Gilbert Newton Lewis

This article was supported by the U.S. Department of Energy under contract No. DE-AC03-765700098.

Chemical Intelligencer 1995(3), 27–37.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

I had the extraordinary good fortune to serve at Berkeley as the personal research assistant to the great physical chemist Gilbert Newton Lewis during the first two years following my receipt there of my Ph.D. in chemistry. I have often been asked to identify the ablest and greatest scientist that I have known personally during my career as a scientist (now extending over 60 years). I unhesitatingly designate Lewis as one of the two best that I have known (the other being the extraordinary physicist Enrico Fermi). Yet Lewis is relatively unknown to the present generation of scientists, including chemists. And, somehow, the Nobel Foundation made one of their rare mistakes by not awarding him the Nobel Prize in chemistry.

Although in this essay I shall focus on my personal contacts and impressions, I begin with a short account of the remarkable College of Chemistry that he developed at the University of California at Berkeley. (His influence transcended the College of Chemistry—he played an important role in turning the Berkeley campus into a world-class university.)

In the fall of 1912, Gilbert Newton Lewis (Figs. 1 and 2), then at the Massachusetts Institute of Technology, accepted the position of Dean of the College of Chemistry and moved to Berkeley. When Lewis arrived, the chemistry faculty already had four members: Edward Booth, who served until he died in 1917; Edmond O’Neill, who retired in 1925; Walter C. Blasdale, who retired in 1940; and Henry C. Biddle, who left Berkeley in 1916. From MIT, Lewis brought with him William C. Bray, Merle Randall, and Richard C. Tolman, together with several graduate students. Bray and Randall were to stay at Berkeley, but Tolman left in 1916. George Ernest Gibson from England and Germany and Joel H. Hildebrand from the University of Pennsylvania joined the faculty in 1913. These proved to be the last non-Berkeley Ph.D.’s appointed to the faculty until Melvin Calvin’s appointment in 1937 (see Table 1). Of the permanent chemistry faculty from 1912 to the present, I have known all but William C. Argo, Booth, O’Neill, and Biddle.

Fig. 1

Gilbert Newton Lewis at the Massachusetts Institute of Technology, circa 1910, two years before he accepted the position of Dean of the College of Chemistry, University of California, Berkeley.

Fig. 2

College of Chemistry Dean Gilbert N. Lewis, circa 1925.

Table 1 University of California, Berkeley—Chemistry Faculty

The photograph in Fig. 3 is one of the few of the members of the College of Chemistry and was taken in front of Gilman Hall in the fall of 1917, at about the time this building was completed. This photograph includes faculty members Ermon D. Eastman, Blasdale, Bray, Randall, Gibson, C. Walter Porter, T. Dale Stewart, O’Neill, Argo, and Lewis (Gerald E.K. Branch was away in the Canadian Armed Services, and Hildebrand was apparently out of town); Lewis’s secretary, Constance Gray, and clerk M.J. Fisher; graduate students Esther Kittredge, Esther Branch (wife of Gerald Branch), Charles S. Bisson, Wendell M. Latimer, Charles C. Scalione, Hoy F. Newton, William G. Horsch, William H. Hampton, John M. McGee, George S. Parks, Parry Borgstrom, Albert G. Loomis, Angier H. Foster, and Axel R. Olson; undergraduate students Carl Iddings, William D. Ramage, Willard G. Babcock, and Reginald B. Rule; assistant George A. Linhart; glassblower William J. Cummings and woodworker James T. Rattray and curator Harry N. Cooper.

Fig. 3

Members of the University of California at Berkeley College of Chemistry. Photo was taken in the fall of 1917, in front of the newly constructed Gilman Hall . Front Row (from the left to right): M. J. Fisher (bookkeeper), Esther Branch, Esther Kittredge, Constance Gray, Gilbert N. Lewis, William L. Argo, Edmond O’Neill, T. Dale Stewart, C. Walter Porter, G. Ernest Gibson, Merle Randall, William C. Bray, Walter C. Blasdale, and Ermon D. Eastman . Ascending Stairs (from left to right): Charles S. Bisson, Wendell M. Latimer, William J. Cummings (glassblower), Carl Iddings, Reginald B. Rule, J.T. Rattray (woodworker), Charles C. Scalione, Hal D. Draper, William G. Horsch, William H. Hampton, Willard G. Babcock, John M. McGee, George S. Parks, Parry Borgstrom, Albert G. Loomis, George A. Linhart, William D. Ramage, and Harry N. Cooper . Seated (from left to right): Axel R. Olson and Angier H. Foster.

