摘要:
Paper Dynamics of cluster anions: a detailed look at condensed-phase interactions Andrei Sanov*a and W. Carl Linebergerb aDepartment of Chemistry, University of Arizona, Tucson, Arizona 85721-0041, USA. E-mail: sanov@u.arizona.edu bJILA and Department of Chemistry and Biochemistry, University of Colorado, Boulder, Colorado 80309-0440, USA. E-mail: wcl@jila.colorado.edu Received 20th September 2002, Accepted 5th November 2002 First published as an Advance Article on the web 19th November 2002 This Perspective reflects on several recent advances in the studies of structure and dynamics of cluster anions, bridging the gap between ‘cluster’ and condensed phases. Applications involving photofragment and photodetachment spectroscopy, as well as femtosecond time-resolved experiments, are described.Special emphasis is given to the effects of microscopic solvation on the electronic structure and reactivity of negative ions in heterogeneous and homogeneous cluster environments. Some recent breakthroughs in experimental methodology are also outlined, in particular the application of photofragment and photoelectron methods and the imaging technique to the studies of molecular cluster anions. 1. Introduction Clusters have long fascinated chemists with the unique opportunities they offer for studies of intermolecular interactions implicated in practically all areas of chemistry, particularly in solvation. Ionic clusters are especially appealing for experimental studies of reactions, as in their well-defined microscopic environments the chemical dynamics can be examined at an unprecedented level of detail.Clusters are used as model micro-solutions that simplify the interpretation of solvation effects at a molecular level.1–4 Such detailed analysis is difficult to achieve in macroscopic media, such as liquids, solids, or high-pressure gases, because of the poorly defined solvent coordination and the very large number of pairwise interactions involved. These characteristics of condensed environments usually require an averaged, i.e. statistical, rather than dynamical, description of bulk properties. Despite their relatively small size, clusters retain many characteristics of the bulk media that make the condensedphase dynamics as rich as they are. One central question in cluster studies is: how much matter is needed for physical laws associated with bulk materials to be applicable? The multitude of intermolecular interactions in clusters affords chemical reactions a variety of pathways and mechanisms that attracted the interest of many generations of chemists.The origin of the field of cluster chemistry can be traced back to the early studies of colloids, aerosols, and nucleation phenomena in the midnineteenth century, followed by an explosive growth during the twentieth century. For decades, many experimental and theoretical studies have been directed at deepening the understanding of detailed mechanisms of reactions within clusters. This paper is not intended to offer a comprehensive review of the vast field. Its scope is limited to recent photodissociation and photodetachment studies of cluster anions, giving a perspective of the dynamics of photoinduced reactions in cluster anions and the bridge between the ‘cluster’ and condensed phases.A broader outlook on the advances in cluster research during the later part of the twentieth century can be found, for example, in the review by Castleman and Bowen.1 In particular, we concentrate on several recent advances in the quest for better understanding of solvent-induced effects on the electronic structure and reaction dynamics. The most obvious effect of solvation is the limitation of physical space available to the solute, whose dynamics are constrained by physical barriers imposed by the presence of other bodies. In addition, the solvent acts as an energy bath, opening a pathway for internal relaxation of the solute.However, the most challenging aspect of solvation is the perturbation of the electronic structure of both the solute and the solvent. We discuss the role of solvent-induced perturbations of the electronic structure in determining the outcomes of chemical reactions. These perturbations are particularly important in reactions involving non-adiabatic transitions. It is emphasized that the strong interactions implicated in ionic solvation often cannot be considered as merely a ‘perturbation’ in the perturbation-theory sense. If the strength of these interactions is comparable to the bonding in the solute or/and the solvent, chemical transformations are possible and the structure of the cluster core may differ greatly from the corresponding unsolvated species.The cumulative ion–solvent interactions in ionic clusters can be easily comparable to chemical bonding, lending the solvent a major role in determining the reaction outcomes. For example, I2 2, one of the most studied ionic chromophores,5–24 has a bond dissociation energy of 1.01 eV, compared to a typical solvent binding energy of y0.2 eV per solvent molecule.21,25 Clearly, the collective effect of the solvent on the electronic structure of I2 2 cannot be viewed as a minor perturbation even in moderate-size clusters. Therefore, although the term ‘perturbation’ is used widely in this Perspective, it is often assumed to imply a considerable effect rather than a minor modification of state potentials and their couplings.Different types of solvent-driven reactions require various degrees of such perturbation. In many cases, the perturbation present within a cluster is rather large, although the reaction could proceed, in principle, even if it were minimal. One example is the solvent-induced recombination, or caging, of photofragments.5,6,17,20–53 The fundamental appeal of this process lies in that it involves both the breaking and remaking of chemical bonds, both occurring under the influence of the solvent. The classic I2 2 caging reactions5–7,21,45,46,49 evolve on potential energy surfaces correlating with the lower I2 2 dissociation limit yielding the neutral I fragment in the ground spin–orbit state. In this case, the I(2P3/2) 1 I2 Perspective DOI: 10.1039/b209239e PhysChemComm, 2002, 5(25), 165–177 165 This journal is # The Royal Society of Chemistry 2002recombination proceeds via a mechanism54–56 common for caging in both neutral and ionic, gas and condensed39,57–59 phase systems.According to this mechanism, the fragment separation is halted by the solvent that absorbs the translational energy. The fragments then recombine following the conversion at large separation to the ground I2 2 electronic state. Subsequently, the recombined I2 2 undergoes vibrational relaxation as the energy is transferred to the solvent. Although the I2 2 electronic structure is greatly perturbed by the solvent,60–62 only modest solvent-induced coupling between electronic states is actually needed, as the states in question are asymptotically degenerate.In the framework of this straightforward mechanism, the solvent effects are overwhelming in magnitude compared to the ‘probe’ used to monitor them (i.e., the caging process). Therefore, it is not surprising that one finds the recombination dynamics not to be overly sensitive to the details of cluster structure. It is possible, however, that a much larger degree of solventinduced perturbation of the solute may be required for the reaction to be possible in principle. A new type of caging reaction was discovered recently,22,24,63 in which the I2 2 chromophore is dissociated via the I*(2P1/2) 1 I2(1S) channel. In this case, a simple reversal of fragment trajectories cannot result in recombination, because the I*(2P1/2) 1 I2(1S) potentials are not bound. Nevertheless, caging on the ground electronic state of I2 2 is observed, which requires electronic quenching of I*(2P1/2) to precede the recombination. This crucial transition involves an energy gap of nearly 1 eV and is known to be extremely slow when occurring via radiative or collisional energy transfer mechanisms.Not surprisingly, this type of caging was found to exhibit more sensitivity to the details of solvent-shell structure and dynamics. Many laboratory tools are available today for the studies of ionic clusters. Among the most basic probes of cluster-ion structure and dynamics are mass-spectroscopy, photoelectron spectroscopy,64,65 photofragment spectroscopy, and several relatively new ultrafast pump–probe techniques. The field made a jump towards better-controlled and more comprehensive experimental studies with the introduction in the 1980s of the pulsed cluster-ion techniques,3 coupled with advances in tandem time-of-flight mass-spectroscopy.Lineberger-type pulsed ion sources, now prevalent in many laboratories around the world, allow the preparation and subsequent mass-selection of internally cool ionic clusters of known composition and often predictable structure. Using a combination of photoelectron and photofragment spectroscopic techniques, the reaction dynamics in cluster ions can be studied on the molecular level, bridging the gap between the properties of isolated molecules and chemistry and physics of condensedphase environments. The next revolutionary experimental tool introduced into the field in the late 1980s and early 1990s is the coupling of ultrafast pump–probe techniques with photoelectron spectroscopy and photofragment measurements.Whereas the experiments in the frequency (energy) domain help characterize the chemical and electronic structure of the cluster anions studied, the addition of the femtosecond pump–probe delay coordinate allows to study the reactions in real time, putting a truly dynamical emphasis on the experiments. The Lineberger group demonstrated the utility of ultrafast photofragment spectroscopy in the studies of cluster ion dynamics,20,66 while Neumark et al. developed a time-resolved variant of negative-ion photoelectron spectroscopy (PES), know as femtosecond PES, which allows observation of reactions through the window of evolving photoelectron spectra.46 These developments became possible with the advent of powerful and versatile femtosecond lasers.Other important trends in the field of cluster anion dynamics involve the move towards more detailed approaches to photoelectron and photofragment measurements, such as detecting several particles (e.g., photoelectrons and photofragments) in coincidence,67–73 infrared spectroscopic characterization of the cluster ion vibrational modes and structures,74,75 the photoelectron imaging76–78 approach to negative-ion PES.79–83 In this Perspective, we discuss several recent benchmark studies of cluster anion structure and dynamics. Section 2 describes the predicted structures of heterogeneous cluster anions I2 2(OCS)n and I2 2(CO2)n. Section 3 explores the caging dynamics in these clusters, with a special emphasis on the solvent-induced spin–orbit relaxation occurring on a remarkably fast timescale.Section 4 presents an outlook on the structure and photochemistry of homogeneous cluster anions, using (OCS)n 2, (CO2)n 2, and others as benchmark systems. Section 5 gives a brief summary of the paper and outlines future directions. 