In memoriam: Walter Kauzmann (1916–2009)

Introduction

Walter Kauzmann came to Princeton in 1946 to replace his former mentor, Henry Eyring, who went back to Salt Lake City and his Mormon roots. He has written about his reminiscences in physical chemistry, proteins, and related topics in Protein Science,¹ and a special issue of Biophysical Chemistry has been published containing tributes of his many students and admirers.² These sources, together with his published work and our own personal experiences, have provided the material for this tribute and review.

In his early days, WK was at once shy and modest and a wondrous communicator. The simplicity of his laboratory–office, where this communication took place, has already been described.³ His reputation for “knowing everything” was quickly perceived by students, faculty members, and visitors. There were no limits on subject matter. Sometimes discussions involved a blistering array of quantum mechanical or statistical mechanical formulas; at other times quite basic ideas for beginning students, for example, partition coefficients or elementary discussions of the entropy. The range of topics was very large and went from mathematical science, to history, geology, paleontology, and many others. A major part of learning in these early days came from observing how a real scientist worked and thought.

In the first 5 years (1946–1951), there were at least 10 students and two postdocs associated with the laboratory. Clearly, they did not fit in his small space and most spent much of their time either in the library, their rooms, or at instruments outside the laboratory, for example, the polarimeter. He was not especially solicitous with his students. You had to have something to say to get his attention. (There is a small percentage of disgruntled students.) A disagreement or a misconception on our part was especially effective as a conversation starter.

His trips to the Carlsberg laboratory in Copenhagen, where science and fun were mixed in a very natural way, made a lasting impression on him.¹ After returning from his first trip, he spent less time in his office, adopted an interest in tennis, not discussed, but obvious to his laboratory mates, and signs appeared of a lady in his life, Elizabeth Flagler, who was to become his wife in 1951.

His life was a very busy one outside of his research. One of his constant lines of interest was the structure and thermodynamics of water. Also, he published a book with David Eisenberg on water and extensive papers with Kuntz and Henn (see section on Water). He was avidly interested in his courses including those for undergraduates. A result of this interest was the publication of two excellent books^4,5 at the undergraduate or beginning graduate level and one at the graduate level on quantum mechanics.⁶

Early Background

Walter Kauzmann was born in 1916 in the town of Mt. Vernon, NY, a small city in Westchester County. He later attended public schools in nearby New Rochelle. In 1933 he entered Cornell as a chemistry major. In his autobiographical sketch,¹ he discusses at some length the effect of Walter Bancroft on his undergraduate education. Bancroft (together with G.N. Lewis and A.A. Noyes) was a major pioneer in establishing physical chemistry in the United States.⁷ Unfortunately, he had a basic mistrust of mathematics and quantum mechanics, which had an unfortunate and negative effect on the development of physical chemistry at Cornell. This caused WK to make a shift in interest to organic chemistry, which by contrast was an excellent part of the curriculum at Cornell.¹

He chose Princeton University for his graduate career and arrived there in 1937. After a few misguided months as a major in organic chemistry, the department head, H.S. Taylor, suggested that he switch to physical chemistry and work with Henry Eyring. At this time Eyring was one of the most prominent and imaginative physical chemists in the nation sharing this distinction with Linus Pauling and John Kirkwood. Eyring was always bursting with ideas and had recently published a paper with Condon and Altar on the one electron theory of optical rotation.⁸ Prior to this, most theories had to do with the dynamic coupling of electronic transitions. The new theory demonstrated optical activity as arising from static, but asymmetric, perturbations from the environment of the chromophore in addition to the process of intramolecular coupling. Another paper by Condon discussed the basic theoretical side of the problem.⁹ These papers broadened the field considerably and generated an obvious choice of study for Eyring and coworkers. WK decided on a thesis on optical activity.

He published a number of papers based on his work as a graduate student. Two were on the fundamental theory of optical rotation,^10,11 two on practical organic problems involving optical rotation, one on viscosity which probably helped to prepare him for his later viscometric studies on unfolded proteins,¹² and another of a more general chemical nature. The two papers on the theory of optical rotation will be dealt with in more detail in the next section.

