Low-energy ionic collisions at molecular solids.

ion mechanism, discussed in section 1.1.2. Figure 27 illustrates the reactive scattering mechanism with four representative snapshots of a Cs scattering trajectory in a classical MD simulation. The abstraction reaction is driven by the ion−dipole attraction force between the Cs ion and an adsorbate molecule. The impinging projectile first releases part of its initial energy to the surface (Figure 27b) even without direct collision with the adsorbate. Subsequently, the projectile pulls the adsorbate gently away from the surface in its outgoing trajectory (parts c and d of Figures 27 in sequence), leading to the formation of a Cs−molecule complex. The velocity of the outgoing Cs must be slow enough to accommodate the inertia of the adsorbate. As a result, adsorbates of low mass and small binding energy are efficiently abstracted. A heavier projectile like Cs transfers more energy to the target surface, and its lower velocity in the outgoing trajectory enhances the efficiency of reactive scattering events. Detailed aspects of Cs reactive scattering and its application for surface analysis have been reviewed. Table 9. Hyperthermal Energy Collisions at Condensed Molecular Solids method (projectile ion) system aim/observations refs reactive scattering and LES (Cs) H2O−D2O rate and activation energy of self-diffusion and H/D exchange of water 462, 476, 479, 496 H3O −water ice affinity of protons for the ice surface and proton transfer mechanism 478−480 H3O −H2O−D2O hydronium ion-mediated proton transfer at the ice surface 495 OH−H2O−D2O hydroxide ion-mediated proton transfer at the ice surface 497 HCl−water ice molecular and ionized states of HCl on ice 457, 477 Na−water ice hydrolysis of Na 484 H3O −NH3−water ice incomplete proton transfer from H3O to NH3 on the ice surface 454, 458 H3O −amine−water ice proton transfer efficiency on ice is reversed from the order of amine basicity 502 CO2−Na−water ice CO2 hydrolysis is not facilitated by a hydroxide ion 463 NO2−water ice NO2 hydrolysis produces nitrous acid 465 SO2−water ice SO2 hydrolysis occurs through various intermediates 511 C2H4−HCl−water ice electrophilic addition reaction mechanism at the condensed molecular surface 466 ethanol/2-methylpropan-2-ol−water ice SN1 and SN2 mechanisms at the condensed molecular surface 505 NH3−water ice and UV irradiation ammonium ion formation 608 CH3NH2−water ice and UV irradiation protonated methylamine formation 483 CH3NH2−CO2−water ice and UV irradiation glycine and carbamic acid formation 464 NaX−water ice (X = F, Cl, Br) surface/bulk segregation and transport properties of electrolyte ions 472−474 reactive scattering (Cs) CO and CO2 on Pt(111) mechanism of Cs + reactive ion scattering 89 Ar, Kr, Xe, and N2 on Pt(111) adsorbate mass effect on the reactive ion scattering cross-section 609 C2H4 on Pt(111) dehydrogenation mechanism of ethylene to ethylidyne 459, 610 C2D4 and H on Pt(111) ethylidene intermediate in H/D exchange reaction with ethylene 80, 610 reactive scattering (H) water ice and alcohol H2 + formation 469 CS (Ar) water ice−chloromethanes (CCl4, CHCl3, CH2Cl2) except CCl4, others undergo diffusive mixing 174 water ice−simple carboxylic acids structural reorganization on the ice film 175 water ice micropore collapse in the top layers of the ice film 176 water ice−butanol 494 Figure 27. Illustration of the reactive scattering mechanism of a Cs ion in four snapshots of a scattering trajectory from a Pt(111) surface: (a) initial positions before impact, (b) impact of the Cs and energy release to the surface, (c) Cs pulling the adsorbate away in its outgoing trajectory, (d) slow outgoing Cs dragging the adsorbate along and forming a Cs−molecule association product. Reprinted with permission from ref 88. Copyright 2004 John Wiley and Sons, Inc. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5388 Figure 28 shows an example of reactive collision mass spectra, which were obtained on a D2O ice film exposed first to 0.5 L of HCl gas and then to varying amounts of NH3 gas at 140 K. The spectra show peaks at higher masses than Cs (m/z 133), viz., CsNH3 + at m/z 150, Cs(D2O)n + (n = 1, 2) at m/z 153 and 173, and CsHCl at m/z 168, indicating the presence of the corresponding molecules on the surface. The intensities of H/D-exchanged species represent their original concentrations on the surface, because H/D isotopic scrambling does not occur during the ion/surface collision time (<1 × 10−12 s). The conversion efficiency of a neutral adsorbate (X) into a gaseous ion (CsX) ranges from ∼10−4 for chemisorbed species to ∼0.1 for physisorbed small molecules. Typical product ion signal intensities for ice film surfaces are much stronger than those for chemisorbed species. Also, it is worthwhile to point out that reactive collisions of Cs are ineffective for detecting large molecules such as polymers or long-chain SAM molecules. The mass spectra in Figure 28 also show LES signals corresponding to pre-existing ions on the surface. The hydronium ions seen are produced by the spontaneous ionization of HCl on the ice surface, and they undergo proton transfer reactions with NH3 to generate ammonium ions. The spectra show characteristic H/D isotopomers of each species produced by H/D exchange reactions with D2O molecules. The LES signals due to preformed hydronium and ammonium ions exhibited sputtering