Index of content:
Volume 102, Issue 7, 15 February 1995

High‐resolution analysis of the fundamental bending vibrations in the Ã ^{2}Π and X̃ ^{2}Σ^{+} states of CaOH and CaOD: Deperturbation of Renner–Teller, spin–orbit and K‐type resonance interactions
View Description Hide DescriptionThe v _{2}=1 bending vibrational levels of the Ã ^{2}Π and X̃ ^{2}Σ^{+} states of both CaOH and CaOD have been rotationally analyzed via laser excitation of the Ã(010) κ ^{2}Σ, ^{2}Δ, μ ^{2}Σ←X̃(010)^{2}Π, (000)^{2}Σ^{+} bands. The estimated measurement accuracy for rotational transitions is 0.0035 cm^{−1}. K‐type resonances and level crossings have been identified between the Ã(010) ^{2}Σ and ^{2}Δ vibronic components, and account for the observation of dramatic intensity anomalies from quantum mechanical interference. An effective Hamiltonian has been constructed to model the Renner–Teller, spin–orbit, and K‐type resonanceinteractions in the Ã(010) vibronic manifold, and to fit all the observed bands simultaneously for each isotopomer. The fundamental bending frequencies and Renner–Teller parameters have been determined: ν_{2}(X̃)=352.9259(9), ν_{2}(Ã)=363.1736(8) and εω_{2}=−36.2634(6) cm^{−1} for CaOH; ν_{2}(X̃)=266.8398(10), ν_{2}(Ã)=274.6475(5), and εω_{2}=−26.9601(8) cm^{−1} for CaOD. The isotope relations between the fitted molecular constants of CaOH and CaOD are examined. The spin‐rotation constant γ and the centrifugal distortion constant A _{ D } of the spin–orbit coupling have been separated in the Ã ^{2}Π(010) level owing to the Renner–Teller effect. The equilibrium bond lengths and force constants in the X̃ ^{2}Σ^{+} state have been derived based on the data of the two isotopomers. The Coriolis coupling constants have been derived from the harmonic force field, and yield calculated l‐type doubling constants that are in excellent agreement with the experimental values for both CaOH and CaOD.

^{17}O hyperfine and quadrupole interactions for water ligands in frozen solutions of high spin Mn^{2+}
View Description Hide DescriptionThe magnetic couplings of ^{17}O in H_{2} ^{17}O coordinated to high spin Mn^{2+} in a frozen aqueous solution were determined using the complementary magnetic resonance techniques of pulsed and continuous wave (cw) ENDOR(electron nuclear double resonance), ESEEM (electron spin echo envelope modulation), and PFSEPR [pulse field sweep electron paramagnetic resonance(EPR)]. Several complications arise from the high electron spin multiplicity of the d ^{5}, Mn^{2+} ion and the high nuclear spin multiplicity, I=5/2, of the ^{17}O nucleus. At the applied magnetic field strengths in 9 GHz EPR studies, the zero‐field splitting of the S=5/2 Mn^{2+} ion in aqueous frozen solution is small relative to the electron spin Zeeman interaction so that the M _{ S }=±1/2,±3/2,±5/2 electron spin states all contribute to the ENDOR spectrum.
This results in a complex spectrum in which the ^{17}O ENDORpowder pattern arising from the M _{ S }=±1/2 manifolds are separately resolved but the powder patterns from the M _{ S }=±3/2,±5/2 manifolds overlap the multiple ^{1}H ENDOR lines arising from all six M _{ S }manifolds [X. Tan, M. Bernardo, H. Thomann, and C. P. Scholes, J. Chem. Phys. 98, 5147 (1993)]. Given this complexity, a combination of complementary spectroscopic techniques and numerical simulations are used to deconvolute the overlapping spectra and to assign the spectral lines. The ENDOR spectra provided an experimental description of H_{2} ^{17}O hyperfine couplings to high spin Mn^{2+} in a frozen solution. The ESEEM results are consistent with the first‐order assignments of the ENDOR lines and demonstrate the feasibility of ESEEM measurements of ^{17}O ligand hyperfine couplings to Mn^{2+}. Simulations of the ^{17}O ENDOR hyperfine patterns of aqueous frozen solutions of Mn^{2+}, especially those near 20 MHz, indicated an A‐tensor anisotropy of A _{⊥}=−6.5±0.5 MHz and A _{∥}=−9.5±0.5 MHz, consistent with couplings observed by single crystalENDOR of H_{2} ^{17}O ligated to Mn^{2+} doped in [LaMg(NO_{2})_{12}⋅24(H_{2}O)]. More detailed simulations of the ENDOR pattern below 10 MHz indicated the need for quadrupole couplings consistent with those measured by single crystalENDOR and with those determined by gas phase measurements on H^{17}OD. Simulations of the ENDOR spectra recorded by the cw and pulsed techniques have delineated important features of the techniques which must be taken into account for a quantitative analysis of the ENDOR amplitudes. It is expected that the general ENDOR conditions employed and the theory developed will be useful in frozen solution studies of ^{17}O involved as a ligand to Mn^{2+} in enzymes.