This photograph was taken at about the time when Lewis published his famous concept of the electron pair for the covalent bond [1]. This was one of the most important ideas of twentieth century chemistry.

I started my graduate work in the College of Chemistry at Berkeley in the fall of 1934. As an undergraduate at UCLA, I had become acquainted with Lewis’s book Valence and the Structure of Atoms and Molecules [2], published in 1923, and was fascinated by it. This book described and elaborated the concept of the electron-pair bond and also his famous and useful “electron dot” depiction of the structure of atoms and molecules. I wanted to meet and become acquainted with this remarkable man, but I could not then have envisioned that I would be working with him on a daily basis.

I was drawn to Berkeley by my admiration for Lewis and by the presence there of Ernest Orlando Lawrence and his cyclotron, for I was intrigued by the relatively new field of nuclear science. When I arrived and started my classes and research, I found the atmosphere and surroundings exciting to an extent that defies description. It was as if I were living in a sort of world of magic with continual stimulation. In addition to Lewis, I met the authors of most of the chemistry textbooks I had used at UCLA—Hildebrand, Latimer, Bray, Blasdale, and Porter. I took classes from Olson, Branch, and William F. Giauque, and I opted to do my graduate research in the nuclear field under Gibson in a laboratory situated in Lawrence’s nearby Radiation Laboratory. In my thermodynamics class with Olson, I was introduced to the classic book Thermodynamics and the Free Energy of Chemical Substances [3], by Lewis and Randall. This book, also published in 1923, was a monumental contribution to placing chemical thermodynamics on an understandable theoretical and practical basis for chemists and chemical engineers. This book was also used, although augmented by more recent material, in Giauque’s more advanced thermodynamics course that I took during the second semester of my graduate work.

Nearly everyone who participated as a member of the College of Chemistry in the Lewis era recalls and comments on the Research Conference presided over by Lewis in his own inimitable style. This was held each Tuesday afternoon during the school year, starting at 4:10 p.m. and lasting until about 5:30 p.m. in Room 102 at the extreme south end on the first floor of Gilman Hall (a building which, miraculously, is still there—the only surviving building of Lewis’s day). Lewis's office was only a few doors away in Room 108, with its door usually open. His and the College of Chemistry’s secretary, Mabel Kittredge (Mrs. Wilson), was located next door to him in Room 110 (Fig. 4). At the Research Conference, Lewis always occupied the same place at the central table—the first chair on the right side facing the speaker and the blackboard. Members of the faculty sat at the table, and the others (graduate students, postdocs, research fellows, etc.) sat in chairs set at two levels at the two sides and back of the room. Lewis always had one of his Alhambra Casino cigars in his hand or mouth and several more in his upper coat pocket (Fig. 5). The first of the two speakers, a graduate student giving a report from the literature, started when Lewis gave his inevitable signal, “Shall we begin!” The second speaker—a faculty member, research fellow, or advanced or finishing graduate student—then reported on research that had been conducted in the College. Although Lewis dominated the scene through sheer intellectual brilliance, no matter what the topic, anyone was free to ask questions or speak his piece; in the latter instance, prudence suggested that the comment had best not be foolish or ill-informed. If Lewis had any weakness, it was that he did not suffer fools gladly—in fact, his tolerance level here was close to zero.

Fig. 4

Entrance door to Room 110, Gilman Hall, University of California, Berkeley, Mabel Kittredge’s (Lewis’s secretary) office, and the official entrance to Gilbert N. Lewis’s office in Room 108.

Fig. 5

Gilbert N. Lewis with his cigar as he was typically seen at the Tuesday afternoon Research Conference in Room 102, Gilman Hall, University of California, Berkeley.

During my three years as a graduate student and the subsequent years until the war, Lewis always attended the Nuclear Seminar held on Wednesday evenings in Room 102, Gilman Hall. This seminar was run by Willard F. Libby, together with Robert Fowler (until he left Berkeley in 1936), and was attended regularly by Latimer, Bray, and Eastman. Lewis also conducted some research with neutrons in 1936 and 1937. He was always highly supportive of my nuclear research, some of which was conducted in my spare time during the period in which I served as his personal research assistant.

With this background in mind, let me now proceed to a description of my work with Lewis as a research associate. I’ll never forget how this got started. I had completed my graduate research in the spring of 1937, my Ph.D. degree had been awarded, and it was time for me to go and find a job someplace. Lewis didn’t recommend me for a position anywhere, which I could have regarded as a bad sign. Actually, in this case, it was a good sign. That meant that I still had a chance to stay at Berkeley in some capacity—which, of course, was my objective. One day in July after the next academic year had actually started (so I was technically without any salary), Lewis called me into his office and asked me if I would like to be his research assistant. Lewis was unique in having a personal research assistant, whose salary at that time was $1,800 per year. Although I was fervently hoping to stay in some capacity, I was flabbergasted to find he thought me qualified for this role, and I expressed my doubts to him. He smiled and indicated that if he didn’t think I could do the job, he wouldn’t have offered it to me. My acceptance of the position he offered was enthusiastic, and thus our two-year intimate association began.

