ARTICLE pubs.acs.org/JPCA
Reaction Products and Kinetics of the Reaction of NO2 with γ-Fe2O3 Brian C. Hixson,† John W. Jordan,‡ Erica L. Wagner,|| and Holly M. Bevsek*,‡ Department of Chemistry, Susquehanna University, Selinsgrove, Pennsylvania 17870, United States ABSTRACT: The reaction of NO2 with Fe2O3 has relevance for both atmospheric chemistry and catalysis. Most studies have focused on hematite, α-Fe2O3, as it is the thermodynamic stable state of iron oxide; however, other forms of Fe2O3 naturally occur and may have different chemistries. In this study, we have investigated the reaction products and kinetics for NO2 reacting with γ-Fe2O3 powder using diffuse reflectance infrared Fourier transform spectroscopy and compared the results to those of previous studies of NO2 reacting with α-Fe2O3. Both α- and γ-Fe2O3 produce surface-bound nitrate at the pressures examined in this study (24212 mTorr); surfacebound nitrite products are observed at all pressures for γ-Fe2O3 whereas nitrite was only observed on α-Fe2O3 at lower pressures. Surface-bound NO+ and Fe-NO products are observed on γ-Fe2O3, which have not been observed with α-Fe2O3. The reaction kinetics show a first-order dependence on NO2 pressure and this is used to support the hypothesis of unimolecular reaction of adsorbed NO2 with the γ-Fe2O3 surface as the slow step in the reaction mechanism. The difference in product formation between NO2 reacting with γ-Fe2O3 and previous studies of α-Fe2O3 illustrate the fact that care must be taken in generalizing reactivity of different polymorphs.
’ INTRODUCTION The discovery of the role of polar stratospheric clouds in the destruction of stratospheric ozone demonstrated the impact that heterogeneous reactions can have on atmospheric chemistry.1,2 Since then, attention has been drawn to heterogeneous reactions in the troposphere—a region in which there are many different types and sources of particulate matter. Specifically, reactions on mineral dust substrates have attracted a great deal of interest.37 This is due in part to the content of mineral dust, which has the composition of the Earth’s crust1,3 and thus consists primarily of silicates, oxides, and carbonates. This variety of chemical species leads to a range of potential chemical reactions of trace gases with the substrate. Furthermore, mineral dust is the most widespread and concentrated tropospheric aerosol1,2 and continued conversion of forested land to agricultural use and subsequent increased erosion implies that mineral dust concentrations should rise.8 Although dependent upon the source, Fe2O3 is generally the most common oxide found in mineral dust after SiO2 and Al2O3.9 Because iron ions may exist in different oxidation states, electron transfer reactions are possible on Fe-containing mineral dust particles that would not occur with main group metal elements. Another interesting feature of the Fe2O3 chemistry is the presence of different polymorphs. α-Fe2O3 (hematite) is the thermodynamic stable phase and therefore the most common form of iron oxide;10,11 because of this most studies of Fe2O3 atmospheric chemistry focus on it. γ-Fe2O3 (maghemite) is another phase of iron oxide formed by oxidation of Fe3O412 that has been much less studied. α-Fe2O3 exists as a corundum crystal structure whereas γ-Fe2O3 exhibits a defect inverse spinel structure where 1/6 of the octahedral holes are vacant, with these vacancies being distributed at random or ordered depending r 2011 American Chemical Society
