5582
J. Phys. Chem. B 2009, 113, 5582–5592
Charge Trapping in Imidazolium Ionic Liquids Ilya A. Shkrob* Chemistry DiVision, Argonne National Laboratory, 9700 S. Cass AVe, Argonne, Illinois 60439
James F. Wishart* Chemistry Department, BrookhaVen National Laboratory, Upton, New York 11973-5000 ReceiVed: December 30, 2008; ReVised Manuscript ReceiVed: January 29, 2009
Room-temperature ionic liquids (ILs) are a promising class of solvents for applications ranging from photovoltaics to solvent extractions. Some of these applications involve the exposure of the ILs to ionizing radiation, which stimulates interest in their radiation and photo- chemistry. In the case of ILs consisting of 1,3-dialkylimidazolium cations and hydrophobic anions, ionization, charge transfer and redox reactions yield charge-trapped species thought to be radicals resulting from neutralization of the constituent ions. Using computational chemistry methods and the recent results on electron spin resonance (ESR) and transient absorption spectroscopy of the ionized ILs, we argue that electron localization in the imidazolium ILs yields a gauche dimer radical cation with the elongated C(2)-C(2) bond. This species is shown to absorb in the near-infrared and the visible regions and accounts for the observed ESR spectra. We suggest that the excess electron in these aromatic ILs is localized as such a dimeric ion, and consider the chemical implications of this attribution. We also suggest that three-electron N-N bonding with the formation of a dimer radical anion occurs for amide anions, such as dicyanamide, when the parent anion traps holes; steric hindrance prevents the analogous reaction for bis(triflyl)amide anion. For another anion of practical importance, bis(oxalato)borate, a pathway involving the elimination of CO2 is suggested. Together, these results indicate the unanticipated tendency of the ILs to localize primary charges as radical ions as opposed to neutral radicals. Thus, it appears that secondary chemistry in the ionized ILs may be dominated by radical ion reactions, similarly to the previously studied conventional organic liquids, depending on the composition of the IL. 1. Introduction Room temperature ionic liquids (ILs) have emerged as a new class of solvents for practical applications due to their unique combination of low volatility, chemical stability, high conductivity, wide electrochemical windows, ability to dissolve organic and inorganic solutes and gases, and tunable solvent properties.1 One of these applications is in the processing of spent nuclear fuel,2-5 where the unusual properties of ILs provide advantages over conventional organic solvents. Since nuclear separations processes are frequently performed in high-radiation fields, an understanding of radiation damage in these IL-based solvent systems is required for their successful use.6-15 The same need exists for understanding the mechanisms of IL degradation in electrochemical applications, including batteries, photoelectrochemical cells,16,17 and electrodeposition of metals.18,19 To address the concerns about the stability of these new solvents, product analyses13,14 and mechanistic studies6-12,15 of radiationinduced IL degradation are currently pursued by several groups, whereas others study photoionization and charge-transfer-tosolvent (CTTS) reactions in these IL solvents.20-22 The latter reactions yield similar products as the radiation-induced ionization and electrochemical redox reactions, as these photoreactions also generate electrons and holes (electron deficiencies) in the solvent. Pulse radiolysis-transient absorption (TA) spectroscopy,6-12 nanosecond flash photolysis,20 femtosecond and picosecond TA laser spectroscopies,21,22 and matrix isolation * Authors to whom correspondence should be addressed. E-mail:
[email protected] (I.A.S.),
[email protected] (J.F.W.).
electron spin resonance (ESR)15 have been used to identify shortlived reaction intermediates in these ionized ILs, and a picture of the early events leading to the degradation of the solvent is beginning to emerge.12 It is already clear that this chemistry is different from both of the reference systems for this class of liquids, viz., molecular organic solvents and inorganic ionic solids. Despite this rapid progress, the insight gained from these spectroscopic methods has been limited because the chemical structure and expected properties of the postulated intermediates are insufficiently understood. Below, we use a computational approach to close this gap. From the outset, ionization in ILs has been expected to be different from the ionization of molecular liquids, as the primary charged species, the electrons and holes (electron deficiencies), can be trapped by the constituent ions (depending on composition) to yield neutral radicals. In organic solvents, the trapped charges are either excess electrons (cavity electrons) or radical ions resulting from electron and hole attachment to the molecules in the solvent (that may undergo subsequent reactions to yield neutral radicals). Due to this difference, the radiolysis is thought to be driven by reactions of these neutral radicals, such as crossrecombination, chain propagation, oligomerization, and disproportionation reactions, in contrast to redox and bond-addition reactions that are typical of molecular ions. Understanding of the structure and properties of these first-generation radical intermediates is required for putting the subsequent chemistry on a firm foundation. In the ILs consisting of aliphatic cations (such as quaternary ammonium, phosphonium, and pyrrolidinium cations), the
10.1021/jp811495e CCC: $40.75 2009 American Chemical Society Published on Web 03/26/2009
Charge Trapping in Imidazolium Ionic Liquids
J. Phys. Chem. B, Vol. 113, No. 16, 2009 5583
SCHEME 1: Structures of the Ions Considered in This Work
electron is localized and trapped as a cavity electron, which is analogous to F-centers in ionic crystals.23 The electron is trapped in anion vacancies and stabilized by Coulomb interaction with several cations forming the cavity and lining it with the aliphatic chains in the side arms.12 Conversely, the holes are trapped by the anions. For example, the ionization9 and/or photo-CTTS reactions of halide anions (X-) in tetralkylammonium halide ILs generate halogen atoms (X•) that react with the parent anion to yield X2-• anions (2Σu) that can be observed using TA.20-22 This reaction is analogous to the formation of hole (VK) centers in alkali halide crystals.23 In the ILs that consist of aromatic cations, such as 1-alkyl-3-methylimidazolium (Rmim+, see Scheme 1 above), the fate of the primary charges, the electron and the hole, is not yet established and theoretical insight is needed to identify the potential structures for the trapped-charge species and assist in their spectroscopic attribution. This is the goal of the present study. As these imidazolium ILs are some of the most widely studied IL solvents for radionuclide extractions2-5 and other practical applications,1,16-19 we have limited this study to such ionic liquids. Although there are many different anions that could be studied, we have selected three representative anions that have received significant attention in recent years: dicyanamide (DCA-), bis(trifluoromethylsulfonyl)imide (NTf2-, also called bistriflimide or bis(triflyl)amide), and bis(oxalato)borate (BOB-). The structures of these anions are given in Scheme 1. Electron trapping in imidazolium ILs presents a greater puzzle than electron trapping in the previously studied aliphatic ILs, because two different modes of such trapping are possible: (i) the formation of the cavity electron and (ii) electron attachment to the aromatic ring of the Rmim+ cation (Figure 1a) with the formation of a C(2)-centered radical (Figure 1b). In dilute (