2. Structures of heterogeneous cluster anions The structure of the solvent shell about the ionic core of the cluster plays a determining role in its dynamics. To illustrate the basic principles involved in shell formation, we discuss the qualitative aspects of building up the first shell of OCS or CO2 molecules around an I2 2 cluster core. The structures of I2 2(OCS)n and I2 2(CO2)n have been examined using molecular dynamics and a model-Hamiltonian approach in a number of publications by the Parson group.22,35,62,84,85 Their findings are in agreement with a number of experimental observations, some of which are highlighted in the following sections.In the I2 2(OCS)n cluster ions the binding of the solvent to the anionic core is dominated by charge–dipole interactions between I2 2 and OCS, whereas in I2 2(CO2)n, in the absence of a permanent CO2 dipole moment, the charge–quadrupole interaction takes center stage.84,86,87 The solvent–solvent interactions also play important roles in both structure and dynamics. For example, the known fact that carbon dioxide forms dry ice implies strong interactions within the CO2 solvent shell. Since the charge in unsolvated I2 2 is equally divided between the two I-atoms, electrostatic considerations dictate that in stepwise solvation the first solvent molecule is expected to bind near the waist of I2 2.Thus, in I2 2?OCS the positively charged S-end of the solvent points toward the I2 2 center of mass. In larger I2 2(OCS)n clusters, solvent–solvent interactions are added and the above solvent orientation motif is modified to accommodate these interactions as well. The most revealing minimum-energy I2 2(OCS)n structures, determined by the Parson group, are shown in Fig. 1.22 The first five OCS molecules were found to arrange themselves around the waist of I2 2 with the sulfur atoms about 3.6 A° from the I2 2 center of mass. The solvent molecules are tilted at about 117u, rather than pointing directly outward, as best seen in Fig.1(a). In I2 2(OCS)5, the five OCS molecules complete a ring around I2 2. The next five molecules form a second ring around one end of the solute, then a single OCS molecule fills the axial site, ‘‘capping’’ that end of the cluster and completing the halfshell structure of I2 2(OCS)11 shown in Fig. 1(b). It is notable that I2 2(OCS)11 corresponds to a prominent ‘‘magic number’’ in the I2 2(OCS)n cluster mass-spectrum, i.e., the n ~ 11 peak clearly stands out in intensity compared to its neighbors.22 This observation supports the particularly stable structure of this cluster predicted by the Parson model. The other side of the I2 2(OCS)n cluster shell is filled in a similar fashion, completing the first solvation shell with 17 OCS molecules [see Fig.1(c)]. In I2 2(CO2)n, the charge–quadrupole interactions with the cluster core result in CO2 molecules lying ‘‘flat’’ with respect to I2 2. Parson and co-workers have shown that the CO2 molecules tend to first pack together on one side of I2 2, as shown in Fig. 2 for I2 2(CO2)5,36,62,84,88 rather than form a ‘ring’ [compare the structure of I2 2(CO2)5 to I2 2(OCS)5 in 166 PhysChemComm, 2002, 5(25), 165–177Fig. 1(a)]. This type of packing maximizes the interactions within the CO2 solvent shell itself. To the contrary, the ring structure of an OCS shell accentuates the strength of solvent– solute interactions, which are the strongest near the waist of the solute. 3. Photofragment caging dynamics in cluster ions The photodissociation and recombination dynamics of I2 2 have been studied in both bulk liquids, such as water and alcohols (Barbara and co-workers),5,6,45 and gas-phase clusters (Lineberger and Neumark groups).20–25,48–50,53,66,89 Kondow and co-workers examined collisions of I2 2(CO2)n cluster ions with solid surfaces and observed a wedge-type splitting of the I2 2 bond by solvent CO2 molecules upon cluster-surface collisions.10,11,90,91 The dynamics of ionic clusters have been compared to the studies of I2 recombination in solid rare-gas matrices by the Apkarian group.39,57–59 Several groups provided extensive theoretical background for understanding the dissociation and recombination in clusters.15,16,34,36,62,63,88,92 In particular, in the analysis of Parson and co-workers, the photofragmentation and caging are mediated by couplings of electronic states with differential charge character that is caused by interactions with the solvent.36,47,62,88,93 It was shown that differential solvation of electronic states can lead to isoenergetic curve-crossing regions playing key roles in the relaxation and recombination dynamics.We provide a perspective of the dynamics involved in caging by considering the photodissociation and recombination of I2 – in clusters of the OCS22–24,53 and CO2 20,21,25,48,49,66,94–96 solvents. Together with the studies of I2 2 in N2O89 and Ar,46,49,50,97 these experiments highlight the structural and dynamical effects of closed-shell solvents with different electrostatic properties. The original studies20,21,25,48,49 of I2 2 caging in clusters utilized the A’ 2Pg,1/2 B X 2S1 u,1/2 transition centered near 790 nm to promote the dissociation of the chromophore.This transition, yielding the I2 1 I(2P3/2) asymptotic products, is indicated by a red vertical arrow in Fig. 3, which shows the unperturbed I2 2 potentials calculated by Parson et al.88 Later, striking new dynamics22,24 were observed in the I2 –(OCS)n and I2 2(CO2)n cluster ions excited at 395 nm via the B 2S1 g,1/2 B X 2S1 u,1/2 transition in the I2 – chromophore (indicated by a blue vertical arrow in Fig. 4). This excitation accesses an electronic state correlating with the I– 1 I*(2P1/2) dissociation limit, where I* indicates a spin–orbit-excited atomic fragment. The dynamics observed at 790 and 395 nm are quite different.In a qualitative (and rather simplistic) description of the reaction following the near-IR excitation (Fig. 3), one might consider the solvent as playing a dual role. First, it acts as a physical barrier blocking the exit channel for Fig. 1 Calculated minimum-energy structures of I2 2(OCS)n, n~5, 11, and 17. Each structure is shown from two different angles, viewed perpendicular and along the I2 2 bond (left and right columns, respectively). These structures correspond to the isomers selected out of many nearly isoenergetic solvent configurations. While the general manner of solvent packing around I2 2 is reproduced in all low-energy isomers, local structural details, such as the relative orientation of neighboring OCS molecules, may vary.Adapted from ref. 22. Fig. 2 Calculated minimum-energy structure of I2 2(CO2)5 as viewed along the axes perpendicular (left) and parallel (right) to the I2 2 bond. Adapted from ref. 36. Fig. 3 Potential energy diagram of I2 2 (from ref. 88) and solvent loss energetics in 790 nm experiments on I2 2(OCS)n or I2 2(CO2)n. Fig. 4 Potential energy diagram of I2 2 (from ref. 88) and solvent loss energetics in 395 nm experiments on I2 2(OCS)n or I2 2(CO2)n. PhysChemComm, 2002, 5(25), 165–177 167photofragment escape from the cluster, thus triggering recombination. Second, it acts as an energy bath, enabling the relaxation of the caged chromophore and the cluster as a whole by evaporation of solvent molecules. The right-hand side of Fig. 3 illustrates the solvent loss energetics involved in the process.As each solvent molecule (OCS or CO2) is bound to the cluster by y0.2–0.25 eV, the complete dissipation of the 790 nm photon energy requires the evaporation of y6 solvent molecules. Alternatively, if the I2 2 bond is not re-formed, only y2–3 solvent molecules are lost by the I2(OCS)k fragments in the uncaged channel. Although the true mechanism has been shown to be more complex,36,62,84,92 this simplified picture does provide an initial understanding of the caging dynamics. This picture’s seeming simplicity stems from the fact that it gives no significant account to the solvent-induced perturbation of I2 2 electronic structure. But at the same time, little perturbation is actually needed to envision the recombination process. The minimum perturbation required is that which would enable transitions from the bright state of I2 2 back to the ground electronic state, on which the recombination process ultimately terminates.Given that the A’ state as well as two other excited states (A and a) are asymptotically degenerate with the X state, the required transitions can occur in the exit channel even with a minimum solvent-induced perturbation providing the state coupling. Thus, one obtains a tutorial (albeit somewhat misleading) picture of the reaction by considering the dissociation of largely unperturbed I2 2 confined inside the ‘‘solvent walls.’’ The situation is dramatically different at 395 nm (Fig. 4). The B 2S1 g,1/2 B X 2S1 u,1/2 transition in bare I2 2 leads to dissociation exclusively on the spin–orbit excited I2 1 I*(2P1/2) asymptote with a translational energy release of about 1.2 eV.No products are formed on the lower I2 1 I(2P3/2) asymptote.24 Thus, given the negligible probability of I* quenching in collisions (y1027–1028 per collision),98 the unperturbed electronic-state diagram in Fig. 4 predicts that practically no I2 2 caging is possible following the near-UV excitation. The experiment, however, testifies to the contrary,22,24 which raises the main question: Given the highly inefficient quenching of I* in collisions, how do the dissociating I2 1 I*(2P1/2) fragments find their way to the lower spin–orbit asymptote, on which not only the recombination, but also the uncaged channel labeled B in Fig. 4 can evolve? The quenching of spin–orbit excitation was found to be surprisingly efficient in the I2 2(OCS)n and I2 2(CO2)n cluster ions.In general, three distinct pathways are observed (as labeled in Fig. 4): (A) the ‘‘uncaged’’ I2(OCS)k products formed in coincidence with the excited I*(2P1/2) fragments, which are ejected from the cluster; (B) the ‘‘uncaged’’ I2(OCS)k products formed in coincidence with the quenched I(2P3/2) fragments; and (C) the ‘‘caged’’ I2 2(OCS)k products. The critical step in channels B and C is the quenching of the spin–orbit excitation. The competition between channels A–C is a delicate probe of the solvent-induced couplings between electronic states, which make the spin–orbit relaxation possible. In all three channels A–C, the excess energy is removed from the cluster by ejecting (n – k) solvent molecules.Additionally, in channel A, almost 1 eV is carried away in the form of I* spin–orbit excitation. As a result, the size distribution of the uncaged I2(OCS)k products is in general bimodal, compared to a single-modal distribution of the caged I2 2(OCS)k fragments.22,24 The typical number of solvent molecules evaporated in each channel is indicated on the right-hand side of Fig. 4. We continue our discussion of the dynamics in clusters by considering the cluster size-dependent caging probabilities. This is followed by the time-resolved dynamics of caging monitored in femtosecond pump–probe experiments. Caging probabilities and their structural implications First, consider the fraction of caged fragments observed in 790 nm dissociation of I2 –(OCS)n. This fraction, referred to as the caging probability, is plotted in Fig.5 (open symbols) as a function of the number of solvent molecules.22 No caged products are observed for clusters with n v 3, while for larger clusters the caging probability increases nearly monotonically with n, until a 100% caging is observed for n ¢ 17. The theoretical simulations described in Section 2 predict that 17 OCS molecules comprise a complete solvent shell around I2 2 [Fig. 1(c)]. This is consistent with the experimental observation that I2 2(OCS)17 is the smallest cluster anion with the OCS solvent for which a 100% recombination of the chromophore is observed. Despite the different nature of the dominant interactions in I2 2(OCS)n and I2 2(CO2)n, the 790 nm results for I2 2(CO2)n, also shown in Fig.5 (filled symbols),49 are disappointingly similar to the findings for I2 2(OCS)n. Complete 790 nm caging in I2 2(CO2)n is observed for n ¢ 16, consistent with the first solvent shell closing at n ~ 16–17, as predicted by the Monte- Carlo simulations.62,84 The monotonic increase in 790 nm caging probability and the lack of significant differences between the OCS and CO2 solvents are consistent with the simplified picture of caging given in the introduction to this section. Even small perturbations of the I2 2 electronic structure would be sufficient to couple the asymptotically degenerate states, correlating to the lower dissociation limit. Therefore, the intricate details of the solvent–solute and solvent–solvent interactions are not too important.The main role of the solvent in this case indeed appears to be that of a physical obstacle in the exit channel of I2 2 dissociation and an energy bath. In this light, increasing the number of solvents bound to the chromophore naturally results in a monotonic increase in the caging probability. Now consider the entirely different dynamics observed in the 395 nm experiment. The corresponding fraction of caged fragments is plotted in Fig. 6(a) for both I2 2(OCS)n and I2 2(CO2)n cluster anions. As expected, the onset of 395 nm caging is observed in larger clusters, compared to 790 nm. However, the caging probability itself is no longer the only parameter characterizing the channel competition. From a dynamical viewpoint, perhaps of even greater interest is the probability of spin–orbit relaxation induced by the solvent.Regarding the diagram in Fig. 4, the spin–orbit relaxation is prerequisite for channels B and C. Therefore, the probability of quenching is given by adding together the respective branching ratios for the spin–orbit relaxed uncaged channel (B) and the Fig. 5 Probabilities of recombination (caging) of the I2 2 chromophore in I2 2(CO2)n and I2 2(OCS)n clusters following photoexcitation at 790 nm, as functions of the parent cluster size. Data from ref. 22 and 24. 168 PhysChemComm, 2002, 5(25), 165–177caged channel (C). The thus obtained spin–orbit quenching probability is plotted in Fig. 6(b) for both I2 2(OCS)n and I2 2(CO2)n as a function of n. The complete caging is not achieved for any of the parent clusters studied (e.g., up to 26 solvent OCS molecules, for which the caging probability is 0.98). Still, recombination is the dominant reaction pathway for parent clusters with n ¢ 17.What sets the 395 nm caging probability curves [Fig. 6(a)] aside from the corresponding 790 nm curves (Fig. 5) is their structured nature. The 395 nm caging curve for the CO2 solvent is not monotonic in the range of n ~ 11–16, while the corresponding curve for the I2 2(OCS)n cluster ions exhibits a plateau in the same approximate range (n ~ 12–16), followed by a step-like nearly three-fold increase at n ~ 17. Notably, n ~ 17 is the smallest I2 2(OCS)n cluster size for which 100% caging is observed at 790 nm and it corresponds to a cluster with a predicted complete solvent shell.The sharp increase in caging upon addition of the 17th OCS molecule was attributed to a steric effect: the 17th molecule occupies the only remaining open site at the end of I2 2 [see Fig. 1(c)]. The occupation of this site closes the last collisionfree escape route for I2 2AI2 1 I dissociation. The presence of solvent on the I2 2 dissociation coordinate also increases the likelihood of non-adiabatic quenching of I*, which is prerequisite for recombination. Another prominent feature of the I2 2(OCS)n caging and spin–orbit quenching probability curves in Fig. 6 is the plateau at n~12–16. No such plateau was observed for I2 2(CO2)n, for which both caging and quenching in the same cluster size range exhibit seemingly erratic behaviors.It is revealing that the I2 2(OCS)n plateau consists of exactly five cluster sizes, reminiscent of the theoretical prediction that the first OCS solvent shell around I2 2 consists of three five-membered OCS rings plus two end molecules ‘‘capping’’ the cluster (Fig. 1). The plateau thus corresponds to the formation of the third solvent ring around the chromophore. The lack of a similar plateau for I2 2(CO2)n is consistent with the qualitatively different structural motif of I2 2(CO2)n, which is not based on solvent rings (see Fig. 2). The suggested mechanism of the spin–orbit quenching and recombination is discussed following the examination of the timescales on which these processes transpire. Time-resolved dynamics of caging The application of time-resolved techniques to chemical processes occurring on a femtosecond time scale has been one of the most important developments in reaction dynamics during the past decade.46,77,99–105 In particular, the application of femtosecond pump–probe spectroscopy to cluster anions allowed examination of the relaxation and energy-transfer processes at unprecedented levels of detail.The dissociation of the I2 2 chromophore within a cluster destroys the cluster’s ability to absorb visible/near-IR light, resulting in transient bleaching, while the ensuing recombination revives the absorption cross-section. A second photon can probe either the A’ 2Pg,1/2 B X 2S1 u,1/2 or the a 2Pu,3/2 B A 2Pg,3/2 transition.21 The transient bleaching and absorption recovery thus provide a way for examination of real-time dynamics of caging by monitoring the delay-dependent yield of two-photon products in a pump–probe experiment.Such measurements using a 720–790 nm pump and probe were performed (among others)21,49,89 on I2 2(CO2)n and I2 2(OCS)n cluster ions.21,23,24,49,53 The experiments revealed picosecond time-scales of the recombination and subsequent relaxation. For example, Fig. 7(a) shows the absorption recovery curves for I2 2(OCS)7 and I2 2(OCS)17 obtained in 790 nm pump– probe experiments. In this case, the positive and negative delays correspond to the reversal of the order of the identical pump and probe photons and therefore convey the same dynamical information. The near disappearance of signal at a zero delay reflects the bleaching of the probe absorption due to the dissociation of the chromophore by the pump photon.The fast (y2 ps) rise is absorption recovery observed following 790 nm excitation of I2 2(OCS)17 is attributed to the initial recombination of the I(2P3/2) and I2 fragments. In this delay range, the probe photon is absorbed by I2 2 in one of the excited electronic states or a highly excited vibrational level of the ground state.48,92 The 2 ps timescale corresponds to the period of the pseudo-vibrational I??I2 motion in the system excited above its dissociation threshold but constrained by the solvent. The bump appearing in the I2 2(OCS)17 absorption recovery curve at 2 ps is characteristic of the coherent I??I2 motion23,48 within the cluster. In I2 2(OCS)7 the solvent cage is smaller, the dynamics are correspondingly slower and not so much coherent, and the 2 ps peak does not appear. The longer timescale dynamics in both I2 2(OCS)7 and I2 2(OCS)17, characterized by the pump–probe signal leveling off after y20 ps, reflect the internal relaxation of the caged chromophore.An alternative perspective of these dynamics is provided by time-resolved photoelectron spectroscopy. While the photofragment measurements, by their very definition, focus primarily on the nuclear degrees of freedom, Neumark and co-workers pioneered an experimental approach that shifts the emphasis to the evolving electronic structure.46 Femtosecond photoelectron spectroscopy of cluster anions is used to probe the transient states and changing environment of the chromophore or its fragments by recording transient spectra of the electrons detached from the excited cluster with a delayed UV probe pulse.When these measurements are carried out on the isolated Fig. 6 (a) Probabilities of recombination (caging) of the I2 2 chromophore in I2 2(CO2)n and I2 2(OCS)n clusters following photoexcitation at 395 nm, as functions of the parent cluster size. (b) Similar curves for the probability of I* spin–orbit quenching. Data from ref. 22 and 24. PhysChemComm, 2002, 5(25), 165–177 169chromophore, the time-dependent