upon sample preparation.1315 This difference leads to different densities16 (α: 5.3 g/cm3; γ: 4.9 g/cm3), different colors (α: red; γ: dark brown), and potentially different chemistries. Although there are a variety of trace gases in the troposphere, NO2 is one of the most commonly studied. The high reactivity of NO2 gives it a rich chemistry in the atmosphere1,2 and the reaction of NO2 with Fe2O3 is of interest not only for its atmospheric relevance but also for its utility as a possible catalyst for NOx reduction.1719 All previous studies involving the reaction of NO2 with iron oxide involve the α-phase;2022 there are studies of the reaction of HNO3 with the γ-phase23,24 but no comparison of α- and γ-reactivity. The most thorough investigation has been that of Underwood et al., in which α-Fe2O3 was exposed to increasing pressures (5350 mTorr) of NO2 with products being identified by transmission FTIR and mass spectrometry.20 It was found that chelating nitrite surface species were formed at the lowest pressures and monodentate, bidentate, and bridging nitrate species (Figures 1 and 2) were formed at higher pressures (∼30 mTorr). The only gas-phase product observed was NO. On the basis of these observations, the authors postulated the following mechanism. After NO2 adsorbs on the α-Fe2O3 surface (1), it reacts with the surface to produce nitrite (2). The nitrite can then either react with another surface nitrite via a Langmuir Hinshelwood mechanism (3a) or react with a gas-phase NO2 via an EleyRideal mechanism (3b) to produce the NO and surface
Received: July 5, 2011 Revised: September 24, 2011 Published: September 27, 2011 13364
dx.doi.org/10.1021/jp206342w | J. Phys. Chem. A 2011, 115, 13364–13369
The Journal of Physical Chemistry A
ARTICLE
Figure 2. Coordination geometries of adsorbed nitrate: (a) monodentate; (b) bidentate; (c) bridging. M represents a metal atom.
Figure 1. Coordination geometries of adsorbed nitrite: (a) monodentate nitrito; (b) monodentate nitro; (c) bidentate nitrito; (d) bidentate nitro-nitrito; (e) bridging monodentate nitrito; (f) bridging nitronitrito; (g) bridging bidentate nitrito. M represents a metal atom.
nitrate species. 2NO2 ðgÞ h 2NO2 ðadsÞ
ð1Þ
2e þ 2NO2 ðadsÞ f 2NO2 ðadsÞ
ð2Þ
NO2 ðadsÞ þ NO2 ðadsÞ f NO3 ðadsÞ þ NOðgÞ þ e
ð3aÞ or NO2 ðadsÞ þ NO2 ðgÞ f NO3 ðadsÞ þ NOðgÞ
ð3bÞ
Alternative mechanisms could involve the dimerization of NO2. For example, it is known that NO2 dimerizes on zeolite surfaces26,27 and subsequently the adsorbed N2O4 disproportionates to form NO+ and NO3. 2NO2 ðadsÞ h N2 O4 ðadsÞ f NOþ ðadsÞ þ NO3 ðadsÞ ð5Þ However, no evidence of nitrosonium was found for the reaction of NO2 with α-Fe2O3 in the Underwood et al. study. In this investigation we examine the reactivity of γ-Fe2O3 with 24, 109, 153, and 212 mTorr NO2 and compare it to that of α-Fe2O3 as described in Underwood et al.20 Using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS), product formation was followed in time, allowing the reaction order with respect to NO2 to be determined and providing insight to the reaction mechanism.
