J. Phys. Chem. 1987, 91, 3150-3157
3150
ARTICLES Structure of the Hydrated Electron H. F. Hameka,’ Frank J . Seiler Research Laboratory, US.Air Force Academy, Colorado Springs, Colorado 80840
G. W. Robinson,*$and C. J. Marsden School of Chemistry, University of Melbourne, Parkville, Victoria 3052, Australia (Received: October 16, 1986; In Final Form: February 4, 1987)
An “intuitive” 4-water structure of the hydrated electron based on recent threshold photoionization studies is analyzed in light of past kinetic, thermodynamic, and spectroscopicmeasurements. It is concluded that the equilibrium state of the hydrated electron may actually be the complex (OH--H3O)(aq). The central feature is an H20- anion with a large distortion of one OH bond caused by the presence of the localized excess electron. A structure for the hydrated dielectron is also suggested. In these structures, thermodynamic stability is gained through hydration of the hydroxyl ion, while spectroscopic properties mainly arise from the oxonium radical. In order to examine these structures more extensively, quantum theoretical computations using GAUSSIAN 82 haw been carried out on (H20)N-with N = 1 and 4. The computations do suggest a propensity to form a single stretched OH bond when a localized excess electron is present. This feature is similar to conclusions reached by Bettendorff, Buenker, and Peyerimhoff from extensive CI calculations on the isoelectronic monomeric species H F , where two structures participate in an “in-out” dual stability: (1) a delocalized electron with a “localized” hydrogen atom (normal H-F bond), or (2) a localized electron with a “delocalized” hydrogen atom (distended H-F bond). In solution, solvent interactions are expected to lower the energy of the localized electron structure compared with the delocalized one.
1. Introduction
Excess free energy processes, such as radiolysis, have been the most popular source of solvated electrons’s2for study. However, if threshold photoionization3 (AG > 0) is used, a much greater specificity can be achieved with respect to interactions of the electron with the surrounding medium. These experimentsM have shown that the solvation of an electron in water, in companionship with its photodetachment from an excited-state precursor molecule, has a rate that is at least 2 orders of magnitude faster than that in the simple alcohols, while radiolytic sources of electrons produce each species almost equally well.’ The reason for this is that for threshold electrons the hydration reaction is fast enough to be able to compete efficiently with conventional radiative and nonradiative transitions from the excited state of the precursor, while in other solvents solvation is too slow to compete. These experiments furthermore have demonstrated that special solvent structural considerations are important in water. Rapid hydration of the electron requires exactly four water molecules except in those cases where a strong structure-perturbing ionic In fact, the kinetics data on group such as -SO3- is threshold electrons in water exactly parallel the data on threshold protons&I0from weak acid precursors dissociating in water, where a distinct molecular species H904+(aq)has long been implicated through c h e m i ~ a l , ~and ~ , ’t~h e o r e t i ~ a l ’studies. ~ Though molecular models for the hydrated electron having stoichiometry e-.(H20), have been postulated,2 up to now theoretical studies16-ls have indicated that such structures, under isolated conditions, are unstable with respect either to fragmentation of whole water molecules from the cluster or to autoionPermanent address: Department of Chemistry, University of Pennsylvania, Philadelphia, PA 19104. ‘Permanent address: Picosecond and Quantum Radiation Laboratory, P.O. Box 4260,Texas Tech University, Lubbock, TX 79409.
