In the Classroom

The Concept of Resonance

A recent article in this Journal advocates substituting the Resonance in Quantum Mechanics
word “delocalization” for “resonance” (1). This recommen-
dation should not be accepted because modern electronic Probability
structure theory shows that “delocalization” and “resonance” If one wishes to discuss a situation involving resonance
are not the same, nor is one a special case of the other. without using the word “resonance”, the following general lan-
guage is one correct way to do so. In quantum mechanics, an
Examples of Delocalization and Resonance approximate description of the ground state of a system cor-
responds to a higher energy than the true energy. A quantum
In electronic structure theory, we must include electronic mechanical wave function is a probability amplitude. For ex-
spin in order to satisfy the Pauli principle. I will neglect spin– ample, an electron in a hydrogen atom has a nonzero prob-
orbit coupling to make the arguments simpler. Then a suit- ability to be near the nucleus and a nonzero probability to be
able electronic wave function for an atomic or molecular far from it. Restricting the ground-state wave function so that
system with N electrons must be antisymmetric under per- any situation with a nonzero probability is excluded would
mutation (transposition) of two electrons and must corre- correspond to a higher energy than the true ground-state en-
spond to a definite value S of total electronic spin. If the ergy and would be a more approximate description.
spatial part of the electronic wave function is formed from Similarly in the H2 molecule there is a nonzero prob-
no more than N orbitals, each of which may be interpreted ability of observing a charge distribution that corresponds to
as the variationally best representation of an electron mov- our usual picture of a covalent bond. There are also nonzero
ing in the average field of the other electrons, then the elec- probability amplitudes for observing structures that could be
tronic wave function including spin must be expressible as a described as H+H− or H−H+ (these zwitterionic structures are
permutation operator acting on product of N spatial orbitals usually called “ionic” in valence bond theory.) Restricting the
times a spin function (2); this kind of approximate wave func- wave function to exclude any of these contributions would
tion is called a spin-coupling-optimized generalized valence correspond to a higher energy than the true ground state and
bond wave function. In such a wave function, the orbitals to a less accurate description.
do not need to transform according to the full symmetry of Sometimes one starts with the model wave function cor-
the molecule (2). Thus, for example, the orbitals can be lo- responding to the restricted wave function with just the co-
calized even in symmetric systems (3). valent component. Then one proceeds to write a more
By restricting the form of the wave function, one can accurate wave function that also includes the ionic contribu-
obtain, as a special case, the Hartree–Fock molecular orbital tions. The covalent and ionic contributions are sometimes
method in which the wave function may be written as a Slater called covalent and ionic configurations (or valence bond
determinant of delocalized symmetry orbitals. In contrast, structures), and this process of adding the ionic contribu-
requiring the orbitals of the more general spin-coupling-op- tions is sometimes called configuration mixing or configura-
timized wave function to be delocalized symmetry functions tion interaction. But the students need to be able to read the
would be a constraint that raises the energy and—by the literature so one should add: this process is also called reso-
variational principle—makes the wave function less accurate. nance stabilization, and the energy lowering in proceeding
This provides a very general example of where a delocalized from the energy corresponding to the restricted wave func-
description of a system is associated with a higher energy than tion to the more accurate energy is called resonance energy.
the localized one; thus it is wrong to consider delocalization Learning any scientific field includes not only learning
as equivalent to resonance stabilization. Alternative but the concepts of the field but also learning the language that
equivalent explanations of delocalization are provided by is used by specialists in that field. This is required in order to
other authors (4, 5). understand lectures, to communicate with colleagues, and to
One can consider another example, namely HCl. This read the literature. “Resonance” is an important and well es-
may be considered a resonance hybrid of a covalent and an tablished technical term in chemistry. Students should learn
ionic structure (6).1 In HCl, the covalent structure has a weight it. Even if it were possible to legislate it away (which seems
larger than 50%, and the ionic structure, therefore, has a impossible after 75 years of use), it would be wrong to re-
weight less than 50%. Resonance stabilization of the domi- place it by “delocalization” since that is a different concept.
nant covalent structure by the ionic structure leads to the va-
lence electrons being more localized on the chlorine and less Wave Functions
delocalized over two centers than in the covalent structure. Reference 1 also asks “What is resonating?” and concludes
These two examples show that delocalization is not the that “there is no oscillation of any kind.” One can give an-
same concept as resonance. other answer to this question, but it requires a more advanced