’ EXPERIMENTAL SECTION All diffuse reflectance spectra were obtained with a DRIFTS accessory (Harrick Scientific, DRP) aligned within a ThermoNicolet 670 FT-IR (DTGS detector, 4 cm1 resolution) spectrometer. Reactions were carried out in a stainless steel chamber (Harrick Scientific, HVC-DRP-2) equipped with CaF2 windows. The reaction chamber had been aligned within the DRIFTS accessory previously using KBr as a reflective medium. Prior to an experiment, γ-Fe2O3 powder (Alfa Aesar, primary particle size 2030 nm) was gently packed by tamping the powder down once in the sample holder of the reaction chamber
and leveling it with a plastic blade. The chamber and sample were then evacuated overnight to approximately 1 107 Torr by a turbomolecular pump (BOC Edwards, EXT70H). A spectrum of the γ-Fe2O3 powder was recorded and used as the background for each subsequent spectrum. The baseline was checked for linearity directly before gas introduction; this spectrum also provided peak height values for time zero (vide infra). NO2 (99.5%, Messer, used as received) was then introduced as a continuous flow over the γ-Fe2O3 powder simultaneously with the onset of data collection. NO2 pressures used in this study were 24, 109, 153, and 212 mTorr, measured using an absolute capacitance manometer (MKS 722A; range 0.0110.00 Torr) and held constant during each experiment. The sample temperature was maintained at 25 °C during each experiment using a temperature controller (Harrick Scientific, ATC-001). Spectra were collected sequentially without delay to follow product formation in time; however, a 5 s delay was used between spectra for the lowest pressure. Spectra corresponding to 24 mTorr NO2 were averaged for 60 s (representing 50 scans), leading to a 65 s total interval between the beginnings of each spectrum, and spectra corresponding to 109, 153, and 212 mTorr NO2 were averaged for 15 s (representing 10 scans). Data were collected until there was no noticeable change in the DRIFT spectra (approximately 20 min for the reactions with 109, 153, and 212 mTorr and 90 min for 24 mTorr). Each experiment was run in triplicate to ascertain reproducibility. Experiments exposing γ-Fe2O3 to NO (vide infra) were performed in the same manner as that described above, with the exception of the treatment of the NO prior to reaction. NO (99.5%, Airgas National Welders) was purified before use following the protocol described by Byl.27 Briefly, the NO and concomitant impurities were condensed in a glass bulb surrounded by a liquid nitrogen bath. The liquid nitrogen bath was then replaced with a liquid oxygen bath and the NO distilled over to a second glass bulb surrounded by a liquid nitrogen bath. This separates the NO from higher boiling point compounds such as NO2 and N2O. This procedure was repeated to ensure a high level of purification, verified by transmission FTIR spectroscopy.
’ RESULTS AND DISCUSSION A. Assignments. Figure 3 shows DRIFT spectra of γ-Fe2O3 after reaction with 24, 109, 153, and 212 mTorr of NO2. These spectra were recorded after the NO2 was evacuated and the vacuum chamber had returned to its base pressure; therefore, signals represent chemisorbed surface species. Three bands are observed at 16501540 cm1 (band 1), 13001110 cm1 (band 2), and 15001320 cm1 (band 3) at all pressures. The first two bands are similar to those previously observed for NO2 reacting with α-Fe2O320 as well as gas-phase HNO3 reacting with γ-Fe2O323,24 and are assigned28 to the v3 mode of adsorbed 13365
dx.doi.org/10.1021/jp206342w |J. Phys. Chem. A 2011, 115, 13364–13369
The Journal of Physical Chemistry A
ARTICLE
Table 1. Vibrational Assignments of Nitrogen Oxide Products from Reaction of NO2 with γ-Fe2O3 Figure 2 assignment
(cm1)
a b
v v3 high
2201 ( 2 1603 ( 2
m
v3 low
1236 ( 3
c
v3 high
1581 ( 2
n
v3 low
1225 ( 2
d
v3 high
1545 ( 2
l
v3 low
1276 ( 2
bridging nitro-nitrito NO2
e
v3
1507 ( 1
bridging monodentate nitrito NO2
p f
v1 v3
1165 ( 2 1490 ( 1
nitrosonium NO+ bridging NO3 bidentate NO3 monodentate NO3
Figure 3. Typical DRIFT spectra of γ-Fe2O3 after reaction with different pressures of NO2.
q
v1
1119 ( 2
monodentate nitrito NO2
g
v3
1437 ( 1
bidentate nitro-nitrito NO2
h
v3
1420 ( 1
monodentate nitro NO2
i
v3
1374 ( 2
k
v1
1341 ( 2
j o
v3 v1
1364 ( 3 1210 ( 3
bidentate nitrito NO2
Figure 4. Vibrational assignments of DRIFT spectra of γ-Fe2O3 after reaction with different pressures of NO2. Assignments are given in Table 1.