thresholds at Cs impact energies of 17 and 19 eV, respectively. On the other hand, on pure H2O and NH3 surfaces, these ions were emitted only above ∼60 eV due to their formation during secondary ion emission. It was also found that ultra-low-energy (a few electronvolts) collision of H with the ice surface can produce H2 +. The reaction proceeds more efficiently on amorphous solid water than crystalline water, reflecting differences in the surface concentration of dangling O−H bonds. Simple alkanols also behave in the same manner. The combined occurrence of reactive scattering and LES provides a powerful means to probe both neutral molecules and ions on surfaces and, therefore, to follow reactions on ice surfaces such as the ionization of electrolytes and acid−base reactions, which are described below. 7.2. Surface Composition and Structure Impurities in ice become concentrated in the quasi-liquid layers in the surface and at grain boundary regions due to the “freeze concentration effect”, and this has important consequences for atmospheric reactions on ice surfaces. However, there appear to be numerous exceptions to this general trend, where the surface segregation behavior of the dissolving species and their bulk solubility are determined by thermodynamic factors specific to individual chemical species. A good example is the formation of stable bulk phases of clathrate hydrates. Chemical specificity in the segregation phenomena can be studied by monitoring the surface populations of the dissolving species during the slow annealing of ice samples. Kang and coworkers examined these propensities in Na and halide ions at the surface and in the interior of ice films. They ionized NaF, NaCl, and NaBr molecules on ice films by the vapor deposition of the salts, and the variation in the surface population of the ions was monitored as a function of the ice temperature for 100−140 K by using LES. As shown in Figure 29, the LES intensities of Na and F− ions decrease with an increase in temperature above ∼120 K, whereas the Cl− and Br− intensities remain unchanged. The results indicate that Na and F− ions migrate from the ice surface to the interior at the elevated temperatures. The migration process is driven Figure 28. Cs reactive scattering and LES spectra monitoring the H3O −NH3 reaction on ice. The D2O film [3−4 bilayers (BLs), 1 BL = 1.1 × 10 water molecules cm−2] was exposed first to 0.5 L of HCl to generate hydronium ions and then to NH3 at varying exposures: (a) 0.02 L, (b) 0.3 L, (c) 0.7 L. The sample temperature was 100 K. The Cs collision energy was 30 eV. Reprinted with permission from ref 454. Copyright 2001 John Wiley and Sons, Inc. Figure 29. Surface populations of Na (□), F− (▲), Cl− (◇), and Br− (●) ions as a function of the ice film temperature measured from LES intensities of the ions. NaF, NaCl, and NaBr were deposited for a coverage of 0.8 ML for each salt on a D2O ice film grown at 130 K. The LES signals were measured at the indicated temperatures of salt adsorption. The LES intensities are shown on the normalized scale with the intensity at 100−105 K as a reference. The Cs beam energy was 35 eV. The figure is drawn on the basis of the data in refs 473 and 474. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5389 by the ion solvation energy, and it requires that surface water molecules have enough mobility to facilitate ion passage at temperatures above 120 K. It is worth noting that such a segregation behavior for ice agrees with the negative adsorption energy of these ions at water surfaces predicted by the Gibbs surface tension equation and MD simulations. An interesting property of hydronium ions observed in recent studies is that they preferentially reside at the surface of ice rather than in its interior. Evidence of this property has come from a variety of experimental observations over the past decade. The adsorption and ionization of HCl on an ice film promotes H/D exchange on the surface. However, vertical proton transfer to the film interior is inefficient. Continuous exposure of HCl gas on the ice film led to saturation in the hydronium ion population at the surface, and the amount of HCl uptake required for this saturation was independent of the thickness of the ice film. These observations suggest that protons stay at the ice surface and hardly migrate to the interior. This behavior can be attributed either to the active trapping of protons at the surface or to the lack of proton mobility to the ice interior. The observation of asymmetric transport of protons at an ice surface and in its interior led to the conclusion that protons have a thermodynamic propensity to reside at the ice surface. The surface preference of the proton allows reinterpretation of the results for the H3O + SL experiment conducted by Cowin et al., who showed that H3O + ions deposited onto an ice film surface remained on the surface over a wide range of temperatures (section 6.9). This observation was initially interpreted to indicate the immobility of protons through the ice films. However, the thermodynamic affinity of a proton for the ice surface and the proton’s known mobility in the interior suggest an alternative explanation of the observation consistent with the well-known proton transport mechanism in ice and the observations of related experiments. More recently, the ice systems used in the H3O + SL experiment