Optical Rotation

About the time he started his graduate studies, two new, complementary theories of optical rotation were developed, which have been cornerstones of interpretation ever since. The first was the one electron theory, which deals with the perturbation of a chromophore by the static field of its asymmetric environment.^8,9 In the original paper, Condon and Eyring simply postulated the asymmetric field on a single electron (hence the name). In later applications, the chromophore was represented by a group wave function with ground and excited states. The field was approximated by calculating the charge distributions of neighboring groups, even of neutral groups.

The other theory was introduced by Kirkwood.¹³ It depended on the dynamic rather than static interactions of groups. In its original form, it started out with a firm basis in the quantum theory of the interaction of light and molecules and was then converted into a classical model. In its modern usage, it is quantum mechanical, and optical rotation arises from the mixed excitations on neighboring groups. It also explains effects like hypochromicity and hyperchromicity, which lead to the transfer of absorption intensity from one absorption band to another.

As a result, the field of molecular optical rotation was ripe for new development by experimentalists and theorists. WK's first paper was a survey of the literature incorporating the latest theoretical views that were made possible by the advances in theory.¹⁰ A second paper established rules for the interpretation of optical rotation and its temperature derivative: the Kauzmann and Eyring rules.¹¹ This dealt with the fact that flexible molecules with conformations governed by statistical distribution functions would, in general, decrease the magnitude of optical rotation as the temperature increased because of the decrease in the average rotational asymmetry about bonds, whereas molecules in a rigid conformation could increase or decrease. As optical rotation is a signed quantity, one must be careful to discuss the magnitude of the rotation on making statements about temperature effects. Unfolded proteins and polypeptide chains show characteristic decreases in the magnitude of rotation as temperature is increased. Indeed in many cases, it is possible to distinguish whether a chain is folded or not by merely looking at the temperature derivative of the magnitude of rotation. Some simple facts: clupein (a highly charged unfolded polypeptide), unfolded proteins, and many peptide hormones have little or no structure. Insulin folds into a compact form but its separated A and B chains are not. This type of distinction is better done nowadays with optical rotatory dispersion (ORD), circular dichroism (CD), or NMR.

On the other hand, molecules that are locked into a conformation can have temperature slopes of either sign, depending on whether increased thermal activation tends to increase or decrease the asymmetry. With proteins, one must be careful that the measurements are not made too near the thermal transition of the protein, which strongly increases the magnitude of negative rotation. WK was quick to recognize these differences, although his work was monochromatic. From the beginning of his experiments at Princeton, he used optical rotation as one of the criteria for folded and unfolded proteins. In the early days of protein work, this correlation was very useful, but is less so now because of the battery of new techniques which are available.

One part of his reminiscence is fascinating, partly because it is so little known. WK mentions his interest in measuring the sources of optical activity in molecular systems by looking at the circular dichroism of the absorption bands.¹ Awareness of this possibility can be seen in the old classical literature but little was done about it because of experimental difficulties. He discusses his interest in this type of study, and even located a unique investigator who had constructed the apparatus for the time that might have made it possible, P.A. Levine of the Rockefeller Institute. Unfortunately, Levine's apparatus had been pirated and was nonfunctional when WK visited him. This possible project was never mentioned by him in any discussions, even when CD and ORD in absorbing regions of the spectrum became popular some 20 years later. He was not interested in establishing credit for ideas that were not brought to fruition.

A last paper on the interpretation of optical rotation appeared long after most of the previous work was done.¹⁴ Although he was an authority on quantum mechanics, very little on this topic was published after his early seminal work on optical rotation and his very important book on quantum chemistry.⁶ Although he did not publish very much, he had a very thorough knowledge of the subject. This was probably acquired from his early connection with Henry Eyring, by teaching the Princeton graduate course in QM, and by working on optical rotation theory during his graduate student days. Regardless of where he obtained the knowledge, many of his hours at the blackboard were spent explaining topics in QM to the sincere appreciation of his own students and any graduate students, visitors, or others, who came to his door. Later, there was a series of papers on hybridization by Sovers and Kauzmann.¹⁵

