Gaston Berthier (1923-2009)
© SocFran.d'Exobiologie
|
Gaston Berthier (1923-2009).
Né en 1923, année du 10e anniversaire de la théorie quantique de Bohr, Gaston Berthier a pris part à toute l’évolution de la chimie théorique et de ses conflits internes.
Père de deux avancées majeures de la discipline, la théorie Hartree Fock sans contrainte de spin (UHF) en 1954, puis, conjointement avec l’un des ses élèves, B. Lévy, le théorème de Brillouin généralisé en 1968, Gaston Berthier est largement responsable du développement de la discipline, tant par ses contributions scientifiques, fondamentales ou applicatives les plus diverses, que par son influence sur des générations de jeunes chimistes théoriciens, en particulier dans le cadre de l’Ecole Normale Supérieure avec J. Serre ainsi que dans celui de la communauté des Chimistes Théoriciens d’Expression Latine (CHITEL), véritable melting pot scientifique dont il était l’un des parrains fondateurs.
Son testament scientifique se trouve probablement dans son dernier livre « Nécessaire de chimie théorique » paru en 2009, livre qu’il était venu nous présenter au dernier workshop du laboratoire, passionnant une assistance de jeunes thésards médusés.
Célibataire, il était un féministe convaincu qui a poussé nombre de jeunes femmes à s’affirmer dans cette discipline théorique. Il fut aussi un défenseur irréductible de la langue française et de la culture classique, au point de publier quelques articles en latin.
La seule affirmation erronée qu’il fit, fut d’avoir dit que personne n’est irremplaçable ; pour sa famille scientifique il l’est.
Les Auteurs: Ses collègues du Laboratoire de chimie théorique - The direct link is presently (Jan. 2010) not available
Another online-link (Jan. 2010) is to be found
here ...
Offline (modified)- here ...
|
A short audio impression of the original interview
here... -
the text-position of this audio is to be seen
here ...
Table of further audio cuts of this interview here ...
| |
An Interview with Professor Gaston Berthier
Paris, ENS, rue L'Lhomond, June 2, 1997 - 15:00
This transcript was authorised by G. Berthier.
Anders: Professor Berthier, your first paper, to my knowledge, appeared in 1948, published jointly with B. Pullman.
Professor Berthier: In the Bulletin of the French Chemical Society - on styrene and divinyl-benzenes (1).
A: How old were you then?
B: I was 25 years, being born in 1923.
A: Were you a chemist or a physicist?
B: A Chemist in fact.
A: That's surprising. Why did a chemist in these day go and work on Hückel Theory?
B: Well, I should say that before starting with Pullman I was working as a chemist in the automobile industry, to be precise in analytical chemistry. After one year I decided to leave because it was not a very interesting job in those days. I met Pullman, who was young at that time, in the Curie Institute and I took my chance.
A: What was the spirit in those days on Hückel calculations? In an academic environment one can do work for two reasons: a) for an (academic) career or b) out of sheer interest alone - -
B: In that time physical chemistry and quantum chemistry were not à la mode in France. Because in post-war France chemistry meant organic chemistry. Now it happened that just at that time the French government started a new small organization which grew into CNRS (Centre National de la Recherche Scientifique) and in which one was - in that time - free to work on anything. Pullman was interested in theoretical chemistry and in the Hückel calculation method. Without the use of computers, of course.
Well, the question is complicated to explain. During the war Bernard Pullman, being a refugee from Poland, was staying in Africa and England. His future wife, Pullman's fiancée, Alberte Bucher, remained in France during the war, working with Daudel. There one day she discovered with him papers from Germany, by Otto Schmidt, who was working at the BASF (Badische Anilin und Sodafabrik) in Ludwigshafen (2). And Otto Schmidt was speaking of relationships between electron densities and biological properties in hydrocarbons like anthracene.
That was the beginning of the story. So Mademoiselle Alberte Bucher began to make calculations first by Pauling's somewhat complicated valence bond method, some time later by the Hückel method.
Mademoiselle Alberte Bucher became married right after the war when Monsieur Bernard Pullman had returned to France. That was in 1945 and from then on Alberte and Bernard Pullman began to work together on this problem which was discovered, as I mentioned already, during the war in German papers published in 1938-1939. It seems to have been in Zeitschrift für Physikalische Chemie or something like that - - . The citation is right here in this book of A. and B. Pullman Cancérisation par les substances chimiques et structure moléculaire of 1955 (3).