were re-examined by careful control experiments, confirming the asymmetric transport behavior of protons. The properties of hydroxide ions were also studied. Hydrolysis of Na atoms on an ice surface produced Na and hydroxide ions at the surface. The LES intensity of hydroxide ions increased as the temperature was raised from 95 to 135 K. The result showed that hydroxide ions have a tendency to float on the ice surface, similar to hydronium ions. The LES intensity of Na ions shows that these ions migrated to the ice interior, in agreement with the observation of the ion segregation experiments with sodium halide salts mentioned above. Studies of the dissolution behavior of alkali-metal ions at liquid water surfaces show that polarizable ions such as large halide anions (Br− and I−) have a propensity to reside at the surface, whereas F− and Na avoid surface residence. These trends agree nicely with the observations made for the ice systems. For the cases of hydronium and hydroxide ions, their interfacial distributions for liquid water are a controversial issue in current investigations of the subject. Various experimental methods (vibrational sum-frequency generation spectroscopy, ζ potentials, and gas bubbles) and theoretical calculations report different results regarding whether water surfaces are acidic or basic relative to the interior. A chemical sputtering method with hyperthermal noble gas ions has been used to probe the structural changes of condensed molecular solid surfaces. It is known that carboxylic acids can exist either in a chainlike crystalline form or as dimers in the solid state, depending upon the preparation temperature. When acetic acid is vapor-deposited onto a metal surface at 110 K, it exists in a dimeric amorphous form, but on a thin ice film, it exists as a chainlike crystalline phase at the same temperature. These two forms of acetic acid were distinguished by detecting the selective emission of acetic acid molecular cation from the amorphous phase using 30 eV Ar collisions. The crystalline form did not produce the molecular cation in the sputtering spectra; instead the formation of the CH3CO + fragment was the characteristic signature of this structure. Formic acid followed a behavior similar to that of acetic acid, but propionic acid suppressed the formation of the molecular cation in the CS spectra since it existed as dimers in its crystalline form also. Hyperthermal projectiles do not closely approach the core of individual surface atoms, but are reflected at a larger distance from the surface so they feel the surface as a relatively flat structure. For this reason, hyperthermal ion scattering may not be an atomistic structural probe of a surface. However, it can be used to monitor changes in the ensemble-averaged structure of surfaces. Cyriac and Pradeep observed that the scattering of ultra-low-energy (∼1 eV) Ar ions is sensitive to the surface morphology of ice films. The Ar scattering intensity from an amorphous solid water film increased by a factor of 2 as the temperature increased from 110 to 125 K (Figure 30). Such a change was absent in the case of crystalline ice films and for other condensed molecular solids. The Ar intensity dropped around ∼160 K due to water desorption and then subsequently increased as the bare copper surface was exposed. These results suggest that an amorphous ice film undergoes a structural transformation at 110−125 K, which is below the onset temperature of the glass transition (136 Figure 30. Scattering intensity variation of 1 eV Ar collisions at bare copper (○), 50 ML of ASW(H2O) (□), 50 ML of ASW(D2O) (■), and 50 ML of crystalline ice (H2O) (●). The continuous gray line shows an approximate representation of the overall behavior of ASW. Inset: typical Ar scattering mass spectra of 50 ML of ASW for three different temperatures and averaged for 50 scans. The collision energy was 1 eV. Reprinted from ref 176. Copyright 2008 American Chemical Society. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5390 K) and well below the appearance of quasi-liquid layers. The observed structural change may involve the collapse of micropores in the top layers of the ice surface. 7.3. Transport Properties Several unique reaction properties found at an ice surface compared to its liquid-phase counterpart are intimately related to the difference in molecular mobility in the two phases. Selfdiffusion at ice surfaces has been studied by measuring the kinetics of the diffusional mixing of H2O and D2O molecules at the surface. In this study, a thin H2O ice film was prepared and then covered with a fractional layer of D2O. The diffusion of water molecules in the top one to two molecular layers of the surface gradually changed the relative populations of H2O and D2O in the outermost layer and was monitored as a function of time by reactive scattering. The study indicated that the interlayer diffusion in the surface took several seconds at 140 K and about 1 h at 100 K. Temperature-dependent kinetic measurements yielded a self-diffusion activation energy (Ea ) of 14 ± 2 kJ mol−1. In comparison, the self-diffusion activation energy in bulk ice (Ea ) was measured to be 71 ± 4 kJ mol−1 in laser-induced thermal desorption experiments. These studies show that surface diffusion occurs significantly faster than bulk diffusion in the temperature regime of 100−140 K, and the gap between the