Martin C. Weisskopf

In September of 1960 I arrived at Oberlin College to start the adventure that would become my professional life. I came from Chicago, where I had grown up, an only child of immigrant/refugee parents who had successfully fled to the USA from Austria in 1939 after the Anschluss. This was a remarkable feat in and of itself and no doubt made possible in part by the amount of wealth that the Weisskopfs had possessed in Vienna and had to leave behind. Also, my father’s younger brother had already immigrated and was able to sponsor my parents to enter the USA. Both my parents had been lawyers in Austria and had PhDs from the University of Vienna. The difference between Austrian and American legal systems meant that they could not immediately exploit their legal background here in America. But the PhDs allowed them to teach, and my father first obtained a college-level teaching position in economics, first in Omaha Nebraska where I was born in 1942, and subsequently in Chicago at the YMCA College which would eventually become Roosevelt University, where he taught until he retired. My mother first concentrated on raising me and took many jobs over the years including being a proctor at the University of Chicago. Eventually, she too ended up at Roosevelt University as a Professor of Romance Languages.

Since we moved from Omaha to Chicago when I was less than a year old, my growing up experience was location stable. Indeed, we lived on the third (top) floor in the same three-story railroad car apartment building at 5555 Blackstone in the Hyde Park neighborhood on the south side of Chicago from the time we arrived from Omaha until I left for Oberlin. I thoroughly enjoyed those years. Academically, I first went to Ray Grammar School for grades 1 through 8. Ray was a public school run by the city. I also received home tutoring in Latin for a year or two during this time as my parents were shocked that neither Greek nor Latin were being taught at Ray. Their classical European education was broader than what I experienced, and they often broke out in gleeful recitation of some poem in Greek or Latin due to some random (to me) stimulus. I made many friends during the grammar school years. My best friend, P. Alving, was the son of a doctor who, if I recall, led a malaria study at the Statesville Penitentiary which included working with the infamous Leopold and Loeb (see below). At home I was subject to piano lessons (which I hated), introduced to the Episcopal Church in case I wanted to be President (which I enjoyed especially since the church located across the street from our apartment had a basketball court), participated in the yearly cycle of sports with my friends (Chicago softball – sixteen-inch diameter and no gloves – tag football, basketball, ice-skating and ice hockey under the stands at Stagg field, etc.)

Because my best friend at the time was going to do so, I asked my parents if I could attend the Harvard School for Boys high school, also in Hyde Park. Harvard was small (I believe my class size was about 16) and the smallest private school in the city’s private school league. Harvard was also infamous as the school that Bobby Franks attended, the victim of a sensational murder by two students (Leopold and Loeb mentioned previously) of the University of Chicago’s Laboratory School also in Hyde Park. I enjoyed my education including 4 years of Latin, made easy because of the tutoring my parents had imposed, and mathematics and science, which I always was good at. During my senior year I was allowed to take a course in set theory at the Illinois Institute of Chicago. The high school years were great and, while I drifted apart from many of my grammar school friends, I made new friends, for the most part older than myself. At various times, I participated in several different school sports (football, tennis, and basketball) and I also played organized basketball at the local Boys Club. I also met and enthusiastically dated my high-school sweetheart, S. Abrams, whose father was an attorney. I held different part time jobs while in grammar school and high school including working as a proctor for University of Chicago college tests. (Of course, I had an in as my mother was a proctor too.) One job, perhaps an omen of things to come, was testing Geiger counters built by the N.J. Wood company.

One is supposed to have some idea of an eventual future after college so, because my parents had been lawyers in Vienna and because my girlfriends’ father was a lawyer, I thought that was what I should plan to be. This meant Government courses at Oberlin, their prelaw program of study. In addition to other required courses, I signed up for the first year of calculus because I had always been good at math and the Poet’s Physics course which did not involve calculus but did require a term paper and thus, I thought, would be easy. By the end of my first semester, several factors changed the direction of my professional life. First, I received a B- in my government course, clearly indicating that law was not for me. Also, while researching my Poets’ Physics term paper, I made two influential discoveries. The first was that my uncle, my father’s younger brother, was a famous physicist! (Manhattan Project, Professor at MIT, first Director General of CERN, etc.) The second was learning from my Poet’s Physics term paper on the development of quantum mechanics in the early part of the 19th century that physics was fascinating and that the scientists were having fun making advances in our understanding. Based on these two factors, I changed my major from Government to Physics. I also began dating V. J. Hanfmann. who would become my first wife while still at Oberlin, and later the mother of my two children, Antonia, and Alexander. The summer after my freshman year, I followed V.J.H. to Berkeley, California where she had taken a summer job. I enrolled in summer school at the University taking third semester Calculus and a course in linguistics. The only word I ever comprehended from the postdoc who taught the calculus course was trivial. I believe the average grade on our tests was 60!

My physics education at Oberlin followed a path somewhat different from that of a typical physics major. This was because of my unusual start with the Poet’s Physics course. In my second year I participated in a National Science Foundation-sponsored intensive summer course based on the first chapter of Leighton’s classic text on Modern Physics. The tensor calculus made special relativity and electricity and magnetism much clearer (and easier) for me. By my junior year, I began to develop an interest in working with hardware and performing experiments, which I realize now may have traced back to my youth in Chicago working on my friend’s (R. Shapiro) cars. I really liked my professors, especially Prof.’s Weinstock, Anderson, and Palmieri, a young experimentalist who started at Oberlin during my time there and effectively provided me with a role model, although I never thought about it at the time. I admit to having perturbed the patience of Prof W. with my cigarettes falling out of the ash tray and thus having twice burned holes in his beautiful dining-room table while my girlfriend/wife (V.J.H.) and I babysat for him. Years later, when I received an honorary Doctor’s degree from Oberlin and was hosted by a new faculty member who had bought Prof W.’s house and furniture, I saw the table and the burn marks again!

During my junior year I was challenged by Prof. P. to align and utilize a Littrow spectrograph which the physics department possessed and, in retrospect, might have been one of the first ones built. I recall that alignment was a major challenge as the various adjustments were coupled to each other. Nevertheless, I persevered. In my senior year I was given an honors project to build a magnetic resonance experiment. This wasn’t easy and I remember finally getting it to work at effectively the last minute, allowing me to graduate with honors in physics. Technically speaking, because of my grades, I graduated magna cum laude with honors in physics. I was even elected to Phi Beta Kappa (PBK), much to the surprise of other students who attended the first PBK meeting my senior year. They told me I had the wrong room! This was probably because they must have remembered me as the person who routinely appeared in the smoking room at the library reading the latest Marvel comics, to which I was addicted. (More than 60 years later I would sell my collection for a tidy sum.)

Graduate school

As a senior at Oberlin, I had received a Woodrow Wilson Fellowship for graduate study, so I felt that I had a very good chance of being accepted at most graduate schools. I was advised by my uncle, the physicist, that I would be better off at a smaller school, like Brandeis University, rather than a large school such as MIT. He argued that I would receive more attention from the faculty at the smaller school. I applied to Brandeis, Stanford, and Columbia and visited Brandeis the summer before I would start graduate school. I liked what I saw and the people that I met, especially Prof. Lipworth who would become my thesis advisor. I was accepted at Brandeis and excited about Prof. L.’s research looking for direct evidence for a time-reversal violation by trying to measure the electric dipole moment of the electron. I began working on this project the summer before the first year at Brandeis. I was even given the responsibility of coming into the laboratory at midnight and on weekends to fill the liquid nitrogen traps on the diffusion pumps! This also showed me that my uncle had been right as I was already started on my thesis research. By way of comparison, there were two other physics majors at Oberlin my senior year who went to Harvard University. If I recall correctly, they did not begin their research until their third year at which point, I had already published papers on new lower limits to the electric dipole moment of the electron.