NO3. In “free” nitrate with D3h symmetry the v3 mode is degenerate; surface adsorption lowers the symmetry and breaks this degeneracy, splitting the band. Structures seen within band 1 at 1545, 1581, and 1603 cm1 and within band 2 at 1276, 1225, and 1236 cm1 (Figure 4) are respectively assigned to monodentate, bidentate, and bridging geometries of iron-coordinated nitrate.20,26,29 The rest of the signals of band 2 are assigned to adsorbed nitrite26 as are the signals in a previously unrecorded third band (band 3) in the 15001320 cm1 region. Due to the similarity of absorption frequencies, differentiating among these surface species is difficult; however, we make preliminary assignments using previously assigned nitrite-coordinated inorganic species26,2830 as a guide. All assignments may be seen in Table 1. As stated earlier, reaction of NO2 with α-Fe2O3 was observed to produce nitrite products only at the lowest pressure of 5 mTorr.20 However, reaction of NO2 with γ-Fe2O3 yields nitrite
frequency
label
surface species
products at all pressures investigated, as is clear from Figures 3 and 4. Interestingly, at the lowest pressure of 24 mTorr two types of chelated nitrite products are formed (at different frequencies than that observed in Underwood et al.20) that are not observed at higher pressures: monodentate nitrito (v3 signal g, v1 signal not resolved) and bidentate nitro-nitrito (v3 signal h, v1 signal not resolved). It is presently unclear why this should occur with γ-Fe2O3 and not α-Fe2O3; however, molecular dynamics calculations could shed light into this phenomenon. A low-intensity signal also appears at 2201 cm1 upon reaction of 24 mTorr NO2 with γ-Fe2O3. Although not reported previously for reaction of either α- or γ-Fe2O3 powders with NO2 or gas-phase HNO3,2024 infrared signals in this frequency region have been observed when NO2 is exposed to various metal oxide powders such as TiO226,31,33 and have been assigned to NO+ adsorbed on the powder surface. Thus, the formation of ironnitrosonium products is a second difference between NO2 reaction on γ-Fe2O3 and α-Fe2O3. As stated previously, Underwood et al.20 have identified gas-phase NO as a product of the reaction of NO2 with α-Fe2O3 and it is likely the source of the NO that is producing the nitrosonium species we observe in our NO2 + γ-Fe2O3 DRIFT spectra. The fact that the nitrosonium signal does not appear at higher pressures of NO2 is likely due to competition with the direct reaction of NO2 with the γ-Fe2O3 surface. It is also possible that the production of NO+ on the surface is facilitated by the formation of the monodentate nitrito and bidentate nitronitrito species that also occur only at the lowest NO2 pressure. A third difference between the infrared spectra resulting from reaction of NO2 with γ-Fe2O3 versus α-Fe2O3 is the appearance of a broad band in the region from 1800 to 1650 cm1 for the latter reaction. Although gas-phase NO2 and N2O4 both have absorptions in this region, the fact that the vacuum chamber has been evacuated prior to recording this spectrum precludes this assignment. Nitrate signals do not extend higher than 1650 cm1, eliminating these species from consideration as well. We tentatively assign this band to various configurations of 13366
dx.doi.org/10.1021/jp206342w |J. Phys. Chem. A 2011, 115, 13364–13369
The Journal of Physical Chemistry A
Figure 5. DRIFT spectra of γ-Fe2O3 after reaction with different pressures of NO.