two diffusion rates widens exponentially as the temperature decreases due to the large difference in the activation energies. This illustrates that, if reactions of ice occur at low temperatures, they will occur preferentially at the surface where molecules have a much higher mobility, rather than in the interior. The diffusion of chloromethane molecules (CCl4, CHCl3, CH2Cl2, and CH3Cl) through ice films has been studied by the CS method. To show the sensitivity of CS, the CS spectra at two different temperatures (125 and 130 K) from a system prepared by depositing 50 ML of CHCl3 followed by 250 ML of H2O at 110 K are shown in Figure 31. The CHCl3 concentration increases at the surface with increasing temperature. The study showed that, except for CCl4, other chloromethanes investigated, viz., CHCl3 and CH2Cl2, underwent diffusive mixing with amorphous solid water (ASW) in the temperature range of 100−150 K. CCl4 was not able to diffuse through more than four overlayers of ASW. The hydrogen bond network of the ASW film restricted the transport of CCl4 molecules. Other molecular solids, D2O and CH3OH, also acted as barriers to diffusive mixing of CCl4. The interaction energy between chloromethanes and water in the solid state was in the order CCl4 < CHCl3 < CH2Cl2 < CH3Cl, which is the reverse order of the liquid-phase interactions. Considering that the overall interaction between chloromethanes and water is based on the atomic charge of chlorine and its molecular polarizability, replacement of a Cl atom by a H atom can have a significant effect on diffusivity in ice. Using Ar sputtering, it was found that 1-butanol undergoes diffusive mixing with water ice. Even after deposition of 1000 ML of ASW over solid 1-butanol, both species are observed on the surface. By contrast, water is not seen when 1-butanol is deposited over ASW. The results suggest that long-chain alcohols may act as barriers to H2O diffusion because of their hydrophobic nature. Proton mobility is a fundamental and important property in the physics and chemistry of ice. As discussed in section 7.2, there is a general consensus that a proton is mobile in ice at elevated temperatures. Also, a proton tends to reside at the surface of ice where it is stabilized. The hyperthermal energy ion probe offers a tool to pursue the proton transfer behavior by looking into the H/D isotopomers of water molecules and hydronium ions at ice surfaces generated by proton-induced H/D exchange reactions. Below 120 K, at which the rotational and diffusional motions of water molecules are frozen, a proton hopping relay (Grötthuss mechanism) is the only possible mechanism of proton transfer in ice. This is evidenced by the LES detection of H3O + at the surface of the H2O/proton/D2O film and also by the absence of D-substituted hydronium ions at the surface. Protons can move only across a limited distance by this mechanism, but nevertheless, this makes protons a unique mobile species in low-temperature ices, whereas water molecules and other foreign species are virtually frozen in position. Upon the activation of molecular rotations which occurs at temperatures above 125 K, the hop-and-turn process starts to occur involving the coupling of proton hopping and water molecule reorientation. All these proton transport processes can occur below the onset temperature (130 K) of water molecule diffusion near the ice surface, indicating that proton transfer can occur more easily than water self-diffusion. Proton transfer along the surface of ice was examined through the measurement of the H/D exchange kinetics of surface H2O and D2O molecules in the presence of excess protons generated from HCl ionization. Protons were transferred from hydronium ions mostly to the adjacent water molecules when the surface temperature was low (70 K), but the rate and propagation range of the proton transfer increased as the temperature increased above 90 K. This finding shows that the proton transfer process at an ice surface is thermally activated, and there exists an energy barrier of substantial magnitude (10 ± 3 kJ mol−1) for the proton transfer. This is in agreement with the thermodynamic affinity of protons for ice surfaces, a conclusion derived from independent observations for the surface segregation of protons (section 7.2). On a pure ice film surface in the absence of externally added protons, the H/D exchange reaction occurrs slowly compared Figure 31. Intensities of the CHCl2 + and CHCl3 + peaks are increased due to the change in concentration of CHCl3 on the surface with increasing temperature. The projectile ion is 30 eV Ar, and the system is 50 ML of CHCl3@250 ML of ASW. With the temperature rise from 125 K (lower trace) to 130 K (upper trace), more CHCl3 diffuses through ice overlayers. Reprinted from ref 174. Copyright 2007 American Chemical Society. