Probing the Interaction of Hydrogen Chloride with Low-Temperature

May 6, 2011 - Gregory A. Grieves, Alex Aleksandrov, and. Thomas M. Orlando*. School of Chemistry and Biochemistry, Georgia Institute of Technology, ...
0 downloads 0 Views 1MB Size
ARTICLE pubs.acs.org/JPCA

Probing the Interaction of Hydrogen Chloride with Low-Temperature Water Ice Surfaces Using Thermal and Electron-Stimulated Desorption Babajide O. Olanrewaju, Janine Herring-Captain,† Gregory A. Grieves, Alex Aleksandrov, and Thomas M. Orlando* School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States ABSTRACT: The interaction and autoionization of HCl on low-temperature (80140 K) water ice surfaces has been studied using low-energy (5250 eV) electron-stimulated desorption (ESD) and temperature programmed desorption (TPD). There is a reduction of Hþ and H2þ and a concomitant increase in Hþ(H2O)n=17 ESD yields due to the presence of submonolayer quantities of HCl. These changes are consistent with HCl induced reduction of dangling bonds required for Hþ and H2þ ESD and increased hole localization necessary for Hþ(H2O)n=17 ESD. For low coverages, this can involve nonactivated autoionization of HCl, even at temperatures as low as 80 K; well below those typical of polar stratospheric cloud particles. The uptake and autoionization of HCl is supported by TPD studies which show that for HCl doses e0.5 ( 0.2 ML (ML = monolayer) at 110 K, desorption of HCl begins at 115 K and peaks at 180 K. The former is associated with adsorption of a small amount of molecular HCl and is strongly dependent on the annealing history of the ice. The latter peak at 180 K is commensurate with desorption of HCl via recombinative desorption of solvated separated ion pairs. The activation energy for second-order desorption of HCl initially in the ionized state is 43 ( 2 kJ/mol. This is close to the zero-order activation energy for ice desorption.

I. INTRODUCTION Heterogeneous processes play an important role in the chemistry occurring in the atmospheric polar regions since many reactions are relatively slow or forbidden in the gas phase.1 Surface interactions can lower the activation energy barriers for reactions that would normally not occur in the gas phase. Some of the most well studied heterogeneous interactions involve polar stratospheric cloud (PSC) particles. These PSCs provide a water/ice surface for reactions involving halogens and acids.2 For example, the reactions HCl þ ClONO2 f Cl2 þ HNO3 and HCl þ HOCl f Cl2 þ H2O are thought to occur on PSCs. Cl2(g) can photodissociate to form Cl atoms,1,3,4 which then react irreversibly with ozone. Since understanding heterogeneous reactions on PSCs will help improve current atmospheric models, determining the dynamics and details of surface and interface reactions on PSCs is critical. An important initial step is to determine whether reactants such as HCl are present in ionized or solvated forms versus chemisorbed molecular states. Low-temperature ices provide dynamic systems that will give insight into reactions on PSCs. Ice surfaces at higher temperatures are complicated by rapid desorption and adsorption, with evaporation rates on the order of 10161018 molecules/s cm2 between 180 and 210 K.5 The surface of the ice at these temperatures is constantly changing, and the structure and dynamic response of the terminal water layer is difficult to describe. Lower temperature and ultrahigh vacuum (UHV) conditions decrease the evaporation rates and allow us to obtain a basic understanding of how small molecules interact with ice, albeit at temperatures lower than those which are typical of the stratosphere. r 2011 American Chemical Society

Several studies have investigated the interaction of adsorbates such as HCl on low temperature ices under high or ultrahigh vacuum conditions. Kang et al. used reactive ion scattering (RIS) and low energy secondary ion mass spectrometry (SIMS) to measure the extent of ionization with temperature and found mostly molecular HCl below 70 K, a mixture of the ionized and molecular forms of HCl from 90 to 120 K, and complete ionization above 140 K.6 Molecular beam experiments7 and temperature programmed desorption (TPD)811 have found two types of HCl adsorption between 100 and 170 K, assigned as molecular HCl adsorption and a second ionized or hydrated species. Elegant theoretical studies by Buch et al. 12 showed that HCl can aid further dissolution of HCl molecules due to the Cl ion interaction with the hydrogen side of the incoming HCl molecule.12 In addition, spectroscopic studies have shown evidence of complete ionization at 80 K on ice films,13,14 while others report molecular bands that exist up to 125 K.15 Nearedge X-ray absorption fine structures (NEXAFS) spectroscopy has been used with photon-stimulated desorption (PSD) to probe the interaction of HCl with ice in the bulk and on the surface,16 respectively. These measurements provide evidence of ionization both in the bulk and at the surface; the molecular HCl signal was not observed when 1 L of HCl was adsorbed on 100 L Special Issue: Victoria Buch Memorial Received: October 28, 2010 Revised: April 14, 2011 Published: May 06, 2011 5936