photoelectron spectra reflect the timescale of the dissociation process.17–19 In the experiments on I2 2 embedded in Ar or CO2 clusters,46,94,97 the transient spectra reveal that after the charge localizes on one of the chromophore fragments, the photoelectron bands exhibit a varying energetic shift due to interactions with the solvent.In small (e.g., n ~ 6) Ar clusters, this shift persists for y1 ps, which is the time required for I2 to escape from the cluster.46 Parson group’s molecular dynamics calculations that model the solvent interactions with the localizing charge predicted shifts in electron affinity of the cluster that agree well with Neumark’s measurements.47,93 In larger clusters, the transient photoelectron spectra reveal the recombination of I2 2 in both ground and excited electronic states, followed by vibrational relaxation and solvent evaporation.94,97 The timescale of these processes are consistent with the timescales of caging observed by Lineberger and co-workers.25,49 To summarize, the fast rise in the I2 2(OCS)17 absorption recovery signal seen in Fig.7(a) during the first couple of picoseconds reflects the timescale for I2 2 recombination on the lower spin–orbit asymptote. This approximate timescale is typical of I2 2 recombination in several molecular solvents (CO2, N2O, OCS). It is concluded that it takes y2 ps for the solvent shell to reverse the I2 1 I dissociation trajectories and direct the fragments towards recombination. Similar photofragment measurements carried out with a 395 nm pump examined the dynamics on the upper spin–orbit asymptote of dissociating I2 2. The qualitative zeroth-order picture of this process is illustrated in Fig. 4. The recovery curves for I2 2(OCS)17 and I2 2(OCS)24 in Fig. 7(b) reflect the cumulative timescales of spin–orbit relaxation and recombination.For I2 2(OCS)24, the initial rise in absorption recovery occurs on a timescale ofy2 ps, similar to the period of solventinduced coherent I??I2 motion observed following 720 or 790 nm excitation.21,23,24,48,49 In the smaller I2 2(OCS)17 cluster, the recovery is slower, similar to the 720–790 nm results for smaller clusters. To emphasize this analogy, compare the absorption recovery curves in Figs. 7(a) and (b). The cluster sizes in (a) and (b) are purposefully different, as they were selected for the similarity of the respective timescales. (Qualitatively, adding extra solvent molecules counter balances the effect of doubling the energy pumped into the cluster at 395 nm, compared to 790 nm.) Despite the differences in detail, the timescales of caging following 395 and 790 nm excitations are very similar, indicating that the spin–orbit relaxation step implicated in Fig.7(b), but not in Fig. 7(a), must be fast on the overall timescale of the reaction. Solvent-mediated charge transfer as a fast spin–orbit quenching mechanism The detailed analysis of 395 nm caging dynamics by the Lineberger and Parson groups shows22,24,63 that the quenching of spin–orbit excitation, followed by I2 2 recombination, is only possible because of the strong perturbation of the I2 2 electronic structure by the solvent. Other mechanisms, failing to consider explicitly the perturbed I2 2 potentials (e.g., radiative decay or collisional quenching), have been ruled out based on experimental and theoretical evidence.24 Here, we outline the solventasymmetry mediated electron transfer model, first suggested by Maslen et al.60 and developed by Delaney et al.,63 that has been accepted as an accurate view of the reaction.The drawback of the collisional mechanism of I* quenching is that it considers the I2 fragment as a mere spectator. This strategy fails in the cluster ion environment. As another clue calling for a different interpretation, I* quenching on a picosecond timescale has not been observed in neutral environments. Thus, the proximity of I2 and the perturbed electronic structure of the I2…I* system are key to understanding the relaxation mechanism. The energy gap between the two spin–orbit asymptotes in Fig. 4 can be bridged by the effects of solvation.Because of the substantial binding energy of OCS and CO2 to a negatively charged cluster (y0.2 eV per molecule),21,22,25 the relative electronic state energies are greatly affected by state-specific charge distributions and solvent asymmetry. In the example in the top portion of Fig. 8, the I2 1 I(2P3/2) and I2 1 I*(2P1/2) electronic states are separated by 0.93 eV in the unsolvated I??I2 system (Fig. 8, top left). These states are degenerate with respect to switching the fragment positions (i.e., I2 1 I versus I 1 I2, and I2 1 I* versus I* 1 I2). In the cluster environment, this degeneracy is lifted by asymmetric solvation of the fragment pair (Fig. 8, top right). If the charge is localized at the more solvated end of the cluster (e.g., X?I2??I, where X denotes the collective solvent), the corresponding state energy is lowered significantly.On the other hand, if the charge is localized on the least solvated end (e.g., X?I??I2), the solvation effect is less significant. If the energetic difference between the two charge/solvent configurations, termed the differential solvation energy (DW), is close to 0.93 eV, the X?I2??I* and X?I??I2 states may come into resonance, and a fast spin–orbit quenching transition becomes possible by an electron hoping from I2 to I*. Parson and co-workers developed this picture and identified the electronic states of I2 2 of different charge-switching character.60-62 Their behavior under differential solvation is illustrated qualitatively in Fig. 8 (bottom). In the states with ‘‘normal’’ charge-switching character, the charge gravitates towards the more solvated end of I2 2.Compared to unsolvated I2 2, the energies of these states decrease with increasing I–I2 separation due to more efficient solvation as the charge becomes more localized. To the contrary, the ‘‘anomalous’’ charge-switching states exhibit a charge distribution favoring Fig. 7 Delay-dependent absorption recoveries of indicated I2 2(OCS)n cluster ions following the excitation at (a) 790 nm and (b) 395 nm. In (b), the relaxation processes leading to I2 2 caging include the spin–orbit relaxation of I*, while in (b) the spin–orbit relaxation step is not involved. Adapted from ref. 24. 170 PhysChemComm, 2002, 5(25), 165–177the least solvated end of the chromophore.The energy of these states increases in dissociation, as the I2–solvent interaction drops with increasing I–I2 distance. The resulting curvecrossings, implicated in Fig. 8 (bottom right), make a X?I2??I* A X?I??I2 transition possible. Following that, the electrostatic attraction between the solvent and I2 will tend to reverse the dissociation trajectory towards I2 2 recombination. Fig. 8 is only a qualitative illustration, oversimplifying the energetics and dynamics of quenching. The number and configuration of solvent molecules in the figure are chosen arbitrarily to satisfy the DW # 0.93 eV requirement. In reality, transient resonances between different electron-transfer states are possible starting from a wide range of initial solvent configurations, including nearly symmetric ones. As the cluster breaks up and the charge localizes on one of the fragments, the initial symmetry, if any, is always destroyed and as long as the number of solvent molecules in the cluster is sufficient, the required curve crossing is possible at some cluster geometry.The dynamics on the ground spin–orbit asymptote following the spin–orbit relaxation are reflected in the relative yields of channels B and C (defined in Fig. 4). In the 790 nm experiment, all dynamics transpire on this asymptote, and in small clusters the dissociation is naturally favored over the recombination. 22,25,49,89 To the contrary, at 395 nm no preference is observed for I2 1 I(2P3/2) dissociation over the recombination, even in the smallest clusters in which the spin–orbit quenching is possible. Evidently, this is due to the restrictions imposed by the spin–orbit relaxation step.The quenching can occur only if the number of solvent molecules in the cluster is sufficiently large, which by itself favors recombination. Additionally, following the electron transfer, the charge in the X?I??I2 state localizes on the escaping fragment, which experiences a backward pull fromthe solvent. Thus, the solvent configurations that are prerequisite for quenching also favor recombination. It is noteworthy that if electron transfer fails during the initial fragment separation on the X?I2??I* state, the electrostatic force acting on the neutral fragment is weak, and the ensuing dynamics will favor cage escape. Consequently, the dynamical window for spin–orbit quenching is limited to the initial fragment separation, consistent with the fast timescale of caging observed experimentally.Since the quenching step does not add extra time to that needed for fragment separation and subsequent recombination, the similar caging timescales are to be expected at both 395 and 790 nm. This prediction is in accord with the remarkably similar behaviors reflected in Fig. 7(a) and (b). The resonance condition for spin–orbit quenching by the solvent asymmetry mediated electron transfer is very sensitive to instantaneous solvent configurations.63 Therefore, the caging reaction involving the spin–orbit relaxation is a sensitive probe of solvation, with the cluster structure playing an important role in the dynamics. In both I2 2(OCS)n and I2 2(CO2)n, the quenching and caging probabilities are rather smooth and monotonic outside the range of n ~ 11–17 (see Fig.6). Only in this mid-size range, where the second half of the first solvent shell is believed to be constructed (e.g., see Fig. 1), the curves deviate from the expected monotonic rise, such as the 790 nm trend seen in Fig. 5. For smaller clusters, the too few available solvent molecules restrict the quenching trajectories because of a limited number of extremely asymmetric solvent configurations that satisfy the requirement DW # 0.93 eV. Each additional solvent molecule loosens this constraint, boosting the quenching and caging probabilities. In the midsize range (n ~11–17), the sufficient degree of differential solvation is achieved without imposing severe dynamical restrictions. In this size range, the details of cluster structure, not the mere number of solvent molecules, become crucial in determining the reaction