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5391 to that observed on the proton-rich ice surface. The activation energy of the reaction was also higher (17 ± 4 kJ mol−1) on the pure ice surface. This is because H/D exchange on this surface requires the formation of ion pairs (H3O + and OH−) which require thermal energy in addition to the occurrence of proton transfers. Similar experiments performed with excess hydroxide ions on ice films showed that hydroxide ions, like hydronium ions, mediate proton transfer at an ice surface and promote the H/D exchange of water molecules. The Arrhenius activation energy of 9.4 ± 2.0 kJ mol−1 was estimated for the proton transfer mediated by hydroxide ions, which is comparable in magnitude to the activation energy for the hydronium ion-mediated proton transfer. These studies indicate that proton transfer at an ice surface involves a substantial energy barrier, regardless of whether the process is mediated by hydronium ion or hydroxide ion. 7.4. Acid−Base Chemistry Available spectroscopic tools to investigate acid−base chemistry at ice surfaces are few, for example, IR spectroscopy, metastable impact electron spectroscopy (MIES), and the hyperthermal ion beam techniques described in this review. Acid−base reactions are important in ice surface chemistry because they are intimately related to two important properties of ice surfaces discussed in the previous sections: (i) thermodynamic affinity of protons and hydroxide ions for ice surfaces and (ii) unique mobility of protons in ice. The simplest and most extensively studied example of an acid−base reaction on ice is the ionization of strong protic acids. Experiments performed using a hyperthermal energy Cs ion probe shows that HCl partially ionizes to hydronium and chloride ions on ice films at temperatures below 120 K and that undissociated HCl also exists. The presence of an undissociated HCl molecule on the surface was evident from the CsHCl signal, while hydronium ions (HD2O + and D3O ) and hydrated clusters (HD4O2 + and D5O2 ) represented the ionized form. The results indicate that, although HCl is a strong acid which ionizes completely in an aqueous solution, it acts as a weak acid on ice surfaces at low temperatures, and the degree of ionization varies with the temperature and morphology of the ice surface. As a model case of acid−base reactions involving proton transfer on ice, Park et al. conducted detailed investigations of the reactions between the hydronium ion and amine molecules, including NH3. For example, the hydronium ion−ammonia system was prepared by doping a D2O ice film with HCl followed by dosing with NH3 gas. The ratio of proton donors to acceptor was changed by varying the NH3 concentration. Reactive scattering spectra given in Figure 28 show the donor and acceptor species present on the surface at varying concentrations of NH3. 454 The absence of NH3 molecules on the surface indicates that proton transfer from hydronium ion to ammonia is complete. At a high coverage of NH3 (spectrum b), however, the CsNH3 + signal shows that a substantial portion of NH3 remains unconsumed despite the coexistence of D3O . In fact, all donor and acceptor species of the reaction (NH3, NH3D , D2O, and D3O ) coexist even beyond the equivalence point of the titration. This shows that the acid−base reaction does not reach a true equilibrium on the ice surface due to incomplete proton transfer, in contrast with its complete occurrence in an aqueous solution at room temperature. The extent of proton transfer was evaluated by measuring the reaction quotient (Q) of the reaction, defined by Q = [H2O][BH ]/[H3O ][B], where B is an amine, as a function of amine exposure. As Figure 32 shows, the Q value on an ice surface is much smaller than the equilibrium constant of the same reaction in the gas phase or in an aqueous solution, and Q decreases with increasing amine exposure. The relative proton transfer efficiency by various amines deduced from these results follows the order NH3 > (CH3)NH2 ≈ (CH3)2NH, but this is opposite the trend in intrinsic basicity of amines or their basicity in aqueous solutions. Thermochemical analysis suggests that incomplete solvation of reactant and product species at the ice surface reduces the proton transfer efficiency and reverses the order of the proton-accepting abilities of amines. The hydronium ions formed by UV irradiation of ice transfer protons to methylamine molecules adsorbed on the film surface to form methylammonium ions (CH3NH3 ), and the proton transfer occurs via a tunneling mechanism (hopping relay) at low temperature (50−130 K). The methylammonium ion was stable at the ice surface, in contrast with its spontaneous deprotonation to a neutral methylamine molecule in aqueous solution. Later, the study was extended to ammonium ion (NH4 ), which was formed through UV photolysis of an NH3− H2O ice mixture. 503 The IR spectrum of NH4 + suggests the possibility that it is formed in interstellar ice particles and contributes to the 6.85 μm band discovered in the astronomical observations of dense molecular clouds using an infrared telescope. 7.5. Chemical Reactions Reactions between hydrogen halide and the alcohols ethanol and 2-methylpropan-2-ol are well-known to follow SN2 and SN1 pathways, respectively, in the liquid phase. The intermediate states of these reactions are protonated alcohols and carbocations, which exist only for transient times and rapidly convert to alkyl halides. However, these SN1 and SN2 intermediates are stabilized when the corresponding reactions occur on solid alcohol films at a low temperature. Figure 32. Reaction quotient (Q) for proton transfer from hydronium ion to NH3 (●), CH3NH2 (□), and (CH3)2NH (◆) measured as a function of amine exposure on an ice surface. The surface hydronium ions were produced by adding 0.3 L of HCl onto the ice film at 60 K and then warming it to 140 K. The donor and acceptor populations were measured from the corresponding LES and RIS signals. Reprinted with permission from ref 502. Copyright 2007 John Wiley and Sons, Inc. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5392 Specifically, HBr reacts with the ethanol surface, exclusively producing protonated ethanol species. The reaction between HBr and 2-methylpropan-2-ol resulted in protonated 2methylpropan-2-ol and the tert-butyl cation in 20% and 78% yields, respectively. Importantly, alkyl bromides, which are the final products of the reactions in liquid solvents, were hardly detected on the molecular films. This indicates that the reactions on the frozen films are kinetically controlled, in contrast with the thermodynamically controlled reactions in the liquid phase. The kinetic barriers on the cold molecular surfaces stabilize the ionic intermediates (protonated alcohols and tertbutyl cation) and effectively block the completion of the SN1 and SN2 reaction pathways by impeding the diffusive encounter between the halide ion and the alcohol counterion. In an analogous study, the electrophilic addition reaction of ethene with HCl was investigated on frozen molecular films. The acid-catalyzed electrophilic addition reactions of alkenes are believed to occur through alkyl cation intermediates, but primary alkyl cations such as an ethyl cation have never been identified using spectroscopy, even for reactions in superacids. On a water ice film, the reaction of ethene with HCl initially produced the π complex of HCl and ethene at temperatures below about 93 K. LES was used to detect a C2H5 + signal on the surface at temperatures of 80−100 K, which indicated the formation and kinetic stabilization of an ethyl cation-like species. This species dissociated into ethene and hydronium and chloride ions at high temperatures, but it did not complete the final step on the potential surface to produce ethyl chloride. The ethyl cation-like species was not formed in the reaction of ethene with hydronium ion or when the reaction of ethene with HCl occurred on a frozen ethene film, indicating that the ethyl cation-like species was formed via direct proton transfer from molecular HCl to ethene in water solvating environments. The study shows evidence that the reaction involves an intermediate species that has an ethyl-like structure with ionic character, but it remains uncertain whether this species is actually an ethyl cation or a structure intermediate between the HCl−ethene π complex and its ionized state. Inorganic reactions studied on ice films to date include the hydrolysis of alkali-metal atoms and the reactions of simple oxide gases, viz., CO2, 463 NO2, 465 and SO2. 511 Na and OH− ions, produced by hydrolysis of Na, were efficiently solvated by water molecules at all temperatures investigated, whereas the sodium hydroxide molecule was found to solvate only at high temperatures. The OH− ions tend to reside at the film surface, whereas Na ions migrate to the film interior. The adsorbed Na atoms completely reacted away without forming neutral Na clusters on the surface when the Na coverage was lower than 1 ML. These observations were complementary to the results of the TOF SIMS and metastable impact electron spectroscopy (MIES) studies of Na hydrolysis on ice films obtained under the conditions of lower temperature (15−100 K) and higher Na coverage (>1 ML) (section 8.6). Reactions of acidic oxide gases at the surfaces of snow and ice particles are important to atmospheric chemistry and environmental sciences. While CO2 gas is unreactive to the ice films regardless of the presence or absence of excess hydroxide ions on the surfaces, SO2 511 and NO2 465 readily react with the ice surface to produce various chemical species even at low temperatures (80−150 K). Figure 33 shows spectra due to reactive scattering and LES from a D2O ice film on which SO2 is absorbed at 140 K. Spectrum a shows that SO2 adsorption produces signals of CsSO2 + (m/z 197), CsDSO2 + (m/z 199), and Cs(D2O)(SO2) + (m/z 217) in reactive scattering experiments. The CsSO2 + and Cs(D2O)(SO2) + signals indicate the presence of molecular SO2 adsorbates. Spectrum b shows the negative ion LES signals from the surface, which include OD− (m/z 18), SO2 − (m/z 64), DSO2 − (m/z 66), and DSO3 − (m/z 82). The spectra show that SO2 is transformed into various molecular anions by hydrolysis on the surface. When the ice film with SO2 adsorbates was warmed slowly from 80 to 150 K, the signals of various SO2-related species appeared and disappeared at different temperatures (see Figure 33c). The results indicate that physisorbed SO2 species (detected as CsSO2 ) sharply decrease during the temperature increase from 100 to 140 K. In the narrow temperature range of 130−150 K, DSO2 (detected as CsDSO2), DSO2, and DSO3 intensities grow at the expense of the decreasing SO2 − intensity, indicating the conversion of SO2 − to these species. Combined hyperthermal energy ion collision and TPD experiments indicate that the reaction of SO2 on deuterated ice produces three types of surface species: a solvated SO2 species with a partial negative charge, a DSO2 species, and an anionic DSO3like species. Figure 33. (a) RIS mass spectrum obtained from a D2O ice film exposed to SO2 gas. (b) LES spectrum of negative ions. The ice film (4 BL thickness) was exposed to 0.2 L of SO2 at 80 K, and the RIS and LES measurements were made at 140 K. The RIS signals at masses above m/z = 190 amu/charge were magnified by the factors indicated. (c) Temperature-programmed RIS and LES measurements for the signals of interest detected on the surface exposed to SO2 at 80 K. The RIS yield on the left ordinate iss defined as the ratio of the RIS product to the Cs signal intensity (CsX:Cs). The temperature ramping rate was 1 K s−1. Data were taken from Figures 1 and 2 of ref 511. Copyright 2009 American Chemical Society. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5393 Table 10. SIMS Data for Condensed Molecular Solids primary ion (keV range) system (15−200 K) some sputtered species aim/observations refs Au water ice (H2O)nH , (H2O)n , (H2O)n−1OH + intensity of (H2O)n + > (H2O)nH + ∼ (H2O)n−1OH 517 Au3 + water ice (H2O)nH , (H2O)n , (H2O)n −1OH + intensity of (H2O)nH + > (H2O)n , when n < 20 517 Au3 , C60 + water ice (H2O)nH , (H2O)n , (H2O)n−1OH + 10 increase in yield compared to that of Au 517 Au3 , C60 + water ice (H2O)nOH − less than 10 times compared to that of positive ions 517 Au3 , C60 + water ice depth profiling 591−593 C60 + water ice (H2O)nH , (H2O)n + intensity of (H2O)nH + > (H2O)n , when n < 20 517 C60 + trehalose film molecular depth profiling 520, 587 histamine−ice molecular depth profiling 589, 590 LB film molecular depth profiling 611−614 Ar/C60 / Ga MX−water (MX = LiCl, KI, NaI, NaCl, etc.) ((H2O)n M ± (M = Li, Na, K, Cs or F−, Cl−, I−) cation water clusters > anion water clusters 563, 564 He/Ar CF4@Pt(111) CFx + (x = 0−3), F 615 He water ice H, H3O , Ni(H2O) + existence of quasi-liquid form and dewetting at 160 K 530, 534 CF4−D2O efficient production of (D2O)D ions 2 orders of magnitude higher than that of a pure D2O film 568 methanol−D2O (CH3OH)H, (CH3OH)D, (CH3OD)D + complete intermixing above 136 K, hydrophilic hydration above 120 K, H/D exchange above 140 K, determined Tg of methanol 515, 530, 558, 565 ethanol−D2O (C2H5OH)H, (C2H5OH)D, (C2H5OD)D , CH2OH + complete intermixing above 140 K 566 acetic acid−D2O (CH3COOH)H, (CH3COOH) D, (CH3COOD)D , CH3CO + complete intermixing above 130 K, determined Tg of acetic acid 487 methylamine−D2O (CH3NH2)H, (CH3NH2)D, CH2ND2 + complete intermixing above 140 K 567 methane−D2O CH5, C2H3, other common ions complete intermixing at 15 K 568 CO−D2O D, D3O, C, O, CO, C3O2 complete intermixing at 15 K 616 CO2−D2O D, D3O, C, O, CO, CO2 no complete intermixing at 15 K, determined Tg of CO2 of 50 K 573, 616 CD3OD−methylamine (CH3NH2)H, CH2ND2, (CH3NH2)D + complete intermixing above 125 K 567 HCOOH−D2O ice (D2O)H, (D2O)D, (HCOOH) H HCOOH stays mainly on the surface due to hydrophilic hydration, determined Tg of HCOOH 516, 572, 581 C3H7OH−D2O ice C2H3, (D2O)D, etc. hydrophilic hydration at 100−145 K, hydrophobic hydration at > 145 K 516 C6H14 /C6F14 -D2O ice CH3 , C2H3 , CF, CF3 , etc. dissolve into the bulk due to hydrophobic hydration 516 C8H18−water ice C2H3, (D2O)D, etc. hydrophobic hydration, incorporate into the bulk at <120 K 557 C8H18−CH3OH C2H3, (CD3OD)D, etc. stays on the surface until 135 K 557 pyridine−D2O (C5H5N)H, (C5H5N)D, C4Hn , C3HnN , etc. D2O dissolves in pyridine film above 110 K, pyridine stays on D2O film until 180 K 559 benzene−D2O C3H3, C4Hn, C6Hn, etc. benzene dissolves in D2O film above 120 K by hydrophobic hydration, benzene stays on D2O film 559 butane−methanol C2H3, CD3, etc. butane is incorporated into methanol below 70 K 535 NH3−HCl@water ice NH4, Cl−, OH−, etc. hydration of NH4Cl above 100 K 560 HCl−NH3@water ice NH4, Cl−, OH−, etc. hydration of NH4Cl and HCl above 40 K 561 NH3−HCOOH−water HCO, other common ions produce NH4HCO2 as reaction product, H/D exchange reaction above 140 K 581 LiX−water (X = Cl, Br, I) Li, X− solubility increases at 160 K, evidence of deeply supercooled water 509, 538, 539, 541 NaCl−water Na, Cl− Na ion hydrated preferentially at low temperatures 542 LiI−ethanol (C2H5OH)Li, (LiI)Li, etc. LiI incorporated into the bulk 546 CCl2F2 CClF2 , CCl2F , CF, F crystallization and dewetting at 57 K 617 n-hexane CmHn + (m = 1−6) glass transition at 110 K, dewetting at 130 K 537 ethylbenzene intermixing at ∼80 K, glass transition at ∼118 K 618 water, ethanol, etc. on p-Si (100) onset of self-diffusion temperature 552−554 Xe−water ice Xe hydrated in the bulk of water up to 165 K 540 D2O−hexane C2H3 dewetting at 165 K due to glass−liquid transition 532 D2O−dipalmitoyl-sn-glycero-3phosphocholine C2H3 , NCH3 , etc. hydrophobic water adsorption starts at 133 K 574, 575 D2O− bis[trifluoromethanesulfonyl]imide ([emim][Tf2N]) [emim], C2H5 , CF3 , [Tf2N] −, F−, etc. phase transition in water ice identified 576 [emim][Tf2N] [emim] , C2H5 , crystallization at 200−220 K 536 Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5394 The study of hydrolysis of NO2 on ice films shows that NO2 adsorbs molecularly on the ice surface at 90 K. Upon heating the surface to 140 K, NO2 adsorbates are readily converted into nitrous acid (HONO) on the surface. The implication of this observation for the atmospheric chemistry of NO2 is that nitrous acid gas will be formed by the heterogeneous hydrolysis of NO2 on condensed ice surfaces even during the night Table 10. continued primary ion (keV range) system (15−200 K) some sputtered species aim/observations refs D2O−CH3NH2 (CH3NH2) H, CH2ND2, (CH3NH2)D + H/D exchange reaction above 140 K 567 CD3OD−CH3NH2 (CH3NH2)H, CH2ND2, (CH3NH2) D + H/D exchange