The technical approach that this research exploited was to search for a linear Stark shift to the hyperfine structure of alkali atoms, especially Cesium. The atoms were released in an atomic beam where they passed through a region of a strong electric and magnetic field. Should the necessary frequency to excite the resonance depend linearly on the electric field, why then, one had detected the effect, and theoretical calculations allowed us to extrapolate from the alkali atom to the electron. I was also impressed, and ultimately influenced, by the way the laboratory was managed. There was one post-doctoral fellow and three graduate students — the senior student, the middle student, and the newbie. It took slightly over three years to move through the ranks to become the senior student. Prof. L. let us run the laboratory and would only intervene when we came to him to clarify our direction or to fix something when it went wrong, and we couldn’t cope. In effect, I was learning how to deal with the organization of a complex research project, training that helped me throughout my career. By the end of my tenure at Brandeis I had already published 10 papers mostly in Physical Review Letters and the Physical Review. Much of our research time was spent hunting down and understanding systematic effects that gave rise to spurious linear Stark shifts. Understanding that just because one measures something doesn’t mean that the signal is the one you are looking for was another piece of the training as an experimentalist and analyst of data that was more than useful throughout my career.

Although it might seem that all was wonderful during my years at Brandeis, including the birth of my two children, and setting an upper limit to the EDM of the electron that would stand for many years, all was not roses and light. For example, preparing for my qualifying exam to be taken after two years of courses, I seriously considered emigrating to Australia to become a chauffeur and my wife a housemaid should I fail the exam. Moreover, the year that I took the exam, the faculty changed the approach – which had been 3 days of written tests followed by a brief oral exam. The new approach would involve only a 3-hour oral exam based on two categories of topics that, if I recall, were up to the student but drawn from a list that the faculty prepared. One of my two topics was quantum mechanics (QM). I had chosen this topic, even though one of the toughest examiners was the professor who taught us QM, as I had done very well in his courses and felt that this might offset any difficulties I might encounter. As the exam progressed, I was gaining more and more confidence until I fielded a question about the density matrix. For the life of me, I couldn’t remember what it was, and I made the classic graduate student’s mistake of attempting to answer the question without knowing what I was doing. This led to hints by the faculty to help me and things only got worse until someone said, who taught him QM? at which point the exam ended. As I was one of the first to take the exam under the new format and had done very well until the density matrix fiasco, I was asked to come back in a few weeks to retake part of the exam. I passed. During my advanced exam, taken nominally before starting my thesis work, I received a similar question as to what was Hund’s rule? This time I answered that I didn’t recall the definition but if someone could clarify the physics involved, I would be able to answer. Once I received that feedback, I answered correctly. I knew the rule, just not the name.

On a personal note, highlights of my graduate school years were the close friendship that developed with G. Chanmugam, an astrophysics theorist, and of course the birth of my children of whom I am so terribly proud.

Where to postdoc?

By the end of graduate school, I was extremely confident if not arrogant. I had already published 10 refereed papers in the most prestigious physics journals. I reasoned that I had participated in one of the most frontier experiments in atomic physics and that it was time to switch fields, something that I would plan to do every five years (but never did!). Accordingly, I applied for a postdoctoral position at three institutions under three scientists who were known to Prof. L.: JILA (Joint Institute for Laboratory Astrophysics with Peter Bender – tests of General Relativity); Yale University with Vernon Hughes – muon studies; and Columbia University with Robert Novick – X-ray Astronomy. I decided to accept the position at Columbia for several reasons. One was that the group at Columbia was comprised of atomic physicists like me undertaking a new field. Another was that one of the Professor N.’s junior faculty (R. Angel) had been working on EDM experiments in England with a colleague of my thesis advisor. A final reason was that Prof. N. came to Boston and took me to a meeting of a group of scientists at American Science and Engineering under the leadership of R. Giacconi where the future of the very new field (the first extra solar X-ray source had been discovered by R.G. in 1962) of X-ray astronomy was being discussed. Although buried in acronyms (Super EXplorer, Principal Investigator Group), the discussions and the personalities fascinated me.

Columbia University and the Columbia Astrophysics Laboratory

I was in the physics department at Columbia from 1969 through the summer of 1977 and specifically in the Columbia Astrophysics Laboratory (CAL), progressing from postdoc to Instructor to Assistant Professor. I was able to expand and refine my experimental skills as I was involved in several sounding rocket experiments. (X-rays do not penetrate the Earth’s atmosphere.) From a hands-on experimentalist’s point of view these were wonderful times. Approximately every 1.5 years we launched a sounding rocket and practically each one represented a new instrument that we developed to further explore the properties of X-ray sources, which themselves turned out to be most fascinating — neutron stars and (the vicinity of) black holes. I would participate in, and contribute to, all aspects of a project: conception, design, construction, calibration, data analysis and interpretation. We were also learning what it meant to work with the space agency. In this regard one of the most important lessons was based on an incident that occurred just before I arrived — a CAL sounding rocket, that featured 100 gold-coated microscope slides arranged to serve as an approximation to a parabolic collector to improve signal to noise, blew up shortly after launch! The sounding rocket comprised a solid rocket Nike booster sitting under the liquid fuel Aerobee that contained the scientific payload. Unfortunately, the lanyard starting the ignition of the liquid fuel had, mistakenly, not been connected to the tower so that the Nike, once fired, destroyed the milk stool which separated it from the Aerobee, drove the Nike into the liquid fuel tank, and caused an explosion. The lesson was that we scientists, having the most to lose, need to be cognizant of all the activities of an experiment or mission, especially those that were single point failures. I also learned to adapt to experimental physics that must be performed on a schedule and required detailed advance planning. Very different from the laboratory research at Brandeis.

My first research task while at CAL involved working with another junior (at the time) faculty member, R.A., to develop a crystal X-ray polarimeter. It was felt that such a device might be more sensitive to linear X-ray polarization than the polarimeters that exploited the polarization dependence of X-ray scattering from blocks of lithium that had also been developed at CAL. With R.A., we entered the world of crystal physics and soon discovered that the mosaic form of a graphite crystal, whose first order Bragg reflection occurred at 2.6 keV, might be an excellent choice for such a polarimeter. The energy response was well placed being above the energies that are highly absorbed from the known X-ray sources at the time, but not too high where the spectra were declining. The mosaic form of the crystal would also be ideal as this guaranteed that the integrated reflectivity would be as large as possible, hence more detected photons. Having established this, we needed to find such crystals. R.A. contacted the Museum of Natural History in New York, which had possibly the world’s largest graphite crystal on display, to see if we could carve off a sample. The answer was no. I am not sure precisely who suggested it, but we then contacted Union Carbide which, it turned out, was manufacturing graphite crystals for neutron research.

After the first graphite crystals arrived, we found that they were not as ideally imperfect reflectors as they could theoretically be and so I spent several months performing both theoretical calculations and experiments to try to improve things. I was ultimately successful, having developed a technique that involved first slicing off a thin layer of crystal, bending it back against itself, causing striations in the thin crystal material, and then compressing, under high pressure, the thin layers back together to form the new, almost ideally imperfect, crystal-based on X-ray measurements. This research led to my first paper in X-ray astronomy written with R.A. and entitled Use of Highly Reflecting Crystals for Spectroscopy and Polarimetry in X-Ray Astronomy." I then went on to design and supervise the construction of a new polarization-sensitive payload designed to observe the Crab Nebula which was thought to be a synchrotron source of X-rays and thus highly polarized. The experiment was flown in an Aerobee 350 sounding rocket in 1971 and detected polarization at the 15 ± 5 percent level, thus confirming the synchrotron origin of the X-rays from the Nebula. My next project served as my introduction to X-ray optics and had me develop the first stage of an X-ray telescope using the Kirkpatrick-Baez approach (and yes, Baez, was the father of the famous folk singer.) We flew the sounding rocket to take the first one-dimensional image of the Cygnus Loop and again to determine the energy spectrum of the Perseus cluster of galaxies.