iron-nitrosyl (Fe-NO) complexes. Vibrational frequencies of nitrosyl complexes occur in this region21,28 and iron nitrosyls specifically have been assigned21,26,34,35 to bands from 1735 to 1900 cm1. To investigate this further, we reacted γ-Fe2O3 powder with purified NO and collected DRIFT spectra as a function of time and pressure. Figure 5 shows the DRIFT spectra collected after repeated exposure of the γ-Fe2O3 powder to increasing pressures of NO. As for NO2 reacting with γ-Fe2O3 nitrates, nitrites, and nitrosonium species are all produced on the powder surface. The low, broad band from 1800—1650 cm1 can also be seen growing with increasing NO pressure. The signal at 1783 cm1, clearly visible at 5 and 10 mTorr until it becomes the high frequency terminus of the 1800—1650 cm1 broad band at 136 mTorr, is unambiguously due to iron nitrosyls, specifically an iron cation with one NO species coordinated to it, as the frequency of 1783 cm1 is consistent with previous iron mononitrosyl assignments.34,35 Further support for this assignment may be seen in Figure 6, which shows the time evolution of spectra resulting from 10 mTorr of NO reacting with γ-Fe2O3. If this signal is in fact due to an iron mononitrosyl species, it should be the first species produced once the NO reacts with the γFe2O3, which is indeed the case. This signal then decreases rapidly thereafter as it is partially converted to nitrate, nitrite, and nitrosonium products. The fact that the band at 1800— 1650 cm1 grows in after the iron mononitrosyl signal has appreciably decayed indicates that the environment of the original mononitrosyl species is no longer accessible. Because after several minutes of reaction the powder surface is (at least) partially covered with NO+, NO2, and NO3 products, nascent Fe-NO species are formed in a different environment compared to the beginning of the reaction. These different environments then manifest themselves in slightly different frequencies of the Fe-NO product, causing the low, broad band. Further complicating the environment are the surface defects present on the powder surface as well as the vacancies present in the γ-Fe2O3 lattice, which all provide additional environments for the coordinated NO, also leading to the observed range of vibrational frequencies. Summarizing the products that are formed from reaction of NO2 with Fe2O3, it is found that both α- and γ-Fe2O3 produce a variety of nitrate and nitrite species, although γ-Fe2O3 produces
ARTICLE
Figure 6. DRIFT spectra of 10 mTorr NO reacting with γ-Fe2O3 as a function of time.
Figure 7. First ten DRIFT spectra of γ-Fe2O3 reacting with 109 mTorr NO2. Spectra were recorded 15 s apart.
nitrites at higher pressures than α-Fe2O3. γ-Fe2O3 powder surfaces, however, also produce FeNO+ and FeNO species. It appears that both of these species are formed as the product NO continues to react with the surface. The vacancies in the crystal structure of γ-Fe2O3—lacking in that of α-Fe2O3— provide reaction sites that both stabilize NO+ and provide a variety of environments for the FeNO species. B. Rate Law for the Reaction of NO2 with γ-Fe2O3. Figure 7 displays the first ten spectra collected during the reaction of 109 mTorr NO2 with γ-Fe2O3; spectra recorded at different pressures show similar features. Control experiments performed by covering the sample holder with Teflon tape show that gas-phase NO2 contributions to the signal are insignificant; however, those of surface-adsorbed NO2 and N2O4 are not. The effect of this is demonstrated in Figure 8, which shows spectra of γ-Fe2O3 with and without 151 mTorr NO2 present at the end of an experiment as well as the difference of these two spectra. The four bands (denoted by asterisks in Figure 8) centered at 1608, 1581, 1277, and 1241 cm1 may be assigned to adsorbed NO2 for the first two bands and adsorbed N2O4 for the last two, on the basis of gas-phase values.36 This significant overlap between adsorbed NO2 and N2O4 signals and the bridging and bidentate nitrate 13367
dx.doi.org/10.1021/jp206342w |J. Phys. Chem. A 2011, 115, 13364–13369
The Journal of Physical Chemistry A
ARTICLE
Table 2. Average Initial Increase in Product Signal as a Function of NO2 Pressure NO12 pressure (mTorr)
average initial increase (au s1) (4 ( 2) 104
25 109 153
(1.9 ( 0.2) 103 (2.6 ( 0.5) 103
212
(3.6 ( 0.4) 103
The amount of active sites on the γ-Fe2O3 powder should be much greater than the number of reacting NO2 molecules early in the reaction and therefore constant, so the reaction rate should only depend upon the pressure of NO2. Therefore, the rate law will take the form Figure 8. Comparison of nitrogen oxide products on γ-Fe2O3 before and after evacuation of NO2: (a) sample in the presence of 151 mTorr flowing NO2; (b) same sample with NO2 evacuated; (c) difference spectrum, (a) (b).