(A short audio impression of the original interview at this point)
here...
A: Streitwieser in his famous book of 1961 (4) mentions you 23 times but these were, more or less, 23 papers. (Streitwieser cites himself 68 times, but those were exactly 3 papers). Were you trying to find out more on the carcinogenic properties?
B: After looking at the biological properties of this type we started with the physico-chemical properties of the conjugated aromatic hydrocarbons in general. Studies of this type continued during the next 20 years.
I belonged to the Pullman group for some time. I did not formally leave the Pullmans, but for commodity reasons I came to this school (ENS, École Nationale Supérieure). We are just now sitting here in the physics lab but formerly I was with the chemistry division. With some young students we were interested in new calculating methods including the so-called ab-initio methods. We left the Hückel method for ab-initio in the sixties. In that time computers became available in France through the IBM company. We were indebted to IBM France for some calculations at this time.
A: You didn't have to pay for it?
B: In the beginning not. We had some kind of a grant with IBM. However, it is a commercial company, set up for selling, so around 1960 we could finally buy ourselves a small IBM 1620 computer. And it was this computer in France that Kutzelnigg in fact used, working at his first paper on natural orbitals (5,6). And in those time a computer meant a lot to us.
A: At first did you use mechanical or electromechanical calculators? How did it start on the calculator side? At first did you do it all exclusively by hand?
B: In the beginning we had electromechanical calculators. It is a curious thing: The story is connected with the general history of that period. In 1945, during the occupation of parts of Germany by the French troops, some calculating machines were confiscated. The CNRS obtained some of these and distributed them in France. These machines originated from a small German town in the Schwarzwald called Sankt Georgen, from a company called Bäuerle, the brandname of the machine being, as I still now recall, Peerless (7). Well, from then on we used electromechanical machines for more than 10 years e.g. American ones like the famous Friden and also a German machine - the name of it was Mercedes, from Jena, I think.
It was around 1961/62 when IBM France installed a really modern computer in the Pullman lab, not far from ENS, 24, rue Lhomond.
A: Working with punched cards?
B: Yes. And also in this time we started with ab-initio calculations.
A: Did your group have to pay to IBM? Did the ENS pay and it was free for you?
B: At this time we had to pay. Later on we obtained from CNRS a certain amount of computer time. But now we have all workstations and all that - - .
A: How would you answer the question as to how closely chemistry and computers are connected?
B: I think that, in France at least, chemists are organic chemists and they are not really interested in physical chemistry. They are dived in true organic chemistry. In France chemistry means organic chemistry, 90 %.
A: One of my main questions is: why did e.g. organic chemists never really believe in quantum chemistry? Streitwieser for instance, who had written that wonderful book on Hückel MO calculations in 1961 (4) does not use the results of older or newer quantum chemical methods in his big organic text book. The Pullmans, on the other hand, tried in the early days of quantum chemistry to really use the semiempirical methods in their books as a way for explaining certain properties, reactions, etc. (8). Why then are the results of quantum chemistry so sparsely incorporated into pure chemistry textbooks?
B: In my view the reason is the following: Chemistry made by organic chemists is an independent domain. Organic chemists have their own form of reasoning which is sufficient for their own business. For them there is no real need and no alternative because they are able to visualize molecules in their mind, in their own fashion. With that they are planning synthetic pathways etc. without using any calculations.
A: What about the use of molecular modeling in chemistry and the pharmaceutical industry?
B: Yes, but you know, commercial products are hardly ever coming from calculations. They don't really use the results. Once they have made their business, they use the theory in order to give some flavor to their publication, if any.
As a consequence of this development quantum chemists are more in touch with physicists, physical chemists, spectroscopists and so on than with true organic chemists. For instance, we are sitting here right now in the department of physics, in the group of astrophysics. Because astronomers are presently concerned with molecules, quantum chemists can help them to identify compounds by computation. Another reason for the connection between quantum chemistry and astrophysics is that it is impossible to make experiments in stars. And we are obliged to -
A: Simulate -
B: Yes, the chemist is not obliged to make simulations. Astrophysicists, on the other hand are detecting molecules very often not yet isolated in labs; they are fugitive, exotic species for which experiments are only in the beginning in our country. The situation is more or less the same elsewhere.
A: What is the present state of the use of quantum chemistry programs?