The rocket experiments also served to acquaint me with an aerospace contractor, Ball Aerospace in Boulder, Colorado. The polarization experiment needed four doors to which the crystals were attached and would open to a 45-degree angle once the payload reached altitude. In addition, the optics for the Cygnus loop experiment required highly polished plates of fused silica which could be tuned to the required parabolic shape and then locked in place. Ball successfully built both systems for us, and I met and learned to respect the highly skilled engineers at Ball involved in our projects. One of these was D. Hicks, who later would become a president of Ball. Establishing such contacts was to prove to be an important ingredient in my career as an experimental scientist.

Of course, I had to use computers to perform the necessary theoretical calculations and data analysis. The Columbia Astrophysics Laboratory had access to the IBM 360 computer at the Goddard Institute of Space studies located around the corner from the restaurant on 112th and Broadway whose exterior would later be featured in practically every episode of the Seinfeld TV show. My programs in those days were all on punch cards and the more sophisticated ones filled at least two boxes. Prior to my first dropped box, I had (thankfully) learned to use a magic marker to indicate the order by making the appropriate lines on the tops of the cards.

It was at this time that I was also introduced to the power of Monte Carlo simulations for performing theoretical calculations, ray traces for optics, and detector performance simulations. I also became interested in trying to characterize the time variability of the black-hole binary X-ray sources. I was able to obtain the data from Cygnus X-1 taken with the Uhuru satellite launched in 1970 (Was I the first Guest Investigator? Note that in the future, e.g., for Chandra and IXPE I pushed to get Guest Investigator changed to General Observer for what seemed to me obvious reasons. For the missions like Chandra, external scientists earned observing time through peer review and were therefore not guests receiving data from a kindly Principal Investigator as I had for the Uhuru data.) Together with P. Sutherland and S. Kahn, whose undergraduate adviser I was, we wrote several papers on the shot-noise nature of this aperiodic time variable source.

My next project was to lead the development of a Bragg crystal polarimeter, which I would design, to fly on the Orbiting Solar Observatory number eight (OSO-8) I had helped write the proposal to fly a Bragg crystal spectrometer on OSO-8 to search for line emission in the spectra from non-solar X-ray sources. The proposal was selected no doubt because of the sentence that pointed out that the device would also record a complete solar X-ray spectrum whenever the satellite was in the day sky and for each 10-second rotation of the wheel section in which the crystal panel spectrometer was housed. Surprisingly, Prof. N. was then able to convince NASA that the addition of two crystal polarimeters was technically and programmatically feasible and scientifically advantageous. It is very unlikely, if not impossible, that this change could have been accommodated in today’s NASA. I designed and supervised the building of the crystal polarimeters. The polarimeters were launched on OSO-8 in June of 1975. The sounding rocket detection of the polarization of the Crab Nebula was confirmed at the 20-standard-deviation level. Unfortunately, only upper limits were obtained for about a half a dozen other of the brightest X-ray sources. The sensitivity was hampered by several factors, especially that the polarimeter could only be pointed at targets a fraction of the time that OSO-8 was on-orbit. There were simply too many instruments on OSO-8 competing for observing time that required different pointing directions. Building a better polarimeter to fly in a dedicated mission became a personal goal of mine, one which I accomplished (see below), but only after 50 years of effort after the original Crab Nebula sounding rocket flight. I always found that persistence was important for achieving success whether for scientific or sports or other reasons. (I played basketball until I was 68, for instance.)

In collaboration with other institutions, the CAL was also involved with the High Energy Astronomy Observatory’s (HEAO) satellite programs both before and after cancellation and restructuring. Originally, four missions were planned, and the CAL had significant roles in three of them. The first featured a 1000 cm2, 1 arc-minute-angular-resolution X-ray telescope to perform an all-sky survey (Principal Investigator, P. Gorenstein, at the Smithsonian Astrophysics Observatory). Unfortunately, the survey mission was a casualty of the cancellation and then restructuring into three missions. The focusing survey mission (would have been decades before ROSAT) became a large-area proportional counter experiment and achieved the support of the Naval Research Laboratory, politically important for realizing the new missions. The lesson of the importance of political collaborations was appreciated. The HEAO cancellation and restructuring also impacted my training as an astrophysics experimentalist in ways that I had not expected. For example, I came to realize that the documentation requirements that NASA imposed were, if approached sensibly, useful in that they required one to make detailed plans, something necessary for projects that had a deadline. I also was asked to undertake the odious task of terminating the employment of those involved in the cancelled missions. No one had told me that was something I would ever have to do, and I did not enjoy it. I began to learn the art of, and the necessity for, writing proposals. This included working with L. van Speybroeck at SAO preparing the final Einstein Observatory proposal in Sept of 1973. Finally, I began to interact with other scientists at SAO including H. Tananbaum with whom I have spent decades in collaboration for Chandra and in a friendship for which I am most grateful.

The Marshall Space Flight Center (MSFC) Years

The bulk of my professional career has been spent at MSFC, both as a Civil Servant scientist and now as an Emeritus continuing to perform research. There is too much to remember so I have organized this section of my story into a few vignettes, rather than attempt a complete biography.

Marshall Space Flight Center (MSFC) in Huntsville, Alabama had been the NASA Center managing the HEAO Project, and I had visited there several times before the original HEAO project was terminated. Of course, Huntsville, especially in the 1970s, was quite different from New York where the CAL was located. Thus, I understood what was meant when, in 1977, R.N. sarcastically asked me something along the lines of You don’t want to live in Huntsville, Alabama do you? (As it happens, Huntsville was a great place to live and raise a family.) I learned that R.G. and H.T. had written a proposal to NASA HQ in 1976 for a follow-on mission to the second HEAO mission. That mission had been nicknamed the Einstein Observatory and had made a profound impact on the realization of the importance of X-ray Astronomy by discovering that every known class of astronomical sources, or a subset thereof, was an X-ray emitter. NASA HQ had probed the interest of the NASA Centers and Goddard, JPL with Cal Tech, and MSFC with SAO had all shown interest in the mission. MSFC was given the assignment and was now looking for a Project Scientist for what was then known then as the Advanced X-ray Astrophysics Facility (AXAF). I saw this as a wonderful opportunity to work on the development of what one of my future colleagues (J.K.) would refer to as a scientific cathedral, and to create my own group! I successfully applied for the position in competition with two others (D. Koch and F. Seward).

Chandra

Key parameters from the original 1976 proposal were a long life (∼10 years achieved by regular servicing), excellent angular resolution (∼0.5″), high sensitivity (1.2 m diameter outer mirror with an effective area over 1000 cm2 for wavelengths around 4 Å, and high detector efficiency). The original broadly stated science objectives included studies of stellar structure and evolution, large-scale galactic phenomena, active galaxies, rich clusters of galaxies, and cosmology. Apart from implementation details, these key parameters stayed constant through the development of the mission and this requirements stability was one of the reasons that Chandra was such a success. We X-ray astronomers worked very hard to maintain this stability, not only because it held us true to our objectives, but also, we knew from our experience that constantly changing requirements led to serious programmatic issues. The interested reader may find an excellent summary of Chandra’s development in the prologue of the book The Chandra X-ray Observatory: Exploring the High Energy Universe, edited by Belinda Wilkes and Wallace Tucker.