Lewis had suffered some disappointment in his previous research with neutrons. In fact, I had played a role in advising him frankly where he was going wrong, an act that took some courage on my part, and this may have influenced him in his decision to undertake the risk of having me as his research assistant. He told me that he had decided to forgo research for a time, during which I would be free to continue the nuclear research that I had under way. As I have already indicated, I continued a rather substantial effort in the nuclear field, with his blessing, during the entire two-year period that I was associated with him.

In the late fall of 1937, Lewis resumed his research. He decided to try to separate the rare earths praseodymium and neodymium using a system involving repetitive exchange between the aqueous ions and their hydroxide precipitates. He employed a long, tubular, glass column extending from the third floor to the basement at the south end of Gilman Hall. The column was constructed with the help of Bill Cummings, the long-time glassblower in the college, and erected with the help of George Nelson, the irascible head of the machine shop. (He was irascible from the standpoint of graduate students but very polite to Lewis and now to me in my prestigious role as the assistant to the “Chief.”) The long column was serviced by a machine-driven system for agitation in order to keep the hydroxide precipitates suspended along the column’s length. It was my duty to keep this operating, which I did with only limited success. Lewis, with no help from me, measured the degree of separation of the praseodymium from neodymium with the spectroscope in the darkroom off Room 301, Gilman Hall. For whatever reason, including possible shortcomings in my performance, no detectable separation of praseodymium from neodymium was achieved.

In the early spring of 1938, Lewis returned to his former interest in acids and bases. If I recall correctly (this was 57 years ago!), he was, at least in part, motivated by the need for an interesting topic, supported by feasible experimental demonstrations, for a talk that he was scheduled to give at the Franklin Institute in Philadelphia in May on the occasion of his receiving a Doctor of Science degree and honorary membership in the Franklin Institute, in connection with the dedication of the Benjamin Franklin Memorial (i.e., the large new building housing the Institute’s activities, including the science museum). In any case, much of our first work in this area was concerned with such demonstration experiments.

Our experiments were directed toward his generalized concept of acids and bases. In his 1923 book Valence and the Structure of Atoms and Molecules, Lewis had proposed a very general definition of acids and bases. According to that definition, a basic molecule is one that has an electron pair which may enter the valence shell of another atom to consummate the electron-pair bond, and an acid molecule is one which is capable of receiving such an electron pair into the shell of one of its atoms. Lewis wanted, with my help, to find a broad base of experimental evidence for this concept.

We worked in Room 119 (Fig. 6), at the north end of the first floor of Gilman Hall, a laboratory that Lewis had used for a number of years previously. It was here that he did his trailblazing work with Ronald McDonald and others during the period 1933–1935 on the isolation of deuterium by the electrolysis of water and the determination of a number of its properties [4]. The apparatus used for this work was still there in the east side of the room, a part of the room that we didn’t use at this time. We used the laboratory bench extending along the west side of the room, flanked in the back by a row of windows. The sink, at which I washed and cleaned our glassware each evening (Fig. 7), was at the extreme right (north) end of the bench, and our writing desk adjoined the opposite end of the laboratory bench against the south wall. Our indicator experiments were performed on the laboratory bench top at the ambient room temperature in ordinary test tubes. For later, more sophisticated (but still basically simple) experiments, which I shall describe presently, we used a low temperature bath that consisted of a large, wide-mouth Dewar filled with acetone which was cooled by the addition of chunks of dry ice. Our vacuum bench, used in later experiments, was in the center of the room, opposite and parallel to the laboratory bench.

Fig. 6

Entrance to Room 119, Gilman Hall, University of California, Berkeley, where Seaborg served from 1937 to 1939 as research associate to Gilbert N. Lewis.

Fig. 7

Sink area in Room 119, Gilman Hall, University of California, Berkeley, where Seaborg serviced the experiments he performed with Gilbert N. Lewis in between moonlighting as a nuclear chemist.

I was immediately struck by the combination of simplicity and power in the Lewis research style, and this impression grew during the entire period of my work with him. He disdained complex apparatus and measurements. He reveled in uncomplicated but highly meaningful experiments. And he had the capability to deduce a maximum of information, including equilibrium and heat of activation data, from our elementary experiments. I never ceased to marvel at his reasoning power and ability to plan the next logical step toward our goal. I learned from him habits of thought that were to aid continuously my subsequent scientific career. And, of course, working—and apparently holding my own—with him boosted my self-confidence, which was not at a very high level at this stage of my life.

