Reactive uptake of acetic acid on calcite and nitric acid reacted calcite aerosol in an environmental reaction chamber

Abstract

The heterogeneous chemistry of gas-phase acetic acid with CaCO₃(calcite) aerosol was studied under varying conditions of relative humidity (RH) in an environmental reaction chamber. Infrared spectroscopy showed the loss of gas-phase reactant and the appearance of a gaseous product species, CO₂. The acetic acid is observed to adsorb onto the calcite aerosol through both a fast and a slow uptake channel. While the fast channel is relatively independent of RH, the slow channel exhibits enhanced uptake and reaction as the RH is increased. In additional experiments, the calcite aerosol was exposed to both nitric and acetic acids in the presence of water vapor. The rapid conversion of the particulate carbonate to nitrate and subsequent deliquescence significantly enhances the uptake and reaction of acetic acid. These results suggest a possible mechanism for observed correlations between particulate nitrate and organic acids in the atmosphere. Calcium rich mineral dust may be an important sink for simple organic acids.

---

1. Introduction

The chemical composition of organic species in the atmosphere is extremely complex and, in addition to methane, includes aliphatic chains, alcohols, aromatics, phenols and organic acids, amongst many others. A significant portion of the non-methane hydrocarbon species, approximately 25%, is comprised of organic acids.¹ The two most abundant organic acids are formic acid, HCOOH, and acetic acid, CH₃COOH (HAc). These acids are often discussed together in the literature as field measurements have shown that their mixing ratios are correlated, although their relative concentrations exhibit marked diurnal and seasonal variations.^1–5 While the field data suggest common sources and sinks for both species, a clear understanding of the relevant reaction pathways for these simple organic acids is currently lacking.^1,4,6

Known anthropogenic sources of atmospheric organic acids include direct emission from motor vehicles^7,8 and biomass burning.^8,9 Primary biogenic emissions from plants^5,10 and soil microbes^11–13 are known to be sufficient to influence global fluxes. Secondary production from the photochemical oxidation of atmospheric hydrocarbons, of both biogenic and anthropogenic origin, is also a major source of organic acids. Photochemical reaction mechanisms leading to the formation of formic and acetic acids have been discussed in the literature.^{1,4,6,14–18}

Acetic acid has been detected in the atmosphere as a gas-phase species, in rain and fog water, and in association with particulate matter. Gas-phase measurements of acetic acid concentrations are highly variable and depend on geographic location, season, and time of day^1,5,6,19 A major removal mechanism for atmospheric acetic acid is via wet deposition. Acetic acid is readily soluble in water, exhibiting a large Henry’s law constant,⁴ 8800 M atm⁻¹, and is observed in precipitation samples worldwide.^{1,6,8,19–25} While a great deal of focus has been given to inorganic acids in rainwater, acetic and formic acids can contribute approximately 25% of the free acidity in the United States and greater than 60% in remote regions.^24,26

An alternative loss pathway involves dry deposition of acetic acid to particulates and surfaces. In a 1989 study, Grosjean estimated that 91% of the total budget for acetic acid deposition in the Southern California air basin was through dry deposition.²⁷ However, it was also pointed out that the estimates for removal fluxes are greater than those for production by a factor of two to four and it is unclear as to whether the discrepancy can be ascribed to overestimation of the deposition rates or if the corresponding rates of production were being underestimated. This study,²⁷ and others,^8,28,29 indicate that the dry deposition pathways for acetic acid are not well understood.

There has been some work addressing acetic and formic acid adsorption on transition metals as these systems have proven to be ideal for fundamental surface studies.^30–35 Investigations of organic vapor adsorption on mineral surfaces at varying RH values have concluded that for most organics the interaction is dominated by adsorption to the surface of the water film present under ambient conditions but that uptake decreases with increasing RH.^36–39 However, these results may not be relevant for small polar organic compounds such as acetic acid. Several studies involving organic acids have focused on adsorption from the aqueous phase to a solid mineral surface, representative of processes occurring in soils and sediments.^40–43 Experiments on the dissolution kinetics of calcite in acidic solutions have also been performed^44,45 and a recent study by Sabbioni et al. analyzed small organic anions in association with damage to monuments and buildings.²⁹