B: You know the problem is the following: Until a few years ago theoretical chemists were unable to calculate molecules which really interested chemists - there were results only on small molecules. Now the situation is different because we can calculate rather big systems, including crystals and solids. And nowadays we can also incorporate the effect of solvation and so on. This is the reason why the situation has changed; our modeling, in a certain fashion, is taking the experimental condition into account. In some cases it is enough to make a good semiempirical calculation. It is not necessary to use ab-initio with big bases, big CI, and so on. I am working with Italian colleagues on the application of semiempirical methods for photochemistry purposes. This kind of calculation is really working (9). But most of the hard, pure quantum chemists do not at all like this kind of calculation. I myself have no opinion on this issue. If the problem has to be treated by semiempirical methods, I will use them. If not, I use another method - maybe ab-initio, why not. And as you probably know, à la mode in chemical modeling in chemistry is the density functional theory (DFT). For me the DFT is not a true ab-initio method, but rather some sort of semiempirical method; between true semiempirical methods and ab-initio ones. It's success is dependent on the stuff.
A: What about the problem of the basis functions?
B: Oh dear!
A: In your early time you once wrote about the problem of the basis functions. Is this still a problem? From which time on did you see that the basis functions problem no longer existed?
B: In the seventies, a little after the dissemination of computers. Before that we had no choice, we had to use rather small yet physically significant basis functions. Afterwards we had the impression that we can use anything as basis functions. This is the reason why the problem still exists. "Anything" is nothing good, as a matter of fact.
We should be able to treat a problem correctly without any problem of basis functions. Either we use reasonable basis functions or we use numerical techniques. It is a problem of informatics, for computer specialists, - all in all a problem for people who are not primarily chemists. And now I hope that we can, that we should be able in the next century to surpass unpleasant gaussian expansions.
There is some progress in this field in the two directions: either the purely numerical way or to abandon gaussians for Slater's true atomic orbitals.
The reason why all the programs used were written in gaussians was facility. But the situation is moving. For instance the group of Pisa has just revived a program in Slater orbitals, initiated by Scrocco more than 20 years ago. This program is running on my workstation, and it is fast enough to avoid having recourse to poorer expansions, even for non-linear molecules.
I would say in jest: Les fonctions gaussiennes - c'est comme les médicaments: il faut les utiliser pendant qu'ils guérissent, mais quand ils ne marchent plus il faut en changer (10). At this point we could be obliged to change our attitude with respect to quantum chemistry of the "good old days".
A: Somebody at one time used a thousand expansions - -
B: One thousand - - O.K. I know calculations like that. Those of the Karlsruhe group using TURBOMOLE: zeolithes, fullerenes for instance.
A: Once in 1975 you wrote about a possibly existing bond Hamiltonian which may not have been discovered yet. Do you still follow up this idea?
B: I believe it is impossible to define a genuine bond Hamiltonian. We can introduce the concept of bond from our Hamiltonian and adjust it, from quantum mechanics. If we decide to calculate a molecule, we fix the nuclei and having fixed the latter, we presuppose the idea of bond connectivity. All this has no importance if we work correctly. It is not that the concept of bond properties is wrong, but that the logical existence of a bond cannot be strictly deduced from Born-Oppenheimer. The question raised by Woolley (11) was considered - but not solved - by us and others (12).
A: The book by Maksic (13)?
B: Sutcliffe devoted some pages to this (14), and also Claverie and Diner (15).
A: Professor Berthier, some other question. Prof. Del Re mentioned that you were at Löwdin's around 1958, 1959?
B: In 1959.
A: From your papers one sees clearly that from then on you started to use linear algebra. Was Löwdin an expert in the usage of linear algebra? Did Löwdin teach you guys linear algebra? What did you learn from Löwdin?
B: Yes, that's right. In fact, I've learned from Löwdin all the background I've not had the occasion to learn in France. I have a large debt to Löwdin from that time. Löwdin was an extraordinary teacher and he was able to state the mathematical problem clearly, often without giving the answer. Indeed many people have learned many things at that time at Löwdin's: the spin problem, the projection problem, the CI problem. The personality of Löwdin was very strong; his influence began in the fifties, it lasted during the next 20 years.
A: Where did Löwdin learn it from? From Roothaan? By whom was Löwdin influenced?
B: In fact Löwdin did not learn it from Roothaan's paper (16) because all the quantum chemists knew Roothaan's work before he published.