dx.doi.org/10.1021/jp110332v | J. Phys. Chem. A 2011, 115, 5936–5942

The Journal of Physical Chemistry A of crystalline ice at 120 K. Changes due to the solvated Cl ion were also observed in the spectra. Several theoretical and modeling studies have attempted to describe solvation at ice surfaces. Gertner and Hynes employed molecular dynamics simulations of HCl on the surface of ice at 190 K and suggested that the HCl was initially incorporated as a molecular species into the ice lattice (not the bulk).17 In this case, there was a barrier for dissociation since the dynamic lattice response was not taken into account. However, more recent calculations by Bolton et. al18 and Buch et. al12 have clearly demonstrated the importance of defect configurations, the amount of HCl, and the dynamic lattice relaxation of the ice. Electron stimulated desorption (ESD) is an extremely sensitive surface specific probe that has been used to study the near-surface structure of ice.1922 The two features that make ESD inherently surface sensitive are the low penetration depth of electrons and the fact that cations must be produced at or near the surface to desorb from the sample. It has been shown in our previous work that the ESD products can be greatly affected by small changes in the local potential and excited state lifetimes of the surface molecules.20 Specifically, we demonstrated that changes in the ESD yield of water cluster ions can be understood in terms of HCl ionization induced restructuring of the ice surface. This leads to increased hole localization in the distorted (elongated) bonds associated with Cl solvation.22 In this paper we expand upon the previous studies and present more details concerning the interactions of low concentrations of HCl with low-temperature porous amorphous solid water (PASW), amorphous solid water (ASW), and crystalline ice (CI) surfaces. New data that corroborates and expands our previous model is presented which includes (i) the ESD cation yields from PASW, ASW, and CI ice as a function of HCl coverage, (ii) isotope exchange data from D2O ices demonstrating HCl autoionization, (iii) proton kinetic energy distributions that reflect HCl induced changes in the dangling bond sites, and (iv) TPD measurements demonstrating thermally activated HCl ionization on low-temperature ice. We also present a detailed description of the effect of HCl adsorption on multihole excited state lifetimes of ice and discuss how this provides insight into the local bonding geometry of the terminal water layers.

II. EXPERIMENT The experimental apparatus has been described in detail previously.22,23 However, there are a few modifications that have been made to improve the measurements and facilitate the study of HCl on ice surfaces. Briefly, the system consists of an ultrahigh vacuum (UHV) chamber with a typical base pressure of 2  1010 Torr equipped with a pulsed low-energy electron beam, a quadrupole mass spectrometer (QMS), a time-of-flight (TOF) mass spectrometer, and a cryogenically cooled sample mount. The substrate was a zirconia sample that could be varied in temperature from 80 to 600 K and rotated to face any port in the UHV chamber. The temperature was controlled via a computer and feedback loop for TPD measurements as well as controlling the morphology of the ice sample. The pulsed electron gun was at a 45° incident angle relative to the sample substrate normal and the electron beam energy could be varied from 5 to 250 eV. The electron flux density during a single 200 ns pulse was typically 6  1011 s1 cm2. The configuration of the electron gun, TOF mass spectrometer, and sample was

ARTICLE

designed to allow field free measurements to be taken with the TOF at a 90° angle relative to the sample plane. A 100 V extraction field on the TOF was typically used to collect the desorbing cations unless otherwise stated, such as during the velocity distribution measurements where field free conditions were implemented. Precautions were taken to eliminate any effects of codosing HCl with H2O and D2O since this may lead to the dosing of ionized species or unwanted isotope exchange. Therefore, the chamber was equipped with two completely separate dosing lines terminated in leak valves with directed dosing tubes. The dosing lines were thoroughly baked and equipped with liquid nitrogen traps. The HCl dosing line was purged to passivate the surface and equipped with a directed glass dosing tube to reduce contamination products from HCl reacting with metal surfaces. Using the QMS, we found the HCl to be free of any contamination. The water samples were purified by several freezepumpthaw cycles prior to dosing. The ice samples were vapor deposited at controlled temperatures designed to specifically grow PASW, ASW, or CI, at 80, 110, and 140 K, respectively. A minimum of 40 ML (ML = monolayer) was found to be sufficient to decouple the surface of the ice from the substrate, and all samples were >50 ML to ensure the measurements were not affected by the substrate. HCl coverages were determined by measuring the pressure change in the chamber during dosing corrected with an enhancement factor of 10 to account for the directed dosing. This is an upper limit to the amount of HCl and assumes unit sticking probability. At low temperatures, the HCl sticking probability is close to unity. However, at higher temperatures (specifically those samples dosed at 140 K) the coverage of HCl is most likely lower than calculated. For the ESD studies, the samples were grown, dosed with HCl and irradiated at the growth temperature of the respective ice substrate; 80 K (PASW), 110 K (ASW), and 140 K (CI).