I knew from the beginning that, despite local optimism, it would take many years to accomplish AXAF, which would ultimately be renamed Chandra and will be referred to as such in what follows. I arrived at MSFC in the fall of 1977 with strong feelings of elation and enthusiasm, not to mention a little bit of fear. I was going to be the boss of my group which at the time was just me! On the other hand, and from time to time, occasional self-doubts would creep in but never take over. When I arrived, the launch was to have been in 1985, but by the end of the year our budget was changed, and the new launch was to be 1986. The progression of the launch date together with a revised budget was to be a constant of the motion until 1992. Thus, part of my scientific expertise required becoming expert in inflation factors and mission replanning, something we grew to be very good at.

I also learned from the Project Manager for Chandra at the time (J. Downey) the importance for a Project Scientist of maintaining strong personal relationships with the relevant infrastructure which included not only the scientific community but also MSFC upper management and NASA Headquarters. Thus, for the entire development of Chandra, I spent many hours on the telephone with H.T. at SAO and the various Program Executives at NASA HQ, especially A. Fuchs However, I could not imitate J.D.’s style, which was along the lines of the Southern country boy — he introduced me to phrases like self-eating watermelon.

I realized that, in addition to Chandra, I needed to develop research programs of my own to maintain the viability and reputation of the X-ray Project Science Team at MSFC. I began this almost immediately and, with the agreement of R.G., a long and fruitful collaboration with J. Grindlay at Harvard University was established. We at MSFC would develop the software for analyzing the timing data from the Monitor Proportional Counter (MPC) aboard the Einstein Observatory. These data were provided by the Time Interval Processor (TIP) of the MPC which measured the time between events to high precision. A negative flaw of the TIP was that the amount of data returned to ground was limited by a buffer which would stop taking data once full and would not start again until empty. Thus, the only data from the source known as the rapid burster was from the time between the bursts, not from the bursts themselves! Despite this limitation, we published several useful papers on the implications of the time variability of X-ray sources that were the targets of the Einstein telescope. Some of these papers also provided general and useful mathematical insights relevant to the analysis of X-ray timing data (e.g., the appendix to Leahy et al. 1983. Astrophysical Journal, Vol. 266, p. 160-170.)

To help me get started at MSFC, NASA HQ allotted me two postdoctoral fellows through the National Research Council (R. F. Elsner and P. Ghosh), both young neutron star theorists from the University of Illinois. R.F.E. would become my first Civil Service hire and serve as an important member of the X-ray Team for many years, analyzing data, interpreting scientific results, and becoming an expert at performing Monte Carlo simulations of the scientific performance of instrumentation until his retirement in 2019.

One of my major duties as a lead or chief scientist for X-ray Astronomy at MSFC, and one that I took great pride in, was bringing in to MSFC additional members of the X-ray team. There were two categories of scientists: post-doctoral fellows and Civil Servants.

In building up the Civil Service Team, I adopted the following strategies: first, attempt to bring aboard people that seemed to me to have the potential of making contributions that I probably could not easily, if at all, make; second, and for the PhD scientists, concentrate on theorists and experimentalists. For the most part I avoided observers. A major exception was A. F. Tennant, who made, and continues to make, outstanding and unique contributions.

Here I recognize (in alphabetical order) the post-doctoral (PD) fellows and the Civil Servants (CS) that at one time or another I brought to join me at MSFC. Note that several of the CS were brought on-board first as PDs.

PD: P.C. Agrawal, R.A. Austin, K.M.V. Apparao, A. Bolotnikov, C.-T. Chen, T. Ebisuzaki, K.K. Ghosh, P. Ghosh, S. Gunjii, T. Hayashi, Y.-D. Jung, K. Kilaru. D. Leahy, R. Manandhar, J. McDowell, L. Minamitani, J. Mir, S. Miyaji, S. Naranan, S.K. Patel, V. Peskov, A.H. Prestwich, H. Sakuri, N. Shibazaki, D.A. Swartz, Y. Takahashi, K. Wu, J. Youngen, M. Yukita, V. Zavlin, C.G. Zirnstein.

CS: C. Bankston, W.J. Baumgartner, S.D. Bongiorno, C.R. Bower, D.M. Broadway, R.W. Bussard, W. Darbro, K. Dietz, J. Davis, S.R. Ehlert, R.F. Elsner, A. Fulton, J.A. Gaskin, M.V. Gubarev, M.K. Joy, P. Kaaret, J.J. Kolodziejczak, S.L. O’Dell, B.D. Ramsey, M.E. Sulkanen, A.F. Tennant, N. Thomas, A.C. Williams.

The people I surrounded myself with played such an important role in my career. My successes partially became theirs and their successes partially became mine. I have been blessed in this regard and owe much to all of them.

Somewhere along the way, I realized that I couldn’t do everything, and I made the conscious decision to give up laboratory work. This was very difficult for me, but I realized that my commitment to Chandra and the responsibilities of the leader of a research group were not compatible with the full-time dedication that laboratory research required. I delegated the responsibility to lead that effort to my colleague B. D. Ramsey.

From the start, the Chandra study was organized with MSFC responsible for overall management and initial system engineering and the Smithsonian Astrophysical Observatory (SAO, the home institution of the Einstein PI (R.G.) and his technical team) providing scientific and technical support to MSFC and to Project Science. My perspective was, and still is, that this teaming agreement together with the establishment of a Project Science team at MSFC provided Chandra with enough end-to-end scientific participation wherein science considerations were integrated with engineering, schedule, and cost across all areas of the project. Quoting the Chandra Program Manager for the development and launch phase, Fred Wojtalik: This culture [the Chandra team] is easier to work with. I’m not sure what it is, but they (the scientists) have helped us tremendously, when we need to make trades and try to make ends meet. They are much more members of a team. (Revealing the Universe: the making of the Chandra X-ray Observatory, Cambridge, MA: Harvard University Press, Tucker & Tucker 2001, pp. 62–63).

With regards to Chandra, I knew that one of the most significant challenges I faced as a NASA scientist was to establish a strong rapport between science and the traditional elements of NASA projects — management and engineering. I knew that to accomplish this would require not only strong technical and scientific support from our colleagues at the Smithsonian Astrophysical Observatory, but also expanding Project Science at MSFC since the job could simply not be accomplished effectively by a single person. At first this proved to be difficult as MSFC was also managing the Hubble Space Telescope Project and had only used a single Project Scientist associated with that program for most of the development. (Even then only a deputy was added.) In addition, and unfortunately, the general MSFC impression of a scientist was often based on a stereotype, some sort of idealized version of Albert Einstein, brilliant, but incapable of designing and building anything. How often, in the subsequent decades, would we have to emulate the model of the great experimentalist Enrico Fermi to disprove the stereotype. Nevertheless, we scientists prevailed, and Chandra has proven to be one of NASA’s outstanding successes and there is no doubt in my mind that this was the result of the scientific involvement in all aspects of the mission — from interactions with Congress to participating in Change Boards.

