1. In Memory of John Pople
  2. Charles Darwin and John Pople
  3. John Pople: Interview

In Memory of John Pople

John A. Pople, Nobel Laureate of the year 1998 (deceased 2004) had a large impact on the research and the development of the CATCO group. Dieter Cremer worked under his supervision from 1972 to 1975. This was the time when Dieter learned John Pople’s approach to scientific problems, was a student in his quantum chemistry classes, and became familiar with the Gaussian70 program. Dieter and John had a very productive time, which culminated in the development of:

  1. the “Cremer-Pople” ring puckering coordinates and the ab initio description of the conformational energy surfaces of medium-seized ring molecules,
  2. the explanation of the conformational features of geminal double rotors in terms of sigma- and pi-electron delocalization,
  3. the development of programs for the geometry optimization in terms of special ring coordinates,
  4. and the description of the strain of small- and medium-sized ring molecules.

If one started as a student of John Pople the first meetings with him were decisive for the future collaboration. Dieter joined the Pople-group in the fall of 1972. Before this time, he had worked, besides scattered studies in journalism, literature, and chemistry, ten years as a part-time journalist with various Cologne newspapers and also as a businessman in the family company after his father had become seriously ill. He remembers that at the first meeting he was dressed as a businessman, which was observed by John with curiosity. “Does the guy really want to work with me?” he must have thought. Actually this was the only day, Dieter wore his business suit during his Pittsburgh years.

Before his time at Carnegie-Mellon, Dieter had had training in Organic Chemistry (work with Emanuel Vogel on the synthesis of bridged annulenes) and NMR spectroscopy (work with Harald Günther on the analysis of spin-spin coupling constants as indicators for electron delocalization). His interest in quantum chemistry was a result of the influence Harald Günther had and from a stay in Uppsala, Sweden, with Per-Olov Löwdin. His move to Pittsburgh and Carnegie-Mellon was triggered by the fact that in the 60ies and early 70ies the Mellon Institute was famous for the work of Aksel Bothner-By, Josef Dadok, and Salvatore Castellano in NMR and Pople’s background in this field leading to the quantum mechanical description of NMR spectra in the famous book “High-resolution Nuclear Magnetic Resonance” by J. A. Pople, W. G. Schneider, and H. J. Bernstein.

John told Dieter that he planned to investigate medium-seized molecules more precisely. In 1972, the newly developed Gaussian70 ab initio program contained just a Hartree-Fock and a CISD program, however had, compared to the ab initio packages of that time, the advantage that all necessary calculations were organized in one single program with a user-friendly input and many options to control the execution of the program. Gaussian70 was the kernel of the Gaussian package that later became so popular and attracted experimentalists in numbers to do for the first time quantum chemical calculations.

John had investigated with Leo Radom the rotational behavior of a large set of organic rotor molecules and several publications on this topic had been published in John’s series “Molecular Orbital Theory of the Electronic Structure of Organic Compounds.“ The description of ring molecules seemed to be at this time rather difficult because of the redundant parameter problem and the unsolved problem of exactly calculating transient ring conformations with a specified puckering along a ring pseudorotation path. “Dieter, look into the problem of calculating cyclopentane conformations of a specific pucker.” This was the task John gave him.

Dieter studied the literature and came up with several solutions all based on the use of atomic position vectors with regard to the geometrical origin of the ring molecule. None of these solutions were without flaws. Then, Dieter red Herzberg’s vibrational mode analysis of cyclopentane in “Infrared and Raman Spectra.” He realized that there were exactly N-3 out-of-plane vibrations that convert, for example, a pentagon (N = 5) into an envelope or twist form. Hence, the description of the out-of-plane vibrations with fixed, finite vibrational amplitude must lead to the envelope and twist conformations of cyclopentane and linear combinations of these two modes must yield any desired conformation along the pseudorotation path of the molecule. Within weeks he was back with a unique mathematical solution of the problem John had given him.

Talking with John was always best on a Saturday. On a Saturday afternoon, Dieter wrote the equations he had worked out for cyclopentane on the blackboard of the office at the 8th floor of Mellon Institute. John looked at them, started to mull over them, and then starred out the window into the courtyard for about 7-8 minutes. In this time both of them did not move, not a single word was spoken, and one could hear only the Saturday afternoon noise of the cars on Fifth Ave. Then John turned to Dieter with the words “This is indeed a splendid solution. If you do a Fourier transform you can easy generalize these equations and have a description for any N-membered ring.” With these words he turned around and left the office. – It took Dieter years to understand that on this Saturday afternoon shortly before 5 pm, he was promoted to become a member of the “Pople-people” who John accepted as eligible to be in his group.

In the time 1972-1975, John and Dieter worked out the concept of homodesmotic reactions for the determination of the strain of ring molecules. They called them “ring strain determining reactions” and evaluated them with Hartree-Fock and later second order Møller-Plesset Perturbation Theory and Pople basis sets such as 4-31G and 6-31G(d,p) (the use of d-functions had become possible thanks to the work of “Harry” P.C. Hariharan in ’73). The calculation of the 5-membered and six-membered ring molecules was slow and on the edge of what was possible (two tapes with 2-electron integrals were filled during the calculation of a 5-membered ring) using a UNIVAC 1108, which was the main computer at Carnegie-Mellon at this time. Therefore, Dieter continued the project at the University of the Cologne from 1975 on with the much faster CDC7600 computer. However, before any results could be published the paper by Phillip George and co-workers on “homodesmotic reactions” appeared (submitted already in the summer of 1975), which focused on the idea of balancing formal reactions so that ring strain and other electronic effects can be determined. Phillip George showed that by just using experimental heats of formation one could calculate the strain energies of ring molecules accurately.