III. RESULTS III.A. ESD of Cations from Ice Surfaces Containing HCl. The cations produced and desorbed during 250 eV electron impact of pristine ice can be seen in the solid lines of Figure 1. A detailed description of the cation yields from pristine ice can be found elsewhere.23 Briefly, the cation yield is dominated by Hþ with a much less H2þ, and an even smaller yield of Hþ(H2O)n. The cluster yield from PASW is the highest and it is typically 56 times larger than for CI, and about 23 times larger than for ASW.23 A very interesting and dramatic change in the cation yields occurs due to the presence of small amounts of HCl.22 This change can be seen in Figure 1 where the dashed gray lines represent the ESD yields after a dose of ∼0.1 ML of HCl. Dosing PASW with small amounts of HCl increased the cluster yield by approximately 1.3 times compared to that of pristine PASW ice while dosing HCl on ASW increased the cluster yield by 35 times. However, the largest relative increase in the cluster yield is observed for HCl on CI. Specifically, the Hþ(H2O)n=1 and Hþ(H2O)n=2 yields increased by over a factor of 5 and 8, respectively. This large enhancement required the presence of ∼0.3 ML HCl on H2O at 140 K. When CI was initially grown at 140 K, then lowered to 80 K for HCl dosing, and finally raised back to 140 K prior to irradiation, approximately the same increase was seen in the clusters for HCl samples that was both dosed and irradiated at 140 K. The CI sample dosed and irradiated at 80 K (data not shown) also showed an increase in 5937

dx.doi.org/10.1021/jp110332v |J. Phys. Chem. A 2011, 115, 5936–5942

The Journal of Physical Chemistry A

ARTICLE

Figure 2. Relative increase in cluster yield with HCl coverage for Hþ(H2O) (filled symbols) and Hþ(H2O)2 (open symbols) for (A) PASW, (B) ASW, and (C) CI. The temperatures indicated on each frame represent both the ice growth and HCl deposition temperatures. Figure 1. Cations produced and desorbed during 250 eV electron impact of pure (solid line) and HCl (dashed gray line) dosed PASW (bottom panel), ASW (middle panel), and CI (top panel). Upon addition of 0.1 ML HCl there is a large decrease in the Hþ and H2þ yields and an increase in the cluster yields. The temperatures indicated in each frame represent both the ice growth and HCl deposition temperatures.

the cluster ion yield, although the change was not as large as with dosing at higher temperatures. As discussed in more detail below, this probably indicates partial ionization of the HCl at 80 K. Figure 2 shows the relative increase in integrated total cluster yields as a function of HCl coverage. The filled symbols correspond to Hþ(H2O), and the dimer (Hþ(H2O)2) is shown as open symbols. CI, ASW, and PASW samples are represented by circles, squares, and triangles, respectively. Dosing each phase of ice at different temperatures provided a response similar to that obtained with increasing amounts of HCl. CI dosed at 140 K shows the largest relative change. The smallest increase is PASW dosed at 80 K, with the clusters approaching twice the value of the bare ice sample. While this relative increase is the smallest of all samples, it is important to note the PASW had the highest overall cluster yields. The cluster yield from ASW dosed with HCl at 110 K is 34 times the value of pure ice, reaching this value at ∼0.5 ML HCl coverage. Similar to the ASW sample, CI dosed at 140 K also reaches an upper limit for clusters between 0.5 and 0.6 ML. In summary, the CI shows the most dramatic increase in the cluster yield, followed by the annealed ASW and PASW ice films. In contrast to the cluster yields, the Hþ and H2þ yields decrease significantly upon adsorption of HCl. For example, after the first 0.1 ML dose of HCl on PASW, the Hþ and H2þ yield initially dropped ∼30%, and then decreased by ∼80% after ∼0.75 ML HCl.