There were many technical development issues that Project Science had to deal with from the time I arrived until the launch in 1999. One of the earliest I encountered took place during the attempts to define the Chandra Mission requirements. The issue involved the pointing stability of the observatory. Based on some at MSFC with HST development experience, highly precise and stable pointing was assumed by the engineers to be the Chandra requirement. This was, for us scientists, obviously not the case. It was very frustrating for me that we couldn’t make our requirements clear and to counter the implication that we X-ray astronomers didn’t know what we were doing. We wanted stable pointing that could be dithered uniformly over a region large compared to the angular resolution of the X-ray telescope — in the case of Chandra a dither of order 10-30 arcseconds. The obvious reason for us was that pixilated detectors, especially those with numerous pixels, would be difficult, if not impossible, to calibrate pixel-by-pixel. Furthermore, depending on the eventual quality of the X-ray telescope and the details of the design of the imaging detectors, highly stable pointing could end up with some, or all, of the incident flux focused onto a dead region!

Another discussion, that occupied us for many years, was the by-product of the way that the US government does business regarding construction of facilities which are done separately from the project that requires them. Since Chandra would represent a major leap forward, not only with respect to angular resolution, but also spectroscopic capabilities, and because of the time variable nature of X-ray sources, we had no standard candles and thus could not calibrate on-orbit. Therefore, we knew we would need to perform detailed calibration of the scientific response prior to launch and this, in turn, meant that we would need an X-ray calibration facility that could do the job. A prototype existed at MSFC built for the HEAO Program, but it required major upgrades to be useful for Chandra. Unfortunately, the cost could not be carried as part of the Chandra Program but would have to be submitted to Congress separately as part of the Construction of Facilities budget. To justify the facility, it was thus necessary to testify that Chandra made no sense without it. This meant that the mission would be under double jeopardy with regards to Congressional approval, once for the test facility and again for the Project. We constantly were asked if we really needed the calibration facility. My answer was always yes, we did. This remained an issue until the launch of HST in 1990 when it was discovered on-orbit that its telescope had not been built correctly. Had HST been ground-calibrated this would have been discovered and rectified prior to launch. Clearly, given a test facility, this would not happen to Chandra. All pressure to exclude the Chandra test facility from the Project vanished from that time forward.

An Aside — A Scrabble Tournament at the Huntsville Library

Some time in 1982 my daughter and I decided to enter a two-day long Scrabble tournament to be held at the Huntsville Public Library. I had been playing this game with my mother since early childhood and thought I was pretty good at it. Needless to say, I was slaughtered! I also met the Huntsville players including the lady who would become my best friend and second wife (M. E. Prince), although I would lose over 160 games to her before becoming proficient.

Another Aside — Supernova 1987A, Proportional Counters, and Australia

During the 1980s a focus of the research program of the X-ray team at MSFC was the development of large area proportional counters with fluorescence gating, an effort led by my colleague B.D.R. Working with J.G. at Harvard, we put together a balloon payload with complimentary X-ray and gamma-ray instrumentation. We first took the payload to Alice Springs, Australia in the spring of 1989 with the intent to look at the Crab and the recently discovered nearby supernova. This project was my first balloon payload and it turned out to be both momentous and a learning experience. There were several payloads there, both from the USA (3) and from Italy (2), and we were under the impression that the order of launch would be based on the order in which the experiments would declare that they were ready. Cal Tech had arrived first, declared ready first, and easily launched with minimal (if any) delays. We declared ready next, but the weather and the winds at the important altitudes refused to cooperate. We spent over a month getting up early to support a possible morning launch, giving up as the launch conditions weren’t right, sometimes returning in the afternoon for a possible evening launch and repeating this process day after day. Apart from the frustration caused by this introduction to ballooning, I spent most of my time working on the analysis software. The other experimenters were also frustrated with the delay, and this caused the other USA Principal Investigator to contact NASA HQ to make the case that his payload should be launched next, not us. We, and a change in the weather, prevailed and our payload successfully launched. The main lesson that I learned was that, if possible, one’s payload should arrive and declare ready first. Sadly, our part of the experiment was not a complete success as the Xenon detector gas, although highly purified, turned out to have had a significant Krypton contamination which increased the background level, hence lowering the sensitivity. Another lesson learned! An unexpected benefit of the trip was the friendship that developed between our team and one of the Italian teams (P. Ubertini & A. Bazzano). This friendship also led to collaborations in the future, including ESA’s Integral mission and IXPE (see below).

Another Aside — INTEGRAL

Following the balloon campaign in Australia, a collaboration with Italy, led by P.U. and the group at Frascati, was formulated. Both groups were in the process of developing fluorescence-gated microstrip xenon-filled proportional counters, and this seemed to be a natural choice for the monitor detectors required by an upcoming ESA M2 mission, INTEGRAL. Our role would concentrate on research and development of prototype detectors (as would Italy), detailed simulations of detector performance including estimating the background, providing technical input to the details of the detector design and use of MSFC facilities for X-ray calibration, and, of course, an equal share in the scientific output. Our proposal, EIKON, was submitted to both ESA and NASA in 1994 and was selected!

Subsequently, a major perturbation on the same scale as the HEAO cancellation / restructuring hit us. England, which was a major partner in the imaging instrument for INTEGRAL, withdrew from the partnership for financial reasons. A major reshuffling of responsibilities followed with Italy taking on England’s role and P.U. becoming the PI of the imager. The monitor detector would now be provided by Denmark that had lost the competition with EIKON.

Another Aside — The Wide Field X-ray Telescope

In 1992, Burrows, Burg, & Giacconi (ApJ, 392, 760) wrote an important, and originally unappreciated, paper on modifying the traditional design of X-ray telescopes to optimize the angular resolution over a broader field of view at the price of the on-axis resolution. Such a design would be ideal for a survey mission, significantly enhancing its sensitivity. R.G. even brought the design up in 1992 for discussion by the Chandra Science Working Group to consider it as a possible modification to Chandra. After much discussion, and partially for programmatic reasons at this tumultuous time in the Chandra development, it was decided not to modify Chandra.

Nevertheless, R.G. and colleagues continued to investigate and promote the wide-field design for possible missions. Their Wide Field X-ray Telescope (WFXT) mission was designed to survey the X-ray sky. It would be orders-of-magnitude more effective than previous and planned X-ray missions in carrying out surveys. (And this statement is still true today!) The number of sources it could detect and characterize in five years and in three proposed fields ranged from hundreds of thousands to millions. It would thus provide a unique legacy data set that would support contemporaneous and planned ground-based and on-orbit astronomical programs, including at the time future flagship missions, such as the James Webb Space Telescope.

Surprisingly, at least to me, the concept of a WFXT did not generate a new mission, either in the USA or abroad. The team had originally targeted an European Space Agency M-class mission. Unfortunately, the Italian group that was to have provided the optics decided on a different priority. I was approached by R.G. and my friend, S. Murray (yes, I did once steal M&Ms from his office) to see if the X-ray Team and MSFC would be interested in playing a more significant role in proposing such a mission for a NASA opportunity. In addition to a modest role in the science, the X-Ray team at MSFC would support the mission by taking the lead responsibility for purchasing and managing the X-ray optics from Italy, to calibrate them using the Chandra X-ray Calibration Facility, perform systems engineering, and manage the development of the spacecraft — all of which were similar roles that MSFC was taking on for Chandra.

Together with S.M. we convinced MSFC management and NASA HQ of the potential importance of a WFXT mission and the benefit to the Center and uses for its workforce. We initiated system studies including a complete design performed together with MSFC’s Advance Concept Office and a technology program that included the purchase of a $1M-class polishing machine designed foe complex surfaces still in use in our optics programs today.

As part of this effort, several science team meetings were held in Sesto, Italy which I enjoyed tremendously. The subsequent deaths of R.G. and then S.M., and the launch of Chandra, and the advent of IXPE (see below) prevented me from further pursuing the WFXT.