A more recent and relevant investigation from our laboratory by Carlos-Cuellar et al. measured the uptake coefficients for representative organic species, including acetic acid, on model mineral compounds under dry conditions.⁴⁶ Comparison of the measured uptake coefficients for acetic acid to the known rates for homogeneous photolytic loss pathways shows that heterogeneous uptake can be competitive with photolysis.⁴⁶ As discussed by Carlos-Cuellar et al., it is suggested that an increase in relative humidity will increase the acetic acid uptake coefficient.⁴⁶ Al-Hosney et al. have conducted similar studies of formic acid reactivity with carbonate.⁴⁷ However, more laboratory studies addressing the heterogeneous reactions of acetic acid with environmental surfaces, such as mineral aerosols, under relevant RH and temperature conditions are clearly still needed as significant questions remain concerning the atmospheric cycling of acetic acid.

Organic acids in the atmosphere have been correlated with mineral aerosol and nitrate in field studies. During the Atlanta SuperSite Project, Lee and co-workers analyzed 380 000 single aerosol particles in the 0.35–2.5 μm size range using laser mass-spectrometry.²⁸ Approximately 40% of the particles analyzed contained fragments associated with organic acids, such as formic and acetic acid. Nitrate and oxidized organic species were found to be more often present in the larger particles measured.⁴⁸ The most common particulates in the size range studied contained a mixture of organics and sulfate. However, mineral aerosols comprised of aluminosilicates, as well as other metal oxides, were also present. The majority of the mineral aerosols (95%) also had contributions from organic acids, in addition to other water-soluble species such as sulfate (95%) and nitrate (80%).²⁸ Another study by Russell et al. using single particle X-ray spectroscopy has observed correlations between large calcium containing particles and fragments indicative of organic acids.⁴⁹

In the current study, we report on our exploration of the reactivity of the organic acid, HAc, towards calcium carbonate (calcite) aerosol, a model mineral dust component. Experiments were performed in an environmental reaction chamber that allowed the effects of varying RH to be assessed while the dust sample was aerosolized, maintaining isolated particle conditions. Reactivity was determined through the use of IR absorption spectroscopy. In addition to studying the effects of RH on the uptake and reaction of HAc, the interaction of acetic acid with processed mineral dust aerosol was investigated using mixed samples of HAc and HNO₃ in the presence of water vapor.

2. Experimental

The experimental apparatus and techniques used in the current study have been discussed in detail in previous publications.^50–52 Briefly, the environmental reaction chamber has an internal volume of 151 L and almost all of the exposed internal surfaces have either been coated with Teflon or are fabricated from glass or Teflon, rendering the chamber relatively inert. The chamber can be evacuated and filled with reagent gases from an attached sample vacuum line. Buffer gas is provided by air from a commercial purge gas generator which is initially very dry, <1% RH, but can be humidified to the desired RH using a water bubbler. The chamber is usually filled with the gases of interest at the selected RH and brought to near atmospheric pressure with the purge gas. The mixture is allowed to sit in the chamber for a period of hours to insure passivation and to characterize wall reactions.

A known mass of mineral dust is placed in a sample cartridge which is connected by stainless steel tubing to a nozzle with a partial impactor plate positioned at the nozzle exit. The dust is generally held under vacuum for a period of several hours to remove moisture in the sample. When the experiment is to be initiated, a slide valve separating the sample from the main chamber is opened and the sample cartridge is rapidly pressurized by a short pulse of inert gas, forcing the powder through the nozzle and into the chamber. The impactor ensures efficient deagglomeration of the powder and facilitates aerosolization. The pulsed introduction of powder rapidly mixes the chamber contents in less than a minute and marks the initiation of the reaction.

The progress of the chemistry is monitored by absorption spectroscopy using a modified Fourier Transform infrared spectrometer (FTIR) where an external IR beam is directed through a chamber side arm sealed with Ge windows and onto an external mid-band HgCdTe (MCT) detector. The instrument was operated at 8 cm⁻¹ resolution and spectra were averaged for 256 scans, yielding a temporal resolution of approximately 52 s. In addition to following the kinetics and identifying products, IR spectroscopy was used for a Beer’s Law analysis to quantitatively determine the partial pressure of water and other reactant/product species (e.g. HAc, HNO₃, CO₂) in the chamber. The chamber temperature and total pressure were also recorded during the experiment.

