close

Вход

Забыли?

вход по аккаунту

?

H. C. Longuet-Higgins and J. A. Pople Theoretical Chemistry

код для вставкиСкачать
Obituary
any calculations.[3, 4] LonLonguet-Higgins had a
guet-Higgins
was
only
31
sharp
mind, but was also
H. C. Longuet-Higgins and
when he took up a posisomewhat sarcastic and
J. A. Pople: Theoretical
tion in Cambridge in
tended to make jokes at
Chemistry**
1954 as the successor of
the expense of others.
Lennard-Jones,
who
had
L?wdin was apparently
Werner Kutzelnigg*
left Cambridge for the
quite angry about Longuet-Higgins's parody of
Within just a few weeks two remarkable recently founded reform
the
“green
technical
theoretical chemists have left us forever. university in Keele.
Longuet-Higgins's
reports” from Uppsala.
Although they had some common backS. F. Boys in particular sufground, they could hardly have been studies on electron-defifered from the way his
more different. Each in his own way cient boron compounds
superior, Longuet-Higgins,
made a noticeable impact on science. were very important for
H. C. Longuet-Higgins
treated him. Longuet-HigBoth were English, from the famous chemistry. Although one
gins did not care for the
schools of Lennard-Jones and Coulson. usually associates the
Longuet-Higgins and Pople collabo- theory of this class of compounds—cer- numerical work of Boys. Perhaps Lonrated on a few publications together, tainly with good reason—with W. N. guet-Higgins was aware that Boys's profor example, on the Renner–Teller Lipscomb, the decisive publication on posal to perform quantum-chemical
effect.[1] Both initiated their careers the electronic structure of B2H6 was computations by using a Gaussian
with studies on the H+ckel molecular written by Longuet-Higgins while he basis[11] would have at least as much
orbital theory (HMO)
was still a graduate stu- impact as his own most important conof p-electron systems,
dent.[5] Mainly by using tributions. Perhaps he sensed that
but their later roles
group-theoretical argu- Boys's work would turn out to be the
were very different. Common starting
ments, he both ex- beginning of a new and powerful
Whereas Longuet-Hig- point: H$ckel MO
plained the stability of branch of theoretical chemistry in
gins favored the kind
CaB6 and predicted the which he, Longuet-Higgins, would neitheory
of theory that could be
ther have, nor wish to have, a share.
existence of B12H122
formulated on the back
before its experimental This attitude may help explain Lonof an envelope, Pople
guet-Higgins's decision to abandon his
preparation.[4, 6]
became a major protagThere is another successful career as a theoretical chemonist of computational chemistry. The context in which Longuet-Higgins was ist in 1967, to leave his chair in Camspectacular success of the latter field given less credit for his contributions bridge after only 13 years, and to
resulted not only from the intelligence than he was due. He was the first to devote himself in Edinburgh to new
and ingenuity of quantum chemists, but apply the concept of correlation dia- challenges in the realm of neuroinforalso from the extraordinarily rapid grams to explain electrocyclic reactions matics and theoretical psychology, for
development of computer technology. in terms of what are
example, the percepPople bet on the winning horse.
tion of music.[12] I was
now known as the
Woodward-Hoffmann
unable to find out
rules, whereby he went
Theory on the back of whether the standing
Hugh Christopher Longuetfar beyond the someLonguet-Higgins achHiggins (1923–2004)
what vague concept of an envelope
ieved in the neurosciences was comparable to
Hugh Christopher Longuet-Higgins the conservation of orbithe prestige he had forstudied with C. A. Coulson in Oxford. tal symmetry.[7]
merly gained in chemisTheir very nice work on HMO theory
Two further discovtry. Six years ago I had
made this theory esthetically appealing eries of Longuet-Higand mathematically elegant,[2–4] for gins will certainly guarantee his lasting the chance to ask him (at an event in
example, through the formulation of reputation. One is the first example of commemoration of S. F. Boys) whether
HMO-based perturbation theory.[2] I what is now known as a Berry phase, he regretted having left theoretical
particularly like one article, in which, namely, the change of sign of the wave chemistry. His only answer was that he
for a special class of compounds, the function on a closed loop around a con- never regretted anything. Longuet-HigMO coefficients are constructed without ical intersection.[8] Another seminal con- gins's legacy as a theoretical chemist
tribution was his theory of the symmetry might be characterized by his Faraday
groups of nonrigid molecules, into which Discussion lecture in Brighton,[13]
[*] Prof. Dr. W. Kutzelnigg
feasible permutations of equivalent during which he proposed to divide
Fakultt fr Chemie der
atoms enter as symmetry elements.[9] It chemists into three classes: experimenRuhr-Universitt Bochum
is impossible to understand the rota- talists, computationalists, and theorists.