Chandra — Galaxy Clusters and the S-Z effect

One technical issue that we faced during the development of Chandra was the need to support the Chandra observations of clusters of galaxies with measurements of the Sunyaev-Zel’dovich (SZ) effect wherein the microwave background radiation is scattered off the hot, X-ray emitting, gas that pervades the great galaxy clusters. The apparent change in the microwave background temperature due to the presence of the cluster scales as nRT, where n is the electron density, R is the cluster radius, and T is the cluster gas temperature. In addition, the X-ray flux scales as n2R1/2T1/2 θ2 where θ is the angular size. Clearly combining these measurements would be of great scientific value. Unfortunately, the technology of the day required both Chandra-like X-ray sensitivity to support this experiment and a significant advance in millimeter wave measurements to cover that portion of the spectrum. To this end I brought M. K. Joy from the University of Texas to MSFC and obtained funding from the MSFC Center Director’s Discretionary fund for him and his colleague, J. Carlstrom, to support their efforts for using millimeter-wave interferometry to measure the S-Z effect. I am proud to say that this work has been extremely successful and has subsequently become a scientific industry onto itself.

Chandra — Prove it or be cancelled

Chandra faced enormous challenges to its completion in the early 1990s. In 1989 the Chandra project was challenged to build and demonstrate the required sub-arcsecond angular resolution in just three years, by the start of the 1992 fiscal year on October 1 of 1991. If the challenge was met, the program would proceed with full funding; if the challenge was not met, the program would be canceled. The challenge implicitly included preparing the X-ray test facility on an even shorter time scale than had been planned. This latter was accomplished, in no small part, as the MSFC Director of the time (J. Lee) took over management of the effort with meetings held in the Director’s conference room at 7:00 a.m. every Thursday morning. In general, it has been my experience that NASA and MSFC works best during a crisis, and, despite the absurdly (for me) early hours of the weekly meetings, the test facility was ready for the optics as per the accelerated schedule. The optics too were ready, despite having to overcome many, also seemingly insurmountable, hurdles which were as subtle as tiny vibrations introduced by a large fan 100 ft away from a metrology station. I and the science team both at MSFC and SAO were deeply involved in both the building of the optics and definition of the test facility. The optics effort was led by the Telescope Scientist (L. vS. at SAO) while I concentrated onleading the effort to be sure that the test facility could accomplish the measurements that Chandra required for the ground calibration.

X-ray testing of the largest optics segments (aka the Verification Engineering Test Article or VETA) began Labor Day weekend 1991. One-dimensional scans across the image immediately showed behavior different from what was expected. Had the optics been polished to the wrong shape? Several days were spent trying to understand the problem until we were convinced that gravity had caused the mirrors to sag ever so slightly causing distortion of the image. The effects of gravity had supposedly been accounted for in the design but mistakes in the computer codes were overlooked although one of the rather shy quiet MSFC engineers (D. Zissa) had predicted the problem. To proceed, the force required to offset the gravity effects was calculated and a set of springs appropriately placed were used to offset the gravity distortions in the horizontal test configuration. When the chamber was pumped down again and the X-rays were turned on, the desired images were seen. Indeed, the image based on the full set of test data had a measured full width at half-maximum of 0.19″, a factor of 2.5 better than the performance required by the mirror challenge. A very important by-product of the challenge was the team building that took place among the industry leads, NASA, SAO, and the Science Working Group, which was so important for what would follow.

Chandra — Met the Challenge, but Failed to get Started

The joy of having met the Congressional Challenge was short lived and in January of 1992, we were informed that new NASA budgets would no longer support the projected cost. We would have to dramatically descope to avoid cancellation. This devastating news led to months of discussions, trade studies, negotiations, pessimism, and at times hard feelings. We considered more than 20 different combinations of mirror pairs and instrument complements, while struggling to find an acceptable balance between science return and cost. Following a bruising battle, we accepted the fact that the mission could not be serviceable. NASA moved to ensure that decision by proposing to place the observatory in a high-Earth orbit thus unreachable by the Shuttle. Achieving a high orbit would require a substantially less-massive payload, which led us to propose two smaller missions in two different orbits — the one for a high orbit featured the imaging properties of the original Chandra mission. The other, for low- Earth orbit, featured much poorer angular resolution and had the X-ray calorimeter, the heaviest instrument, at the focus. After about a year, this second mission was canceled by the US Senate and the calorimeter was ultimately to be placed on a Japanese satellite. But the basic plan for the imaging mission was accepted and, 15 years after the initiation of the project, we were finally allowed to proceed! A critical element of the program was an unwritten pledge that the project would receive its funds at the levels we had estimated and on our schedule. From 1992, until launch in 1999, we never had to annually replan the mission as was the norm in the previous 15 years. This was a tremendous relief to all of us. I often thought that this was the way all missions should proceed, with strict cancellation policies if the project could not live within its own plans. I realize that this is perhaps too simplistic a solution and clearly does not consider the complications of the NASA budgeting process.

Of course, there were several bumps in the road along the way. For example, during the alignment and assembly of the eight mirror shells into a single telescope at Eastman Kodak in 1996, an unexpected shift in the alignment was detected for one of the shells with respect to the first pair, which had been aligned and bonded to begin the assembly. The shift was of order 1/3-arcsec and within specifications, but substantially larger than expected. We felt we had to understand the origin to be confident in the measured results. Ultimately, it was determined that the 40W fluorescent lights in the vertical assembly tower had heated the support structure and then the surrounding air by a fraction of a degree. The resultant small change in the air density changed its index of refraction, which in turn introduced a tiny change in the path of the laser beam used for the alignment. From then on, we made the final alignment measurements of the bonding of the elements with the lights off.

At the same time, MIT’s Lincoln Laboratory was developing the flight charge-coupled devices (CCDs) for the ACIS instrument. The laboratory had little difficulty producing front-side illuminated (FI) chips, but the backside illuminated (BI) chips, with higher quantum efficiency for lower energy X-rays, were much more challenging. Ultimately, only three good BI chips were produced, and the best-performing chip was lost in testing. The two others were placed into the six-chip linear array serving as a readout for the High Energy Transmission Grating (HETG), with the better BI chip at the prime focus. The rationale was that the moderately worse energy resolution (compared to the FI devices) would not really impact either the high-resolution grating spectroscopy or on-axis images over an 8’ field of view for this one CCD. This CCD would be very useful as a low-energy supplement to the larger ACIS-I 4-chip (all FI) array. The perseverance of the ACIS team in producing BI chips and the placement of this BI CCD at the prime focus turned out to be even more fortuitous than we had thought. Early after launch, we detected degradation of the charge transfer efficiency, and thus the energy resolution, of the FI chips due to bombardment by protons focused by the X-ray optics with the ACIS detector at the prime focus during passage through the radiation belts. The BI chips were far less sensitive to proton hits and sustained much less damage. Of course, once we understood what was happening, the Chandra detectors were routinely moved out of the focal plane during passage through the belts.

Calibrating Chandra at the X-ray Calibration Facility

The ground calibration of Chandra was the most ambitious and thorough calibration of an X-ray mission ever. For me, the process began with the creation of a document, MSFC-RQMT-2229 titled Scientific Requirements for the Calibration of Chandra, completed with the blessing of the Chandra Science Working Group on August 2, 1995. This 121-page tome took several of us over one year to produce and addressed, not only the measurements that needed to take place at the X-ray Calibration Facility at MSFC, but also calibration requirements for scientifically important spacecraft subsystems, the equipment required for performing the X-ray calibrations, the accuracy extrapolated to on-orbit performance, etc.