Femtosecond solvation dynamics determining the band shape of stimulated emission from a polar styryl dye
View Description Hide DescriptionSpectra of transient absorption and stimulated emission are recorded for the styryl dye DASPI, after excitation at 470 nm, with experimental resolution of 100 fs. The evolution of the S_{1}→S_{0} transition energy distribution is obtained for the solvents methanol and acetonitrile at several temperatures. It is described by the dependence of the mean (first moment), width, and asymmetry (second and third central moments) of the distribution on time. The observed time‐dependence of the mean transition energy is simulated by appropriate models for the solvation dynamics. In both methanol and acetonitrile an ultrafast component is observed. Width and asymmetry change most rapidly and characteristically during this initial part of solvation. In the evolution of the higher moments, different relaxation contributions apparently are better distinguished than in the evolution of the first moment. For methanol at 50 °C, an oscillatory evolution is observed mainly in the higher moments which may indicate underdamped coherent solvent motion.

FeC_{ n } ^{−} and FeC_{ n }H^{−} (n=3,4): A photoelectron spectroscopic and density functional study
View Description Hide DescriptionPhotoelectron spectra of the title molecules are reported at 3.49 eV photon energy. Vibrational structures are resolved in the spectra of FeC^{−} _{3} and FeC_{3}H^{−}. The FeC^{−} _{4} spectrum is unusually broad, indicating a large equilibrium geometry change from the anion to the neutral states. The FeC_{4}H^{−} spectrum exhibits a single strong feature. Theoretical studies using the density functional theory are carried out to determine the structures and bonding of these clusters. All the molecules in the anion ground states are found to be linear with the Fe atom bonded at one end. The Fe and C bonding involves strong Fe 4s and C sp interactions as well as considerable Fe 3d and C π interactions. The n=3 species can be best characterized by cumulenic types of bonding with FeC_{3}H also having an acetylenic isomer. The n=4 species in the linear structures can be approximately described by diacetylenic types of bonding. Mulliken charge analyses indicate that the extra charge in all the anions enters mainly into the Fe 4s antibonding orbital, in agreement with the assignment that the threshold detachment takes place from the σ* orbital mainly between the Fe and C atoms. The vibrational structure resolved in the FeC^{−} _{3} spectrum yields a Fe–C stretching frequency of 700 (150) cm^{−1} for the first excited state of FeC_{3}, in agreement with the Fe–C multiple bonding character.

Cavity ring‐down spectroscopy for quantitative absorption measurements
View Description Hide DescriptionWe examine under what conditions cavity ring‐down spectroscopy (CRDS) can be used for quantitative diagnostics of molecular species. We show that CRDS is appropriate for diagnostics of species whose absorption features are wider than the spacing between longitudinal modes of the optical cavity. For these species, the absorption coefficient can be measured by CRDS without a knowledge of the pulse characteristics provided that the cavity ring‐down decay is exponential. We find that the exponential ring‐down decay is obeyed when the linewidth of the absorption feature is much broader than the linewidth of the light circulating in the cavity. This requirement for exponential decay may be relaxed when the sample absorption constitutes only a small fraction of the cavity loss and, consequently, the sample absorbance is less than unity during the decay time. Under this condition the integrated area of a CRDS spectral line approximates well the integrated absolute absorption coefficient, which allows CRDS to determine absolute number densities (concentrations). We determine conditions useful for CRDS diagnostics by analyzing how the absorption loss varies with the sample absorbance for various ratios of the laser pulse linewidth to the absorptionlinewidth for either a Gaussian or a Lorentzian absorption line shape.

