# Heisenberg’s chemical legacy: resonance and the chemical bond

- 221 Downloads
- 1 Citations

## Abstract

Heisenberg’s explanation of how two coupled oscillators exchange energy represented a dramatic success for his new matrix mechanics. As matrix mechanics transmuted into wave mechanics, resulting in what Heisenberg himself described as “…an extraordinary broadening and enrichment of the formalism of the quantum theory”, the term resonance also experienced a corresponding evolution. Heitler and London’s seminal application of wave mechanics to explain the quantum origins of the covalent bond, combined with Pauling’s characterization of the effect, introduced resonance into the chemical lexicon. As the Valence Bond approach gave way to a soon-to-be dominant Molecular Orbital method, our understanding of the term resonance, as it might apply to our understanding the chemical bond, has also changed.

### Keywords

Resonance Chemical bond Valence bond Molecular orbital Quantum chemistryThere is hardly a word in the scientific, or at least in the chemical, lexicon more fraught with peril than resonance. While the resonance of atomic spectroscopy denotes first and foremost a phenomenological observation, for which the matrix mechanics of Heisenberg provided a compelling explanation, the use of the term resonance in chemistry is much more problematic. Stemming from the characterization by Pauling of the Heitler-London wave function (vide infra) a resonance structure in a chemical context usually refers to one of the constituent canonical forms of any given wave function. Within organic chemistry the terminology is most commonly used to denote discrete, localized structures that when taken together, or hybridized, describe a delocalized π-system. Further complicating the matter is the fact that quantum chemistry has long been served by two, competing or complementary depending on one’s perspective, models of chemical bonding, namely those provided by the valence bond (VB) molecular orbital (MO) approaches. Resonance, as a chemical concept, is closely allied with the former and its use has been decried by some MO theorists as an unnecessary fiction. By tracing the origin, and evolution in meaning, of this term it is hoped that some clarity and perspective can be added to the ongoing discussion regarding the nature of chemical bonding.

## Resonance and matrix mechanics

*q*(

*mn*) can be related to the amplitude corresponding to resonance between the two oscillators. For the energy difference between the two oscillators,

*E*

_{m}−

*E*

_{n}, Planck’s relationship yields a frequency corresponding to this resonance of \( \upsilon_{mn} = {\frac{{E_{m} - E_{n} }}{h}}. \) The operator

*q*(

*mn*) can be decomposed to \( \sum\nolimits_{k} {q(mk) \cdot q(kn)} , \) which is equivalent to a summation of frequencies, υ

_{mk}+ υ

_{kn}since \( \upsilon_{mk} = {\frac{{E_{m} - E_{k} }}{h}} \) and \( \upsilon_{kn} = {\frac{{E_{k} - E_{n} }}{h}} \). Heisenberg demonstrated that this relationship requires that

*E*

_{m}=

*mh*υ

_{0}and

*E*

_{n}=

*nh*υ

_{0}, where m and n now correspond to different quantum states for a system of coupled oscillators with a resonant frequency of

*υ*

_{0}. Heisenberg also considered this resonance in terms of “Schrödinger’s formulation”, and through the application of a canonical transformation he obtained “eigenfunctions for the perturbed system” of the form:

*a*and

*b*designating the individual oscillator, and m and n designating the oscillator state (Heisenberg 1926). This explanation of the resonance phenomenon in atomic spectroscopy, defined as the exchange of energy between states described by the same eigenfunction series, represented one of the first major successes of Heisenberg’s matrix mechanics.

## From matrix mechanics to wave mechanics

*gedankenexperiment*, he presented its reformulation as the Dirac-Jordan relationship (Dirac 1927; Jordan 1926):

**p**and

**x**are the momentum and position matrices, respectively. That this relationship is both equivalent to Heisenberg’s original form for uncertainty, and also that the Dirac-Jordan relationship can be used to demonstrate the equivalence of the matrix and wave formulations of quantum mechanics, is shown in “Appendix”. Indeed in acknowledging his other colleague being honored that evening Heisenberg, in his Nobel address, also spoke of how Schrödinger “…had shown the mathematical equivalence of wave mechanics, which he had discovered, with quantum mechanics”. It was Schrödinger’s thesis however that in applying this new quantum Hamiltonian he would achieve “…a harmonic union” between the extremes of considering electrons and protons as “…nothing but wave systems” or as “…mass-points of definite mass and charge” (1926). As would become apparent over time such a unity was not to be achieved, due mainly to the unavoidable constraints imposed on Ψ by the uncertainty principle. Historically it would be Schrödinger’s formalism that again and again would be utilized to demonstrate the power, and the oddities, of this new quantum revolution.