outcomes. This trend continues until the first solvent shell is filled at n ~ 16–17. From there on, additional solvent molecules do not introduce significant energetic or structural changes, and the dynamics revert to a monotonic increase in caging probability with n.4. Structure and photochemistry of homogeneous cluster anions Some cluster anions, particularly the homogeneous clusters Xn 2, where X ~ CO2, OCS, CS2, H2O, and others, pose additional questions concerning their structure that have to do with the chemical identity of the cluster core. The properties of the core are fundamental to the reactivity and dynamics.The first question that must be addressed is often whether the excess electron in the cluster anion is localized on a single monomer or shared between two (or more) monomer moieties.1,106–115 In the limit of electron solvation,116,117 the excess electron wave function is delocalized to such extent that the concept of a cluster core is no longer applicable. These questions did not arise in the preceding discussion of solvated I2 2. It was assumed implicitly that during the entire course of the reaction the charge stays localized on the chromophore or one of its fragments. This assumption is justified, to a degree, in heterogeneous cluster anions, such as I2 2?Xn, when there is a significant difference in the electron affinities (EA) between the species composing the cluster. Even Fig.8 Top: A qualitative energy diagram illustrating the mechanism of spin–orbit relaxation by solvent-mediated electron transfer from I– to I* in an asymmetrically solvated cluster. The required resonance of the X?I–??I* and X?I??I– electronic states occurs when the differential solvation energy (DW) is equal to the spin–orbit energy gap in the I atom (0.93 eV). Adapted from ref. 24. Bottom left: A qualitative I2 – potential energy diagram (unsolvated anion). Bottom right: A qualitative illustration of the behavior of ‘‘normal’’ and ‘‘anomalous’’ charge-switching states of I2 – under the conditions of asymmetric solvation. The curve-crossings promote the X?I–??I*AX?I??I– electrontransfer transitions, quenching the spin–orbit excitation of the I fragment.PhysChemComm, 2002, 5(25), 165–177 171so, the electronic wave functions of negative ions tend to be diffuse, allowing for substantial overlap with the surrounding solvent. For example, even in such a small heterogeneous cluster anion as I2?CO2, Neumark and co-workers observed a 175u bending of CO2, attributed to a small amount of charge transfer from the I2 to the CO2.118–120 The experiments by Johnson and co-workers on hydrated cluster anions indicate the profound effect that the charge density has on the structure adopted by a water network bound to an ion.121–126 Charge delocalization can be more important in larger clusters. In addition, charge-transfer-to-solvent excited states are available in both bulk solutions116,127 and clusters.128–133 In the homogeneous water cluster anions (H2O)n 2, n ¢ 2, the neutral solvent network deforms to trap a diffuse excess electron, forming microscopic precursors of the hydrated electron. Water cluster anions have long served as a favorite system for the studies of electron solvation and the transition between gas-phase (or cluster) and bulk properties.134 The variety of interaction available in these cluster ions, including, but not limited to, hydrogen bonding and delocalized charge– dipole interactions, leads to interesting structure variations.74,75,134,135 Significant molecular rearrangements upon electron attachment to a neutral cluster have been implicated in the formation of different structural isomers.134 Different puzzles pertaining to charge localization and structures are presented by cluster anions of CO2, OCS, and CS2.Since the mid-1980s, photofragment and photoelectron spectroscopies, as well as theoretical studies of (CO2)n 2, posed questions of size-dependent alternation between cluster structures adopting either the monomer or dimer anion cores.107,108,112,136 Photoelectron spectroscopy revealed sharp discontinuities in the n-dependence of the vertical detachment energy of (CO2)n – cluster anions between n ~ 6 and 7 and between n ~ 13 and 14.107,108 These discontinuities have been attributed to ‘‘core switching’’: a transformation of the charged cluster core from a delocalized-charge covalent (CO2)2 2 structure for n v 6 to CO2 2 for 7 ¡ n ¡ 13, and back to (CO2)2 2 for n w 13.Fleischman and Jordan predicted,112 based on electronic-structure calculations, that the global minimum of (CO2)2 – corresponds to a structure of D2d symmetry with the charge equally divided between the two CO2 moieties. The (CO2)2 2ACO2 2 core switching in (CO2)n 2 at n ~ 6 was attributed107,108 to a more favorable solvation of the monomer anion, compared to the covalent dimer due to the more localized charge distribution in the latter. The reverse switch occurs between n ~ 13 and 14,108 in order to accommodate the dimer-based ‘‘magic number’’ structure of (CO2)14 2, i.e. (CO2)2 2?(CO2)12. Photofragmentation studies provide an alternative perspective of the structural properties and photochemistry of cluster anions. Alexander et al.investigated the photofragmentation of (CO2)n – cluster ions136 and found that they break up exclusively to smaller species of similar composition (CO2)k –, k v n. This behavior may give an impression that no bond breaking or chemical rearrangements are taking place and the fragmentation proceeds merely via the loss of solvent molecules. This is not necessarily the case. The ionic core of these clusters is formed by adding an electron to either the LUMO of a CO2 monomer or to the combined LUMO of two CO2 molecules (i.e., a van der Waals dimer).112,137 Thus, these cluster ions can be viewed as ensembles of closed-shell molecules with an access electron either localized on one of them or shared between two. The only additional covalent bond that may be formed in the cluster anion is the weak (order of 1/2) C–C bond in the D2d (CO2)2 2 dimer anion predicted by Jordan and coworkers.112 This bond is likely to be broken after absorption of a photon, and yet the resulting photofragments are described as (CO2)k –, indistinguishable from those formed by solvent evaporation.Given the isovalency and structural similarity of OCS and CO2, one might expect the properties of (OCS)n 2 to be similar to those of (CO2)n –. However, the photochemistry of (OCS)n 2 proved to be a striking deviation from this expectation.113 The studies of carbonyl sulfide cluster anions began in 1998 with the Lineberger group identifying several types of ionic photofragments of (OCS)n 2. In addition to (OCS)k 2, which could be expected by analogy with (CO2)n –, the observed products included S2 2 and S2/OCS2 2 based photofragments.113 Such variety of fragmentation channels suggested that extensive bond-breaking and chemical rearrangements take place in the cluster fragmentation process.These observations were attributed113 to the existence of electronic isomers of (OCS)n 2.114,115 In particular, the abundance of the S2 2-based fragments hinted at the role of a covalently bound dimer anion cluster core with S–S bonding. In collaboration with Jordan, the Lineberger group examined the properties of three (OCS)2 2 isomers, whose calculated structures and relative energies are shown in Fig. 9.113,137 All (OCS)n 2 cluster ions are believed to be based on one of the core species shown in Fig. 9. The first (OCS)2 2 isomer shown in Fig.9(a) is an electrostatically bound cluster of OCS2 with one OCS solvent molecule. The (OCS)n 2 cluster ions with an OCS2 core, adapting the structural motif of Fig. 9(a), are analogous to the CO2 2 based cluster ions of (CO2)n 2, n ~ 7–13.107,108 The covalently bound isomer of C2 symmetry shown in Fig. 9(b) has a C–C bond of the order of 1/2 and is similar to the D2d dimer structure of (CO2)2 2 predicted by Jordan and co-workers.112 Fig. 9(c) shows a cyclic dimer anion of C2v symmetry, whose likely photochemical signature is the formation of the S2 2 fragments. A species of this kind was first observed113 in (OCS)n 2, followed by the discovery of similar anions of (CS2)2 2.106,109,138 In particular, the existence of such structures was revealed in a photodetachment study of (CS2)n 2 by Tsukuda et al.109 and further confirmed by the observation of C2S2 2 products in the photodissociation of (CS2)n 2.106 In this light, it may appear intriguing that no cyclic structures were observed in the experiments on (CO2)n –.107,108,136 As described by Jordan and co-workers,113,137,138 the Fig.9 Equilibrium geometries and relative energies of three (OCS)2 2 species optimized at the MP2/6-311G(d) level of theory. (a) Electrostatically bound OCS2?OCS cluster anion (planar structure). (b) C2 symmetry structure with a C–C bond of the order of 1/2. (c) C2v symmetry structure corresponding to the global potential minimum of (OCS)2 2 with a 2B2 symmetry electronic wave function and C–C and S–S bond orders of 1 and 1/2, respectively.Some important bond lengths are indicated in a°ngstro�ms. The energies given were calculated relative to the OCS 1 OCS2 limit and include harmonic zero-point vibration energy corrections determined at the HF/6-311G* level). Data from ref. 113. 