0022-3654/87/2091-3 150$01.50/0
ization of the electron. Furthermore, there are no experimental data on isolated, negatively charged water cluster^'^-^^ that particularly favor an H804- form. In fact, this is one of the prominent clusters never observed by Haberland et a1.,20though the negatively charged dimer was easily detected by their methcds. (1)Hart, E. J.; Anbar, M. The Hydrated Electron; Wiley-Interscience: New York, 1970. (2)Stein, G.In Hydrogen-Bonded Solvent Systems; Taylor and Francis: London, 1968,pp 87-97. (3)Robinson, G. W.; Lee, J.; Moore, R. A. In Ultrafast Phenomena IV Auston D. H., Eisenthal, K. B., Eds.; Springer-Verlag: Berlin, 1984;pp 313-316. (4) Lee, J.; Robinson, G. W. J. Chem. Phys. 1984,81, 1203. (5) Moore, R. A.; Lee, J.; Robinson, G. W. J. Phys. Chem. 1985,89,3648. (6)Lee, J.; Robinson, G. W. J. Am. Chem. SOC.1985,107, 6153. (7)Kenney-Wallace, G.A. In Photoselective Chemistry Jortner, J., Ed.; Wiley: New York, 1981;Part 2,pp 535-577. Kenney-Wallace, G. A.;Jonah, C. D. J. Phys. chem. 1982,815,2572. (8)Lee, J.; Griffin, R. D.: Robinson, G.W. J . Chem. Phys. 1985,82,4920. (9)Lee, J.; Robinson, G.W.; Webb, S. P.; Philips, L. A,; Clark, J. H. J . Am. Chem. Soc. 1986,108, 6538. (10)Robinson, G.W.; Thistlethwaite, P. J.; Lee, J. J . Phys. Chem. 1986, 90,4224. Lee, J.; Robinson, G. W.; Bassez, M.-P. J . Am. Chem. SOC.1986, 108, 7477. (1 1) Triolo, R.;Narten, A. H. J. Chem. Phys. 1975,63,3624. (12)Lau, Y. K.; Ikuta,S.; Kebarle, P. J. Am. Chem. Soc. 1982,104,1462. (13)Eigen, M.; De Maeyer, L. Proc. R. SOC.A 1958,247, 505. Eigen, M.; De Maeyer, L. In The Structure of Electrolytic Solutions Hamer, W. J., Ed.; Wiley: New York, 1959;pp 64-85. (14)Conway, B. E. In Modern Aspects of Electrochemistry; Bockris, J. O’M., Conway, B. E., a s . ; Butterworths, London, 1964;No. 3, Chapter 2. (15)Newton, M. D.J . Chem. Phys. 1977,67,5535. (16)Newton, M. D.J. Phys. Chem. 1975,79, 2795. (17)Rao, B. K.; Kestner, N. R. J. Chem. Phys. 1984,80, 1587. (18)Kestner, N. R.; Jortner, J. J. Phys. Chem. 1984,88,3818. Also see references cited here. (19)Armbruster, M.; Haberland, H.; Schindler, H . 4 . Phys. Reu. Lett. 1981,47, 323. (20)Haberland, H.; Ludwigt, C.; Schindler, H.-G.; Worsnop, D. R. J . Chem. Phys. 1984.81, 3742. (21)Coe, J. V.;Worsnop, D. R.; Bowen, K. H., to be published.
0 1987 American Chemical Society
The Journal of Physical Chemistry, Vol. 91, No. 12, 1987 3151
Structure of the Hydrated Electron These facts have kept deliberations about specific structural aspects of H804-to a minimum. Of course,in bulk liquid water, conditions for stability are undoubtedly far different than they are in an isolated small cluster. This paper reviews the fundamental kinetic and thermodynamic properties that specifically pertain to the structure of the hydrated electron and dielectron. It then makes a suggestion concerning these strucures, and it is shown how the suggested structures are consistent with the fundamental properties. Finally, results of quantum structural calculations on H20- and H804-that help support the proposed intuitive model are presented.
structures seems awkward, since a nearby, energetically unfavorable cavity must be created by a fluctuation each time the electron moves. Furthermore, these localized cavity structures could never display the exact parallelism discovered between the hydrated proton and the hydrated electron in the threshold ionization experiments, nor would it be quite as easy to imagine such structures leading to the rates and molecular features of the fundamental chemical reactions of the hydrated electron, e.g., reactions 1-4.