Schrödinger’s unease concerning the fundamental tenets of quantum mechanics, most specifically the limitations seemingly placed on the measurement of phenomena subject to the uncertainty relationship, was shared by other prominent physicists, most notably Einstein. If *x* is limited to being one of many probabilities within the “wave packet”, then what can be said about the fundamental nature of position when the measurement of a specific value for *p*_{0} requires \( \Updelta x \to \infty \), and *x* becomes indeterminate? Surely this means that position, as a fundamental and objective property of the system, is now called into question. And the situation becomes worse when another measurement is made, but this time of position, such that \( \Updelta x \to 0 \). If position does not have an objective reality then how does a measurement like this yield such a value for *x*? The implication that measurements cannot be made without somehow perturbing the system and influencing the value of the very phenomenon being determined is unavoidable. This reflection was the subject of the Einstein–Podolsky–Rosen (EPR) paradox (Einstein et al. 1935), the resolution of which by Bohr ushered in the now famous Copenhagen interpretation (1935). In 1964 physicist Bell demonstrated that the effects of uncertainty could be measured (1964), and through such measurements, by Alain Aspect and others, quantum mechanics and uncertainty has endured (Healy 2010).

## The valence bond

The revolutionary power of Schrödinger’s formulation was exemplified almost immediately by it’s extension to molecular systems as presented by Heitler and London in their seminal paper of 1927 (1927). This development gave rise to what is now called the Valence Bond (VB) approach to quantum chemistry, somewhat to the chagrin of the authors themselves as witnessed by a statement of Fritz London. In decrying the chemists’ use of the term “valence”, London is reputed to have said that “…the chemist is made out of hard wood and he needs to have rules even if they are incomprehensible” (Gavroglu 1995).

_{2}, was simply to begin by assuming that \( \Uppsi = \phi_{A}^{1} \phi_{B}^{2} \) where \( \phi_{A}^{1} \) is the wave function for electron 1 of constituent hydrogen atom A and \( \phi_{B}^{2} \) is the wave function for electron 2 of constituent hydrogen atom B. However due to the indistinguishability of the electrons an equally valid form of Ψ, at least at short distances, is \( \phi_{A}^{2} \phi_{B}^{1} \), giving the two possibilities of \( \Uppsi = \phi_{A}^{1} \phi_{B}^{2} \pm \phi_{A}^{2} \phi_{B}^{1} \). Recognizing that \( H\phi = E_{0} \phi \), where

*E*

_{0}is the readily calculated energy, or eigenvalue, for the isolated hydrogen atom they calculated the two energies available for the H

_{2}molecule to be:

*E*

_{α}− 2

*E*

_{0}, allowed for the prediction of what the authors termed “homopolar binding” was an impressive chemical debut for the new wave mechanics. The authors noted that the wave function \( \Uppsi = \phi_{A}^{1} \phi_{B}^{2} + \phi_{A}^{2} \phi_{B}^{1} \), describes two electrons of the same energy but in different wave functions exchanging position. Such a description bore a marked resemblance to the resonance phenomenon described by Heisenberg 1 year earlier. However, Heitler and London, after remarking on this similarity, insisted that “…[in their paper] we cannot speak of resonance”. As noted by the authors this representation lacks a spin

*eigenfunction*, inclusion of which is required since electrons are classed as fermions and any wave function describing fermionic systems must be antisymmetric. In addition to this requirement, often referred to as the Pauli principle, the linear combination must also be normalized so as to reflect the certainty of finding the electrons somewhere within the molecule. As such a complete description of the VB wave function for the bond in H