(Germany)
[**] The author thanks Paul von Ragu&
tion–vibration spectra of fluxional mole- This sharp distinction between theory
Schleyer for his help in preparing the cules, especially in the millimeter and computational chemistry did not
gain many adherents.
region, without this theory.[10]
English version of this manuscript.
2740
* 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200460423
Angew. Chem. Int. Ed. 2004, 43, 2740 – 2743
Angewandte
Chemie
systems. It is a “one-electron” theory,
as the electron–electron interactions
John A. Pople studied with Lennard- are not treated explicitly. Pople invented
Jones in Cambridge. Pople was a mathe- a straightforward improvement of HMO
matician who felt (accurately, as it theory, which, akin to an SCF theory,
turned
out) treats electron interaction (SCF: selfthat he could consistent field).[15] At about the same
make
more time, Pariser and Parr[16] proposed an
of a mark HMO theory in which a semiempirical
in chemistry. correction of the electron interaction
Although he was an important part. The two treatstarted
with ments merged into the famous PPP
studies in stat- (Pariser–Parr–Pople)
approximation,
istical mechan- which dominated the theory of UV/Vis
ics,
Pople's spectroscopy of conjugated p-electron
main achieve- systems for more than two decades.
ments were in
For a remarkably long time only the
three
fields: so-called “mobile” p electrons in conjuJ. A. Pople
the theory of gated systems could be described semithe magnetic empirically.[4] One had to wait for semiproperties of molecules, semiempirical empirical schemes for all electrons
quantum chemistry, and ab initio quan- (including the largely localized s electum chemistry. His theoretical contribu- trons) until R. Hoffmann[17] proposed
tions to NMR spectroscopy, based the extended H+ckel theory (EHT; a
mainly on the HMO method, were sum- one-electron theory, though with inclumarized in his monograph with sion of the overlap between the atomic
Schneider and Bernstein.[14] For over a orbitals)
and
Pople
suggested
CNDO[18, 19] as a generdecade, this book had
considerable impact on
alization of the PPP
the developing field of
method (that is, with
From semiempirical
high-resolution NMR
electron
interaction,
methods …
spectroscopy.
but without overlap).
It is reported that
The notation CNDO
Pople, then head of the
(complete neglect of
basic-physics division
differential
overlap)
of the National Physical Laboratory at has created some confusion, because
Teddington, was appointed to a royal the neglect of overlap is hardly legiticommission charged with finding out mate, especially for s electrons. Howwhy so many British scientists were ever, it can be justified indirectly if one
moving to the USA. Apparently he did regards orthogonalized atomic orbitals
find out, because shortly afterwards, in (AOs), rather than the original AOs, to
1964, he accepted a position in Pitts- be the basis functions.[4] The EHT
burgh, PA, where he pursued most of method has probably been more popuhis scientific career at Carnegie-Mellon lar in inorganic chemistry, whereas
University. He spent his last years at CNDO has found most applications in
Northwestern University, near Chicago. organic chemistry. The literature is rich
Until the late 1950s, molecules of in modifications of CNDO, mainly
chemical interest with more than a few from the school of M. J. S. Dewar. The
atoms could only be treated by drasti- most sophisticated of these, by W.
cally simplified methods. These methods Thiel,[20] is still used successfully, in parcould, in principle, be derived from ticular for very large conjugated sysquantum mechanics.[4] However, such tems. Methods such as PPP and CNDO
crude approximations were required were usually regarded—often with a
undertone—as
extremely
that useful results were only possible if polemic
certain quantities were treated as empir- crude simplifications of the exact
ically adjustable parameters. Originally theory. However, one can regard them
the HMO method was the most impor- in a more positive light as refinements
tant semiempirical scheme, but it was of the Hubbard model, which is still
limited to the description of p-electron extremely popular in solid-state theory.
John A. Pople (1925–2004)
Angew. Chem. Int. Ed. 2004, 43, 2740 – 2743
www.angewandte.org
The essential aspect of this model and
its refinements is the competition
between (overlap-dependent) shortrange forces and long-range electrostatic (Coulomb-like) interactions in a
molecule or a solid.
When Pople decided to abandon
semiempirical theory[21] (for which
Dewar, his companion in its development, apparently never forgave him)
and to join the ab initio branch, the
latter field had already undergone significant development and was quite successful for smaller molecules. Therefore,
Pople's role was not to initiate an
entirely new methodology. Although
he was first regarded as an outsider by
the inner circle of ab initio theorists, or
even as someone who wished to reinvent
the wheel, to organic chemists interested
in theory he soon became the uncontested authority on ab initio theory.