The developments of Roothaan were embedded in the previous publications of Mulliken and his associates of Chicago. All these things were very well known to quantum chemists all over the world who were in connection with Mulliken. And in France we were familiar with Mulliken's work because of the famous Mulliken paper of 1949 (17), written in French.
A: Mulliken published in French!?
B: Yes. But this paper was in fact translated by E. Bauer, professor at the Sorbonne, from an English text. As a matter of fact, I was probably one of the first users of the Mulliken-Roothaan technique in solving an SCF wave equation by the LCAO method.
A: Was it the anthracene paper? Which paper was it?
B: It was for fulvene, an isomer of benzene, in 1953 (19,20). The story starts with Parr, who was a fellow of the Chicago group at that time and the first paper on the Mulliken/Roothaan technique was published not by Roothaan himself, but by Parr and Mulliken. I mean the paper on the p-system of butadiene (14-19) which did appear a little earlier.
So, I had discovered the LCAO SCF technique in the Parr/Mulliken paper (usually referred to Roothaan {author's note}) and I have duplicated the calculations of butadiene onto fulvene - with exactly the same notations. In short, I made the same thing. Actually it could have been difficult to understand what was exactly the content of the Mulliken paper of 1949 because it was a very thick paper containing many things and many footnotes, but most of us knew his works very well through the unclassified research contracts of the Chicago lab with the US Naval Office. At each term, Mulliken put together all the unpublished papers of the Chicago group in a red report which was freely distributed, it was sufficient to write to Mulliken to obtain a copy. This explains the fast diffusion of the Roothaan technique in the States and also in France. The situation in England was somewhat different because the Lennard-Jones group had independently suggested the same treatment. This, by the way, is the reason why some people call these equations the Roothaan-Hall equations.
In the post-war time, French quantum chemistry was more and more connected with English-speaking countries, but it is differently structured.
A: There are the strong schools in France of Daudel's and Pullman's. How did they develop? Do you know of any others?
| Early French schools of theoretical chemists |
| B. Pullman (21) |
R. Daudel |
L. Salem |
J. Barriol (22) |
| First generation: |
G. Berthier
A. Julg
Mrs. J. Serre
A. Veillard
...
... |
S. Bratos
O. Chalvet
R. Lefebvre
Mrs. H. Lefebvre
C. Moser
... |
X. Chapuisat
A. Devaquet
...
...
...
... |
J.L. Rivail
...
...
...
...
... |
| Second generation: |
P. Claverie (23)
J.P. Malrieu
Mrs. M. Suard
... |
M. Allavena
J. Maruani
...
... |
Ph. Hiberty
C. Leforestier
...
... |
A. Cartier
D. Rinaldi
...
... |
B: Let me think. In fact in France Theoretical Chemistry is linked with the names of Pullman and Daudel. In 1945 there was only one group. Then the group separated into two and you have in France theoreticians originating from the Pullman or the Daudel labs. I could give to you their family tree.
In 1964 Salem coming back from the States founded a new group at the University of Paris-Sud (Orsay). Now it is in a center for the popularization of knowledge, still in Orsay.
Another old-established group was in Nancy with Barriol and now Rivail who succeeded me as an editor in THEOCHEM. Nancy has shifted from electric phenomena to solvation effects.
The group of Barriol was founded just before 1945. Barriol was prisoner of war in Germany as a French officer without almost anything to do. So he made with his colleagues, also prisoners of war, a course of quantum mechanics. Barriol wrote a work in quantum mechanics using just the lectures he gave (24). The history of the French quantum chemistry has plenty of windings.
A: Daudel once said in a discussion in 1971:
DAUDEL: - J'aimerais presenter ici quelques récents résultats de l'emploi de la théorie des loges à laquelle Monsieur Berthier fait allusion, en vue de préciser le concept de liaison chimique et les concepts annexes. Monsieur Berthier a signalé que l'analyse de la population électronique à partir d'une base d'orbitales pose des problèmes et qu'en particulier les charges atomiques habituellement définies ne sont pas toujours invariantes par rapport aux transformations unitaires des elements de la base. A ces difficultés s'ajoute le fait qu'au fur et à mesure que l'on utilisera de meilleures fonctions d'onde les bases auront sans doute de moins en moins de sens physique et qu'à la limite, il n'y aura plus de base du tout. Imaginez la géneralisation de la méthode de James et Coolidge ou mieux encore une méthode totalement numérique (Un des rêves de notre collègue Berthier !).