_{2}is

At the end of the paper the authors note the additional possibilities of \( \Uppsi = \phi_{A}^{1} \phi_{A}^{2} \pm \phi_{B}^{2} \phi_{B}^{1} \), or states describing the wave functions for ions H^{−} and H^{+}. The utility of this extension to the molecular Ψ was recognized by London in a subsequent paper, where he demonstrated that the degree to which expanding Ψ to include these ionic states improved the calculated values for the bond distance, R, for any given diatomic molecule, served as a predictor of the polarity of the bond in question (London 1928). Pauling’s extension of this observation to explain properties such as hydrogen bonding, polarization and electron affinity quickly followed (1928). Interestingly Pauling states here, in speaking of Heitler and London’s seminal calculation, that “…this [bonding] potential is due mainly to a resonance effect, which may be interpreted as involving an interchange in position of the two electrons”. This terminology, despite the admonition of the earlier authors, would soon become widely adopted, and today it is to be found throughout the chemical literature. The terminology used in Pauling’s brief, only three page, 1928 paper, “The Shared-Electron Chemical Bond”, in many ways heralds the arrival of new era in chemistry. Given that the timeline from Heisenberg’s original paper on uncertainty to this paper of Pauling is barely 3 years, one can only marvel at the speed at which the revolution that began as quantum physics was giving rise to the field of quantum chemistry.

## Molecular orbital theory

_{2}this results in a molecular wave function that, ignoring normalization, has the form \( \Uppsi = (\phi_{A} + \phi_{B} )^{1} (\phi_{A} + \phi_{B} )^{2} \). While construction of the wave function does not require any formal placement of the electrons, the requirement that Ψ must be antisymmetric with respect to the interchange of the coordinates of any two electrons, or the Pauli principle, does result in the filling of these

*molecular orbitals*by electron pairs, except of course in the case of radical species. In MO calculations this requirement is enforced through the use of a Slater determinant, ironically a technique developed by the VB proponent JC Slater with the express purpose of obviating the need for a group-theoretical approach to VB wave function construction (Slater 1929). For H

_{2}this Slater determinant yields a normalized MO wave function, represented as:

The 5 year span from 1948 to 1952 has been previously identified, based on the changing nature of the content of the *Journal of Chemical Physics*, as a critical transition period in the switch by many quantum chemists from VB to MO methodology (Brush 1999). As has been noted by others this shift was driven in part by the increasing interest on the part of chemical community in molecular spectroscopy. While the MO method could be used to adequately describe electron excited states with relative ease, equivalent VB calculations were substantially more involved, and not always practical. Likewise the failure of the VB method to consistently account for simple diatomic paramagnetism mitigated against its use for the study of magnetic phenomena. However, the a key event in the emergent dominance of MO theory among quantum theorists must be the 1931 paper in which Erich Hückel applied both VB, what he called method 1, and MO, method 2, in an effort to account for the aromaticity of benzene (1931). He concluded that method 2 was clearly superior based on its ability to provide a general explanation for the unique stability of molecules with 4n + 2π electrons. In acknowledging this superiority Pauling himself, in a paper published with Wheland in which they develop a quantum mechanical framework to explain orientation effects in aromatic systems (Wheland and Pauling 1935), states that “In this paper we shall make use of the latter [MO method], since it is simpler in form and is more easily adapted to quantitative calculations”. However, the authors are careful to state, in a footnote, that “For purely quantitative arguments, on the other hand, the former [VB method] is quite convenient”. A more striking contrast between the conceptual frameworks of those working on each side of this divide is to be found by comparing the focus of the Nobel lectures by Pauling in 1954, and Mulliken in 1966. The title in 1954 was “Modern structural chemistry” and the survey of valence and bonding concepts spanned a century of theory, from Frankland and Van’t Hoff in the middle of the nineteenth century, through Lewis and Langmuir and, of course, Heitler and London, to the modern model of peptide coiling developed by Pauling and Corey. Mulliken’s presentation, on “Spectroscopy, molecular orbitals, and chemical bonding”, features a far different cast, comprised early on of physicists such as Schrödinger, Bohr and Bloch and later on the British theoretical chemists Longuet-Higgins and Dewar. In reporting, at the close of his presentation, on the SCF-MO calculations undertaken by Clementi at IBM’s Yorktown laboratory Mulliken states his “…belief that the era of computing chemists, when hundreds if not thousands of chemists will go to the computing machine instead of the laboratory for increasingly many facets of chemical information, is already at hand”. The approach that achieved near-universal adoption in the age of the *computing machine* was the MO method.