Detection of vibrational‐overtone excitation in water via laser‐induced grating spectroscopy
View Description Hide DescriptionIn this paper we describe a method, based on the laser‐induced grating technique, for studying the spectroscopy of vibrational overtone‐excited gas‐phase water. Two phase‐coherent visible laser beams whose frequencies are in the range of the third overtone of the OH stretch in water are crossed in the gas‐phase sample. As the wavelength of these excitation beams is scanned through individual rovibrational OH overtone transitions, vibrational energy is deposited into the water in a spatially sinusoidal pattern. A fixed‐frequency 266 nm probe laser beam is diffracted from the resultant transmission diffraction grating in water. We show that under collision‐free conditions, probe laser diffraction is observed from the initially excited grating, which is a necessary condition for using this technique to study the absorption spectroscopy of the vibrationally excited molecules. Under multiple collision conditions, a probe laser wavelength‐independent refractive indexgrating is formed within the bulk sample. In addition, we observe temporal oscillations in the grating diffraction efficiency arising from excitation of standing acoustic waves.

Electronic properties and geometric structures of Li_{4}H and Li_{9}H from optical absorption spectra
View Description Hide DescriptionOptical absorptionspectra of Li_{4}H and Li_{9}H clusters have been recorded by depletion spectroscopy in the visible range. From comparison with ab initio calculations, geometries of both clusters are identified. The hydrogen atom assumes a peripheral position bridging two and three Li atoms in the planar and three‐dimensional structures of Li_{4}H and Li_{9}H, respectively. Na_{4}F and Na_{9}F clusters are also theoretically studied and it is shown how the strong electronegativity of the F atom leads to different geometries than in lithium hydrids. Finally, the metallic character of these clusters is discussed and in both cases, the hydrogen or fluorine atom localizes one valence electron. However, the optical absorptionspectra are much broader than in pure Li_{ n } and Na_{ n } clusters due to the lower symmetry.

Dissociative attachment in hot CH_{3}Cl: Experiment and theory
View Description Hide DescriptionThe dissociative attachment (DA) cross section of hot CH_{3}Cl has been measured in a crossed electron–molecule beam apparatus at temperatures up to 750 K and electron energies from 0–0.5 eV. The results are compared to cross sections computed using a mixed ab initio‐semiempirical approach, treating CH_{3}Cl as a quasidiatomic molecule. The theoretical treatment requires an anion potential curve in the stable region as a portion of the input data. Computations with three different basis sets show the results to be sensitive to the size of basis set from which the potential is determined. At high temperatures, the experimental DA cross sections are found to be in very good agreement with those derived from theory using the potential curve computed with the most flexible of the basis sets. At room temperature the theory suggests that the measured DA cross section is still limited by the presence of impurities.

Excitation transfer from Kr(5s’,^{3} P _{0}) and Kr(5s,^{3} P _{2}) atoms to ^{12}CO and ^{13}CO
View Description Hide DescriptionEmission spectra have been used to characterize the excitation‐transfer reactions from Kr(5s’,^{3} P _{0}) and Kr(5s,^{3} P _{2}) metastable atoms to ^{12}CO and ^{13}CO at 300 K. The most important products from the Kr(^{3} P _{0}) reactions are ^{12}CO and ^{13}CO(b ^{3}Σ^{+},v’=0 and 1) and ^{12}CO(a’ ^{3}Σ^{+},v’=34 and 35) and ^{13}CO(a’ ^{3}Σ^{+},v’=35 and 36). The rotational distributions of the CO(a’ ^{3}Σ^{+}) and CO(b ^{3}Σ^{+},v’=1) levels are cold, but the CO(b ^{3}Σ^{+},v’=0) distribution is rotationally excited. The populations in the ^{12}CO(a’,v’=34 and 35) levels are transferred to CO(b,v’=0) by collisions with He and the rate constants are 0.4–1.0×10^{−10} cm^{3} s^{−1}. Emission spectra from the Kr(^{3} P _{2}) reaction identified ^{12}CO(a’,v’=23–26) and ^{13}CO(a’,v’=24–27) and CO(d ^{3}Δ,v’=20 and 21) for both ^{12}CO and ^{13}CO as important products; the CO(d ^{3}Δ,v’=20 and 21) states previously were identified by Tsuji and co‐workers. The vacuum ultraviolet spectra from the Kr(^{3} P _{2}) reaction with ^{12}CO and ^{13}CO showed that CO(A ^{1}Π) is a primary product and that it also is formed from CO(d ^{3}Δ) and CO(a’ ^{3}Σ^{+}) by collisions with He and Ar. The Kr(^{3} P _{2})+CO reaction also generates some unassigned CO triplet state emission. The propensity for formation of ^{3}Σ^{+} states rather than the e ^{3}Σ^{−} state of CO is discussed. An improved transition dipole function for the CO(b ^{3}Σ^{+}–a ^{3}Π) transition is presented in the Appendix.