Dieter and John worked in the 70ies and 80ies on various projects where Paul von Rague Schleyer who moved 1975 from Princeton to the University of Erlangen-Nürnberg, was actually organizing and influencing this collaboration with his ideas. For example, Paul had the idea of designing anti-van’t Hoff molecules with “planar carbon”, orthogonal ethene, or other inversions of the van’t Hoof rules. This triggered later the investigation of carbon-boron rings, the extension of aromaticity rules, and other topics.

Dieter and John stayed in contact when Dieter moved to Sweden. The CATCO group adopted distinctive forms there and focused on the development of electron correlation methods. John and Dieter saw each other at conferences and it was at one of these occasions that John made Dieter familiar with the newly developed QCI (quadratic configuration interaction) method as a size-extensive CI method having the nature of a simplified coupled cluster method. Despite of all the success John had with the new QCISD method, QCI also led to some backlashes. Fritz Schaefer, Rodney Bartlett, Josef Paldus, and others proved that i) QCI in its extension to triple excitations was no longer size-extensive and ii) QCI did not save much computer time compared to the more complete Coupled Cluster method. Dieter and his co-worker Zhi He investigated the general concept of a quadratic CI correction and came to the conclusion that QCISD as designed by John is a single not further extensible method. When adding triple excitations and making the method size-extensive it does hardly differ from CCSDT. For quadruple (Q) excitations, size-extensive QCISDTQ is identical with CCSDTQ thus proofing that QCISD actually does not belong to an independent group of methods positioned between CI and CC methods. Dieter and Zhi He also showed that a group of size-extensive QCI methods halfway between CI and CC could only be obtained by deleting certain disconnected terms of the CI equations and including just the connected part of appropriate quadratic correction terms. These QCIc (c for connected) methods could be extended to triple, quadruple, and higher excitations without loosing their size-extensive character.

John and Dieter met in Jerusalem at the WATOC4 conference in 1996. Dieter presented there his work on sixth order Møller-Plesset perturbation theory, which at this time and also today is the limit of what theoreticians can develop in this area of theory. After his talk, Dieter and John met at the dinner table. “I guess you said everything one can say about MP6.” This was John’s comment and it was the strongest word of appreciation Dieter ever got from John. For the rest of the evening, the discussion never returned to MP6 and quantum chemistry. On the contrary, John started to talk about the last years of the war, his entry to Cambridge, his Ph.D. time, and his encounters with Paul Dirac, Freeman Dyson, Fred Hoyle, James Watson, and others in the years after the War. At this evening in Jerusalem he was very relaxed and seemed to be happy to share his memories with Dieter.

A supervisor with the overwhelming knowledge and brilliance of John Pople had a deep impact on the scientific carrier of most if not all of his co-workers, hence also on that of Dieter. When Dieter became Associate Professor at the University of Cologne and a couple of years later Heisenberg Professor, his previous association and frequent collaborations with John Pople played an important role. When asked for recommendation letters, John never wrote flaming letters of support. He normally used short, nevertheless clear statements such as “X is a good student.” In a time of exaggerations, a “good” does not count much. In Dieter’s support letter it said: “Dieter is a very good student. I gave him the problem to ... and he solved it.” Dieter had received from other supervisors and teachers much stronger letters of support and therefore he asked Mrs. Severn (John’s secretary and soul of the group in the 70ies) how to read the recommendation letter: “Dieter realize that this is a strong letter of support. When Dr. Pople says that someone is very good it means that the student is outstanding. You should be very proud of the letter.”

In the 90ies, there were rumors that after having been considered for the Nobel price already a couple of times, there was a good chance that he would be the next Nobel price winner in Theoretical Chemistry. From the sixties on, there had been theoreticians such as Robert Mulliken, Ilya Prigogine, Kenichi Fukui, Roald Hoffmann, Dudley Herschbach, Yuan Lee, or John Polanyi to win a Nobel price and that had occurred in intervals of five to eight years. Rudi Marcus had been the last Theoretician in 1992 and therefore there was a good chance of having in 1998 another Theoretician as the next Nobel price winner. In view of these considerations, Dieter invited John to Sweden to visit the major Universities in a 2-week round-trip. Joyce, John’s wife, accompanied him on the first part of this trip and they stayed with the Cremers when visiting Gothenburg.

On the trip, John met many colleagues from all parts of chemistry and a general comment afterwards was “Well he is quite different from what we thought.” This was an aspect that accompanied John all his life. For many he seemed to be arrogant where in reality he was very friendly, however somewhat withdrawn, not to say a bit shy. When he started to give his comments on a topic these were very clear, very short, and mostly without any concession for not so well thought through alternatives. This made people believe that “John Pople is arrogant,” a statement that however could not be confirmed by any of his co-workers or those who knew him for a longer time.

In 1998, John Pople received the Nobel price (together with Walter Kohn) for “his development of computational methods in quantum chemistry.” Dieter was the first non-member of the Nobel committee to learn about this because he had to prepare a eulogy on John Pole to be given on the day of the announcement before the faculty of the Natural Sciences at Gothenburg University. Dieter fulfilled the request of the Nobel committee to sustain from notifying anybody and especially not the recipient before the committee would inform him by phone call. John got his call 6 days latter when sitting with Joyce at the breakfast table of a hotel in Houston, Texas.

John was full of plans and energy after he received the Nobel price. However, in 2002 Joyce died who was definitely the most important person in John’s life. She took many loads from John thus providing him the free space he needed for research. In 2004, John died in Chicago leaving behind a huge gap in the community of quantum chemists that still has not been filled up to this date.