Starting at this time, I worked with Lewis on a daily basis, interspersed with intervals when he was otherwise occupied and during which I pursued my nuclear research. He would arrive each day between 10 and 11 a.m. in his car, a green Dodge, which he would park on the road, South Drive (Fig. 8), between the chemistry buildings and The Men’s Faculty Club (where I was living at that time). When I spotted his car, I knew that it was time to join him in Room 119. We then usually would work together until about noon or 1 p.m., when he went to The Faculty Club to play cards with his friends (he didn’t eat any lunch) while I went to lunch. He usually returned to our laboratory at about 2 p.m., and we would work together until late afternoon. This gave me time to work on my other research projects before he came, during the noon break, and after he left. However, he often gave me assignments to assemble materials, prepare solutions, etc., over the noon hour, or overnight, or when he left town for a day or two. These assignments were usually unrealistically demanding for such a time scale, and I had to scramble to meet his demands. This was done not for the purpose of keeping me busy, but because he underestimated the size of the tasks. Sometimes we worked in the laboratory during the evening after dinner, often on Saturday morning, and occasionally on Sunday. We did most of the writing up of our work for publication on Sunday afternoons.

Fig. 8

 The Campanile, Le Conte Hall, Gilman Hall, and Chemistry Building on the campus of the university of California, Berkeley, circa 1941. South Drive, where Lewis would park his green Dodge every morning, is in the foreground.

Lewis gave his talk at the Franklin Institute in Philadelphia on Friday morning, May 20, 1938, as scheduled. During his talk, he performed the demonstration experiments that we had developed. So far as I know, his talk was well received. However, the main impact came from his publication, based on the talk, which appeared in the September issue of the Journal of the Franklin Institute [5]. In the preparation of this paper, which was written entirely by Lewis without my help, he used additional data that we developed in subsequent experiments. However, the main thrust of the paper was his beautiful exposition of his concept of generalized acids and bases, which had a worldwide impact and became the “bible” for workers in this field. His primary acids and bases are characterized by their instantaneous neutralization reactions, which occur without any heat of activation. In this paper, he also introduced his concept of secondary acids and bases, whose neutralization requires a heat of activation. I soon found that I was destined to work with him on a program of experimental verification of this idea.

I helped Lewis pack his equipment for his travel by train to Philadelphia for his demonstration lecture at the Franklin Institute. I was pleased to see him bring into our laboratory and place on the bench two suitcases because I felt this would give me ample room to pack the material for his demonstration experiments. However, he told me that he would need much of this space for his cigar boxes. He filled one entire suitcase and part of the other with cigar boxes, which meant that I had to exercise some ingenuity in order to get the equipment, chemicals, etc. into the remaining space.

Lewis and I resumed our experiments on generalized acids and bases during June and early July 1938, after he returned from his trip to Philadelphia. We found many cases where, with one solvent and one indicator, the colors obtained seemed to be dependent only upon the acid or basic condition of the solution and not at all upon the particular acid or base. By means of the color changes, the solutions could be titrated back and forth as in aqueous solution. For example, with thymol blue dissolved in acetone, the color was yellow with either pyridine or triethylamine, while the acids SnCl₄, BCl₃, SO₂ and AgClO₄ gave an apparently identical red color. With crystal violet in acetone, the color changed successively from violet to green to yellow upon the gradual addition of SnCl₄ or BCl₃, after which the original violet color could be restored upon the addition of an excess of triethylamine.

Because similar effects could also be obtained with HCl, and since we had been working in the open with reagents that had not been especially dried, we were afraid that some of the similarities in color produced by the different acids could be due to small impurities of H-acids in the reagents. We therefore conducted experiments with very dry solvents, given to us by Dr. C.H. Li, with indicators that themselves contain no labile hydrogen, such as butter yellow, cyanin, and crystal violet, and upon the vacuum bench to prevent the pickup of water. These experiments gave the same results as those performed in the open with ordinary reagents.

Toward the end of June, Lewis gave me leave to go to San Diego to give a talk on my nuclear work at a meeting of the American Physical Society. During my absence, he conducted vacuum bench experiments to observe the color changes when SnCl₄ and triethylamine were added to a solution of crystal violet in thoroughly dried chlorobenzene, when SnCl₄ or HC1 were added to a solution of butter yellow in chlorobenzene, etc. I reproduce in Fig. 9 his notes covering one of these experiments as he recorded them in my notebook.