Other Early Interests

The glassy state

Readers of Protein Science and other biochemical journals primarily think of WK as a part of their family in terms of his seminal work on the hydrophobic interaction, protein unfolding, and other major contributions. This is probably true, but leaves out his interest in a variety of physical topics early in his career and a switch toward geophysics at the end. This overly narrow description of a man with very broad interest is easily remedied. Simply “google” the name Kauzmann. This provides just two results: the “Kauzmann temperature” and the “Kauzmann paradox,” nothing on proteins. The references are to his early work on the glassy state.¹⁶ Admittedly, the prominence of these topics comes from the fact that workers on glasses always use the Kauzmann surname in referring to his work. Nevertheless, this is evidence that he has made his mark also in fields that are quite distant from his protein work.

In a paper entitled “Kauzman(n)ia,” Cherayil discusses the sudden increase in interest in Kauzmann's work on the glassy state after some 40 years of neglect.¹⁷ Those in WK's laboratory at the time of the original publication of this paper (1948) were stunned that there was a new and large topic that we would have to learn if we were to keep up with our professor. Most of us gave up the policy of keeping up at this point. The denaturation work had switched into high gear, and the laboratory was preoccupied with other activities. The glassy state work will not be discussed in further detail, because it lies outside the interests of most readers and because of the lack of experience of the authors in this field. Modern discussions of the work will be found in the Kauzmann Festschrift.²

Another topic, which will not be taken up in detail, is the creep of soft metals.^18,19

Life in the Laboratory

A real advantage to the small laboratory was the presence of the research leader right in our midst. As most of us began with no idea of what constituted real research or the nature of professors, this was a great benefit. Scientific visitors were especially interesting because the discussions were audible to all who cared to listen. Is there any better way to get to know your professor and his attitude toward research? It was especially interesting when someone like George Halsey, an Eyring student from the distant University of Washington, would implode into the laboratory, vociferous and ready for a lively discussion. Talks about science almost invariably took place at the blackboard. Those blackboard discussions added reality to the classwork acquired in courses. Princeton's graduate school was especially advanced even for present times. There were no course requirements. If you knew your material, it did not matter whether you had a course in the subject.

WK's method of introducing concepts to beginning students was especially interesting. In the early days he liked humanistic descriptions of molecular events. Molecules were not only very real but also had personalities. Many would talk of the “free energy of transfer” but Kauzmann's molecules would go into a lipid phase and cry out “Whee, I like it here.”

On the other hand this close contact was not always advantageous to the professor, who had no privacy at all during working hours. WK was not a complaining type of person, and we never heard what he thought of his overcrowded laboratory.

At this time there were no grants for faculty and no outside research fellowships for students. Apparently, there were just two Navy grants in the entire department, which supported work on the measurement of dipole moments and chemical kinetics. Many students received support from the GI Bill, which paid for tuition, books, and provided a small stipend. There were a few prestigious sounding, but slim-paying fellowships, which were reserved for the lucky. The “Harvard” fellowship paid about $700 per year. For students on the GI bill the total was more than adequate. By 1951 postdoctoral fellowships became available, and there were two in WK's laboratory by 1951. Postdoctoral fellowships for studies abroad were uncommon, but started about 1953.

But beyond these considerations, it was a great period. Students, who never expected to go to college, were moving into research. The period from 1946 to the early 1950s was one of the most exciting and galvanizing periods in American science. It became truly democratized and was the beginning of its postwar expansion. The graduate life at the Princeton Graduate College, the graduate dorm, was intellectually exciting and a wonderful experience. In his single days, WK was a frequent guest at the dinners there.

Protein Denaturation

Prior to the 1940s the word protein denaturation could mean any physical or chemical change, which would render the protein or enzyme nonfunctional.²⁰ The idea thus included a wide variety of physical, chemical, and dissociation phenomena that are not related to structure or conformation. Neurath was one of the figures who helped to clarify the subject. One of the outcomes of the early studies of protein denaturation in the Kauzmann laboratory was to limit the concept of denaturation to the very specialized reaction of proteins, which lose their ordered, native structure (disclosed in the 1930s by the extraordinary number of X-ray reflections in studies of protein crystals, outlined by Bernal²¹) and become random chains. By 1948 the interesting form of denaturation was established as an unfolding reaction from the detailed specific structure and conformation of a native protein, to a random form, resembling an ordinary polymer chain. Kauzmann's laboratory played no small role in the crystallization of this idea. This specialization of the word “denaturation” greatly simplified the concept and has been the standard ever since.