dnðNO3 Þ ¼ k0 PðNO2 Þx dt where n(NO3) is the number density of monodentate nitrate products on the surface, k0 is the product of the rate constant and the surface area of the γ-Fe2O3 powder, P(NO2) is the partial pressure of NO2, and x is the reaction order. To determine x, a logarithmic plot of initial signal increase versus pressure was made; this showed the reaction of NO2 with γ-Fe2O3 to be approximately first order with a value of 0.96 ( 0.01. The rate law for reaction of NO2 with γ-Fe2O3 may then be written as dnðNO3 Þ ¼ k0 PðNO2 Þ1 dt Some insight on the mechanism can be gained by this firstorder behavior. The rate law associated with the Langmuir Hinshelwood mechanism is37 rate ¼
Figure 9. Peak height of the monodentate nitrate signal as a function of time. NO2 pressure is 151 mTorr.
signals of band 1 as well as the entire region of band 2 indicates that care must be taken when signal areas are interpreted. To determine the rate law for reaction of NO2 with γ-Fe2O3, the peak height of the v3 (high) monodentate nitrate signal at 1545 cm1 in band 1 was measured as a function of time using Thermo-Nicolet’s OMNIC software. As stated previously, it would have been preferable to use the area of the entire band 1 region but the overlap of these signals with those of surfaceadsorbed NO2 and N2O4 prevented this. As the only other region free of overlap, the area of band 3 was also measured as a function of time to determine the reliability of the band 1 analysis; however, the signal-to-noise ratio was sufficiently poor that this was impractical. Due to these conditions, we relied on the monodentate NO3 signal for the following analysis. The peak heights of the monodentate NO3 signal were plotted as a function of reaction time (Figure 9) and the initial rate of signal increase (which is proportional to the initial reaction rate) at each NO2 pressure was then found via linear regression of the first 45 s of reaction time (the first 195 s were utilized for 25 mTorr NO2 reactions). Results may be found in Table 2.
kK 2 PðNO2 Þ2 1 þ 2KPðNO2 Þ2
where k is the rate constant and K is the equilibrium constant associated with step 3a in the mechanism. Note that in the low pressure limit the rate is second order in NO2 whereas at high pressures the rate is constant. In contrast, the rate law associated with the EleyRideal mechanism is37 rate ¼
kK 2 PðNO2 Þ2 1 þ 2KPðNO2 Þ
Again the rate is second order at low pressures; however, at high pressures it is first order. Because the rate is determined early in the reaction in this study, one can assume low NO2 coverage/pressure. However, a first-order rate law is not predicted then for either mechanism. We therefore have proposed that in this mechanism the rate-limiting reaction is step 2: unimolecular reaction with the powder surface. In this case, the rate law is37 rate ¼
kKPðNO2 Þ 1 þ KPðNO2 Þ
which yields a first-order rate law in the limit of low NO2 pressure and hence is in agreement with experiment. The reaction then appears to follow a LangmuirHinshelwood mechanism after the initial NO2 f NO2 conversion, on the basis of the fact that NO2 remains on the surface in the presence of a continually refreshed supply of NO2. 13368
dx.doi.org/10.1021/jp206342w |J. Phys. Chem. A 2011, 115, 13364–13369