172 PhysChemComm, 2002, 5(25), 165–177(OCS)2 2 anion in Fig. 9(c) differs from the other covalent species shown in Fig. 9(b) in that its electronic structure is not derived directly from ground-state OCS or its van der Waals dimer. In the molecular-orbital picture, the removal of the excess electron from the (OCS)2 2 anion shown in Fig. 9(b) yields two OCS molecules in the ground electronic states. To the contrary, the detachment from the HOMO of the cyclic dimer is predicted to access a metastable state of (OCS)2, whose electronic configuration is doubly excited with respect to that of two ground-state OCS molecules.Its electronic configuration arises from a singlet coupling of two OCS molecules excited to the lowest triplet states. Most, but not all, of the combined singlet–triplet excitation energy for two OCS molecules is recovered by the strong bonding in the doubly excited neutral dimer, in which both the C–C and S–S bonds are of the order of 1. Adding an electron to its low-lying LUMO yields the ground (2B2) state of (OCS)2 2, whose equilibrium geometry is shown in Fig. 9(c).113,137 In similar ways, the energetics of the (CO2)2 2, (OCS)2 2, and (CS2)2 2 cyclic anions are dependent, in part, on the singlet–triplet splitting in the respective neutral monomers. In CO2, this splitting is significantly larger than in OCS or CS2, and as a result, the cyclic dimer anion of CO2 is less stable.137 The properties of (OCS)2 2 raise general questions about the reactivity of negative ions in homogeneous versus heterogeneous cluster environments.In particular, to characterize the electronic and structural isomers, one needs to discriminate between the covalent and electrostatically bound species. Heterogeneous clusters offer opportunities to study the interactions of the anion (e.g., OCS2) with the solvent under conditions when the charge localization is known and unambiguous. A case in point is OCS2?H2O. Because of the slightly negative EA of OCS, OCS2 is believed to be metastable139 and cannot be formed efficiently in a standard ion source.In hydrated clusters, the anion is stabilized by interactions with the solvent and detailed examination of OCS2 properties is possible. To illustrate this point, Fig. 10 displays a negative-ion mass spectrum139 obtained using standard pulsed ion source techniques 3 with the OCS precursor seeded in Ar with trace amount of water. In comparison to other ions, almost no OCS2 is detected. On the other hand, there is an intense progression of peaks corresponding to OCS2(H2O)k with k ranging from 1 to at least 5. Another intense peak in the spectrum corresponds to (OCS)2 2. Thus, while OCS2 is not formed efficiently by itself, it is stabilized by either electrostatic of chemical interactions. A new approach to negative-ion photoelectron spectroscopy is based on photoelectron imaging.The imaging approach to gas-phase dynamics was originally developed by Chandler and Houston76,140 as a tool of photofragment spectroscopy for studying the photodissociation of neutral molecules.140,141 Several recent breakthroughs in imaging technology142–145 led to an explosive growth in the field and made the application of imaging to negative ions very compelling. Among the recent advances in imaging are: velocity-map imaging,142,143 which results in resolution comparable to that of other spectroscopic techniques; event counting,144,145 which makes it possible to carry out measurements with very low signals; and the Basis Set Expansion method of Reisler and co-workers, which revolutionalized the data analysis.146 The application of imaging to negative-ion photoelectron spectroscopy was implemented recently by the groups of Broyer,80,81 Continetti,83 Neumark,147 and Sanov.82,115,148 The coincidence measurements of photoelectron–photofragment angular correlation and energy partitioning in dissociative photodetachment by Continetti and co-workers demonstrated the ground-breaking capability of giving insights into the molecule-fixed photoelectron angular distributions.70,73 Photoelectron imaging yields three-dimensional distributions of the velocity vectors in the laboratory frame, including the photoelectron speed and angular distributions.140 The former are converted into photoelectron spectra, while the latter reflect the electronic wave function symmetry,149–153 serving as a portal103,154–161 for observing the dynamics from the electronic perspective.Here we describe the application of photoelectron imaging to OCS2?H2O and (OCS)2 2. The monohydrated anion of carbonyl sulfide was predicted to have a straightforward electrostatically bound structure.139,162 The electrostatically and covalently bound isomers of the homogeneous dimer anion, i.e., OCS2?OCS and (OCS)2 2, are discussed above. Comparison of the OCS2?H2O and OCS2?OCS/(OCS)2 2 photoelectron imaginfingerprints gives insights into the electronic structure, as well as the intimate interplay between the gas-phase and bulk properties of matter. Fig. 11 shows a series of photoelectron images of OCS2?H2O and (OCS)2 2 obtained by Surber and Sanov at three different laser wavelengths.82,115 The OCS2?H2O images feature a single electronic transition with a broad Franck–Condon envelope, as characteristic of a bent-to-linear detachment transition in the OCS2 cluster core.82 The angular anisotropy in these images has been compared to that observed for CS2 – and found in agreement with the expected electronic structure of the anion. A qualitative s & p model was proposed that describes the observed photoelectron angular distributions in terms of the lowest l-components (in many cases, limited to the s and p partial waves) of the free (photodetached) electron.82,148 A comparison of the OCS2?H2O and (OCS)2 2 images in Fig.10 Negative-ion mass-spectrum obtained with the OCS/Ar precursor with a trace of water. The magnified (610) spectrum corresponds to experimental conditions optimized for the formation of OCS2 and shows the best OCS2 signal that could be achieved in the experiment.Adapted from ref. 139. Fig. 11 Photoelectron images of OCS2?H2O (top) and (OCS)2 2 (bottom) recorded at 400, 530, and 800 nm, shown on arbitrary velocity and intensity scales. The (OCS)2 2 signal inside the dashed circles is attributed mainly to covalent (OCS)2 2 isomers. The signal outside the dashed circles is attributed to the OCS2?OCS cluster anion (electrostatically bound). Data from ref. 115. PhysChemComm, 2002, 5(25), 165–177 173Fig. 11 clearly reveals signatures of different (OCS)2 2 isomers. 82,148 The diffuse lobes at large radii, polarized in the direction of the laser polarization (vertical in the plane of the figure), are attributed to the electrostatically bound OCS2?OCS cluster anion, based on the similarity of these parts of the images to OCS2?H2O.The photoelectron signatures of OCS2?H2O and OCS2?OCS are expected to be similar, because in both cases the electron is ejected from a clustermolecular orbital with primarily the OCS2HOMOcharacter. The signal inside the dashed circles marked over the (OCS)2 2 images is attributed to the covalently bound structures of the dimer anion.115 The OCS2OCS and covalent (OCS)2 2 anions behave differently in photoexcitation. The former species absorb light via direct photodetachment, resulting in the photoelectron imaging signatures described above. The covalent dimer anion, on the other hand, was predicted to possess a number of lowlying excited anionic states;113,137,138 therefore, it can be either photodetached directly or promoted to an excited state.The excited-state decay, in turn, involves the competition between the autodetachment and fragmentation.114,115 In Fig. 11, the direct photodetachment of covalent (OCS)2 2 is manifest as the diffuse anisotropic feature just inside the dashed circle marked over the 400 nm image. The 530 and 800 nm photon energies are not sufficient to access this transition. The autodetachment, yielding characteristically slow electrons, is seen as the intense isotropic spots at the centers of all three (OCS)2 2 photoelectron images. The autodetachment spots originate from either the excited state of (OCS)2 2 or the internally excited anionic photofragmentation products (e.g., OCS2).In either case, the autodetachment can be modeled115 as a gas-phase analog of thermionic emission,80 an effect usually associated with bulk materials. Thus, the bulk statistical model assuming strong electronic-vibrational couplings and a highly mixed nature of the excited anionic state, was found applicable to such a small system as (OCS)2 2, which appears to combine both molecular and ‘‘bulk’’ properties.115 The case of (OCS)2 2, as well as that of larger (OCS)n 2 cluster anions,114 is just one example that demonstrates that the answer to the question posed at the beginning of this Perspective (‘‘How much matter is needed for physical laws associated with bulk materials to be applicable?’’) is: It is not the size of the microscopic object that determines the applicability of the bulk description, but the details of the electronic structure, including the availability of low-lying, mixed excited states.The one-photon, or ‘static’, photoelectron images, as those shown in Fig. 11, provide insights into the electronic structure of cluster anions. A natural extension of this approach is probing the evolution of the electronic structure in photoinduced chemical reactions using femtosecond pump–probe photoelectron imaging spectroscopy. The power of this approach has been successfully demonstrated for neutral molecules by the groups of Hayden, Continetti, and Suzuki.77,78,163–165 Femtosecond imaging experiments on molecular and cluster anions are now underway in the Neumark group at Berkeley and the Sanov group at the University of Arizona.As with traditional femtosecond photoelectron spectroscopy, the applicability of femtosecond photoelectron imaging is not limited to clusters. However, the imaging approach is particularly promising in the cluster case, as it allows examining the transformations of electronic wave functions under the effects of microscopic solvation. Cluster anions also provide unique opportunities for detailed gas-phase studies of bimolecular encounters, which can be viewed, using photoelectron and fragment imaging, fromboth the electronic and nuclear perspectives. 