It was the philosophy of the laboratory that the only sure way of establishing the nature of this transformation was by careful and complete measurements that would provide a reasonable picture of events taking place during the unfolding reaction. WK's analogy was taking apart a watch and inspecting its mechanism, if you wanted to know how it worked, even though you might not know yet how to build one. Optical rotation and viscosity were chosen as the two physical methods that would allow him to follow changes in both the conformation of the native state (optical rotation) and would allow him to detect and determine the properties of the unfolded state (viscosity). Later, he discussed in general terms the use of long-range physical properties (e.g., viscosity, sedimentation, and diffusion),²² which yield information on the extent and flexibility of unfolded molecules, and short-range properties (optical rotation, molar volume, and absorption spectra), which are sensitive to local conformation.²³ At this time, NMR was a long way in the future for this type of application.

He made use of several proteins, principally bovine serum albumin and ovalbumin, with some experiments on β-lactoglobulin and pepsin. These were about the only proteins feasible at the time, at least for physical chemists who were far separated from biochemical preparation laboratories. The BSA and pepsin were purchased; ovalbumin and lactoglobulin were prepared in the laboratory. These proteins are sufficiently easy to prepare that even budding physical chemists can do it. Parameters, which were varied in the studies, were temperature, urea and guanidinium chloride concentration, pH, and added ionic compounds. Sulfate ion turned out to be especially interesting.

WK had been a Princeton graduate student and as such he had a very considerable background in chemical kinetics. Because of this a great deal of the work on proteins was done on ovalbumin, which unfolded irreversibly with reaction rates that depended on denaturant concentration, temperature, pH, and the concentration of various added salts or sulfhydryl agents. These effects were studied in great detail, and the results stand as a landmark in our knowledge of the effect of chemical and physical parameters on protein reaction rates. The results of these studies appeared in 1953 as a monumental series.^24–28

It was not known at the time that most simple proteins can be unfolded in a reversible reaction, which permits the study of the folding and unfolding reaction and the investigation of the thermodynamics of folding (ΔH, ΔS, ΔG, and ΔV) as functions of temperature and pressure. Serum albumin is a reversible system, but in the early studies was looked upon with a certain amount of suspicion, because it clearly contained organic contaminants in its commercial form. WK liked to demonstrate this by putting the protein into concentrated urea, whereupon a strong odor of organic compounds would develop. These were apparently buried tightly in the folded protein. Pepsin was also studied, and it was shown that this protein is not unfolded by urea under ordinary conditions of pH and temperature.

Between the start of these studies to the published results in 1953, 7 years passed. They were given a great deal of thought because they were probably the most elaborate series of studies on protein denaturation that had been produced, a record that might still stand. The students involved were not damaged by the delay. These were more leisurely times and it was not of vital importance that a new Ph.D. student should publish immediately. The “old boy” system, direct contact between research professors, was still active. Despite the serious disadvantages of this procedure, it could work well for the placement of able students who had research ability.

The importance of these studies was generally known to the world of protein and physical chemistry long before their publication. I can recall visiting the Cal. Tech. Chemistry Department in 1951. As soon as it was learned that I was a Kauzmann student, I was immediately asked to give a seminar on the laboratory's latest results on protein unfolding. With the series of papers cited above, WK had established himself as a leader in the field of protein physical chemistry. His studies made it clear that having SH groups and SS groups in a protein chain added complicated interchanges in the unfolded protein chain.²⁹ WK used this discovery to study the nature and kinetics of disulfide interchange, , where the R factors are protein chains or protein moieties. Later, there was a considerable effort to find unfolding reactions that were reversible and directly led to the thermodynamics of unfolding. Kauzmann was a pioneer in this subject also with his studies of the reversible unfolding in other proteins (see the series of papers quoted above).