Next, we performed what we referred to as a pre-calibration rehearsal using the Technology Mirror Assembly (TMA) that the project had built by Perkin Elmer under the supervision of the Chandra telescope Scientist L.vS. earlier in the 1980s to demonstrate that the Chandra imaging requirements could be met. The TMA was a two-thirds scale version of the innermost paraboloid–hyperboloid pair of Chandra to allow X-ray testing at MSFC’s X-ray facility which, at that time, had been built and used for the Einstein Observatory calibration. At 1.5 keV, the TMA delivered 90% encircled energy within a 1″ radius, well within Chandra specifications. The TMA is still the world’s best focusing X-ray optic made for astrophysics.

During the period when the High-Resolution Mirror Assembly (HRMA, aka the Chandra optics) was present at MSFC, over 3000 X-ray measurements were taken. A thorough technical overview of the calibration may be found in Weisskopf & O’Dell (1997) and O’Dell & Weisskopf (1998) and references therein. One of the most compelling aspects of this activity which, with few exceptions (such as requiring us to leave the facility because of a tornado warning), operated 24 hours per day, seven days a week, was the camaraderie that developed among the 12 organizations and about 100 people involved on a day-to-day basis. I personally was present for a 10-hour shift daily and met with all participants at each shift change. (Shifts were nominally 8 hours). I suffered a heart attack requiring 4 bypasses at the end of the calibration period and after most of the flight hardware had moved on. I was a heavy smoker prior to this event but lost my physical addiction once I left the hospital. It took several years however, to lose the psychological addiction but, with the help of my wife (M. E. Weisskopf), I succeeded.

An Aside — ART XC

In January of 2008 and at the same time we first proposed IXPE we also proposed participation as a Mission of Opportunity in the Russian Spektrum-X-Gamma Mission. B.D.R. was the PI, S. L. O’Dell the Project Scientist, and I was a Co-Investigator. Our potential participation had been sought and encouraged by scientists at the Space Research Institute (IKI) in Moscow. Our technical contribution was to have been the 8 X-ray telescopes required at the time for the Russian Astronomical Röntgen Telescope (ART) experiment. Unfortunately, our proposal was not selected. However, this did not end our participation in the mission. We were then approached to build four sets of optics under a contract to the Russian Space agency which we accepted. Of course, no data rights were associated with this task. Nevertheless, such a project would be useful to our optics program’s infrastructure and reputation. Subsequently, we were told that if we could donate the other four telescopes, why then we could share in the data. We were able to accept this challenge. A principal reason was that by far the largest cost in manufacturing the telescopes was the cost of the mandrels from which shells were replicated. We already had the necessary mandrels for the first set of optics. The remaining small costs were supported by MSFC.

SRG also involved the Max Planck Institute for Extraterrestrial Physics (MPE) in Gärching, Germany. MPE was not terribly enthusiastic about our participation as they were concerned if we had troubles delivering on schedule, we would delay the mission. It is with some pride that I note the MSFC-provided optics was the first of the SRG scientific instrumentation delivered!

The Launch of Chandra

In July of 1999 I was with my family and colleagues at the Cape for the launch of Chandra, 22 years after coming to MSFC to work on the project. Obviously, my emotions were high, yet I knew that this was just the first of many crucial steps before we knew that the observatory would work, let alone work as advertised. The tension of the launch was eased somewhat as we went through three launch attempts, each set for just after midnight with the first scheduled for July 20, 1999. The countdown proceeded until about 10 seconds before launch when a possible fuel leak was detected, and the launch was aborted. Two nights later the launch was cancelled due to the weather. Finally, the launch took place at 12:31 AM on July 23. I was overjoyed at the success of the launch, but then travelled to the control center for a more than a month of commissioning to find out whether all our efforts were successful or in vain.

The First Chandra Field

I thoroughly enjoyed initiating and leading a Chandra project that was more of historical, rather than ground-breaking, scientific, interest — namely, describing the first Chandra field of which I nick-named the brightest source Leon X-1 in honor of the Chandra Telescope Scientist. On 1999 August 12, commands were sent to open the Chandra telescope’s sunshade door, the last barrier between the Chandra focal plane and the X-ray sky. The attitude control system was not yet in its normal operating mode and the chosen field was arbitrary. The Advanced CCD Imaging Spectrometer (ACIS) was set at the nominal focus with the aim point of the telescope set on one of the two back-illuminated CCDs. Although the aspect camera was acquiring data, it was not yet an active part of the attitude control system. Rather, the Chandra gyroscopes, which themselves have very small drift, were controlling the spacecraft’s attitude. Further, the dither mode was not yet active. Nevertheless, the pointing was already very stable, even without the aspect camera data. We were nervous and excited watching the monitor of the CCD to see the first X-rays to reflect from the Chandra optics. We obtained a 9 ksec image of the field shown in Figure 1. It was even immediately clear from the real-time display that the image included at least one tightly clustered collection of X-rays within a diameter measured in arc-seconds — Chandra worked! The years of effort would pay off! I and my colleagues identified that brightest source in the field source as an AGN at a red shift of 0.3.

Fig. 1.—X-ray image of the first Chandra field on the ACIS S3 (back-illuminated) CCD. A 5-arcsec radius circle encloses each of the 10 detected sources. The brightest, S3-5, nick-named by me as Leon X-1, is an AGN at a red shift of 0.3207. Although obtained prior to on-orbit focus adjustment, fine attitude control, etc. the image already demonstrated the excellent angular resolution of Chandra, and thus that the optics were working well.

The Crab Nebula

From my earliest involvement with X-ray astronomy, observations with various instruments of the Crab Nebula and its pulsar have been a mainstay of my scientific interest. By the time of my retirement in 2022, I had been the lead author or co-author of 22 refereed journal articles on this system covering everything from the initial discovery of polarization from the Nebula, to precision measurements of the X-ray and radio polarization of the pulsar, to the discovery with Chandra of the shock front known as the inner ring, to the discovery that the pulsar was visible even at pulse minimum, to the discovery that the pulsar is displaced from the inner ring, to the discovery of Gamma-ray flares from the Nebula and the corresponding implications for gamma-ray emission and particle acceleration, to the determination of the performance of phase-resolved spectroscopy of the pulsar, to the characterization and implication of the optical and X-ray properties of the northern wisps in the Nebula, to the characterization of the feature known as the inner knot, etc.

Other Science Projects

In addition to my study of the Crab, I, together with my colleagues, performed studies of many interesting objects with Chandra and other observatories. These included: the Soft Gamma Repeater SGR 1900+14, the Anomalous X-Ray Pulsar 1E 2259+586, the X-ray hot spot on Jupiter, the soft X-Ray emission from Io, Europa, and the Io plasma torus, the Anomalous X-Ray Pulsar 4U 0142+61, the Globular Cluster M28 and Its Millisecond Pulsar PSR B1821-24, the X-Ray Environs of SN 1998bw/GRB 980425, the Nature of the Bright Short-Period X-Ray Source in the Circinus Galaxy Field, A Multiwavelength Search for a Counterpart of the Brightest Unidentified Gamma-Ray Source 3EG J2020+4017 (2CG 078+2), the X-Ray Emission Processes of Old Rotation-powered Pulsars, the Possible Dark Galaxy VIRGO HI 21, the Globular Cluster M71, the Unusually Bright Cataclysmic Variable in the Globular Cluster M3, the X-Ray Counterpart to PSR J2021+4026, the X-Ray Counterpart to PSR J2021+4026, the IC 443 Pulsar Wind Nebula and Environs, PSR J1741- 2054, and Geminga.