Acetic acid (EMD Chemicals, glacial, 99.9985%) and water (Fischer Chemical, Optima) samples for the bubbler were placed in glass bulbs and then passed though several freeze–pump–thaw cycles before use. Gaseous nitric acid samples were produced in the laboratory by mixing H₂SO₄(96.0%) and HNO₃(79.5%) in a 3 ∶ 1 ratio. The resulting solution was then passed through several freeze–pump–thaw cycles. Typical HAc pressures used in these studies ranged from 138–140 mTorr while lower pressures, <10 mTorr, were employed in some experiments. When nitric acid was included as a reagent, its initial partial pressure was approximately 118 ± 2 mTorr, although the pressure just prior to dust introduction had a larger variation due to RH dependent wall loss. The CaCO₃ powder was obtained from a commercial source (OMYA, UF). We measured the BET specific surface area as 10.1 ± 0.3 m² g⁻¹ using a BET apparatus (Quantachrome, Nova 1200 Multipoint). The crystal phase of the calcium carbonate samples was confirmed to be calcite by X-ray powder diffraction. Calcium acetate powder (Alfa Aesar, Puratronic) was also used in one experiment. Typical powder loadings were in the range of 0.7–0.8 mg.

3. Results

The reactive uptake of HAc onto calcite aerosol was studied in the environmental reaction chamber using initial pressures of 138–140 mTorr and varying the RH, as detailed above. Typical results from IR spectroscopic probes of the reaction progress are contained in Fig. 1. For reference, a spectrum of acetic acid only is shown in Fig. 1a. Some absorption features due to HAc are indicated in the figure and others are summarized in Table 1. Bands due to acetic acid dimers were assigned on the basis of pressure dependence studies and literature references.^53–56 The spectra of Fig. 1b and c show the results of exposing calcite aerosol to HAc after a reaction time of 1 h under dry conditions, <1% RH, (b), and with moisture present, 53% RH, (c). These spectra have the prominent carbonate band at 1445 cm⁻¹ labeled. Other characteristic particulate carbonate absorptions are listed in Table 1. Both spectra have had gas phase water absorptions subtracted.

Fig. 1 FTIR spectral data for the acetic acid–CaCO₃ system showing a reference spectrum for acetic acid in the reaction chamber, (a), and spectra of acetic acid–CaCO₃ mixtures after a reaction time of 1 h under conditions of <1% RH, (b), and 53% RH (c). The initial acetic acid pressure, ≈135 mTorr, is the same in all cases. Gas phase water absorptions have been subtracted from all spectra and the data in (b) and (c) has been offset for clarity. The positions of various bands are indicated.

Table 1 Summary of IR absorption features due to various species relevant to the described experiments. The labels for the normal mode descriptions are, as follows: s (stretch), as (asymmetric stretch), ss (symmetric stretch), b (bend), ipb (in-plane bend), opb (out-of-plane bend)

Species	Absorption^a/cm^−1	Mode description
a Frequencies determined from the current work.
CH3COOH (g, monomer)	3582	s (OH)
	1790	s (C=O)
	1399	b (CH3)
	1179	b (COH)
	990	b (CH3)
HNO3(g)	3550	ss (OH)
	2995	as (NO2)+ ss (NO2)
	1708	as (NO2)
	1316	ss (NO2)
	887	ipb (NO2)
CO2(g)	2349	as
	668	b
CaCO3(particulate)	2510
	1795	ss + ipb
	1445 (1600–1200)	as
	878	opb
	713	ipb
Ca(NO3)2(particulate)	1436, 1358	as
	1045	ss
	820	opb
Ca(CH3COO)2(particulate)	1551	as (COO)
	1455	ss (COO)
H2O (a)	3430 (3660–2900)	ss, as
	1645	b

---

As can be observed in the data of Fig. 1, significant HAc remains even after long exposure to calcite aerosol. However, detailed analysis reveals that some of the HAc has been lost due to reactions with the dust. In addition, a clear feature at 2349 cm⁻¹ assigned to CO₂ is present and appears to increase in intensity with longer reaction times and elevated RH values. Spectral subtraction of gas phase HAc, water, and particulate carbonate bands shows that carbon dioxide is the only product species that could be identified under these reaction conditions.