The Carlsberg Laboratory

In 1949, WK informed his laboratory group that he would be taking a half-sabbatical leave for a visit to the mountains and glaciers of Switzerland, a stay in Cambridge, England to attend a meeting and visit scientists, and a working visit to the Carlsberg laboratory in Copenhagen. This trip and its various adventures are recounted in Kauzmann's later reminiscences.¹

At this time, none of us had heard of the Carlsberg laboratory, but his discussion of its three leaders from Kjeldahl, to Sorensen to Linderstrøm-Lang convinced us that it was a good place for him to go. He would first arrive in the summer for a trip to Switzerland and a walking tour of its mountains and glaciers.

At this point, it is worth mentioning that although the Carlsberg laboratory was supported by the Carlsberg Brewery, it was by way of the Carlsberg Foundation, a nonprofit organization established about 1900. Its main mission was the support of scientific research, the support of a number of Danish art and cultural museums, and the education and support of young scientists and scholars on their way to their careers. It is true that there was an extensive selection of free beer in the laboratories, but that was the only direct connection.

He must have had a wonderful time at the Carlsberg laboratory. They were certainly enthusiastic about him. Lang, the laboratory Professor, was genial and amusing as well as being a great scientist. He played the violin, painted in oils, was a born raconteur, a connoisseur, a writer and had an incredible string of jokes at his disposal, not always high-minded. Lang and the laboratory made a deep impression on WK. The same can be said for the heritage his visit made on the Carlsberg laboratory. A number of years later, Martin Ottesen, the assistant director and later the director of the laboratory, was still reminding visitors of Kauzmann's statements that one should never forget the excluded volume in presenting results on solution thermodynamics.

He had two visits to the Carlsberg laboratory, one in 1949 and the other in 1957. In both instances, remarks by Linderstrøm-Lang led him to fruitful research. During the first, he mentioned the precipitates that formed when serum albumin was unfolded by urea. Lang remarked that the addition of kerosene to urea solutions led to massive precipitation. As a result, WK turned to experimental work. His conclusion after a study of every available chain compound in the laboratory was that urea would form a crystalline product with all straight-chain organic liquids, including those with heteroatoms like diethyl ether, but would not form crystals with branched chains. The addition of urea solutions thus provided a way of separating branched from unbranched chains. But, alas, Max Møller of the Carlsberg laboratory, found a 1949 paper on the subject, which reported the phenomenon in detail including a crystal structure of the adduct.³⁰ WK was of course disappointed that things turned out this way, but not overly so. He had gotten hold of an experimental problem and solved it all by himself. In his reminiscences he goes out of his way to mention the predecessors of his work on this problem.¹

Evidently, WK returned from his first visit to the Carlsberg laboratory with the intention of broadening his activities outside the laboratory and research. Within roughly a year of his return, he had taken up tennis and bought an automobile (the new and pace-setting Studebaker). Not long afterward, he became engaged and married to Elizabeth Flagler, a research associate of the Biology department, who had already made her mark in research at the university.

The Hydrophobic Bond

The 1954 paper

In the 40s and early 50s, it was well known that the nonpolar parts of protein chains must play a part in the overall structure of proteins. They were included in every list of interactions. Also, well known was the old rule from organic chemistry that neither aliphatic chains nor aromatic molecules “liked” to be in an aqueous environment. It was apparent that bringing nonpolar chains together in the structure of a protein would lower the free energy and promote the folding reaction. The attractive energies were assumed to come from London dispersion forces, which had provided good theories of the interaction of hydrocarbons with one another in the gaseous phase and nonaqueous solution phases. This was about as far as theories went for the attractive forces of nonpolar groups in proteins: London forces between nonpolar groups contributed to the stability of proteins but presumably in a conformation-dependent way, as a leucine sidechain can interact with other nonpolar and polar groups via many pairwise conformations.

This is far different from the situation with hydrogen bonds. By this time, Pauling et al. had done their work and Linderstrøm-Lang had introduced the concept of the primary, secondary, and tertiary structure of proteins, which is now extensively used. The primary structure is essentially the sequence of amino acids, the secondary structure is a listing of the presence, location, and extent of the Pauling-Corey structures: mainly α-helices and pleated sheets,³¹ and later various “turns.” The tertiary structure is everything else seen in a successful determination of protein structure: the location and structure of disordered regions, the location of SS and other cross bridges, the location of all interactions amongst the amino acids sidechains including their pairwise geometry, and so forth. Solution methods were available to demonstrate the presence of secondary structures in proteins. The determination of tertiary structure requires a complete structure determination, and information has slowly accumulated over the years by means of X-ray diffraction.