reaction above 125 K 567 Ar Cl2 /Cl2O/ClONO2−water ice Cl1−5, Cl1−5, HOCl−, NO, H2NO3 , water clusters, etc. above 120 K, Cl2 undergoes reaction with ice surface to form HOCl, water, and HCl 569, 570, 583 ClONO2−HCl−water ice Cl1−3, H2OCl, NO2, NO3 formation of nitric acid at the surface 570 Co-depositing SO3−H2O (H2SO4)H, (H2SO4)H2O, (H2SO4)(H2O)1−3H + formation of sulfuric acid monohydrate and tetrahydrate 582 Figure 34. Positive ion SIMS spectrum of water ice using Au primary ions. Three series of cluster ions, (H2O)n , (H2O)nH , and (H2O)n−1OH , were observed; the first one is the most intense. Reprinted from ref 517. Copyright 2010 American Chemical Society. Chemical Reviews Review dx.doi.org/10.1021/cr200384k | Chem. Rev. 2012, 112, 5356−5411 5395 without the absorption of sunlight, which is in accordance with field observations. The examples shown above demonstrate that ice surfaces can be used to take a “frozen snapshot” of a reaction and to capture the reaction intermediates. Characterization of reaction intermediates trapped on the ice surface using ion scattering and sputtering methods reveals important clues for the reaction mechanism. Such investigations may also be useful for understanding the corresponding reactions at gas/liquid interfaces. 8. SIMS ANALYSIS OF CONDENSED MOLECULAR SOLIDS Although organic SIMS was an active area of research in the 1980s, the literature available on SIMS or fast atom bombardment (FAB) analysis of frozen molecular materials is mainly due to Michl and co-workers. More recently, SIMS has become a valued tool in low-temperature studies (see Table 10). Sputtered species and their yields from condensedphase molecular solids provide molecular information as well as the composition for a specific chemical environment. These aspects are important in fundamental understanding related to upper atmospheric chemistry. Inclusion of temperatureprogramming capabilities in TOF SIMS, viz., temperatureprogrammed (TP) TOF SIMS, is a remarkable development in the analysis of condensed molecular solids. It provides unique information on compositional changes/reactions with respect to temperature. A notable number of such investigations have been carried out by Souda and co-workers. The use of cluster ion SIMS provides higher ion yields without significant chemical damage of condensed molecular solids. In some cases, yields were 10 times higher compared to that of an atomic ion source. For example, Au3 + or C60 + is better than Au primary ions for generating protonated clusters. The yield was equivalent to ∼1830 molecules per C60 + collision at 15 keV, while these numbers were ∼1200 and 94 for Au3 + ion and atomic Au ion projectiles, respectively. It is suggested that the C60 + sputter yield is about 3000 amu/keV for organic or ice films and 800 amu/keV for metallic targets. Protonated water clusters, (H2O)nH , and normal water cluster ions, (H2O)n + (n up to several tens), are abundant in the sputtering spectra of pure water ice films. See Figure 34 for the positive ion spectrum of an untreated water ice film at ∼100 K by 15 keV Au bombardment. Clusters with up to n = 50 have been observed. The protonated form of the water clusters is formed predominantly with cluster ion sources C60 + and Au3 + when the value of n is lower than 20. The increased density of surface protons leads to high yields of protonated species. The case is reversed, i.e., water cluster radical ions are abundant compared to the protonated form, when n is larger. Some long-range damage to the crystal structure as a result of sputtering explains the low abundance of protonated water clusters. Negatively charged water clusters, (H2O)nOH −, were also observed by both atomic and cluster primary ions, but the yield is far smaller (∼10 times) compared to that of its positive counterpart. Dosing HCl onto an ice film doubles the yield of protonated water clusters. It was found that the yield of protonated water clusters is reduced significantly in the presence of adenine and alanine. This suggests a suppression effect, with the biomolecules taking up some of the available protons that form (H2O)H + in the pure water ice. In the case of solid methane, cluster ions CnHx + up to n = 20 are desorbed by low-energy He ion bombardment. MD simulation and experimental results on the nature of the neutrals ejected; solid C6H6 surfaces using atomic projectiles Ar and H2 + suggest that more than one mechanism operates. At submonolayer coverages of benzene molecules deposited on Ag{111} surfaces, the kinetic energy of the ejected neutrals usually ranges from 0.25 to 1.00 eV, and these are ejected as a consequence of collisions with substrate particles. For multilayer coverage, the energy of the neutrals shows more of a thermal nature. A peak corresponding to extremely low kinetic energy (0.04 eV) becomes dominant. The thermal emission may be due to exothermic chemical reaction of fragments formed in a molecular collision cascade of C6H6 molecules. Another aspect to be considered in SIMS analysis of condensed molecular solids is the surface charge. These materials may behave similarly to insulator surfaces and charge up, but there is limited support for this so far. It is shown that the charge pattern generated using a primary ion source on frozen (T < 188 K) ionic liquid [emim][EtSO4] was stable and could be viewed in a negative ion map of the surface.