Même dans ce cas la notion de loge permet de sauvegarder le concept de liaison.
...... (25; translation under (26)).
B: That was not exactly my thinking. I did not think in terms like those of James and Coolidge but of a method without any basis and that is definitely not the case of James and Coolidge. They used elliptical functions. Here Daudel was - - say, a little speculative - -
A: Sandorfy was using a simple s
-method, he called it "C"-approximation, in 1955 (27,28). His last paper using that method in 1969 could only be published in a Hungarian journal (29). How do you view Sandorfy's efforts in their time?
B: I like Sandorfy's idea for saturated hydrocarbons. The difficulty was that he used some sort of local bond method not easily transferable to a computer of that time. If Sandorfy would have offered a black-box program, he might have been more fortunate. Nobody had the courage of going back to his approach, as soon as the automatic codes became available in quantum chemistry.
A: Feyerabend could be applied here. And I mention only Åsbrink with his HAM method (30). Many ideas popped up in order to fade away fairly quickly.
B: The problem is that in quantum chemistry all the original, non conventional things vanish quickly because of the influence of standard programs. If you don't use them, you take the risk of being rejected because they are weighing heavily in quantum chemistry. Now we have the choice between several codes: GAUSSIAN HONDO, CADPAC, MOLPRO. There are others, of course, but you know, even TURBOMOL has not broken the software market for workstations in spite of its merits. Money is money. We say in French: La mauvaise monnaie chasse la bonne (31).
I don't mean that gaussians are poor; they are so-and-so. If you know them very well, you can obtain good chemical information, but this may be quite a business for organicists.
A: What then will organic chemists use in the future. Still semiempirical methods?
B: Not easy to say. It is still possible to use semiempirical methods for education, for speaking with experimentalists, and so on. But the problem is to publish and where. On the other hand, an ab-initio paper, even if not very original, is hardly rejected. Generally speaking, however, it's difficult to find new applications of ab-initio methods which are not time-consuming.
That is the reason why we have been interested in special semiempirical methods relating conformation and spectra (9). This problem is not trivial at the ab-initio level, due to the lack of efficient ab-initio programs for excited states. On the other hand, semiempirical methods often fail in accounting simultaneously for both conformation and spectroscopic features of large molecules. Our colleagues in Italy have succeeded to adapt the zero differential overlap approximation of old molecular orbital methods for these purposes. There are people, however, who refuse such an approach.
A: So the community of scientists deletes concepts when they are still useful - -
B:Not always. Although in this lab, for most of the problems concerning interstellar space, e.g. reactivity problems (32) we use ab-initio potential energy surfaces, we have occasionally made semiempirical calculations (33). We tried to see whether some lines of UV spectra could be assigned to definite big aromatics, an impossible task by ab-initio. Similar questions were posed to us by people interested in aromatic carbenes, with divalent carbon atoms. The method used is the same as fro photochemistry problems I have already mentioned.
All in all, I have no objection in principle against semiempirical methods. The choice is dependent on the problem and the people. Young people sometimes prefer the ab-initio ones - why not! I am prepared to accept EHT calculations as well as ab-initio treatments with big bases and big CI - - - the difficulty remains with the editors of the journals.
A: PPP nowadays is not much used. Was it important in it's time?
B:It was important. After all, it was the first method which was able to introduce correlation and so on in large molecules. It is out-of date for that, but we have still used it in this lab always for questions of interstellar lines (33). If the properties to be studied, here electron spectra, are clearly connected to the p-system we are justified to ignore the s-system and to treat the molecule as a p-system. There are methods looking like to PPP very much, for instance for clusters (34). The PPP method has an important historic place, and those which came next have profited from the developments made in their context. Even recent gaussian codes contain parts which are borrowed from old PPP ones. However, attempts of justifying the PPP method from ab-initio considerations seem to be a little forced (35).
A: Professor Berthier, I thank you very much for this interesting and inspiring interview!!
References and Notes
Click to enlarge |
|
Amazing: G. Berthier was 86 years old when he published this book
(US$ 62.00 in 2009) |
(1) G.Berthier and B. Pullmann,
Sur la conjugaison simultanée de plusieurs groupments extracycliques avec un noyau aromatique.
Bull. Soc. Chim. (France), 1948, M 554-558.