2. Some Clues
It is unlikely that an electron could strongly attach itself to an otherwise undistorted water m o l e c ~ l e ,even ~ - ~when ~ the stabilizing influence of hydration is considered. However, consistent with the H80[(aq) structure, and the other considerations described above, is a highly distorted H20- anion radical, hydrated by three other water molecules in a manner resembling the first hydration shell of OH-.10934A connection of this type was also made some time ago by Stein.2 In fact, such a structure would strongly resemble a hydrated OH- ion with the extra H atom bonded weakly to one of the neighboring water molecules as in the neutral oxonium radical. Considering the large hydration enthalpy, AHhyd = -1 13 kcal mol-’, and free energy, AChyd= -103 kcal mol-’, of the hydroxyl this similarity to OH- would be a great asset to the thermodynamic stability of the hydrated electron. The oxonium radical H 3 0 has been inferred previously40 as a transient state of the H atom in water. Quite recently, experimental data4’ on H30.(H20), ( N I3) and their deuteriated analogues have been reported. From these experiments it would appear that the formation of hydrated H30would not significantly decrease the thermodynamic stability of the proposed structure. These experiments and various theoretical calculations4* indicate that the ionization potential of oxonium lies between 3.4 and 4.68 eV, with an expected 3s 3p transition in the range 1.6-2.3 eV. Suggestively, this is very close to the intense absorption peak attributed to the hydrated electron (1.7 eV).27 Could the equilibrium state of the hydrated electron actually be a hydrated OH--.0H3 complex? Figure 1 shows a possible structure based on this kind of intuitive reasoning. This proposed structure fits nearly into, and becomes an integral part of, the liquid water network structure without the apparent energy cost of cavity models. The thermochemistry of the hydrated electron can be interpreted as conforming to the proposed structure, with much of the stability being gained from hydration of the OH- ion, which may be only slightly perturbed in the complex. In fact, a thermodynamic cycle can be constructed (kcal mol-‘, 298 K) illustrating this point:
One clue to the identity of H804-(aq) is the mobility22of the cm2 V-’ s-*, which is about 60% hydrated electron, 1.98 X of the mobility of H+ and is virtually identical with the mobility of OH-.23 Therefore, with the exception of H+ and OH-, the mobility of the hydrated electron is greater by a factor of at least 2.7 than the mobility of any other ion in water [cf. /l(K+) = 7.6 X 1O4 cm2 V-’ s-’] .24 The rapid diffusion of the elementary ions OH- and H+ in water is thought to proceed by a Grotthuss transfer25of H+ during rotational fluctuations of water molecules, which bring local H bonds into favorable configurations. Fundamental chemical reactions of the hydrated electron, and their corresponding rates2qZ6(q1-4) also offer clues to the identity
-
+ H30+ H + H,O + H 2 0 + H + OHe- + H H, + OHee-
e‘
+ e-
--
H2 + 2 0 H -
k l = 2.1
1
X 1Olo
,- 16 M-’ s-I
M-I s-I
1.8 x 107 M-1 k , = 2.5 X IO’O M-’ k4 = 0.6 X 1OIo M-’
(1) (2)
s-1
s-I s-I
(3) (4)
of H804-. All species in the equations are hydrated (aq) and k2/ is the rate constant for the reverse reaction 2. The optical spectrum2’ of the hydrated electron must always be considered in any definitive identification of its structure. The absorption spectrum has a maximum at 71 5 nm with a width at half-maximum of nearly 400 nm. The oscillator strength of the transition is 0.71, which represents a fully allowed transition. The spectrum at 715 nm is known to be sensitive to temperature, shifting to the red by about 23 cm-’ O C - ’ throughout the temperature range of normal liquid water.28 The spectrum shifts to the blue with increasing pressure (Av = 400 cm-I, 1 kbar).29 The optical spectrum is also sensitive to perturbations by dissolved ions.30*31These various results have been held up as evidence in support of a localized cavity model (