www.JCE.DivCHED.org • Vol. 84 No. 5 May 2007 • Journal of Chemical Education 781

understanding of quantum mechanics than the material dis- state electronic energy when one improves a quantum me-
cussed above. When one discusses molecular structure and chanical model containing a single valence bond structure
energetics one is usually referring to the structure correspond- in the wave function to include other significant valence bond
ing to the ground electronic state (7). Kerber’s article (1) makes structures, where “structure” is defined in Note 1. It is in-
the important point that historically Pauling often used time- structive to refer to several research publications where the
dependent language to discuss resonance. Since the ground concept of resonance is used correctly in the context of mod-
state is a stationary state, it has a time-independent probabil- ern valence bond theory (5, 11).
ity distribution (8), and therefore time-dependent language is
incorrect. However, all wave functions (that is, all solutions of Note
the time-dependent Schrödinger equation) are not necessarily
stationary states. Superpositions of stationary states correspond- 1. The word “structure” is used here in the valence bond sense
ing to different energies are called wave packets. A valence bond of an electronic configuration state function corresponding to a single
wave function corresponding to a single covalent or ionic struc- set of spin orbitals with a spin coupling corresponding to a particu-
ture is a wave packet, as is a linear combination of two such lar covalent or ionic bonding scheme. A configuration state func-
wave functions ψA and ψB. In principle one could create a la- tion is an antisymmetric many-electron trial function composed
ser pulse that would excite a state corresponding to such a wave (usually) of the minimum number of Slater determinants required
packet. In the absence of decoherence, and if we assume that to obtain the correct spatial and spin symmetry.
ψA and ψB are orthogonal and that ψA is only strongly coupled
to ψB and vice versa, such a wave packet would physically os- Literature Cited
cillate between the two valence bond structures with a fre-
quency ν (in cycles per second) given by 1. Kerber, R. C. J. Chem. Educ. 2006, 83, 223–227.
1 2. Ladner, R. C.; Goddard, W. A., III J. Chem. Phys. 1969, 51,
2 ∆2
2
2 1073–1087.
ν = + ψ A H ψB (1) 3. Wadt, W. R.; Goddard, W. A., III J. Am. Chem. Soc. 1975, 97,
h 4
2034–2047.
where h is Planck’s constant, Ĥ is the Hamiltonian, and 4. (a) Hall, G. G.; Lennard-Jones, J. E. Proc. R. Soc. London, Ser.
A 1950, 202, 155–165. (b) Anderson, W. P.; Burdett, J. K.;
∆ = ψB H ψB − ψ A H ψ A (2) Czech, P. T. J. Am. Chem. Soc. 1994, 116, 8808–8809.
5. Voter, A. F.; Goddard, W. A., III Chem. Phys. 1981, 57, 253–
If the two structures are degenerate, then ∆ is zero, and 259.
this is the frequency mentioned on page 224 of ref 1. In prac- 6. Pauling, L. The Nature of the Chemical Bond, 3rd ed.; Cornell
tice, even if one could devise such a laser pulse, and even if University Press: Ithaca, NY, 1960; pp 67–75.
the assumptions mentioned above were valid, the perfectly 7. Herzberg, G. Molecular Spectra and Molecular Structure. III. Elec-
resonating superposition would not last very long because the tronic Spectra and Electronic Structure of Polyatomic Molecules;
electronic wave function in a nonstationary state is subject von Nostrand Reinhold: New York, 1966; p 9.
to decoherence (which is defined here as the evolution of a 8. Atkins, P. W.; Friedman, R. Molecular Quantum Mechanics,
pure state of the electronic subsystem of the molecular sys- 4th ed.; Oxford University Press: New York, 2005; p 25.
tem to a mixed state; ref 9); even for a molecule isolated in 9. Gottfried, K.; Yan, T.-M. Quantum Mechanics Fundamentals;
space in an ultrahigh vacuum, the motion of the molecule’s 2nd ed.; Springer: New York, 2003; pp 40–50.
own nuclei causes decoherence of the electronic wave func- 10. Zhu, C.; Jasper, A. W.; Truhlar, D. G. J. Chem. Theory Comput.
tion (10). Nevertheless, at least in this idealized case, there 2005, 1, 527–540.
does exist a physical resonating state, although it is not the 11. (a) Voter, A. F.; Goddard, W. A., III J. Chem. Phys. 1981, 75,
same as the ground-state wave function one is usually dis- 3638–3639. (b) Cooper, D. L; Gerratt, J.; Raimondi, M. Topics
cussing in the context of molecular structure and energetics. Current Chem. 1990, 153, 41–55. (c) Paldus, J.; Li, X. Israel
J. Chem. 1991, 31, 351–362. (d) Hiberty, P. C. J. Mol. Struct.
Summary 1998, 451, 237–261. (e) Shaik, S.; Shurki, A. Angew. Chemie
Int. Ed. 1999, 38, 586–625. (f ) Dijstra, F.; van Lenthe, J. H.;
The meaning of resonance energy as it appears in va- Havenith, R. W. A.; Jenneskens, L. W. Int. J. Quantum Chem.
lence bond theory is the lowering of the calculated ground- 2003, 91, 566–574.

782 Journal of Chemical Education • Vol. 84 No. 5 May 2007 • www.JCE.DivCHED.org