Fig. 9

Entry in Gilbert N. Lewis’s handwriting in Seaborg's laboratory notebook describing experiments an acid/base systems carried out in Seaborg's absence. June 23, 1938.

In addition to taking some vacation during the summer of 1938 with his family at their cottage in Inverness, he spent a good deal of time on his paper “Acids and Bases,” which he was getting ready to send to the Journal of the Franklin Institute. The process of formulating his thoughts and setting them down on paper suggested to him many little confirmatory experiments, which we then performed. I reproduce in Fig. 10 a sample page from my journal (notebook) of this period.

Fig. 10

Sample page from Seaborg's laboratory notebook during his collaboration with Gilbert N. Lewis. August 23, 1938.

In September, Lewis turned to his next project—experiments related to his concept of secondary acids and bases—and from the latter part of September until Christmastime, I worked with him on a daily basis on much the same schedule as I outlined earlier. We did some broadly based experiments, which led to the publication of our background paper “Primary and Secondary Acids and Bases” [6], and a detailed investigation of a specific secondary and primary base, which was published as a companion paper entitled “Trinitrotriphenylmethide Ion as a Secondary and Primary Base” [7].

It was in the course of this detailed investigation of this secondary and primary base that I was to see firsthand a master researcher at work and to be privileged to be a participant in his work. Here was a prime example of simple experiments leading to interesting and fascinating interpretations, and in my description I shall do my best to capture the flavor of the process. As background for understanding these experiments, we should recall that Lewis had suggested that there is a large group of acids and bases, called primary acids and bases, which require no energy activation in their mutual neutralization, and there is another group, called secondary acids and bases, which do not combine with each other (nor does a secondary base combine with a primary acid nor a secondary acid with a primary base) except when energy, and frequently a large energy, of activation is provided.

As the experiments that we conducted are illustrative of the Lewis method, I shall describe them in some detail. Based on the results of some preliminary experiments and Lewis’s intuition and analysis, we decided that the intensely blue 4,4′,4″-trinitrotriphenylmethide ion should be a base that could exist in the primary and secondary forms and be a good material for experimentation to give support for and information on this concept.

Lewis soon deduced that our first interest should be in the secondary base ${\mathrm{B}}_s^{-}$ in the blue form that requires a heat of activation to be converted to the primary form $\left({\mathrm{B}}_p^{-}\right)$ in which it reacts instantaneously with acetic acid. Thus, he deduced that the two forms would have the formulas shown below.

We launched into a series of kinetic experiments to measure the rate of the fading of the blue ${\mathrm{B}}_s^{-}$ upon the addition of acetic acid or other acids which combined instantaneously with the small proportion of ${\mathrm{B}}_p^{-}$ that was present. This mechanism, for any acid HY, can be summarized as follows:

$$ {\mathrm{B}}_s^{-}\rightleftarrows {\mathrm{B}}_p^{-} $$

(1)

$$ {\mathrm{B}}_p^{-}+\mathrm{H}\mathrm{Y}\to {\mathrm{B}\mathrm{HY}}^{-} $$

(2)

$$ {\mathrm{BHY}}^{-}\to \mathrm{H}\mathrm{B}+{\mathrm{Y}}^{-} $$

(3)

Lewis suggested that reaction (2) is the rate-determining step and that the concentration of ${\mathrm{B}}_p^{-}$ depends upon the concentration of ${\mathrm{B}}_s^{-}$, the temperature, and the difference in energy between ${\mathrm{B}}_p^{-}$ and ${\mathrm{B}}_s^{-}$. On this basis, the reaction should be bimolecular, and the measured heat of activation should be the same with all acids (HY) of sufficient strength.

To test this, we measured the rates of reactions (rates of fading of the blue color) over a range of temperatures in order to determine the heat of activation. The experimental method was simplicity itself. The first experiments were performed in open test tubes, but it was found that trinitrotriphenylmethane was sensitive to oxygen under the conditions used, and therefore the reaction vessels were evacuated. Our solvent was 85 % ethyl alcohol and 15 % toluene, and our first series of experiments were with acetic acid. The reaction vessel, in the form of an inverted Y, with the alkaline blue methide ion solution in one limb and the acid in the other, was placed in the low-temperature bath (of acetone cooled with dry ice). When temperature equilibrium was attained, the vessel was tipped rapidly back and forth until the contents were thoroughly mixed. The reaction (rate of fading of the blue color) was then followed by comparing the color with a set of standard color tubes. (The set of standard color tubes consisted of solutions of crystal violet, which had blue colors nearly identical to those of the blue methide ion, made by successive two fold dilutions to cover the entire range of diminishing blue color.) After the experiments had indicated that the reaction was always of first order with respect to the colored ion, the procedure was simplified further. The time was taken merely between the mixing and the matching of a single color standard, which corresponded to l/16th of the original concentration of the blue methide ion (i.e., the color standard was made by four twofold dilutions of the original matching crystal violet solution).