(2) The BASF was contacted in 1997 by this website in this respect. The BASF presently sustains a history/archive division. Neither in their archives nor in the past records of the BASF-personnel division a Dr. Otto Schmidt could be located. According to their further communication several Otto Schmidt existed during that period in the telephone book of Ludwigshafen. On the other hand, in the years 1933-1945 the German compound of companies existed under the name of 'IG-Farbenindustrie' existed - a conglomerate of diverse chemical companies put together, including BASF, Höchst and others. In his publications 1938-1939 Schmidt had given as his address of publication as 'IG-Farbenindustrie'. Yet in 1941 Otto Schmidt gave in a further article in Naturwissenschaften 29, 146-150 (1941) an address in Heidelberg. After the end of the second world war in 1945 the IG-Farbenindustrie was split up again - some information may have disappeared in that time. - - Other tracks e.g. with the help of the archives of the town of Heidelberg are presently pursued in order to find biographical dates about Dr. Otto Schmidt.
(3) Alberte Pullman, Bernard Pullman,
Cancérisation par les substances chimiques et structure moléculaire. Masson, Paris, 1955.
(4) A. Streitwieser,
Molecular Orbital Theory for Organic Chemists.
John Wiley & Sons, New York, 1961.
(5) W. Kutzelnigg, Die Lösung des quantenmechanischen Zwei-Elektronenproblems durch unmittelbare Bestimmung der natürlichen Einelektronenfunktionen. I. Theorie.
Theoret. Chim. Acta, 1, 327-342 (1963).
(6) W. Kutzelnigg, Résolution du problème à deux électrons en mécanique quantique par détermination directe des orbitales naturelles. II: Application aux états fondamentaux de l'hélium et des ions isoélectroniques.
Theoret. Chim. Acta, 1, 343-352 (1963).
(7)
The Story of the PEERLESS Electromechanical Calculator (36):
The company of Mathias Bäuerle of St. Georgen, situated in the heart of the Black Forest in southwestern Germany, was founded in 1863 by Mathias Bäuerle (1839-1916). In this area of the Black Forest clocks were and still are being produced in large quantities since the mid-seventeenth century. Later, around 1750, the historic sources (36) mention a production of 75000 clocks for the township of St. Georgen alone. The clock company of Mathias Bäuerle expanded over the years, its products were improved and at the world exposition in Paris in 1900 Bäuerle obtained a gold medal for one of his clocks.
From the expertise of producing and handling sprocket wheels and the like the innovative idea formed that a calculator could be produced. This part of the country had, since ages, always supported new ideas in the field of mechanics. However, contrary to the trend in technical innovations, the financing was far less dynamic: it had always been difficult for the companies in this area to obtain financing due to the somewhat conservative thinking of the surrounding banks.
Nevertheless, in 1903 Bäuerle had enough capital to dare the production of a hand-driven mechanical calculator, under the name of PEERLESS, which already in 1904 won a prize at St. Louis, USA. By 1914 the great success of its calculator business led the company to produce, under the technical guidance of a certain Andreas Abele (1904-1957), a keyed calculator in modern appearance, traded in Germany as Badenia, thereby referring to the county of its origin. For export the name PEERLESS was retained, and it was registered as an international trademark in 1922.
According to the municipal archive (36) the invading French troops in 1945 confiscated 25-40% of the machinery of all companies of St. Georgen, including those of the Bäuerle company. In this historic process the calculating machines were later redistributed in France, among others to the CNRS.
The Bäuerle company continued to produce electromechanical calculators with newer models in 1957 and 1963. But by then the electronic computers had gained wide success. As a consequence, the Bäuerle line of electromechanical calculators was discontinued soon after.
Anonym.,
100 Jahre Stadterhebung St. Georgen im Schwarzwald, Festschrift
1891-1991. Hrsg. und Verlag: Stadt St. Georgen im Schwarzwald, 1991.
(8) Bernard Pullman and Alberte Pullman,
Quantum Biochemistry.
Interscience Publishers, New York, 1963.
(9) A. Despres, V. Lejeune, E. Migirdicyan, A. Adamasu, M.S. Platz, G. Berthier, O. Parisel, J.P. Flament, I. Baraldi and F. Momicchioli,
Study of the Electronic Structure and Spectra of Diphenylcarbene Conformers in Their Ground State and Lower Excited States.
J. Phys. Chem. 97, 13358-13367 (1993).
(10) Gaussians are like medicine - you have to use them as long as they are healing, but once they don't work any more, you must change them.