Diffusion‐controlled reaction rate to an active site in an external electric field
View Description Hide DescriptionThe effect of an external electric field on the rate constant for diffusion‐controlled reactions between asymmetric reactants described by the model of Solc and Stockmayer is investigated. We propose an easy method to calculate the rate constant with any necessary accuracy and derive some analytical formulas which approximate it well for any angular size of the active site or either weak or strong electric fields. We also calculate the rate constant by applying the constant‐flux method of Shoup, Lipari, and Szabo and compare it with our results. It is shown that in the case of a weak electric field the rate constant slightly depends on the active site location but this dependence becomes significant for a strong electric field.

Evidence for quantum interference in collision‐induced intramolecular energy transfer within CO singlet–triplet mixed states
View Description Hide DescriptionThe quantum interferenceeffect associated with one type of radiationless transition, the collisional energy transfer between singlet–triplet mixed molecular states, is studied. The experiments are conducted on CO(A ^{1}Π,e ^{3}Σ^{−})–He(Ar) via the ultrasensitive optical–optical double resonance multiphoton ionization (OODR‐MPI) technique which measures state‐to‐state cross sections to an accuracy of ±10%, irrespective of the lifetime of the excited state. Distinct evidence of a quantum interferenceeffect on the transfer rate has been obtained for mixed state CO intramolecular energy transfer processes. A simple, explicit expression for the cross section for mixed state energy transfer, based on the first order Born approximation of time dependent perturbation theory, is derived. The use of a transition phase angle θ_{ST} is incorporated in the expression to describe the phase angle difference between singlet and triplet channels. This greatly refines the existing theory of quantum interference for collisional processes in that it successfully calculates the energy transfer cross sections and interprets the observed quantum interferenceeffect under various quantum transitions. Experimental values of θ_{ST} obtained for the CO–He system are 66° for J=9 and 73° for J=13. These results indicate that the infinite‐order‐sudden approximation which calls for θ_{ST}=0 cannot satisfactorily account for the quantum interferenceeffect. In addition, measured θ_{ST} values for CO–Ar system are consistent with our theoretical expectations.

Vibrational energy transfer from the 6^{1} level of S _{1} (^{1} B _{2u }) benzene in a supersonic expansion. I. Monatomic collision partners
View Description Hide DescriptionVibrational energy transfer has been monitored from the 6^{1} level of ^{1} B _{2u }(S _{1}) benzene seeded in rare gas supersonic free jet expansions at X/D=5, where the temperature is calculated to be ∼10 K. The monatomic collision partners helium, neon, argon, and krypton form the subject of this study. Consequently, transfer is limited to one mechanism, transfer of vibrational energy in benzene to translational energy of the collision pair. The vibrational energy transfer is followed using time resolved, dispersed fluorescence spectroscopy. While there are five possible destination levels, only three are found to be important. These are transfer to the 16^{2} level and transfer to the spectrally unresolved 11^{1} and 16^{1} levels. Negligible transfer is observed to both of the remaining two accessible levels, 0^{0} and 4^{1}. It is found that the branching ratio for the two destination channels is insensitive to the identity of the collision partner. The branching ratios are reproduced by calculations based on SSH‐T theory. The calculations suggest that the insensitivity of the branching ratio to the collision partner is fortuitous: while the combined 11^{1}/16^{1} channel retains approximately the same ratio to 16^{2} for all collision partners, the relative importance of the 11^{1} and 16^{1} levels themselves is collision partner dependent. Evidence is presented suggesting that there is significant rotational excitation accompanying the vibrational energy transfer in the case of heavy collision partners. This study, by establishing the behavior of vibration to translation transfer, forms the basis for further studies of the role of vibration to rotation and, subsequently, vibration to vibration mechanisms in vibrational energy transfer in benzene at low temperatures.