In the beginning there was very little consideration of the interaction of nonpolar groups with an aqueous environment. This can be seen in the Kauzmann paper published in 1954,²² where the entropy of unfolding is derived mainly from the liberation of high entropy random chains. Apparently, at the time, WK was only thinking about the consequences of the removal of the aqueous shell from nonpolar molecules, which occurs simultaneously with folding. His first idea was that the interaction of a nonpolar group with an aqueous environment was energetically repulsive because the nonpolar group prevents the formation of the full complement of H-bonds in its solvation shell. In the 1954 paper, no entropic effects are attributed to the hydration shell, which plays a purely energetic role: the breaking of H-bonds in the shell.

As WK states in his reminiscences, this idea was tried out on Linderstrøm-Lang at a time when the latter was very ill. Lang, as usual, looked at the problem in the most direct and simple manner possible. According to Kauzmann, he said, “That's strange. When you mix alcohol with water, heat is given off!” This indicated that the energy of this system was lowered, not raised, on the contact of hydrophobic molecules and water, A distinct call back to the drawing board.

The 59 paper

WK found the answers he was seeking in the older literature, especially in the work of Frank and Evans³² and Butler.³³ His earlier notion was not correct. Addition of nonpolar groups to water causes a decrease in the energy and the entropy of the contact region. Both vary with temperature. The thermodynamic force is essentially entropic as ΔH goes to zero for many hydrocarbons at temperatures near to 20°. There, it is purely entropic, ΔG = −TΔS (ΔH = 0). At other temperatures, ΔG is both entropic and enthalpic with entropy dominating in the temperature range where the hydrophobic interaction is most effective. This is a reasonable summary of the results, which can differ depending on whether the aliphatic molecule is transferred from the gas phase, the pure liquid, or a solution in a different solvent.²³

WK's approach is mainly qualitative in this pioneering paper. If nonpolar surfaces in contact with water are sources of positive free energy, then the elimination of such surfaces by the folding of a protein will lead to negative free energies, which thermodynamically contribute to the folding free energy. As is well known, there are hundreds of papers that have followed this first one on the hydrophobic interactions and a number of them are discordant. We have preferred to stick to the early and simple version of the initial paper to provide a guide to WK's thought processes in their creative stage. A long review paper on the hydrophobic interaction was published a number of years ago.³⁴ When asked what he thought of this review paper with its 371 references, Kauzmann replied, “Do you know, sometimes I think that I no longer understand the hydrophobic bond.”

The concept of hydrophobia now seems reasonably clear and intuitive but details vary; the current situation is complex and not entirely resolved.

Water

Since his reading as a graduate student in the late 1930s, Kauzmann had developed a life-long interest in water and other hydrogen-bonded liquids. Understanding the unusual thermodynamic and transport properties of liquid water became even more important to him as he studied proteins and formulated his ideas on the hydrophobic bond (later referred to as the hydrophobic interaction). He used to say that we could never expect to understand the chemistry of proteins completely without understanding aqueous solutions and the structure of water—the universal biological solvent and liquid of life.

Starting in the 1960s, WK's group undertook more specific studies directed toward understanding water and aqueous solutions, doing basic work, for instance, on NaOH solutions,³⁵ solute–solute interactions in water,³⁶ effect of pressure on spectral solvent shifts in water,³⁷ the effects of temperature and pressure on the solubility of 4-octanone in water,³⁸ and solutions of water in organic solvents.³⁹

In the mid-60s, David Eisenberg joined WK's group as a postdoc and collaborated in the writing of the well-known monograph on the structure and properties of all the physical states of water.⁴⁰ Their book, originally published in 1969, was printed in several languages and continued to be printed 30 years after its initial release.