Polarimeters and IXPE

Coming to MSFC and spending my first 22 years there working on Chandra, did not mean that I had lost interest in X-ray polarimetry, a field that I helped pioneer while at Columbia. For more than a decade after I arrived at MSFC, we collaborated with R.N. at Columbia in all aspects of a polarimetry mission that was to have flown on a Russian mission (Spektrum X-Gamma). I, and my colleagues, helped R.N. in many ways including providing a complete Monte Carlo simulation of the experiment. This mission was something of a Christmas tree of experiments and the polarimeter was only one of several experiments included in the payload. Planning for the first year of the mission only allowed for 11 days of observing with the polarimeter, grossly inadequate, but supposedly better than nothing. Unfortunately, Spektrum X-Gamma never launched due to the collapse of the Soviet Union in 1989. This setback, however, did not deter me from pursuing the dream of eventually performing meaningful X-ray polarimetry. Furthermore, it was already clear to me that to do this successfully, polarimetry should be performed on a dedicated mission as its long integration times would put it at a disadvantage with respect to other instrumentation onboard. Ultimately, the ideal polarimeter should also be capable of imaging since X-rays may also be found in extended sources such as supernova remnants and jets in active galaxies.

In 1994, 1995, 1997, and 2003 I proposed as Principal Investigator different versions of a disk scattering polarimeter to various NASA Explorer opportunities. Apart from attempting to advance X-ray polarimetry, our strategy was based on proposing extremely simple, very-low-risk, inexpensive devices in the hope that this would be appealing. Unfortunately, it was not. The feeling in the community was that opportunities were so rare that one should always maximize the possible scientific yield given the available funding.

In the mid-1990s, an extremely important technical advance was made by my colleague, B.D.R., who developed at MSFC the first X-ray polarimeter that exploited the polarization dependence of the photoelectron. He realized that the light emitted by electron avalanches in a parallel plate proportional counter can be used to image the tracks of charged particles initiated by the photoelectron. Since the initial direction of the liberated photoelectron is determined by the polarization of the incident radiation, this opens the potential for polarimetry. An Argon-filled test chamber was built which successfully imaged 54 keV photoelectron-induced electron tracks using an intensified CCD camera. The successful demonstration of the polarization sensitivity for imaging electron tracks implied that one not only had a new polarimeter, but also, that when coupled to an X-ray telescope, one could perform spatially resolved polarimetry on extended sources. The challenge would then be to develop such devices to operate in the standard X-ray astronomy band, from 2-10 keV, where photons are most plentiful. For a few reasons, especially the intense activity leading to Chandra’s launch in 1999, I failed to recognize at first just how important this development was and did not devote precious resources to develop the photoelectric polarimeter to operate at lower energies.

However, at the turn of the century, three Italians (E. Costa, R. Bellazzini, and P. Soffitta) began working on a low energy polarimeter also exploiting the photoelectric effect. By 2006 they had made significant progress and had demonstrated the feasibility of such devices including developing an ASIC ideally suited to their detector design. We had previously worked with E.C. and P.S. on the ill-fated SXRP mission, so it was natural for us to combine forces to propose for the next NASA Small Explorer mission opportunity in 2007. Since we built X-ray optics at MSFC, and the Italians built the polarization sensitive detectors in Italy, it was an ideal scientific marriage. Thus, the Imaging X-ray Polarimetry Explorer (IXPE) concept was born. I proposed it as the Principal Investigator in response to a NASA Small Explorer announcement of opportunity and submitted the proposal in 2007. We knew of at least two other polarimeters being proposed for this opportunity and, because of this, we thought that a polarimeter would likely be selected and that ours, being the only instrument capable of performing imaging polarimetry, would be the one. One of our competitors also proposed to exploit the photoelectric effect and track the electron cloud, but with a technique that, as proposed, had not been shown to work and we felt it would have to be further developed and possibly revised. We were partially right in our predictions — a polarimeter was selected but it wasn’t ours: it was the rival photoelectric polarimeter. After several years of development effort, the rival polarimeter mission was cancelled.

In the years prior to the cancellation of the rival mission, we at MSFC were busy refining the techniques for building lightweight, moderate spatial resolution (<30 arcsec) X-ray optics. Indeed, we entered a collaboration with Russia and delivered 7 X-ray telescopes for the ART-XC experiment aboard the revitalized Spectrum-X mission as discussed previously. Simultaneously, we were working with our Italian colleagues proposing a version of IXPE abbreviated as XIPE (X-ray Imaging Polarimetry Experiment) for an ESA mission and their development of what we called a Gas Pixel Detector (GPD) had continued, providing a more complete understanding of its polarization response, and studying possible systematic effects with unpolarized X-rays. (This latter was very important. Showing that a polarimeter is not sensitive to unpolarized X-rays is critical.) Using the 2007 proposal as a first draft, we reproposed for the Small Explorer opportunity in 2014, and we were selected as one of three missions to compete for the flight opportunity. Our resulting concept study report was submitted in July 2016 and on January 3, 2017, I received a phone call at my home from the Associate Administrator (T, Zerbuchen) for Science at NASA stating that IXPE had been selected to proceed if I would agree to no exclusive data rights. Since no restrictions had been placed on exclusivity of data rights in the Announcement of Opportunity, I was caught off-guard. Nevertheless, I agreed to the condition as a flight opportunity was too rare to reject although I was against the policy as it gave advantage to large groups who could afford to devote the resources to learn how to process the data on a rapid basis. To try to ensure that those who had spent years on the development of IXPE would also get to participate in the science that came from it, I created a Science Advisory Team (SAT) with Technical Working Groups for 7 categories of X-ray sources from blazars to magnetars. The SAT ultimately involved over 150 scientists from 13 countries. This proved to have been a successful strategy.

On December 9, 2021, IXPE launched successfully on a Falcon 9. Almost 2 years later it is safe to say that it has been an outstanding success, detecting polarization from dozens of galactic and extragalactic X-ray sources. IXPE is the polarimetry analog of the UHURU satellite which increased our view of the X-ray Universe from the handful of sources discovered in the sounding rocket era.

By the end of the first year of operation, 39 sources have been observed, half of which have had polarization detected at >99.99% confidence. In some cases, the results were unexpected and surprising, leading to a re-examination of previous theoretical models E.g., the pulsating X-ray binaries were sources for which strong polarization was anticipated, yet only low values were detected. Magnetars have shown fascinating results: 4U 0142+61 has an energy-dependent polarization falling to a minimum at a medium energy and then rising again at higher energies including a 90-degree phase-shift that needs to be explained. Another magnetar, 1RXS J170849.0-0-400910, exhibited no minimum and no phase shift, simply an increase in polarization with energy.

Observation of the Compton-thick active galactic nucleus (AGN), the Circinus galaxy, exploited IXPE’s imaging capability and finds the source to be significantly polarized with a position angle roughly perpendicular to the radio jet, confirming the basic predictions of AGN unification models. This first X-ray polarization detection of a Seyfert galaxy also demonstrated the power of IXPE in probing the geometry of the circumnuclear regions of AGNs.

In summary, IXPE has provided a new tool for astrophysics research and the first results are guiding changes to our current understanding. Thus, more than 50 years after my rocket flight detecting polarized X rays from the Crab Nebula, a new window on the X-ray sky has finally been opened with a dedicated mission that I was honored to lead. At this writing, Chandra is entering its 25th year of operation and still provides unique and outstanding scientific results. I am blessed and proud that my career could come to such a satisfying point when I retired in 2022.

My advice to young scientists — determine and plan for what you want to be doing 5 years in the future. Act on those plans and, if you believe in it, never give up.