This popularity was due not so much to
Pople's praiseworthy improvements to
methods and programs (e.g. in the
form of an accelerated integral program), as to his concept of computational chemistry, which he had developed during his experiences with semiempirical theory. A “traditional” ab
initio theorist would study one system
at a time (such as the potential hypersurface of a chemical reaction) and try to
apply the best method available for a
particular problem. The same theorist's
investigation of the next system would
likely involve a different theoretical
level, again chosen as the best available
for the new purpose. Hence, it was difficult for a non-expert to compare different publications, even by the same
author, to say nothing of the results of
different authors. Pople, on the other
hand, published computations based on
a well-defined method, which was used
uniformly for 100 or more molecules,
rather than just a single species. He
also insisted on making careful and comprehensive comparisons with the corresponding experimental data. His notion
of a “model chemistry”,[22] defined by a
chosen approximation (such as SCF)
and a particular basis set (such as STO3G or 4-31G*), was central to this concept. Each model chemistry was thoroughly documented, especially with
regard to its reliability and the limits of
its validity, based on comparisons with
experiment. This strategy was like the
* 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2741
Obituary
application of standardized empirical for chemical purposes. The situation
dramatically
when
A.
methods, such as CNDO and INDO, changed
which were parameterized to reproduce Becke[28] and J. Perdew[29] proposed soexperimental data. The concept of a called second-generation density funcmodel ab initio chemistry also implied tionals, which included “gradient correcstandardization and, consequently, a tions” and made chemical accuracy posloss of flexibility, but it had the advan- sible. Thereafter Pople, ever a pragmatic
tage of portability. Authors around the scientist, did not hesitate to implement
world could not only perform computa- density functionals, such as B3LYP, in
tions fully consistent with one another his program packages.[30]
on different machines, but they could
Without Pople's contributions, the
also build upon the published computa- unexpected triumph of numerical ab
tions of others. Major support for this initio quantum chemistry, which has
concept came from the decision by made it an irreplaceable tool for most
Pople and his co-workers to develop a chemists, would very likely not have
series of user-friendly programs (key- occurred. It was extremely helpful that
word
GAUSSIAN)
Pople found competent
and to make them genpartners, who helped to
erally available (even
bridge the gap to “real
the source code) at … to ab initio theory
chemistry”.
P. von R.
affordable prices. The and GAUSSIAN …
Schleyer was an imporauthors of other protant convert.[31] For
grams for quantum
example,
he
used
chemistry were much
theory to reveal entire
more reticent in these
domains of chemistry
respects. Pople's goal was to develop that were unknown from experiment.
and distribute a tool for general use by
It was entirely deserved, and came as
chemists. Consequently, GAUSSIAN no surprise, that John Pople was
became the market leader, and gained awarded the Nobel Prize in Chemistry
for Pople authority on the ab initio in 1998 for his life's work.[32] The prize
nomenclature.
was shared by Walter Kohn in recogniWhen Pople decided to perform tion of his original proposals of density
computations including electron corre- functionals,[27] which had been devellation (that is, the effects ignored in the oped rapidly[28, 29] and implemented into
SCF approximation),[23]
quantum-chemical program packages.[30] Howhe might have had
recourse to existing
ever, its “marriage”
methods
that
had … to the Nobel Prize in with density functional
already been developed Chemistry
theory was not necesand were being applied
sary for ab initio quansuccessfully in Gertum chemistry to achmany, such as IEPA
ieve its big break(independent electron
through. Density funcpair approximation)[24] or CEPA (cou- tional formalism is just one stone
pled electron pair approximation).[25, 26] (albeit a very bright one) in the rich
However, he preferred to establish his mosaic of ab initio quantum chemistry.
own (not very different) hierarchy of Other stones include the evaluation of
approximations for electron correlation, energy gradients and of relativistic corwhich then largely eclipsed IEPA and rections.
Together with many others—who,
CEPA. Much later, in private conversations, he admitted that it might not willingly or not, remained in the backhave been fair to completely ignore the ground—Pople contributed greatly to
the fulfillment of the goal, which has
debt to these earlier approaches.
Like most ab initio theorists, Pople now even been surpassed: that theoretiinitially regarded density functional cal chemistry should attain uncontested
theory with skepticism, especially importance in chemistry. Although I
approaches of the LDA (local density am sincerely pleased about this achieveapproximation) type,[27] for the simple ment, I wonder whether my own
reason that these were too inaccurate dream[33] (shared by other theoretical
2742
* 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
chemists) was really only the reduction
of theory to a black box, which can be
used routinely like a scientific instrument, even by someone who has no
knowledge of or interest in what is happening inside. Pople himself shared
these misgivings. Chemistry is much
more complex than most of its practitioners appreciate.
Sir John, as he was entitled to call
himself in the last year of his life, was a
pleasant person and always enjoyable
to talk to. His arguments were clear
and lucid. His wit was pithy. Unlike
many less famous colleagues, he never
took himself to be very important. I
liked him.
[1] J. A. Pople, H. C. Longuet-Higgins, Mol.