We made measurements with acetic acid at four temperatures: −53 °C, −63 °C, −76 °C, and −82 °C. From these measurements, we could calculate that the reaction was first-order with respect to the acid and that the heat of activation for the reaction of fading the blue methide ion was 8.6 kcal. According to our interpretation, then, this is the energy difference between the secondary form B _s and the primary form B _p of the methide ion. We next measured the heat of activation for the same reaction for five additional acids, for which the reaction also proved to be bimolecular, and found the same value for the heat of activation within the limits of our experimental uncertainty—an average of 9.1 kcal. Such a result is to be expected from our interpretation that the heat of activation should be equal to the difference in energy between the primary and secondary forms of the base. If the activation occurred only at the moment of collision between the reacting molecules, it would be hard to explain why the heat of activation, or, in other words, the potential barrier in the activated complex, should be the same for such very different substances as alcohol (for which we also measured the heat of activation, indirectly, as described below) and our other acids—chloroacetic, furoic, α-naphthoic, lactic, and benzoic—as well as acetic acid.

I have recounted here in some detail only the central conclusions from this research. Lewis made many other deductions that are too involved to be easily described here, but which can be enjoyed by reading the paper reporting this work. I shall merely sketch some—by no means all—of these conclusions. From some of our other measurements, he was able to deduce the equilibrium constant for the reaction in which the blue methide ion is formed from the reaction of the hydroxide (or ethylate) with the trinitrotriphenylmethane and the heat of activation, from which he found that the heat of activation for the reverse reaction (${\mathrm{B}}_s^{-}$ plus ethyl alcohol), corresponding to the difference in energy between the primary and secondary forms of the base, is 8.9 kcal, in good agreement with our direct determination for the six acids (9.1 kcal). He could deduce from our measurements that only one-eighth of the trinitrotriphenylmethane was in the form of the blue methide ion under the conditions of the kinetic experiments. He also concluded that our kinetic measurements with such weak acids as phenol and boric acid suggest that these displace the solvent alcohol from the nitro groups in the blue methide ion to an extent depending upon their concentration and that the ion with the phenol attached is less reactive than the corresponding alcohol compound.

We found that an orange color was produced immediately upon the addition of the strong acid HCl to a solution of the blue methide ion. We also found this upon the addition of the relatively strong trichloroacetic acid. Lewis found a ready explanation for this. When the blue ion has been formed and the central carbon has lost its power of acting immediately as a base, the basic power has, in a certain sense, been transferred to the three nitro groups. Therefore, a sufficiently strong acid should attach itself at one or more of the nitro groups, and in this process the blue ion should act as a primary base.

We finished these experiments just before Christmas time in 1938. After a diversion in January to test another of his ideas experimentally, we began in February the process of writing our two papers [6, 7] on primary and secondary acids and bases for publication in the Journal of the American Chemical Society. Writing a paper with Lewis was a very interesting process. We did most of our work on these papers, sporadically over several months, on Sunday afternoons in our laboratory, Room 119 in Gilman Hall. The process consisted of Lewis, pacing back and forth with cigar in hand or mouth, dictating to me. I recorded his thoughts in longhand. However, his output was interspersed with discussions with me and even with experimental work when he wanted to check a point or simply wanted a break. His sentences were carefully composed, and the result was always a beautiful and articulate composition.

After we had finished with the two papers up to the point of the summary of the second paper, he said to me that he was tired of this process and suggested that I write this summary by myself. By this time, I was familiar enough with his thought processes to make this feasible. I wrote the following, which he accepted after no more than a glance at it and without changing a word:

Trinitrotriphenylmethide ion was expected and has proved to be a secondary base. In alcohol when this blue ion is added to any weak acid at temperatures between −30 and −80° the formation of the corresponding methane is slow and can be followed colorimetrically. The rate of neutralization was studied with numerous acids and under like conditions the rates diminish with diminishing acid strength. With the weakest acids the rates are not proportional to the concentration of acid, and this fact is explained. With the six acids of intermediate strength the rates were found proportional to the concentrations of blue ion and of unionized acid, and unaffected by neutral salts. In these cases the heat of activation was calculated from the temperature coefficient of the rates and was found approximately constant with a mean value of 9.1 kcal. By indirect methods the rate of neutralization by alcohol itself was determined. Here the heat of activation is found to be 8.9 kcal. The constancy of the heat of activation over the great range from chloroacetic acid to alcohol can hardly be explained by the theory of an activated complex. The value obtained is taken as a measure of the difference in energy between the primary and secondary forms of the base. The small departures from this constant value are attributed in part to experimental error, but especially to differences in the actual composition of the reacting ion. Several kinds of evidence are adduced to show that the actual composition of the blue ion depends not only upon the solvent but in several cases upon the presence of other solutes.