5. Summary This Perspective reflected on several recent advances in the studies of cluster anion structure and dynamics. We discussed the effects of the solvent on the electronic structure and reactivity of negative ions in homogeneous and heterogeneous solvation environments.Some recent breakthroughs in experimental methodology were outlined, in particular the application of photofragment and photoelectron methods and the imaging technique to the studies of molecular cluster anions. In the future, we should expect to see an increased emphasis on ultrafast coincidence dynamics in the studies of cluster ion reactivity. As more sophisticated experimental tools become available, physical chemists will be able to describe the reaction dynamics in more detailed and less averaged ways. The trend towards better-resolved and less averaged observables is seen in many recent developments, including the growing popularity of the imaging technique (complementary energy and angular distributions), a variety of coincidence methods (correlated product distributions), and time-domain measurements.Another trend that should not be overlooked by experimentalists is the developments in theoretical chemistry, in particular the explosive growth in the computational capabilities. As demonstrated by several examples in this Perspective, the interpretation of experimental results is often dependent on extensive theoretical work. As the theoretical and experimental capabilities reach new levels of elegance and sophistication, chemists acquire the ability to tackle more intriguing questions of structure and dynamics. Many ab initio problems that a decade ago would have required the computing power of a supercomputer today can be solved at minimal expense using a personal machine in one’s office.Thanks to this development, experimentalists can now routinely use the power of computational chemistry to obtain the initial interpretation of results enabling them to navigate more efficiently in search of scientific answers. Such skillful navigation becomes increasingly important, as the sheer dimensionality of information provided by modern state-of-the-art experiments makes their success more dependent than ever on one’s ability to ask the right questions. Acknowledgements We thank Prof. Robert Parson and Prof. Kenneth D. Jordan for their collaborations and insights into several projects discussed in this Perspective. A. Sanov acknowledges support from the National Science Foundation grants Nos.CHE- 9982057 and CHE-0134631, the Beckman Young Investigator Award, the Research Corporation Research Innovation Award No. RI0515, and the ACS PRF grant No. 35589-G6. W. C. Lineberger is pleased to acknowledge support from the National Science Foundation grants Nos. CHE-0201848 and PHY-0096822 and the Air Force Office of Scientific Research grant No. F49620-02-1-0371. References 1 A. W. Castleman and K. H. Bowen, J. Phys. Chem., 1996, 100, 12911. ƒ An extensive review of the field of cluster research highlighting the significant advances made during the later part of the twentieth century. Published in the Centennial Issue of J. Phys. Chem. 2 J. M. Farrar, in Current Topics in Ion Chemistry and Physics, ed. C. Y. Ng and I. Powis, Wiley, New York, 1992. 3 M.A. Johnson and W. C. Lineberger, in Techniques for the Study of Ion Molecule Reactions, ed. J. M. Farrar and J. W. Saunders, Wiley, New York, 1988, p. 591. ƒ The original review of the pulsed methods of cluster ion spectroscopy extensively used by many research groups in the negative-ion community throughout the world. 4 A. W. Castleman, Jr., in Clusters of Atoms and Molecules, ed. H. Haberland, Springer-Verlag, New York, 1992. 5 J. C. Alfano, Y. Kimura, P. K. Walhout and P. F. Barbara, Chem. Phys. Chem., 1993, 175, 143. ƒ Experimental study of I2 2 photodissociation and vibrational relaxation in water and ethanol. 6 D. A. V. Kliner, J. C. Alfano and P. F. Barbara, J. Chem. Phys., 1993, 98, 5375. 174 PhysChemComm, 2002, 5(25), 165–1777 P. K.Walhout, J.C. Alfano, K. A. M. Thakur and P. F. Barbara, J. Phys. Chem., 1995, 99, 7568. 8 U. Banin and S. Ruhman, J. Chem. Phys., 1993, 98, 4391. 9 E. Gershgoren, U. Banin and S. Ruhman, J. Phys. Chem. A, 1998, 102, 9. 10 H. Yasumatsu, S. Koizumi, A. Terasaki and T. Kondow, J. Chem. Phys., 1996, 105, 9509. 11 H. Yasumatsu, A. Terasaki and T. Kondow, J. Chem. Phys., 1997, 106, 3806. 12 F. G. Amar and L. Perera, Z. Phys. D, 1991, 20, 173. 13 B. J. Gertner, K. Ando, R. Bianco and J. T. Hynes, Chem. Phys., 1994, 183, 309. 14 R. Bianco and J. T. Hynes, J. Chem. Phys., 1995, 102, 7864. 15 R. Bianco and J. T. Hynes, J. Chem. Phys., 1995, 102, 7885. 16 I. Benjamin, P. F. Barbara, B. J. Gertner and J. T. Hynes, J. Phys. Chem., 1995, 99, 7557. 17 B. J. Greenblatt, M. T. Zanni and D.M. Neumark, Chem. Phys. Lett., 1996, 258, 523. 18 M. T. Zanni, T. R. Taylor, B. J. Greenblatt, B. Soep and D. M. Neumark, J. Chem. Phys., 1997, 107, 7613. 19 M. T. Zanni, V. S. Batista, B. J. Greenblatt, W. H. Miller and D. M. Neumark, J. Chem. Phys., 1999, 110, 3748. 20 D. Ray, N. E. Levinger, J. M. Papanikolas and W. C. Lineberger, J. Chem. Phys., 1989, 91, 6533. 21 J. M. Papanikolas, V. Vorsa, M. E. Nadal, P. J. Campagnola, H. K. Buchenau and W. C. Lineberger, J. Chem. Phys., 1993, 99, 8733. ƒ Experimental study of I2 2 photodissociation and recombination dynamics in size-selected CO2 clusters. 22 S. Nandi, A. Sanov, N. Delaney, J. Faeder, R. Parson and W. C. Lineberger, J. Phys. Chem. A, 1998, 102, 8827. ƒ The first report of I2 2 caging in clusters of OCS involving fast spin–orbit relaxation.23 A. Sanov, S. Nandi and W. C. Lineberger, J. Chem. Phys., 1998, 108, 5155. 24 A. Sanov, T. Sanford, S. Nandi and W. C. Lineberger, J. Chem. Phys., 1999, 111, 664. ƒ Time-resolved study of photofragment caging involving fast quenching of spin–orbit excited iodine by solvent-asymmetry mediated electron transfer. 25 J. M. Papanikolas, J. R. Gord, N. E. Levinger, D. Ray, V. Vorsa and W. C. Lineberger, J. Phys. Chem., 1991, 95, 8028. 26 J. Troe, Annu. Rev. Phys. Chem., 1978, 29, 223. 27 B. Otto, J. Schroeder and J. Troe, J. Chem. Phys., 1984, 81, 202. 28 H. Kunz, J. G. McCaffrey, R. Schriever and N. Schwentner, J. Chem. Phys., 1991, 94, 1039. 29 J. Xu, N. Schwentner and M. Chergui, J. Chem. Phys., 1994, 101, 7381. 30 K. H. Godderz, N.Schwentner and M. Chergui, J. Chem. Phys., 1996, 105, 451. 31 V. E. Bondybey and L. E. Brus, Adv. Chem. Phys., 1980, 41, 269. 32 P. S. Dardi and J. S. Dahler, J. Chem. Phys., 1990, 93, 242. 33 G. N. R. Tripathi, R. H. Schuler and R. W. Fessenden, Chem. Phys. Lett., 1985, 113, 563. 34 V. S. Batista and D. F. Coker, J. Chem. Phys., 1997, 106, 7102. 35 B. M. Ladanyi and R. Parson, J. Chem. Phys., 1997, 107, 9326. 36 N. Delaney, J. Faeder, P. E. Maslen and R. Parson, J. Phys. Chem. A, 1997, 101, 8147. 37 X. Xu, S. Yu, R. Lingle, H. Zhu and J. B. Hopkins, J. Chem. Phys., 1991, 95, 2445. 38 A. L. Harris, J. K. Brown and C. B. Harris, Annu. Rev. Phys. Chem., 1988, 39, 341. 39 R. Zadoyan, Z. Li, C. C. Martens and V. A. Apkarian, J. Chem. Phys., 1994, 101, 6648. 40 R.Zadoyan, Z. Li, P. Ashjian, C. C. Martens and V. A. Apkarian, Chem. Phys. Lett., 1994, 218, 504. 41 Q. Liu, J. K. Wang and A. H. Zewail, Nature, 1993, 364, 427. 42 E. D. Potter, Q. Liu and A. H. Zewail, Chem. Phys. Lett., 1992, 200, 605. 43 Q. L. Liu, J. K. Wang and A. H. Zewail, J. Phys. Chem., 1995, 99, 11321. 44 J. K. Wang, Q. L. Liu and A. H. Zewail, J. Phys. Chem., 1995, 99, 11309. 45 A. E. Johnson, N. E. Levinger and P. F. Barbara, J. Phys. Chem., 1992, 96, 7841. 46 B. J. Greenblatt, M. T. Zanni and D. M. Neumark, Science, 1997, 276, 1675. ƒ A landmark study of dynamics of I2 2Ar6 and I2 2Ar20 subsequent to photodissociation of the I2 2 chromophore using femtosecond photoelectron spectroscopy. 47 R. Parson and J. Faeder, Science, 1997, 276, 1660. ƒ A theoretical perspective on the Neumark group’s study of I2 2Arn dynamics by femtosecond PES.48 J. M. Papanikolas, V. Vorsa, M. E. Nadal, P. J. Campagnola, J. R. Gord and W. C. Lineberger, J. Chem. Phys., 1992, 97, 7002. 49 V. Vorsa, S. Nandi, P. J. Campagnola, M. Larsson and W. C. Lineberger, J. Chem. Phys., 1997, 106, 1402. ƒ Timeresolved study of photodissociation and recombination dynamics of I2 2 in size-selected Ar and CO2 clusters. 50 V. Vorsa, P. J. Campagnola, S. Nandi, M. Larsson and W. C. Lineberger, J. Chem. Phys., 1996, 105, 2298. ƒ Observation of metastable solvent-separated (I–…I) Arm photofragments, where the I2 2 bond has not reformed. 51 J. M. Papanikolas, P. J. Campagnola, V. Vorsa, M. E. Nadal, H. K. Buchenau, R. Parson and W. C. Lineberger, in The Chemical Dynamics and Kinetics of Small Radicals, ed.K. Liu and A. Wagner, World Scientific Publishing Co., Singapore, 1995, vol. 6, p. 616. 52 M. E. Nadal, P. D. Kleiber and W. C. Lineberger, J. Chem. Phys., 1996, 105, 504. 53 A. Sanov and W. C. Lineberger, Proc. SPIE, 1998, 3271, 188. 54 E. Rabinowitch and W. C. Wood, Trans. Faraday Soc., 1936, 32, 547. 55 E. Rabinowitch and W. C. Wood, Trans. Faraday Soc., 1936, 32, 1381. 56 J. Franck and E. Rabinowitch, Trans. Faraday Soc., 1934, 30, 120. 57 J. Xu, N. Schwentner, S. Hennig and M. Chergui, J. Chim. Phys. Phys. Chim Biol., 1995, 92, 541. 58 Z. Li, R. Zadoyan, V. A. Apkarian and C. C. Martens, J. Phys. Chem., 1995, 99, 7453. 59 V. S. Batista and D. F. Coker, J. Chem. Phys., 1997, 106, 6923. 60 P. E. Maslen, J.M. Papanikolas, J. Faeder, R. Parson and S. V. ONeil, J. Chem. Phys., 1994, 101, 5731. 61 P. E. Maslen, J. Faeder and R. Parson, Chem. Phys. Lett., 1996, 263, 63. ƒ Ab initio study of I2 2 electronic states. 62 J. Faeder, N. Delaney, P. E. Maslen and R. Parson, Chem. Phys., 1998, 239, 525. 63 N. Delaney, J. Faeder and R. Parson, J. Chem. Phys., 1999, 111, 651. ƒ Presentation of the solvent-asymmetry mediated electron transfer model of spin–orbit quenching involved in the near-UV caging of I2 2. 64 J. H. D. Eland, Photoelectron Spectroscopy, Butterworths, London, 1984. ƒ A classic book on photoelectron spectroscopy. 65 K. M. Ervin and W. C. Lineberger, in Advances in Gas Phase Ion Chemistry, ed. N. G. Adams and L. M. Babcock, JAI Press, Greenwich, 1992, vol. 1, p.121. ƒ A review of photoelectron spectroscopy of negative ions. 66 W. C. Lineberger, M. E. Nadal, P. J. Campagnola, V. Vorsa, P. D. Kleiber, J. M. Papanikolas, P. E. Maslen, J. Faeder, R. Parson and O. Poplawski, in Proceedings of The Robert A. Welch Foundation 38th Conference on Chemical Research: Chemical Dynamics of Transient Species, R. A. Welch Foundation, Houston, TX, 1994, vol. 38, p. 175. 67 J. H. D. Eland, Mol. Phys., 1987, 61, 725. 68 J. H. D. Eland, J. Electron Spectrosc. Relat. Phenom., 2000, 112, 1. ƒ A presentation of novel coincidence techniques for the study of single photon double ionization. 69 K. A. Hanold, C. R. Sherwood and R. E. Continetti, J. Chem. Phys., 1995, 103, 9876. ƒ A landmark study of photoelectron– neutral–neutral coincidence dynamics of dissociative photodetachment of O4 –.70 K. A. Hanold, M. C. Garner and R. E. Continetti, Phys. Rev. Lett., 1996, 77, 3335. ƒ A beautiful coincidence measurement of photoelectron–photofragment angular correlation and energy partitioning in the dissociative photodetachment of O4 2 providing insights into the molecule-fixed photoelectron angular distribution of a negative ion. 71 T. Baer and Y. Li, Int. J. Mass Spectrom., 2002, 219, 381. 72 T. Baer, Int. J. Mass Spectrom., 2000, 200, 443. ƒ Review of photoelectron–photoion coincidence (PEPICO) spectroscopy. 73 R. E. Continetti, Annu. Rev. Phys. Chem., 2001, 52, 165. ƒ An extensive review of coincidence techniques in the studies of the reaction dynamics of isolated molecules. 