As most readers know, there has been a longstanding debate, going back to Roentgen in 1892, about the structure of water. Does this unusual liquid consist of an equilibrium mixture of discrete, variously bonded molecules, or is it made up of molecules possessing a more or less uniform, continuous distribution of energies (and configurations), as is true of other liquids.

Notwithstanding the fact that his graduate advisor and mentor, Henry Eyring, had published on a mixture-type model of water,⁴¹ WK was firmly in the camp that water did not consist of a mixture of species. He preferred a random network with distorted hydrogen bonds as a more accurate model. His thinking on the subject probably goes back to his reading of the classic paper of Bernal and Fowler.⁴² It was also clear that WK was an admirer of Pople's pioneering work on a distorted, H-bonded network to account for the properties of water.⁴³

In 1976, WK published a highly cogent, but underappreciated paper,⁴⁴ that basically shows how the mixture models of liquid water cannot account for its unusual thermodynamic properties. For instance, to explain the temperature and pressure dependences of water's compressibility and thermal expansion coefficient, one would need a large cluster of x molecules, while to explain the temperature dependence of the heat capacity and Raman spectrum, a small cluster of y molecules would be required. The paper was published in a journal not widely read by physical chemists, and probably did not gain the audience it deserved.

As he neared retirement, WK phased out all projects requiring experimentation. Indeed, he relinquished all of his laboratory space in 1980 and focused the last of his academic research solely on theoretical studies and analysis of existing data. Now, when trying to understand and explain physical phenomena theoretically, probably the majority of WK's students would agree that he preferred to start with a heuristic model, and then build from there to a theoretical explanation of physical phenomena. This is how he approached liquid water.

WK had his last grad student (ARH) review almost all the models, mixture and otherwise, that had been published on water up to that time (1978), not to mention all the computer simulations that had been carried out. Their initial effort attempted to expand Pople's model and quantify some of the particularly interesting PVT properties of water, such as the volume dependence of the heat capacity. However, it became apparent that too many parameters had to be introduced to fit the data. This was a criticism of many previously proposed models, and it was something that WK wanted to avoid.

When Rice and Sceats introduced their random network, distorted ice lattice-type of model for water around the same time (see ref.45 and references therein for a review), WK recognized it as being in line with his own thinking and considered it a major contribution to the field. Rice and Sceats, however, had not attempted to explain the pressure–volume properties of water, and it fell to us to modify and expand the distorted ice lattice model to encompass these unique properties. This was accomplished by considering a small, spherical sample volume of water whose density varied according to the H-bond distortion variables of stretching and bending. With a complete expression of the model's Helmholtz free energy in terms of the H-bond distortion variables now in hand, the free energy could be minimized with respect to those variables over a range of volumes and temperatures and the thermodynamic properties derived therefrom. WK's long-held view that water is more likely a random network of distorted hydrogen-bonded molecules culminated, after his retirement, in the publication of a random network, continuum model that successfully addressed numerous thermodynamic properties of water.^46,47

Other Activities

This section will examine a few of his WK's activities outside his research and teaching.

About the time of his marriage in 1951, the Kauzmanns initiated their annual trips to Cape Breton Island in Nova Scotia for the summer. He accepted no remuneration for this annual trip. He spent much time caring for the property, construction, and doing odd jobs. A large fraction of his book writing took place during these periods.

In addition to his research and teaching at Princeton, he was Chairman of the Chemistry Department and later of Biochemical Sciences. He was a strong recruiter for science at Princeton. He had a special interest in bringing the new biology and accomplished biological scientists to his University.

His honors include membership in the National Academy of Sciences and the American Academy of Arts and Sciences, receipt of the first Linderstrøm-Lang Award, an honorary degree from the University of Stockholm, the Stein and Moore Award, the annual Kauzmann Lecture at Princeton, a festschrift in his Honor with 48 papers by students and admirers,² and key official roles in many conferences on biophysical chemistry.

One often sees his deepest thought and interpretation in his reviews: the glassy state¹⁶ (also see Cherayil¹⁷ for a much later and very positive review), his set of five papers in 1953 on protein unfolding and denaturation,^24–28 protein denaturation and interactions,²² the hydrophobic interaction,²³ and quantum chemistry.⁶ He has summed up his own history.¹