Persistence and magnitude of microscopic structure-dependent corrections to the dielectric constant of dilute solutions of polar molecules in non-polar solvents

Van Vleck (Molecular Physics, 1972) showed by an explicit, idealized microscopic calculation that the dielectric constant of an infinitely dilute solution of a polar molecule in a non-polar solvent depends on the packing geometry of the solvent molecules: second-order dipolar-coupling corrections, governed by a dimensionless lattice sum Q that differs between simple, body-centred and face-centred cubic gratings and between substitutional and interstitial solute placement, are invisible to the continuum (Onsager, Buckingham) cavity theories, which reproduce the microscopic result only after the discrete sum is replaced by a continuum integral roughly twice as large. The open problem is whether these structure-dependent ('topological') corrections persist in real liquids and what their magnitude is: generalizing the expansion to structurally disordered, correlated solvents with practicable numerical estimates of the Q-analogues for realistic liquid structures; quantifying the changes caused by quantum-mechanical response and realistic intermolecular interactions; extending beyond second order and to finite solute concentration; and thereby establishing the quantitative reliability of the Debye and Onsager formulas, and of the dipole moments inferred from them, for dilute solutions in real non-polar solvents. Later literature settled the idealized lattice case (exact solutions show only few-percent deviations) and built correlation-function frameworks for fluids, but never delivered the quantitative completion for realistic liquids; the surviving core remains open.

Background

The static dielectric constant $\epsilon$ of a liquid solution is the macroscopic signature of how its molecules polarize and orient in an applied electric field. Measuring $\epsilon$ as a function of solute concentration and extrapolating to infinite dilution has been, since the 1920s, the standard experimental route to the permanent dipole moment $\mu$ of a molecule in solution: at extreme dilution the solute's orientational polarization contributes the familiar Langevin term $N_{\mathrm{s}}\mu^{2}/3kT$ (with $N_{\mathrm{s}}$ the solute number density and $kT$ the thermal energy) on top of the induced-polarization contributions $N\alpha + N_{\mathrm{s}}\alpha_{\mathrm{s}}$ of solvent and solute polarizabilities. The classical Debye dilute-solution formula combines these contributions additively through the Clausius–Mossotti ratio $(\epsilon-1)/(\epsilon+2)$; more refined treatments — Onsager's 1936 reaction-field model, in which the solute dipole sits in a spherical cavity carved out of a dielectric continuum and is amplified by the field of its own polarization, and later cavity-model refinements, most notably A. D. Buckingham's dilute-solution formula — modify the additive formula by a cavity-dependent factor. All of these are semi-macroscopic: the discrete solvent molecules are replaced by a continuum, so nothing in them can depend on how the molecules are actually packed.

Van Vleck (Molecular Physics 24, 341 (1972)) asked what a genuinely microscopic calculation gives. His model was deliberately idealized: classical mechanics; solvent molecules on a perfect cubic grating (simple, body-centred, or face-centred); molecular polarizabilities represented by isotropic classical harmonic oscillators (bound charges, so each polarizability is a fixed scalar $\alpha$); only dipole–dipole intermolecular interactions; a single polar solute, either substituting for a solvent molecule at a lattice site or sitting at a symmetric interstitial position; a spherical sample boundary; and a series expansion truncated at second order in the dipolar coupling. The result is that the dilute-solution dielectric relation acquires second-order corrections proportional to a dimensionless lattice sum

$$Q = 2N^{-2}\sum_{j} r_{ij}^{-6},$$

where $N$ is the solvent number density and the sum runs over all other lattice sites $j$ at distance $r_{ij}$ from a reference site. The corrections depend on the packing: $Q \approx 16.8$ for a simple cubic grating, $Q \approx 14.5$ for body-centred cubic, $Q \approx 14.4$ for face-centred cubic, and on the solute's placement (substitutional versus interstitial). Two symmetry cases make the Debye additivity exact: a substitutional solute whose polarizability equals the solvent's ($\alpha_{\mathrm{s}} = \alpha$), and a non-polarizable ($\alpha_{\mathrm{s}} = 0$) solute at a symmetric interstitial site of a simple or face-centred cubic lattice. Most strikingly, in the source's own words, 'the formulas of Onsager and Buckingham are recovered if a lattice sum is arbitrarily replaced by a continuum integral which actually is about twice as large' — the integral in question being $2(4\pi/3)^{2} \approx 35.1$ against lattice values of roughly 14 to 17 — so that continuum theories both mask the geometry dependence and overestimate the second-order corrections. The source's own abstract summarizes the point: these are 'topological effects which the usual semi-macroscopic theories employing continuum approximations are inherently incapable of revealing', demonstrated on an idealized lattice.

The idealization is the issue. A real non-polar solvent is not a cubic crystal: its local structure is disordered and is described statistically by pair correlation functions; its molecules interact through repulsion, dispersion and multipolar forces, not only point-dipole coupling; their polarizabilities are quantum-mechanical response functions, not classical oscillators; and real solutions have finite solute concentration, where solute–solute interactions matter. Whether the geometry-dependent corrections survive in real liquids — with what sign and magnitude, relative to the few-percent accuracy routinely attributed to the Debye and Onsager formulas — is the open question that the source's calculation bequeathed. The dielectric increment at infinite dilution, i.e. the concentration derivative of $\epsilon$ at zero solute mole fraction, $(d\epsilon/dx)_{0}$, is the natural observable in terms of which the corrections and the reliability of continuum formulas are to be quantified.