The Journal of Physical Chemistry A
’ CONCLUSIONS The nitrate products formed on the surface of γ-Fe2O3 powder as a result of reaction with NO2 are similar to those formed on the surface of α-Fe2O3, as evidenced by this and previous studies.21 Differences occur in the formation of nitrite products, though. Underwood et al.20 observed nitrite formation only at 5 mTorr whereas we observe formation at all pressures investigated, with the exception of monodentate nitrito and bidentate nitro-nitrito forming only at 24 mTorr. In addition, our studies indicate permanent nitrosonium formation at low pressures and nitrosyl formation at all pressures, both of which have not been observed previously. These differences originate in the crystal structures of α- and γ-Fe2O3; specifically, vacancies in the γ-Fe2O3 lattice stabilize nitrosonium formation and permit a range of nitrosyl geometries to form. Theoretical studies and similar experiments with single crystal surfaces are needed to discover the role of the crystal structure in these reactions. We have also made the first determination of the rate law for NO2 reacting with γ-Fe2O3 and have found this reaction to be first order in NO2. This leads us to believe that the slow step in this reaction is surface reaction of NO2 to NO2. On the basis of the observation that NO2 exists on the surface in the presence of excess NO2, we further hypothesize that NO2 reacting with γ-Fe2O3 occurs through a LangmuirHinshelwood mechanism after the rate limiting step. Clearly, the NO2 pressures used in this experiment are much greater than those found in the atmosphere so these results are directly applicable to catalysis rather than to atmospheric chemistry. Furthermore, the dry conditions under which the experiments were executed is another important difference from atmospheric conditions and it is known that adsorbed water can play a significant role in heterogeneous reactions.29 However, it would be interesting to investigate whether the differences observed at these high pressures and dry conditions are obtained at atmospherically relevant conditions as well. ’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Present Addresses †
)
Department of Chemistry, Vanderbilt University, Nashville, TN 37235. ‡ Department of Chemistry, The Citadel, Charleston, SC 29409. Tel: (843) 953-7790. Fax: (843) 953-7795. Vaccine and Biologics Lab, Pharmaceutical Product Development, Wayne, PA 19087.
’ ACKNOWLEDGMENT This research was supported by a faculty start-up grant from Susquehanna University. H.M.B. also acknowledges a University Research Grant from Susquehanna University for the purchase of the DRIFT accessory and The Citadel Foundation for funding the NO studies. B.C.H. and E.L.W. acknowledge Summer Research Partner grants from Susquehanna University for support of their work. J.W.J. acknowledges partial summer stipends from The Wideman Foundation and The Star of the West Foundation.
ARTICLE