While the trinitrotriphenylmethide ion is a secondary base with respect to addition of acid to the central carbon, it is a primary base with respect to addition of acid to the nitro groups. In the presence of strong acids an orange substance is thus formed which contains more than one free hydrogen ion per molecule. The very slow rate of fading of the orange compound is studied, and an explanation is suggested for the large catalytic effect of water. Mono- and dichloroacetic acids give mixtures of the orange and blue substances and the rate of fading in these solutions leads to some of the conclusions already mentioned.

During January 1939, Lewis and I worked to make an experimental test of an old, rather far-out, idea of his. This is far afield from acids and bases but is, I believe, worth mentioning as a further illustration of the breadth of his intellect and interests. A number of years before (1930), he had published an article in Science magazine on the “The Symmetry of Time in Physics” [8]. He had already revealed this idea in his third book, The Anatomy of Science, the published account of his philosophical Silliman Memorial Lectures, published in 1926 by the Yale University Press [9], A consequence of this theory, as it applies to radiation, is that we must assign to the emitting and the absorbing atom equal and coordinate roles with respect to the act of transmission of light. A consequence of this, Lewis told me, is that the receiver or observer of the light (for example, the apparatus used for this purpose) is of importance equal to that of the emitter of the light and exerts its own influence upon how the light manifests itself.

Lewis told me he wanted to test this hypothesis by setting up a Michelson interferometer to detect the interference fringes with different receivers or detectors of radiation and to thus determine if some properties of the radiation depend on the receiver or detector as it should if it conformed with his theory on the symmetry of time. He asked me to set up a Michelson interferometer in the dark room off Room 301 at the southwest comer of the third (attic) floor of Gilman Hall. This room contained a spectrograph with which Lewis had made his spectrographic measurements mentioned earlier on rare-earth samples.

I went to the Department of Physics and borrowed a Michelson interferometer that was ordinarily used for demonstration experiments in some of the physics lecture courses. In order to make this operate correctly, I had to prepare some “half-silvered” surfaces on glass with a silver layer of such thickness that about one-half of the incident light would be reflected and the other half transmitted through the layer. Since Professor Axel Olson had some experience with this “half-silvering” process, I enlisted his help. Lewis and I detected the interference fringes with each of a number of different types of photographic film in order to see if we could detect any gross differences in the way the films reacted. We found some peculiar effects, which excited Lewis for a time, but my skepticism prevailed when I was able to explain these as due to rather prosaic failures in our techniques and to show how we could eliminate the effects by correcting our techniques. These negative results then convinced Lewis to go on to something else.

During the period from January to June 1939, Lewis and I did scouting experiments with a wide range of indicators, acids, and bases. Many interesting observations were made that are not susceptible to summarization in a reasonably brief fashion. As always, there were moments of excitement. I recall a series of experiments, conducted with test tubes immersed in our acetone-carbon dioxide bath, on the development of color when trinitrobenzene and sodium phenol-late were reacted in absolute ethyl alcohol over a range of temperatures below room temperature. We found that large excesses of NaOH were needed to produce the indicator color. This elicited some bizarre interpretations from Lewis. However, when these experiments were repeated on the vacuum line, the action of NaOH was more reasonable. Apparently, in our open test tube experiments, large amounts of CO₂ were absorbed in the alcoholic solution from our ${\mathrm{CO}}_2$-cooled acetone bath!

Our research during this period did result in one coordinated project from which some interesting conclusions could be drawn. We reacted each of the bases ammonia, methylamine, dimethylamine, triethylamine, and hydroxide with each of the acids m-dinitrobenzene (DNB) and symmetrical trinitrobenzene (TNB), trinitrotoluene (TNT), trinitroxylene (TNX), and trinitromesitylene (TNM)—25 combinations in all—and made observations on the degree of development of color (a measure of the degree of reaction between these acids and bases). At any point in Table 2 corresponding to a given base and a given nitro compound, the sign + indicates the formation of color.