The rate of collisional quenching of N_{2}O^{+} (B ^{2}Σ), N^{+} _{2} (B ^{2}Σ), O^{+} _{2} (b ^{4}Σ), O^{+} (3d), O (3p), Ar^{+} (4p’), Ar (4p,4p’) at the temperature ≤200 K
View Description Hide DescriptionRate constants of collisional quenching of the states N_{2}O^{+} (B ^{2}Σ^{+} _{ u }), N^{+} _{2} (B ^{2}Σ^{+} _{ u }), O^{+} _{2} (b ^{4}Σ^{−} _{ u }), O i (3p ^{5} P), O ii (3d ^{4} F), Ar ii (4p’ ^{2} F ^{0}) in their parent gases and the states Ar i (4p,4p’[1/2]) in oxygen have been measured by studying Stern–Volmer dependencies in the temperature range from 20 to 200 K. Radiating states were excited by the electron beam. A flow on free jet axis was used as a gas target. Negative dependencies of rate constants on temperature which were obtained are consistent with the available experimental data on rate of internal energy and charge transfer at low energy collisions.

Improved quantum Monte Carlo calculation of the ground‐state energy of the hydrogen molecule
View Description Hide DescriptionWe report an improved Green’s functionquantum Monte Carlo calculation of the nonrelativistic ground‐state energy of the hydrogen molecule, without the use of the Born–Oppenheimer or any other adiabatic approximations. A more accurate trial function for importance sampling and the use of the exact cancellation method combine to yield an energy which is a factor of 10 more accurate than that of previous quantum Monte Carlo calculations. The energy is less accurate than that of recently improved analytic variational calculations. The calculated energy is −1.164 0239 ±0.000 0009 hartree. Expressed as the dissociation energy and corrected for relativistic and radiative effects, the result is 36 117.84±0.20 cm^{−1}, a value in agreement with the most recent experimental value 36 118.11±0.08 cm^{−1} obtained by Balakrishnan et al.

Density functional calculation of nuclear magnetic resonance chemical shifts
View Description Hide DescriptionA current‐density functionaltheory for the calculation of nuclear magnetic resonancechemical shifts is presented. If the Kohn–Sham orbitals are expanded in a finite basis set, one of the main problems is the strong dependency of the results with respect to a shift of the gauge origin of the vector potential which describes the external magnetic field. Two computational schemes implementing both the individual gauge for localized orbitals (IGLO) and gauge including atomic orbitals (GIAO) concepts, which overcome this problem by introducing distributed gauge origins, are presented in detail. A comparison of the density functional IGLO and GIAO schemes shows that IGLO is much more efficient if one neglects the current‐dependent part of the density functional (as is done in ‘‘uncoupled’’ density functional theory), but that this advantage is less pronounced in the full current‐density functional treatment.

Direct Monte Carlo simulation of chemical reaction systems: Dissociation and recombination
View Description Hide DescriptionWe report direct Monte Carlo simulations of a chemical reaction system with bimolecular and termolecular dissociation and recombination reactions of the type M+AB■M+A+B. The simulations are carried out at the molecular level using a simple flexible reaction model for termolecular reactions satisfying all the requirements of momentum and energy conservation, microscopic reversibility, and equilibrium. Energy transfer among reactants and products is included. The method is especially useful for treating reaction systems with nonequilibrium distributions and coupled gas dynamic‐reaction effects. For systems with thermally equilibrated reactants the observed behavior is identical to that predicted by conventional methods.

Semiclassical behavior at a quantum avoided crossing
View Description Hide DescriptionFor a polynomial potential with resonant fundamental frequencies (1:2 and 1:3 resonances), quantum avoided crossings can occur when quantum eigenvalues are plotted versus a parameter in the Hamiltonian. In the present paper, primitive (EBK) semiclassical behavior at the quantum avoided crossing is reinvestigated, using the exact analytical calculation of the action integrals, which was devised recently [Chem. Phys. 185, 263 (1994)] for an approximate resonance Hamiltonian that can be deduced from the exact polynomial Hamiltonian by low order perturbation theory. The previously reported behavior, that is semiclassical levels passing through the intersection instead of avoiding each other, is shown to happen if there exist two superimposed branches in the plot of the second action integral I_{2} as a function of the energy. These results are interpreted in terms of semiclassical diabatic basis and of quantum dynamical tunneling. In contrast, if the semiclassical system enters the (anti)crossing region with semiclassical quantum numbers I_{2} which do not lie on superimposed branches of the plot, it is shown that at least one, and possibly two, level(s) must cross the separatrix, that is pass from the inside to the outside of the resonance region (or conversely) in order to adapt to the quantum avoided crossing. This causes (i) corresponding semiclassical quantum number I_{2} to change (ii) the close correspondence between quantum and semiclassical mechanics to break down.