Problem Statement

Determine whether the discrete-structure ('topological') corrections to the dielectric constant of an infinitely dilute solution of a polar molecule in a non-polar solvent — the second-order dipolar-coupling corrections encoded in the lattice sum $Q$ of the source's cubic-grating model — persist in realistic liquids, and quantify their magnitude. The source model's own stated idealizations mark four linked directions of generalization serving this single objective: (1) generalize the second-order microscopic expansion from a perfect cubic grating to a structurally disordered, short-range-correlated liquid solvent, and construct a practicable procedure yielding reliable numerical estimates of the resulting structural sums (the analogues of $Q$) for experimentally relevant liquid structures; (2) quantify how the corrections change when classical harmonic-oscillator polarizabilities and pure point-dipole coupling are replaced by quantum-mechanical response and realistic intermolecular interactions; (3) extend the treatment beyond second order in the dipolar coupling and to finite solute concentration, where solute–solute interactions cannot be neglected; and (4) from these, determine the sign and size of the deviations of the Debye and Onsager (cavity/reaction-field) formulas for the dielectric increment at infinite dilution, $(d\epsilon/dx)_{0}$ with $x$ the solute mole fraction, and the resulting reliability of dipole moments inferred from dilute-solution dielectric data in real non-polar solvents. The perfect-lattice case is settled — exact lattice solutions show only few-percent deviations from the Debye and Onsager formulas — so the open question is the carry-over of this four-direction program to real fluids.

References

1. J. H. Van Vleck, 'On the theory of the dielectric constant of dilute solutions of polar molecules in non-polar solvents', Molecular Physics 24(2), 341–348 (1972). DOI: 10.1080/00268977200101491, https://doi.org/10.1080/00268977200101491
2. Lars Onsager, 'Electric Moments of Molecules in Liquids', Journal of the American Chemical Society 58(8), 1486–1493 (1936). DOI: 10.1021/ja01299a050, https://doi.org/10.1021/ja01299a050
3. André G. Bellemans and M. Plaitin, 'On the dielectric constant of dilute solutions of polar molecules in non-polar solvents', Bulletin de la Classe des Sciences, Académie Royale de Belgique 61, 324–335 (1975). DOI: 10.3406/barb.1975.57922, https://doi.org/10.3406/barb.1975.57922
4. J. Biemond, C. MacLean and M. Mandel, 'Alignment of polarizable dipoles on a lattice', Physica A 79(1), 52–64 (1975). DOI: 10.1016/0378-4371(75)90087-4, https://doi.org/10.1016/0378-4371(75)90087-4
5. J. Biemond, C. MacLean and M. Mandel, 'On the theory of the alignment of a polar molecule in a non-polar solvent', Physica A 83(2), 358–370 (1976). DOI: 10.1016/0378-4371(76)90041-8, https://doi.org/10.1016/0378-4371(76)90041-8
6. J. Biemond, 'Dielectric alignment of a non-polar molecule substituting a polar molecule in a lattice of molecular dipoles', Physica A 88(2), 385–394 (1977). DOI: 10.1016/0378-4371(77)90012-7, https://doi.org/10.1016/0378-4371(77)90012-7
7. Robert L. Fulton, 'Optical properties of a guest molecule in a discrete lattice', The Journal of Chemical Physics 61(10), 4141–4155 (1974). DOI: 10.1063/1.1681712, https://doi.org/10.1063/1.1681712
8. S. J. Sheng and D. M. Hanson, 'Dipole moments of the lowest excited singlet and triplet states of 2,4,5-trimethylbenzaldehyde in a durene host crystal', The Journal of Chemical Physics 60(2), 368–374 (1974). DOI: 10.1063/1.1681051, https://doi.org/10.1063/1.1681051
9. Freeman P. Chen, David M. Hanson and David Fox, 'The origin of Stark shifts and splittings in molecular crystal spectra. I. The effective molecular polarizability and local electric field: Durene and naphthalene', The Journal of Chemical Physics 63(9), 3878–3885 (1975). DOI: 10.1063/1.431829, https://doi.org/10.1063/1.431829
10. P. Bordewijk, 'On the separation of long-range and short-range dipole correlations in polar fluids', Chemical Physics 33(3), 451–464 (1978). DOI: 10.1016/0301-0104(78)87095-5, https://doi.org/10.1016/0301-0104(78)87095-5
11. Léo Degrève, 'On the structure and dielectric constant of infinitely dilute fluid mixtures constituted of spherical molecules', Molecular Physics 40(4), 837–856 (1980). DOI: 10.1080/00268978000101921, https://doi.org/10.1080/00268978000101921
12. C. J. F. Böttcher and P. Bordewijk, 'The experimental determination of permanent dipole and quadrupole moments', chapter in Dielectrics in Time-dependent Fields (Theory of Electric Polarization, Vol. 2), pp. 387–422 (1978). DOI: 10.1016/B978-0-444-41579-0.50011-2, https://doi.org/10.1016/B978-0-444-41579-0.50011-2
13. Axel Thuresson and Gunnar Karlström, 'Dielectric response from lattices of dipoles with fixed orientation', The Journal of Chemical Physics 141(23), 234505 (2014). DOI: 10.1063/1.4903926, https://doi.org/10.1063/1.4903926
14. Roberto Cammi, 'The Role of Computational Chemistry in the Experimental Determination of the Dipole Moment of Molecules in Solution', Journal of Computational Chemistry 40(26), 2309–2317 (2019). DOI: 10.1002/jcc.26009, https://doi.org/10.1002/jcc.26009
15. Dmitry V. Matyushov, 'Multiparticle orientational correlations are responsible for the nonlinear dielectric effect: Analysis of temperature-dependent measurements for glycerol', The Journal of Chemical Physics 157(16), 164501 (2022). DOI: 10.1063/5.0106766, https://doi.org/10.1063/5.0106766