K. Fukui
11
The “orbital” concept, which was established and developed by many
scientists, such as Pauling, Slater, Mulliken, Roothaan, Löwdin, Hückel, Parr
and so on, had till then been employed to construct the wave function of a
molecule, through which molecular properties were usually interpreted.
7
It
seemed that the electron distribution in an orbital was directly connected to
chemical observations and this fact was certainly felt to be interesting by many
chemists.
But the results of such a rather “extravagant” attempt was by no means
smoothly accepted by the general public of chemists. That paper received a
number of controversial comments. This was in a sense understandable, be-
cause, for lack of my experiential ability, the theoretical foundation for this
conspicuous result was obscure or rather improperly given. However, it was
fortunate for me that the paper on the charge-transfer complex of Mulliken
was published in the same year as ours.
The model of Mulliken et al. for protonated benzene was a good help.
9
Our
work in collaboration with Yonezawa, Nagata and Kato provided a simple and
pointed picture of theoretical interpretation of reactions,” as well as the
“overlap and orientation” principle proposed by Mulliken with regard to the
orientation in molecular complexes.
11
Subsequent to the electrophilic substitu-
tion, the nucleophilic substitution was discussed and it was found that in
this case the lowest energy vacant orbital played this particular part.
12
In
reactions with radicals, both of the two orbitals mentioned above, became
the particular orbitals.
There was no essential reason to limit these particular orbitals to
π orbitals,
so that this method was properly applied not only to unsaturated compounds
but also to saturated compounds. The applicability to saturated compounds
was a substantial advantage in comparison with many theories of reactivity
which were then available only for
π electron compounds. The method dis-
played its particular usefulness in the hydrogen abstraction by radicals from
paraffinic hydrocarbons, the S
N
2 and E2 reactions in halogenated hydrocar-
bons, the nucleophilic abstraction of
α−hydrogen of olefins, and so forth
-13
These two particular orbitals, which act as the essential part in a wide range
of chemical reactions of various compounds, saturated or unsaturated, were
referred to under the general term of “frontier orbitals”, and abbreviated
frequently by HOMO (highest occupied molecular orbital) and LUMO (low-
est unoccupied molecular orbital).
In this way, the validity of the theory became gradually clearer. The vein of
ore discovered by chance was found to be hopefully more extensive than
expected. But it was attributed to the role of the symmetry of particular orbitals
pointed out in 1964 with regard to Diels-Alder reactions
14
that the utility of
our studies was further broadened. It was remarked that as is seen in Fig. 2, the
symmetries of HOMO and LUMO of dienes and those of LUMO and HOMO
of dienophiles, respectively, were found to be in a situation extremely favoura-
ble for a concerted cyclic interaction between them.
This signified the following important aspects: First, it pointed out a possible
correlation between the orbital symmetry and the rule determining the sub-
Chemistry 1981
stantial occurrence or non-occurrence of a chemical reaction, which may be
called the “selection rule”, in common with the selection rule in molecular
spectroscopy. Second, it provided a clue to discuss the question concerning
what was the “concertedness” in a reaction which forms a cycle of electrons in
conjugation along the way.
In 1965 Woodward and Hoffmann proposed the stereoselection rules which
are established today as the “Woodward-Hoffmann” rules.
15,16
An experi-
mental result developed in Havinga’s important paper” was extended im-
mensely. It is only after the remarkable appearance of the brilliant work by
Woodward and Hoffmann that I have become fully aware that not only the
density distribution but also the nodal property of the particular orbitals have
significance in such a wide variety of chemical reactions. In fact, we studied
previously the noted (4n+2) rule proposed by Hückel
18
and noticed that the
sign of the bond order in the highest energy electron orbital of an open-chain
conjugation should be closely related to the stabilization of the corresponding
conjugated rings.
19
We did not imagine, however, on that occasion that the
discussion might be extended to the so-called Mobius-type ring-closure!
20
K. Fukui
13
By considering the HOMO-LUMO interactions between the fragments of a
conjugated chain divided into parts,
21
the frontier orbital theory can yield
selection rules which are absolutely equivalent to those obtained from the
principle called “the conservation of orbital symmetry” by Woodward and
Hoffmann. One point that I may stress here is, as was pointed out by Fujimoto,
Inagaki and myself,
22
that the electron delocalization between the particular
orbitals interprets definitely in terms of orbital symmetries the formation and
breaking of chemical bonds which, I believe, should be a key for perceiving
chemical reaction processes.
In the cycloaddition of butadiene and ethylene shown in Fig. 2, both the
interaction between the HOMO of diene and the LUMO of dienophile and
that between the LUMO of diene and the HOMO of dienophile stabilize the
interacting system. If one is interested in the local property of interaction,
however, one may recognize the clear distinction between the roles of the two
types of orbital interactions. The HOMO of ethylene and the LUMO of
butadiene are both symmetric with regards to the symmetry plane retained
throughout the course of cycloaddition. This signifies that each of the carbon
atoms of ethylene are bound to both of the terminal carbons of butadiene. The
chemical bonding between the diene and dienophile thus generated may be
something like the one in a loosely bound complex, e.g., protonation to an
olelinic double bond. On the contrary, the HOMO of butadiene and the
LUMO of ethylene are antisymmetric. The interaction between these orbitals
leads, therefore, to two separated chemical bonds, each of which combines a
carbon atom of ethylene and a terminal carbon atom of butadiene. Needless to
say it is the interaction between the HOMO of diene and the LUMO of
dienophile that is of importance for the occurrence of concerted cycloaddi-
tion.
22
In this way, it turned out in the course of time that the electron delocaliza-
tion between HOMO and LUMO generally became the principal factor deter-
mining the easiness of a chemical reaction and the stereoselective path, irre-
spective of intra- and intermolecular processes, as illustrated in Fig. 3. Besides
our own school, a number of other chemists made contributions. I want to refer
to several names which are worthy of special mention.
First of all, the general perturbation theory of the HOMO-LUMO interac-
tion between two molecules was built up by Salem.
23-25
One of Salem’s
papers
25
was in line with the important theory of Bader,
26
which specified the
mode of decomposition of a molecule or a transition complex by means of the
symmetry of the normal vibration. Furthermore Pearson
27
investigated the
relation between the symmetry of reaction coordinates in general and that of
HOMO and LUMO.
The discussion so far may seem to be an overestimation of these selected
orbitals, HOMO and LUMO. This point was ingeniously modified by Klop-
m a n .
28
He carefully took into account the factors to be considered in the
perturbation theory of reacting systems and classified reactions into two cases:
the one was “frontier-controlled” case in which the reaction was controlled by
the particular orbital interaction, and the other was the “charge-controlled”