74 P. Ayotte, G. H. Weddle, C.G. Bailey, M. A. Johnson, F. Vila and K. D. Jordan, J. Chem. Phys., 1999, 110, 6268. ƒ A benchmark study of negatively charged water clusters by infrared spectroscopy. 75 C. G. Bailey, J. Kim and M. A. Johnson, J. Phys. Chem., 1996, 100, 16782. 76 D. W. Chandler and P. L. Houston, J. Chem. Phys., 1987, 87, 1445. ƒ This pioneering paper describes the original photofragment ion imaging experiment. The entire field of photofragment and photoelectron imaging developed from this work. PhysChemComm, 2002, 5(25), 165–177 17577 J. A. Davies, J. E. LeClaire, R. E. Continetti and C. C. Hayden, J. Chem. Phys., 1999, 111, 1. ƒ A pioneering application of femtosecond time-resolved photoelectron–photoion coincidence imaging to studies of neutral dissociation dynamics. 78 L.Wang, H. Kohguchi and T. Suzuki, Faraday Discuss., 1999, 113, 37. ƒ The first demonstration of femtosecond photoelectron imaging used to probe dynamics in neutral molecules. 79 C. Bordas, J. C. Pinare, B. Baguenard and M. Broyer, J. Phys. IV, 2000, 10, 55. 80 B. Baguenard, J. C. Pinare, C. Bordas and M. Broyer, Phys. Rev. A, 2001, 63, art. no. 023204. ƒ Photoelectron imaging study of tungsten cluster anions. 81 B. Baguenard, J. C. Pinare, F. Lepine, C. Bordas and M. Broyer, Chem. Phys. Lett., 2002, 352, 147. ƒ Photoelectron imaging study of carbon cluster anions. 82 E. Surber and A. Sanov, J. Chem. Phys., 2002, 116, 5921. 83 H. J. Deyerl, L. S. Alconcel and R. E. Continetti, J. Phys. Chem. A, 2001, 105, 552. 84 J. M. Papanikolas, P. E. Maslen and R. Parson, J.Chem. Phys., 1995, 102, 2452. 85 J. Faeder, PhD Thesis, University of Colorado, 1998. 86 M. R. Battaglia, A. D. Buckingham, D. Neumark, R. K. Pierens and J. H. Williams, Mol. Phys., 1981, 43, 1015. 87 A. D. Buckingham, C. Graham and J. H. Williams, Mol. Phys., 1988, 49, 703. 88 J. Faeder, N. Delaney, P. E. Maslen and R. Parson, Chem. Phys. Lett., 1997, 270, 196. 89 M. E. Nadal, PhD Thesis, University of Colorado, 1996. 90 H. Yasumatsu, U. Kalmbach, S. Koizumi, A. Terasaki and T. Kondow, Z. Phys. D, 1997, 40, 1. 91 H. Yasumatsu, T. Tsukuda, T. Sugai, A. Terasaki, T. Nagata and T. Kondow, Surf. Rev. Lett., 1996, 3, 901. 92 N. Delaney, J. Faeder and R. Parson, J. Chem. Phys., 1999, 111, 452. 93 J. Faeder and R. Parson, J. Chem. Phys., 1998, 108, 3909. 94 B. J. Greenblatt, M.T. Zanni and D. M. Neumark, J. Chem. Phys., 2000, 112, 601. 95 A. V. Davis, R. Wester, A. E. Bragg and D. M. Neumark, J. Chem. Phys., 2002, 117, 4282. 96 R. Wester, A. V. Davis, A. E. Bragg and D. M. Neumark, Phys. Rev. A, 2002, 65, art. no. 051201. 97 B. J. Greenblatt, M. T. Zanni and D. M. Neumark, J. Chem. Phys., 1999, 111, 10566. 98 D. Husain and R. J. Donovan, in Advances in Photochemistry, ed. J. N. Pitts, Jr., G. S. Hammond, and W. A. Noyers, Jr., Wiley- Interscience, New York–London–Sydney–Toronto, vol. 8, 1971. 99 A. H. Zewail, in Chemical Reactions and Their Control on the Femto-second Time Scale: XXth Solvay Conference On Chemistry, John Wiley & Sons Inc, New York, 1997, vol. 101, p. 892. 100 A. H. Zewail, J. Phys. Chem. A, 2000, 104, 5660.ƒ An anthology adapted from the Nobel Lecture, which gives an overview of the field of femtochemistry from the author’s perspective. 101 A. H. Zewail, Pure Appl. Chem., 2000, 72, 2219. ƒ A historical perspective of the development of femtochemistry. 102 J. S. Baskin and A. H. Zewail, J. Chem. Educ., 2001, 78, 737. 103 S. Lochbrunner, J. J. Larsen, J. P. Shaffer, M. Schmitt, T. Schultz, J. G. Underwood and A. Stolow, J. Electron Spectrosc. Relat. Phenom., 2000, 112, 183. ƒ A detailed discussion of femtosecond time-resolved photoelectron spectroscopy as a technique for investigating excited state dynamics. 104 V. Blanchet, M. Z. Zgierski, T. Seideman and A. Stolow, Nature, 1999, 401, 52. ƒ A popular description of applications of timeresolved photoelectron spectroscopy to discerning vibronic molecular dynamics. 105 G.Gantefor, S. Kraus and W. Eberhardt, J. Electron Spectrosc. Relat. Phenom., 1998, 88, 35. 106 T. Maeyama, T. Oikawa, T. Tsumura and N. Mikami, J. Chem. Phys., 1998, 108, 1368. 107 M. J. DeLuca, B. Niu and M. A. Johnson, J. Chem. Phys., 1988, 88, 5857. ƒ The original demonstration of core-switching in (CO2)n 2 cluster anions. 108 T. Tsukuda, M. A. Johnson and T. Nagata, Chem. Phys. Lett., 1997, 268, 429. 109 T. Tsukuda, T. Hirose and T. Nagata, Chem. Phys. Lett., 1997, 279, 179. 110 K. H. Bowen and J. G. Eaton, in The Structure of Small Molecules and Ions, ed. R. Naaman and Z. Vager, Plenum, New York, 1988, p. 147. 111 K. Hiraoka, S. Fujimaki, G. Aruga and S. Yamabe, J. Phys. Chem., 1994, 98, 1802. 112 S.H. Fleischman and K. D. Jordan, J. Phys. Chem., 1987, 91, 1300. ƒ The first theoretical prediction of the covalently bound dimer anion of CO2. 113 A. Sanov, S. Nandi, K. D. Jordan and W. C. Lineberger, J. Chem. Phys., 1998, 109, 1264. ƒ The original experimental study of (OCS)n 2 photochemistry. 114 E. Surber, R. Mabbs and A. Sanov, in preparation. ƒ This paper describes photoelectron imaging experiments on carbonyl sulfide cluster anions, focusing on the isomer coexistence and competition of excited-state decay mechanisms. 115 E. Surber and A. Sanov, Phys. Rev. Lett., submitted. ƒ This letter addresses imaging of direct photodetachment and delayed autodetachment of (OCS)2 2, emphasizing the excited-state dynamics of the covalent dimer anion. 116 J. Jortner, M.Ottolenghi and G. Stein, J. Phys. Chem., 1964, 68, 247. 117 P. Ayotte and M. A. Johnson, J. Chem. Phys., 1997, 106, 811. 118 D. W. Arnold, S. E. Bradforth, E. H. Kim and D. M. Neumark, J. Chem. Phys., 1992, 97, 9468. 119 Y. X. Zhao, C. C. Arnold and D. M. Neumark, J. Chem. Soc., Faraday Trans., 1993, 89, 1449. 120 A. Sanov, J. Faeder, R. Parson and W. C. Lineberger, Chem. Phys. Lett., 1999, 313, 812. 121 C. G. Bailey, J. Kim, C. E. H. Dessent and M. A. Johnson, Chem. Phys. Lett., 1997, 269, 122. 122 P. Ayotte, G. H. Weddle, J. Kim and M. A. Johnson, Chem. Phys., 1998, 239, 485. 123 P. Ayotte, G. H. Weddle and M. A. Johnson, J. Chem. Phys., 1999, 110, 7129. 124 P. Ayotte, S. B. Nielsen, G. H. Weddle, M. A. Johnson and S. S. Xantheas, J. Phys. Chem. A, 1999, 103, 10665.125 S. A. Corcelli, J. A. Kelley, J. C. Tully and M. A. Johnson, J. Phys. Chem. A, 2002, 106, 4872. 126 E. A. Woronowicz, W. H. Robertson, G. H. Weddle, M. A. Johnson, E. M. Myshakin and K. D. Jordan, J. Phys. Chem. A, 2002, 106, 7086. 127 S. E. Bradforth and P. Jungwirth, J. Phys. Chem. A, 2002, 106, 1286. 128 I. Becker, G. Markovich and O. Cheshnovsky, Phys. Rev. Lett., 1997, 79, 3391. 129 I. Becker and O. Cheshnovsky, J. Chem. Phys., 1999, 110, 6288. 130 L. Lehr, M. T. Zanni, C. Frischkorn, R. Weinkauf and D. M. Neumark, Science, 1999, 284, 635. 131 M. T. Zanni, C. Frischkorn, A. V. Davis and D. M. Neumark, J. Phys. Chem. A, 2000, 104, 2527. 132 A. V. Davis, M. T. Zanni, C. Frischkorn and D. M. Neumark, Abstr. Pap. Am. Chem. Soc., 1999, 218, 186. 133 C. Frischkorn, M. T. Zanni, A. V. Davis and D. M. Neumark, Faraday Discuss., 2000, 49. 134 J. Kim, I. Becker, O. Cheshnovsky and M. A. Johnson, Chem. Phys. Lett., 1998, 297, 90. 135 G. H. Lee, S. T. Arnold, J. G. Eaton, H. W. Sarkas, K. H. Bowen, C. Ludewigt and H. Haberland, Z. Phys. D: At., Mol. Clusters, 1991, 20, 9. 136 M. L. Alexander, M. A. Johnson, N. E. Levinger and W. C. Lineberger, Phys. Rev. Lett., 1986, 57, 976. 137 A. Sanov, W. C. Lineberger and K. D. Jordan, unpublished. 138 A. Sanov, W. C. Lineberger and K. D. Jordan, J. Phys. Chem. A, 1998, 102, 2509. 139 E. Surber, S. P. Ananthavel and A. Sanov, J. Chem. Phys., 2002, 116, 1920. 140 A. J. R. Heck and D. W. Chandler, Annu. Rev. Phys. Chem., 1995, 46, 335. ƒ A classic review of the photofragment ion imaging technique. 141 P. L. Houston, Acc. Chem. Res., 1995, 28, 453. ƒ An excellent tutorial review of applications of imaging to studies of chemical reaction dynamics. 142 A. T. J. B. Eppink and D. H. Parker, Rev. Sci. Instrum., 1997, 68, 3477. ƒ The pioneering introduction of velocity-map imaging. This paper revolutionized the applications of imaging to chemical dynamics. 143 D. H. Parker and A. T. J. B. Eppink, J. Chem. Phys., 1997, 107, 2357. 144 L. J. Rogers, M. N. R. Ashfold, Y. Matsumi, M. Kawasaki and B. J. Whitaker, Chem. Phys. Lett., 1996, 258, 159. 145 B. Y. Chang, R. C. Hoetzlein, J. A. Mueller, J. D. Geiser and P. L. Houston, Rev. Sci. Instrum., 1998, 69, 1665. 146 V. Dribinski, A. Ossadtchi, V. A. Mandelshtam and H. Reisler, Rev. Sci. Instrum., 2002, 73, 2634. ƒ Introduction of the Gaussian BAsis-Set EXpansion (BASEX) approach to Abel transformation used in image analysis. 147 D. M. Neumark, Abstr. Pap. Am. Chem. Soc., 2002, 223, 137. 176 PhysChemComm, 2002, 5(25), 165–177148 E. Surber, R. Mabbs and A. Sanov, in preparation. 149 J. Cooper and R. N. Zare, J. Chem. Phys., 1968, 49, 4252. 150 J. Cooper and R. N. Zare, J. Chem. Phys., 1968, 48, 942. 151 J. Cooper and R. N. Zare, in Atomic Collision Processes, ed. S. Geltman, K. T. Mahanthappa, and W. E. Brittin, Gordon and Breach, Science Publishers, New York–London–Paris, 1968, vol. XI-C, p. 317. 152 S. Geltman, Phys. Rev., 1958, 112, 176. 153 K. J. Reed, A. H. Zimmerman,H. C. Andersen and J. I. Brauman, J. Chem. Phys., 1976, 64, 1368. 154 T. Seideman, J. Chem. Phys., 1997, 107, 7859. 155 T. Seideman, J. Chem. Phys., 2000, 113, 1677. 156 T. Seideman, Phys. Rev. A, 2001, 6404, art. no. 042504. 157 T. Seideman, Annu. Rev. Phys. Chem., 2002, 53, 41. ƒ A detailed review of the utility of photoelectron angular distributions as probes of molecular dynamics. 158 V. Blanchet, S. Lochbrunner, M. Schmitt, J. P. Shaffer, J. J. Larsen, M. Z. Zgierski, T. Seideman and A. Stolow, Faraday Discuss., 2000, 115, 33. 159 V. Blanchet, M. Z. Zgierski and A. Stolow, J. Chem. Phys., 2001, 114, 1194. 160 M. Schmitt, S. Lochbrunner, J. P. Shaffer, J. J. Larsen, M. Z. Zgierski and A. Stolow, J. Chem. Phys., 2001, 114, 1206. 161 S. Lochbrunner, T. Schultz, M. Schmitt, J. P. Shaffer, M. Z. Zgierski and A. Stolow, J. Chem. Phys., 2001, 114, 2519. 162 M. A. Johnson, personal communication. 163 J. A. Davies, R. E. Continetti, D. W. Chandler and C. C. Hayden, Phys. Rev. Lett., 2000, 84, 5983. 164 T. Suzuki, L. Wang and H. Kohguchi, J. Chem. Phys., 1999, 111, 4859. 165 M. Tsubouchi, B. J. Whitaker, L. Wang, H. Kohguchi and T. Suzuki, Phys. Rev. Lett., 2001, 86, 4500. PhysChemComm, 2002, 5(25), 165–177 177
ISSN:1460-2733
DOI:10.1039/b209239e
出版商:RSC
年代:2002
数据来源: RSC