## Resonance and the chemical bond

In his Nobel lecture Mulliken states unequivocally that the VB method, or what he terms the atomic orbital method, “…is a valid alternative to the MO method”. While no comparable characterization of MO theory is to be found in Pauling’s Nobel address 12 years earlier, in reality none was required since ease of application and dramatic predictive successes had combined early on to ensure the primacy of MO theory. Indeed the asymmetry of the relationship persists to this day, with the legitimacy of the VB approach being called into question for reasons as varied as its dependence “…on great flocks of resonance structures” (Mulliken 1975) to successes being ascribed “…to its [resonance] correspondence with MO theory rather than to the validity of its own premises” (Dewar and Longuet-Higgins 1952). Nevertheless one advantage of VB theory that even its harshest critics usually admit to is its heuristic power, mainly due to the conceptual power and ease of representation of the canonical resonance structures. While it might be tempting to dismiss this advantage as being necessitated by the intellectual “hard wood” of the chemist, the persistence of VB-derived models throughout both the research and pedagogical literature is testament to its utility.

Even at the algorithmic level the two treatments have a similar methodology (Vermulapalli 2008). In MO theory the wave function \( \Uppsi = \prod\nolimits_{i} {n_{i} } \psi_{i} \), where ψ_{i} are the molecular orbitals expressed as linear combinations of AOs, \( \psi_{i} = \sum\nolimits_{j}^{atoms} {n_{i} c_{j} \phi_{j} } \). In the Valence Bond treatment \( \Uppsi = \sum\nolimits_{i} {c_{i} \chi_{i} } \), where χ_{i} are the canonical resonance forms, constructed as the product of occupied atomic orbitals, \( \chi_{i} = \prod\nolimits_{j} {n_{j} } \phi_{j} \). Where the product is over “overlap-enhanced” orbitals rather than pure atomic orbitals, i.e. \( \chi_{i} = \prod\nolimits_{j} {n_{j} } \gamma_{j} \), where for example \( \gamma_{j} = c_{1} \phi_{1} + c_{2} \phi_{2} \), the method is often termed hybridization.

It is at the more fundamental and epistemological level that serious issues arise in connection with the use of the term resonance. As has been noted by others the initial use of the term by Pauling was somewhat haphazard (Kerber 2006), with resonance denoting both a dynamic phenomenon, the exchange of electron position, and also being used as a static descriptor for the constituent chemical structures. Eventually Pauling was to clarify this ambiguity, most notably through revisions of his seminal text “The Nature of the Chemical Bond”. In a 1977 paper, written to refute what he termed were “…wrong and misleading statements” that appeared in a biographical memoir for Sir Robert Robinson, Pauling states that “the theory of resonance…was a direct consequence of quantum mechanics” because it is based “on the principle that the correct wave function for a state of a system can be expressed as a sum of functions constituting a complete set” (1977). For two such functions, describing “two unperturbed states”, he states that “…the true one is to be found by linear combination, so that the real situation is intermediate between the two”, and from such a combination “…there will be a resonance effect in the energy which will generally bring the real energy lower”. Gone was the mention of resonance forms and any attendant implication that these forms, as his longtime collaborator Wheland wrote, “imply the actual existence of distinct substances. Wheland actually went further in clarifying, and narrowing, the use of the term by stating that “…we shall never speak of resonance as a phenomenon” (1955). But the assumption that resonance, or exchange, energy is the stabilizing force responsible for the chemical bond remained, and in a 1992 paper on the nature of the chemical bond Pauling categorically defends the idea of a localized and directional bond, and lauds the role that resonance has played in keeping the concept “alive and well” (1992).