(11) R.G. Woolley,
J. Amer. Chem. Soc. 100, 1073 (1978) and the article in ref. (8).
(12) G. Berthier, M. Defranceschi, F. Momicchioli, FCTL (Folia Chimica Theoretica Latina) 20, 1-20 (1992).
(13) Z.B. Maksic (ed.),
Atomic Hypothesis and the concept of Molecular Structure.
Springer-Verlag, Berlin etc., 1990. Vol I. Altogether 3 volumes.
(14) B.T. Sutcliffe,
The concept of Molecular Structure.
Ref. (8), pages 1-28.
(15) P. Claverie and S. Diner,
The concept of molecular structure in quantum theory. Interpretation problems.
Israel J. Chem. 19, xxx-xxx, 1980.
(16) C.C.J. Roothaan,
New Developments in the Molecular Orbital Theory.
Revs. Mod. Phys. 23, 69-89 (1951).
(17) R.S. Mulliken,
Quelques aspects de la théorie des orbitales moléculaires.
J. Chim. Phys. 46, 497-713 (1949).
(18) G. Berthier,
LCAO Self-Consistent Field Calculations of the p-Electrons Energy Levels and Electronic Structure if Fulvene.
J. Chem. Phys. 21, 953-954 (1953).
(19) G. Berthier,
Structure électronique du fulvène: étude par la méthode du champ moléculaire self-consistent.
J. Chim. Phys. 50, 344-351 (1953).
(20) R.G. Parr, and R.S. Mulliken,
LCAO Self-Consistent Field Calculations of the p-Electrons Energy Levels of cis- and trans-1,3-Butadiene.
J. Chem. Phys. 18, 1338-1346 (1950).
(21) Deceased in 1996.
(22) Deceased in 1989.
(23) Died of cancer, aged 43, in 1988.
(24) J. Barriol,
Elements of Quantum Mechanics with Chemical Applications. Barnes and Noble, New York, 1971.
The original:
Éléments de Méchanique Quantique.
Masson, Paris, 1966.
(25) Colloques Internationaux du CNRS, N° 195, p. 70, 1971.
(26)
DAUDEL: I would like to present here some recent results on the use of the theory of loges, which Mr. Berthier refers to, in order to specify the concept of the chemical bond and similar concepts. Mr. Berthier has communicated that the electronic population analysis, starting from an orbital basis, poses problems and that in particular the atomic charges, as usually defined, are not always invariant with respect to unitary transformations of the basis. To these difficulties adds the fact that as one will want to use the best wave functions possible, the basis will undoubtedly have less and less physical significance up to the limit when there will be no basis whatsoever. Just imagine the generalization of the method of James and Coolidge or better even a totally numeric method (one of the dreams of our colleague Berthier!).
But even in this case the notion of a lodge will allow us to save the concept of 'bonds ...
(27) C. Sandorfy and R. Daudel,
Comptes Rend. 238, 93 (1954).
(28) C. Sandorfy,
LCAO MO Calculations on Saturated Hydrocarbons and their Substituted Derivatives.
Can. J. Chem. 33, 1337-1351 (1955).
(29) C. Sandorfy,
On the Structure and Spectra of s-Electron Systems.
Acta Phys. Hung. 27, 151-160 (1969).
(30) E. Lindholm and L. Åsbrink,
Molecular Orbitals and their Energies, Studied by the Semiempirical HAM Method.
Springer-Verlag, Berlin, 1985.
Lecture Notes in Chemistry - Series. Nr. 38.
(31) Bad money pushes the good one away.
(32) F. Pauzat, Y. Ellinger, G. Berthier, M. Gérin and Y. Viala, Theoretical study of a basic process in interstellar nitrogen chemistry: reaction of N with OH.
Chem. Phys. 174, 71-79 (1993).
(33) O. Parisel, G. Berthier and Y. Ellinger,
New clues for ionized polycyclic aromatic hydrocarbons as possible carriers of diffuse interstellar bands.
Astron. Astrophys. 266, L1-L4 (1992).
(34) A. Julg, G. Del Re and V. Barone,
A theoretical study of distortions induced by finite size in regular clusters.
Phil. Mag. 35, 51-53 (1977).
(35) e.g. the work of Karl Freed, Chicago.
|
The original Interview, mp3 - not public
1of3
2of3
3of3
|
Last updated : Jan. 31, 2010 - 15:47 CET
|
|