Phys. 1958, 1, 372.
[2] a) C. A. Coulson, H. C. Longuet-Higgins, Proc. R. Soc. London Ser. A 1947,
191, 39; b) C. A. Coulson, H. C. Longuet-Higgins, Proc. R. Soc. London Ser.
A 1947, 193, 456.
[3] H. C. Longuet-Higgins, J. Chem. Phys.
1950, 18, 275.
[4] W. Kutzelnigg, Einfhrung in die Theoretische Chemie, Vol. 2., Verlag Chemie,
Weinheim, 1978/1994, reprinted 2002.
[5] H. C. Longuet-Higgins, R. P. Bell, J.
Chem. Soc. 1943, 250.
[6] H. C. Longuet-Higgins, M. de V. Roberts, Proc. R. Soc. London Ser. A 1955,
230, 110.
[7] H. C. Longuet-Higgins, E. W. Abrahamson, J. Am. Chem. Soc. 1965, 87, 2646.
[8] G. Herzberg, H. C. Longuet-Higgins,
Discuss. Faraday Soc. 1963, 35, 77.
[9] H. C. Longuet-Higgins, Mol. Phys. 1963,
6, 445.
[10] P. R. Bunker, Molecular Symmetry and
Spectroscopy, Academic Press, New
York, 1979.
[11] S. F. Boys, Proc. R. Soc. London Ser. A
1950, 200, 542.
[12] a) H. C. Longuet-Higgins, Nature 1976,
263, 646; b) H. C. Longuet-Higgins,
Mental Processes: Studies in Cognitive
Science, MIT Press, Cambridge MA,
1987.
[13] H. C. Longuet-Higgins, Faraday Discuss.
Chem. Soc. 1977, 62, 347.
[14] J. A. Pople, W. G. Schneider, H. J. Bernstein, High Resolution Nuclear Magnetic
Resonance, McGraw-Hill, New York,
1959.
[15] J. A. Pople, Trans. Faraday Soc. 1953, 49,
1375.
[16] R. G. Pariser, R. G. Parr, J. Chem. Phys.
1953, 21, 466; b) R. G. Pariser, R. G.
Parr, J. Chem. Phys. 1953, 21, 767.
[17] R. Hoffmann, J. Chem. Phys. 1963, 39,
1397.
Angew. Chem. Int. Ed. 2004, 43, 2740 – 2743
Angewandte
Chemie
[18] J. A. Pople, D. P. Santry, G. A. Segal, J.
Chem. Phys. 1965, 43, 129.
[19] J. A. Pople, D. L. Beveridge, Approximate
Molecular Orbital
Theory,
McGraw-Hill, New York, 1970
[20] W. Thiel, Adv. Chem. Phys. 1996, 93,
703.
[21] W. J. Hehre, R. F. Stewart, J. A. Pople, J.
Chem. Phys. 1969, 51, 2657.
[22] “Theoretical Models in Chemistry”:
J. A. Pople in Proceedings of the
summer research conference on theoretical chemistry, structure and reactivity
Angew. Chem. Int. Ed. 2004, 43, 2740 – 2743
[23]
[24]
[25]
[26]
[27]
(Ed.: D. W. Smith), Wiley, New York,
1973.
J. S. Binkley, J. A. Pople, Int. J. Quantum
Chem. 1975, 9, 229.
M. Jungen, R. Ahlrichs, Theor. Chim.
Acta 1970, 17, 339.
W. Meyer, Chem. Phys. 1973, 58, 1017.
W. Kutzelnigg, Modern Theoretical
Chemistry, Vol. 3 (Ed.: H. F. Schaefer
III), Plenum, New York, 1975.
W. Kohn, L. S. Sham, Phys. Rev. A 1965,
140, 133; b) J. C. Slater, Phys. Rev. 1951,
81, 385.
www.angewandte.org
[28] A. D. Becke, Phys. Rev. A 1988, 38, 3098.
[29] J. P. Perdew, W. Yue, Phys. Rev. B 1986,
33, 8800.
[30] B. G. Johnson, P. M. W. Gill, J. A. Pople,
J. Chem. Phys. 1989, 98, 5612.
[31] W. J. Hehre, L. Radom, P. v. R. Schleyer,
J. A. Pople, Ab-initio Molecular Orbital
Theory, Wiley, New York, 1986.
[32] J. A. Pople, Angew. Chem. 1999, 111,
2014; Angew. Chem. Int. Ed. 1999, 38,
1894.
[33] W. Kutzelnigg, Theor. Chem. Acc. 2000,
103, 182.
* 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2743
Документ
Категория
Без категории
Просмотров
2
Размер файла
271 Кб
Теги
chemistry, theoretical, pople, longuet, higgins
1/--страниц
Пожаловаться на содержимое документа