Rate coefficients for the N+O_{2} reaction computed on an ab initio potential energy surface
View Description Hide DescriptionAccurate calculations of the potential energy surface of the N+O_{2}reaction have been performed at complete active space self‐consistent field (CASSCF) and multireference single–double configuration interaction (MR‐SDCI) levels. Features of the calculated potential energy values are analyzed and compared with those of previous ab initio calculations and experimental data. The comparison has been extended to kinetic properties of the reaction.

High‐level ab initio prediction of the structure and infrared spectra of formaldehyde–water radical‐cation complexes
View Description Hide DescriptionIn a previous work we have identified two possible structures for the radical cation obtained by ionization of hydrogen‐bonded formaldehyde–water complexes [Coitiño et al., J. Am. Chem. Soc. 115, 9121 (1993)], a hydrogen‐bonded and an addition‐like complexes. We observed that the results were highly dependent on the method of calculation employed. Inclusion of correlation was crucial for obtaining the correct structures of some of the complexes. In this work we used high‐level ab initio calculations in order to predict the equilibrium structure of these two complexes, the possibility of its existence in gas phase, and the infrared spectrum to be expected in each case.
A series of progressively more sophisticated basis sets was used to assess the effect of the quality of the calculations on the expected results. Also, full geometry optimization with each basis set was performed at the second‐order Mo/ller–Plesset level, and correlation energy was calculated at the fourth‐order Mo/ller–Plesset level to assess the contribution of this factor to the global result. Confirming our previous results, we found that correlation affects the hydrogen‐bonded radical‐cation complex more than the addition one, due to the different bonding patterns in each of them. Both complexes are stable—toward decomposition to the fragments or to CO+H+H_{3}O^{+}—by several kcal/mol at all levels of theory. The hydrogen‐bonded complex is more stable than the additional one by a respectable amount (13 kcal/mol at the highest level used here), lending support to our previous analysis of the reactions of the former as the main channels for evolution of the formaldehyde–water radical cation. The H‐bonded complex [H_{3}O^{+}...HCO^{⋅}] presents two characteristics, very intense absorptions which should allow identification of this radical cation if present in the experimental setup. These transitions are also present in the HCO^{⋅} radical but their intensity is enhanced by an order of magnitude due to the coupling with the proton in H_{3}O^{+}. We conclude that the combination of stability and characteristic infrared transitions should make this radical‐cation complex a relatively easy target for experimental determination.

Charge–cavity and charge–dipole effects in ionic fluids
View Description Hide DescriptionThe contribution of charge–dipole interactions to the thermodynamics and structure of ionic solutions is considered from the standpoint of two models—a quasicontinuum cavity model of charged spheres of dielectric constant ε_{0} in a continuum of dielectric constant ε as well as a molecular model consisting of charged‐sphere solute particles and dipolar‐sphere solvent particles. With suitable scaling of the interaction potentials one expects to obtain the cavity‐model description from the molecular model in the limit σ_{ d }/σ_{ i }→0, where σ_{ d } and σ_{ i } are the ion and dipolar‐particle diameters, respectively. We recover for the first time the full r ^{−4} ‘‘cavity‐term’’ of the ion–ion cavity‐model pair potential in such a limit as well as a corresponding screened r ^{−4} term in the solvent‐averaged direct correlation function, which is screened by the factor e ^{−2λr } and multiplied by the prefactor (1+λr)^{2}, where λ is the ion–ion inverse screening length. Here, r is the distance between a pair of ions. We also show that if one makes the usual approximation of pairwise additivity of the n‐ion potentials of mean force at infinite dilution, one loses both the screening and the prefactor, recovering an earlier result in G. Stell, Phys. Rev. A 45, 7628 (1992). This is thermodynamically significant; if there were no screening of the r ^{−4} term, there could not be a critical point in the ionic fluid if ε_{0}<ε, while if ε_{0}≳ε, the critical point would be mean‐field‐like if there were no screening, as pointed out in the earlier work.