Table 2 Color Production in Mixing Several Bases with Nitro Compounds

We found that with trinitrobenzene the intensity of color is least with triethylamine, greater with dimethylamine, and still greater with methylamine and ammonia. For the direct addition of the base to one of the ring carbons that is not attached to a nitro group, there is the possibility of double chelation of hydrogen atoms to nitro groups in the case of methylamine and ammonia, thus strengthening the acid-base combination. With the weaker acid m-dinitrobenzene, methylamine and ammonia—which are capable of double chelation—give good colors, while the two stronger bases dimethylamine—which is capable of only one chelation—and triethylamine—where no chelation is possible—give no color at all. Thus, our conclusion was that the stability of the colored compounds is greatly enhanced by chelation, in which the hydrogens of an aliphatic amine are attached to oxygens of the nitro groups. Similarly, we could deduce that the chief effect of introducing methyl groups into symmetrical trinitrobenzene is to diminish resonance between the nitro groups and the ring and that this effect, which is very strong when the nitro group is ortho to two methyl groups, as in symmetrical trinitroxylene, becomes weak when only an ortho methyl is present, as in symmetrical trinitrotoluene. Trinitromesitylene, in which each nitro group lies between two methyl groups, showed no color with any base.

Lewis and I didn’t write up this work for publication until about a year later owing to the press of our other activities. When we did, of course, it was done by the same method of dictation with me serving as a scribe. Our publication, which included explanations for all of our observations, was entitled “The Acidity of Aromatic Nitro Compounds toward Amines. The Effect of Double Chelation” [10].

During my last months with Lewis, April, May, and June 1939, he turned part of his attention toward spectroscopic observations on light absorption and the observations of fluorescence and phosphorescence in various colored organic substances. For this we used the spectrograph in Room 310, Gilman Hall, where Ted Magel, then a graduate student, was working. Lewis was now beginning his experimentation on the relation of energy levels in molecules to their emission of light and was already beginning to think in terms of the triplet state. Besides Magel, Otto Goldschmid, a volunteer research fellow, and Ed Meehan, an instructor in the College of Chemistry helped us in these measurements. Melvin Calvin, David Lipkin, Jacob Bigeleisen, and Michael Kasha were active in this program (Fig. 11).

Fig. 11

Gilbert N. Lewis at vacuum line during work on spectroscopic observations.

Also during this time, Lewis was working with Calvin, putting the finishing touches on their review paper “The Color of Organic Substances,” which they mailed in August for publication in Chemical Reviews [11]. Lewis had been interested in the color of chemical substances for a long time, and, in fact, this was the subject of his acceptance address in New York on May 6, 1921, when he received the Nichols Medal of the New York Section of the American Chemical Society. He had been working with Calvin, off and on, during much of the last year. I can recall looking in on them in Room 102, where they had their writing sessions, and finding them totally immersed in their piles of reference journals and notes.

During all of the time that I was working with Lewis, he was, of course, serving as dean of the College of Chemistry and chairman of the Department of Chemistry. These positions would ordinarily entail heavy administrative duties, but he did not allow himself to be burdened by them. Nevertheless, I believe, he discharged his responsibilities very well (Fig. 12). He was efficient and decisive, highly respected by the faculty members in the college, and eminently fair in his dealings with them. To a large extent, he ran the college from his laboratory. I recall that his efficient secretary, Mabel Kittredge, would come into our laboratory, stand poised with her notebook until she commanded his attention, and then describe clearly and briefly the matter that required his attention or decision. Lewis would either give his answer immediately or ask her to come back in a little while, after he had given the matter some more thought. This system worked very well in those days but might not be adequate today, and, in any event, certainly could only function with a person of Lewis’s ability.

Fig. 12

Gilbert N. Lewis at work in his Gilman Hall office. Room 108, University of California, Berkeley.

Lewis never addressed me with a harsh word although there were times when he might have been justified in so doing. I recall one occasion when, together with friends, I had overindulged in alcohol the previous evening to the extent that on the following morning, I had to steady my right hand with my left hand in order to turn a stopcock on our vacuum line. With a grin on his face, he recommended that I should return to my room in The Faculty Club “to rest” awhile; I took his advice and was able to return to my duties in the afternoon.

Sometime in June 1939, Lewis told me that he was putting me on the faculty of the College of Chemistry as an instructor. In his whimsical way, he expressed the opinion that he had been taking up “too much of my time.” This was a revealing comment considering that I was supposed to be serving as his full-time research assistant. However, I have good reason to believe that he was not at all unhappy with my additional research and writing projects. He told me my salary would be $2,200 per year, that of a third-year instructor. Thus, to my delight, he was giving me full credit for my two years in the capacity of his research assistant.

In conclusion, I want to say that I regard it as extraordinarily good fortune that I was granted the privilege of spending this time working so closely with Gilbert Newton Lewis, an extraordinary scientist of the twentieth century.