(2) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics; John Wiley and Sons, Inc.: New York, 1998. (3) Usher, C. R.; Michel, A. E.; Grassian, V. H. Chem. Rev. 2003, 103, 4883–4940. (4) Ndour, M.; Conchon, P.; D’Anna, B.; Ka, O.; George, C. Geophys. Res. Lett. 2009, 36, L05816–L05820. (5) Zhang, X.; Zhuang., G.; Chen, J.; Wang, Y.; Wang, X.; An, Z.; Zheng, P. J. Phys. Chem. B 2006, 110, 12588–12596. (6) Ullerstam, M.; Johnson, M. S.; Vogt., R.; Ljungstr€om, E. Atmos. Chem. Phys. 2003, 3, 2043–2051. (7) Dentener, F. J.; Carmichael, G. R.; Zhang, Y.; Lelieveld, J.; Crutzen, P. J. J. Geophys. Res. 1996, 101, 22869–22889. (8) Sheehy, D. P. Ambio 1992, 21, 303–307. (9) Goudie, A. S.; Middleton, N. J. Earth-Sci. Rev. 2001, 56, 179–204. (10) Al-Abadleh, H. A.; Grassian, V. H. Surf. Sci. Rep. 2003, 52, 63–161. (11) Muan, A. Am. J. Sci. 1958, 256, 171–207. (12) Housecroft, C. E.; Sharpe, A. G. Inorganic Chemistry; Prentice Hall: Harlow, U.K., 2001; p 511. (13) Belin, T.; Guigne-Millot, N.; Caillot, T.; Aymes, D.; Niepce, J. C. J. Solid State Chem. 2002, 163, 459–465. (14) Schulz, D. L.; McCarthy, G. J. Powder Diff. 1988, 3, 104–105. (15) Greaves, C. J. Solid State Chem. 1983, 49, 325–333. (16) Anthony, J. W.; Bideaux, R. A.; Bladh, K. W.; Nichols, M. C. Handbook of Mineralogy (Vol. III: Halides, Hydroxides, Oxides); Mineral Data Publishing: Tucson, AZ, 1997; p 628. (17) Kureti, S.; Weisweiler, W.; Hizbullah, K. Appl. Catal. B 2003, 43, 281–291. (18) Ferretto, L.; Glisenti, A. J. Mol. Catal. A 2002, 187, 119–128. (19) Ibusuki, T.; Takeuchi, K. J. Mol. Catal. 1994, 88, 93–102. (20) Underwood, G. M.; Miller, T. M.; Grassian, V. H. J. Phys. Chem. A 1999, 103, 6184–6190. (21) Busca, G.; Lorenzelli, V. J. Catal. 1981, 72, 303–313. (22) Contour, J. P.; Mouvier, G. J. Catal. 1975, 40, 342–348. (23) Frinak, E. K.; Wermeille, S. J.; Mashburn, C. D.; Tolbert, M. A.; Pursell, C. J. J. Phys. Chem. A 2004, 108, 1560–1566. (24) Goodman, A. L.; Bernard, E. T.; Grassian, V. H. J. Phys. Chem. A 2001, 105, 6443–6457. (25) Marie, O.; Malicki, N.; Pommier, C.; Massiani, P.; Vos, A.; Schoonheydt, R.; Geerlings, P.; Henriques, C.; Thibault-Starzyk, F. Chem. Commun. 2005, 28, 1049–1051. (26) Hadjiivanov, K. I. Catal. Rev.—Sci. Eng. 2000, 42, 71–144. (27) Byl, O.; Kondratyuk, P.; Yates, J. T., Jr. J. Phys. Chem. B 2003, 107, 4277–4279. (28) Nakamoto, K. Infrared Spectra of Inorganic and Coordination Compounds; Wiley: New York, 1970. (29) Baltrusaitis, J.; Schuttlefield, J.; Jensen, J.; Grassian, V. Phys. Chem. Chem. Phys. 2007, 9, 4970–4980. (30) Hadjiivanov, K.; Bushev, V.; Kantcheva, M.; Klissurski, D. Langmuir 1994, 10 (2), 464–471. (31) Hitchman, M. A.; Rowbottom, G. L. Coord. Chem. Rev. 1982, 42, 55–132. (32) Davydov, A. A. Infrared Spectroscopy of Adsorbed Species on the Surface of Transition Metal Oxides; Wiley and Sons: Chichester, U.K., 1990. (33) Hadjiivanov, K. I.; Klissurski, D. G.; Bushev, V. P. J. Chem. Soc., Faraday Trans. 1995, 91, 149–153. (34) Rethwisch, D. G.; Dumesic, J. A. J. Phys. Chem. 1986, 90, 1625– 1630. (35) Yuen, S.; Chen, Y.; Kubsh, J. E.; Dumesic, J. A.; TopsFe, N.; TopsFe, H. J. Phys. Chem. 1982, 86, 3022–3032. (36) Melen, F.; Herman, M. J. Phys. Chem. Ref. Data 1992, 21, 831–881. (37) Steinfeld, J. I.; Francisco, J. S.; Hase, W. L. Chemical Kinetics and Dynamics; Prentice Hall: Englewood Cliffs, NJ, 1989; p 195.
’ REFERENCES (1) Finlayson-Pitts, B. J.; Pitts, J. N., Jr. Chemistry of the Upper and Lower Atmosphere; Academic Press: San Diego, 1999. 13369
dx.doi.org/10.1021/jp206342w |J. Phys. Chem. A 2011, 115, 13364–13369