From its very first appearance in the chemical literature in 1866 (Frankland 1866) the concept of the chemical bond continues to be resistant to a single, universally accepted definition. While many VB theorists invoke resonance in their battle to preserve the localized and directional character of the traditional two electron, two-center covalent bond, others, including Wheland himself, establish a correspondence between resonance energy and the stabilizing effects of delocalization (1955). With the non-existence of discrete resonance structures conceded, and with all attempts at describing resonance as a real, dynamic event eschewed, it might be considered, not just useful, but also necessary to correlate the resonance effect manifest in the VB wave function with a specific chemical phenomenon. While the rather obvious answer of identifying the resonance energy as a measure of bond energy avoids any epistemological characterization of a resonance form or of an electron exchange, it is not borne out by theoretical analysis. In his landmark paper of 1962 Ruedenberg identified the driving force of covalent bond formation as a quantum mechanical *interference* effect, that contracts the electronic wave function, and lowers the potential energy by lowering the electron kinetic energy (1962). Any explanation of chemical bonding as due to electron sharing or overlap, or based on an electrostatic or nonkinetic concept “…misses the very reason why quantum mechanics can explain chemical binding, whereas classical mechanics cannot”. Ruedenberg’s analysis represents essentially a *deus ex machina* analysis of the quantum origins of chemical bonding, and precludes a discrete correspondence between any resonance stabilization and chemical bonding. Yet VB theory endures, not just as a heuristic model but as the basis for computational program suites competitive in accuracy and speed with MO-based programs. Additionally recent VB approaches to the construction of CASSCF (Complete Active Space Self-Consistent Field) wave functions have succeeded in “reclaiming” the electron-pair bond from MO-derived wave functions (Malrieu et al. 2007), introducing the intriguing proposition that Lewis’ original electron pair might be an emergent consequence of electron correlation.

At the end one can’t help but wonder at what might have been the result if, instead of separate Nobel prizes in Chemistry for VB in 1954 and MO development in 1966, the prize had been awarded successively, with the lectures delivered together, as happened with Heisenberg, Dirac and Schrödinger in 1933. Instead we are at a point today where those speculating on the nature of the chemical bond have to, as one veteran of the struggle writes, “…function with split consciousness between the localized and delocalized worlds” (Shaik 2007).

## Notes

### Acknowledgments

The author is grateful to the Educational Advancement Foundation (EAF) and the W. M. Keck Foundation grant for their generous support, and also the Welch Foundation (Grant # BH-0018) for its continuing support of the Chemistry Department at St. Edward’s University. The author also wishes to acknowledge the contributions of Brian Healy, whose commitment to the intellectual process served as a catalyst for this formulation.