III.B. Isotope Exchange Due to HCl Autoionization. To probe autoionization of HCl, we used D2O ice and examined the isotope exchange in the desorbing cluster ions. Figure 3 shows the protonated water dimer ions desorbed from HCl dosed D2O CI and ASW. Since exchange in the dimer ion is most extensive, we concentrate on this product. The data points for pure ice and HCl dosed ice are shown as circles and triangles, respectively. The solid and dashed thin lines show Gaussian fits for each mass, and the thick dark solid line shows the overall fit. Isotope exchange is most drastic for CI, with certain isotopes growing 18 times larger with HCl dosing (Figure 3A). This large increase is partially due to the initially low cluster yield from CI but demonstrates the importance of proton incorporation into the clusters. The singly hydrogen substituted water clusters shows the largest relative increase for both n = 1 and n = 2, growing approximately a factor of 10 with ∼0.4 ML HCl and further increasing a factor of 18 with ∼0.8 ML HCl. The yield of purely deuterated clusters also increased in response to HCl dosing, growing over a factor of 5 and 8 with ∼0.4 and ∼0.8 ML, respectively. While the percentage increase is largest for the singly substituted clusters, these H-containing clusters comprise only ∼30% of the total cluster yield for CI. The observed ratio of the purely deuterated and singly substituted clusters cannot be approximated with a statistical distribution of the D and H atoms. Similar to the case with CI, isotope exchange is evident when PASW (not shown) and ASW samples are dosed (Figure 3B). When the ASW sample grown at 110 K was dosed with HCl at 85 and 90 K, singly hydrogen substituted clusters (Hþ(D2O)2) were enhanced by a factor of over 4 and the doubly substituted hydrogen clusters (Hþ(HDO)2 increased by a factor of 10 with 5938

dx.doi.org/10.1021/jp110332v |J. Phys. Chem. A 2011, 115, 5936–5942

The Journal of Physical Chemistry A

Figure 3. Time-of-flight spectra of the protonated water dimer produced during pulsed 250 eV electron bombardment of D2O ice (50 ML) containing 0.8 ML of HCl dosed at 85 K. Frame A is CI grown at 140 K and Frame B is ASW grown at 110 K. The data points for pure ice and HCl dosed ice are shown as circles and triangles, respectively. The thin solid and dashed lines show Gaussian fits for each mass and the thick dark solid line shows the overall fit.

∼0.8 ML HCl. Dosing HCl on ASW at 110 K shows an increase comparable to the lower temperature dosing; however, the isotope exchange is less extensive in the doubly hydrogen substituted clusters when dosing at 110 K. This may indicate that the HCl forms a contact ion pair at the lower temperature and is available at the surface. III.C. Hþ Kinetic Energy Distributions with/without HCl. Due to the large proton yields, TOF distributions could be measured under field-free conditions (no extraction potential). The TOF data taken under these conditions can be converted into a kinetic energy distribution by using the Jacobian transformation I(E) = I(t)t3/(mL2), where I(t) is the intensity as a function of time (t), m corresponds to the mass of the ion, and L is the length of the flight path traveled by the ion. The kinetic energy distributions of protons from CI and PASW are shown in Figure 4. The solid squares are the Jacobian transformation of the data for pure ice samples and the open circles are for HCl dosed ices. The dotted and dashed lines in Figure 4A (CI) represent the Gaussian curves fit to the data with the sum shown as a solid line. The kinetic energy distribution from CI can be represented by two Gaussian curves with peak kinetic energies of ∼4 and ∼8 eV, indicating that at least two channels are responsible for proton production and desorption. The solid and dashed lines in 4B are fits to the data and the Gaussian curves centered at ∼8 eV describe the data reasonably well. Note that there is no shift in the kinetic energies of the protons with HCl dosing for either CI or PASW, indicating that the surface is not charging.

ARTICLE

Figure 4. Hþ kinetic energy distribution from pristine (squares) and HCl dosed (circles) (A) CI and (B) ASW ice samples collected under field-free conditions. The dotted and dashed lines in are the Gaussian fit to the fast and slow velocity components for pristine and HCl dosed ice, respectively. The solid lines are the sum of the Gaussian fit, which accurately describes the data.

III.D. TPD of HCl: Ice at Several Growth Temperatures. TPD was performed after ESD-TOF measurements as well as on nonirradiated samples to ensure the irradiation was not causing the ionization of the adsorbed HCl. Since the effects of adsorbed HCl on the ESD yields saturates at coverages >0.5 ML (see Figure 2), HCl doses g0.5 ML were used in many of our TPD studies. Figure 5 shows typical TPD traces from HCl (0.5 ML) dosed ice. The dashed line corresponds to water desorption (divided by 5) and HCl desorption is shown as the solid line. There are three main desorption features that occur under various dosing conditions for HCl coverages g0.5 ML. The first is feature I, which is coincident with water desorption. The second feature (II) is a shoulder (nonresolved peak) that can start between 110 and 130 K depending on HCl dosing temperature and leads into the first feature. The lowest temperature feature (III) occurs when HCl is dosed below 110 K. All of the TPD spectra (not shown) for the corresponding ESD-TOF spectra have only one HCl TPD peak. This peak dominates for HCl coverages