### References

- Bell, J.S.: On the Einstein Podolsky Rosen paradox. Physics
**1**, 195–200 (1964)Google Scholar - Bloch, F.: Heisenberg and the early days of quantum mechanics. Phys. Today
**29**, 23–27 (1976)CrossRefGoogle Scholar - Bohr, N.: Can quantum-mechanical description of physical reality be considered complete? Phys. Rev.
**48**, 696–702 (1935)CrossRefGoogle Scholar - Brush, S.G.: Dynamics of theory change in chemistry: part 2. Benxene and molecular orbitals, 1945–1980. Stud Hist Phil Sci
**30**, 263–302 (1999)CrossRefGoogle Scholar - Dewar, M.J.S., Longuet-Higgins, H.C.: The correspondance between the resonance and molecular orbital theories. Proc Roy Soc
**214**, 482–493 (1952)CrossRefGoogle Scholar - Dirac, P.: The physical interpretation of the quantum dynamics. Proc. Roy. Soc. Lond.
**A113**, 621–641 (1927)Google Scholar - Einstein, A., Podolsky, B., Rosen, N.: Can quantum-mechanical description of physical reality be considered complete? Phys. Rev.
**47**, 777–780 (1935)CrossRefGoogle Scholar - Frankland, E.: Contributions to the notation of organic and inorganic compounds. J. Chem. Soc.
**19**, 372–395 (1866)CrossRefGoogle Scholar - Gavroglu, K.: Fritz London: A Scientific Biography, p. 85. Cambridge University Press, Cambridge (1995)CrossRefGoogle Scholar
- Healy, E.F.: In defense of a heuristic interpretation of quantum mechanics. J. Chem. Educ.
**87**, 559–563 (2010)CrossRefGoogle Scholar - Heisenberg, W.: Multi-body problem and resonance in the quantum mechanics. Zeitschrift für Physik
**38**, 411–426 (1926)CrossRefGoogle Scholar - Heitler, W., London, F.: Interaction between neutral atoms and homopolar bonding.
*Zeitschrift für Physik***44**, 455 (1927) English translation in Hettema, H. “Quantum Chemistry, Classic Scientific Papers”, World Scientific, Singapore (2000)Google Scholar - Huckel, E.: Quantum-theoretical contributions to the benzene problem. I. The electron configuration of benzene and related compounds. Zeitschrift für Physik
**70**, 204–286 (1931)CrossRefGoogle Scholar - Hund, F.: Zur Frage der chemischen Bindung. Zeitschrift für Physik
**73**, 1–30 (1931)CrossRefGoogle Scholar - Jordan, P.: Uber eine neue Begr¨undung der Quantenmechamik I. Zeitschrift für Physik
**40**, 809 (1926)CrossRefGoogle Scholar - Kerber, R.C.: If it’s resonance, what is resonating? J. Chem. Educ.
**83**, 223–227 (2006)CrossRefGoogle Scholar - Lennard-Jones, J.E.: The electronic structure of some diatomic molecules. Trans. Faraday Soc.
**25**, 668–686 (1929)CrossRefGoogle Scholar - London, F.: On the quantum theory of homo-polar valence numbers.
*Zeitschrift für Physik***46***,*455 (1928) English translation in Hettema, H. “Quantum Chemistry, Classic Scientific Papers”, World Scientific, Singapore (2000)Google Scholar - Malrieu, J.-P., Guihery, N., Calzado, C.J., Angeli, C.: Bond electron pair: its relevance and analysis from the quantum chemistry point of view. J. Comput. Chem.
**28**, 35–50 (2007)CrossRefGoogle Scholar - Mulliken, R.S.: The assignment of quantum numbers for electrons in molecule. Phys. Rev.
**32**, 186–228 (1928)CrossRefGoogle Scholar - Mulliken, R.S.: Selected Papers, ed. D.A. Ramsay, J. Hinze (University of Chicago Press), p. 8 (1975)Google Scholar
- Pauling, L.: Shared-electron chemical bond. Proc. Natl. Acad. Sci.
**14**, 359–362 (1928)CrossRefGoogle Scholar - Pauling, L.: The theory of resonance in chemistry. Proc. R. Soc. Lond. A.
**356**, 433–441 (1977)CrossRefGoogle Scholar - Pauling, L.: The nature of the chemical bond–1992. J. Chem. Educ.
**69**, 519–521 (1992)CrossRefGoogle Scholar - Ruedenberg, K.: The physical nature of the chemical bond. Rev. Mod. Phys.
**34**, 326–376 (1962)CrossRefGoogle Scholar - Schrödinger, E.: An undulatory theory of the mechanics of atoms and molecules. Phys. Rev.
**28**, 1049–1070 (1926)CrossRefGoogle Scholar - Schrödinger, E.: The present situation in quantum mechanics.
*Naturwissenschaften*,**23**, 807–812; 823–828; 844–849 (1935) English translation “Quantum Theory and Measurement”, eds. Wheeler, J.A., Zurek, W.H., Princeton Univ. Press (Princeton, NJ, 1983)Google Scholar - Shaik, S.: Is my chemical universe localized or delocalized? Is there a future for chemical concepts? New J. Chem.
**31**, 2015–2028 (2007)CrossRefGoogle Scholar - Slater, J.C.: The theory of complex spectra. Phys. Rev.
**34**, 1293–1312 (1929)Google Scholar - Wheland, G.W., Pauling, L.: A quantum mechanical discussion of orientation of substituents in aromatic molecules. J. Am. Chem. Soc.
**57**, 2086–2095 (1935)CrossRefGoogle Scholar - Wheland, G.W.: Resonance in Organic Chemistry, pp. 7, 75. Wiley, New York (1955)Google Scholar
- Vermulapalli, G.K.: Theories of the chemical bond and its true nature. Found. Chem.
**10**, 167–176 (2008)CrossRefGoogle Scholar