The raw IR spectral data indicate that gas-phase acetic acid is lost upon introduction of calcite aerosol into the atmospheric reaction chamber and gaseous CO₂ is produced. In order to assess the time-dependent partial pressure of these species, IR spectra are collected over the course of an experiment and analyzed. To monitor HAc, the strong feature at 1790 cm⁻¹ was isolated using a spectral subtraction procedure to remove calcite and water contributions which was necessary due to spectral overlap of the relevant features. The net absorbance (peak height) of the acetic acid band centered at 1790 cm⁻¹ was measured for each spectrum and compared to a previously determined Beer’s Law calibration. The resulting HAc partial pressure, before and after aerosol addition, can be plotted as a function of time. The upper panel of Fig. 2 shows the results for a series of experiments at different RH values. A similar analysis procedure was used to quantify CO₂ production through the integrated absorbance of the 2349 cm⁻¹ band and the resultant measurements are summarized in the lower panel of Fig. 2. In order to account for wall reactions, separate experiments at similar acetic acid pressures and RH values but without the calcite aerosol were conducted. A fit to the wall loss data was then used to correct the HAc and CO₂ pressures for times, t > 0 (i.e. after the CaCO₃ dust has been introduced into the chamber).

Fig. 2 Partial pressure of reactant acetic acid, (A), and product CO₂, (B), as a function of time under conditions of <1% RH, (a), 21% RH, (b), and 53% RH, (c). All of the data for times after introduction of the CaCO₃ aerosol, t > 0, have been corrected for wall reactions. The solid lines show fits to the data in the case of the higher RH experiments, (b) and (c).

The HAc pressure manifests a rapid drop upon exposure to the calcite aerosol on a time scale too fast to resolve with the current instrument. The initial acetic acid consumption is approximately the same for each RH value. Subsequent to the initial drop in the dry experiment, <1% RH, Fig. 2a, little or no additional HAc decay is observed. However, for the higher RH experiments, Fig. 2b and c, further HAc consumption can be seen. The acetic acid time course data appears to decay exponentially before approaching a linear decay at longer times. To facilitate quantification of the observed HAc pressure, the data were fit to an exponential plus linear decay function. The fitting function, which is empirical in nature, was selected to describe the approach to zero-order kinetics (vide infra) for the slower reaction pathway. The resultant fits are shown as solid lines overlying the experimental data in Fig. 2.

Similarly, production of CO₂ shows a rapid increase in the first 5–10 min after dust exposure followed by a slower, longer time production process. In the case of dry conditions, <1% RH, Fig. 2a, the CO₂ pressure barely increases beyond this initial value but, in the case of the higher RH experiments, Fig. 2b and c, continued production of CO₂ is readily measured. Analogous to the HAc data, we have fit these experiments to an exponential plus linear growth function and the fits are indicated by the solid lines in Fig. 2.

Several experiments were conducted at lower initial acetic acid pressures, approximately 7–9 mTorr, in an attempt to further characterize the kinetics of the heterogeneous reaction in terms of the uptake coefficient, γ. The results shown in Fig. 3 indicate that we were still unable to resolve the initial fast decay. The amount of HAc adsorbed is similar to that observed for the initial drop of acetic acid at higher starting pressures in Fig. 2 and does not appear to change greatly under different RH conditions. Due to the spectral interference detailed above, vide supra, we could not accurately measure the HAc loss at longer times with these relatively low concentrations. We can, however, estimate a lower limit for the uptake coefficient, γ, from our data using,^51,52,57

where C_mass is the powder mass loading in the chamber, S_BET is the measured BET surface area of the calcite sample per unit mass, c̄ is the mean speed of HAc and τ is the characteristic decay time of the reactant. Assuming τ≈ 1 min, corresponding to the scan time of our FTIR, we then estimate a lower limit for the uptake coefficient of γ≥ 5 × 10⁻⁶ for the initial loss of HAc.

Fig. 3 Time dependence of the acetic acid pressure for lower initial pressures under conditions of (a) <1% RH, (b) 21% RH, and (c) 53% RH, (c).

Representative IR data for the mixed experiments, whereby both HAc and HNO₃ were present in the chamber and reacted with calcite aerosol, are shown in Fig. 4. In the upper panel we have plotted a reference spectrum, (a), of an acetic/nitric acid mixture in the chamber. Characteristic HNO₃ absorptions, in addition to the previously discussed HAc features, are observed and the band frequencies are listed in Table 1. We have previously studied the interaction of HNO₃ with calcite aerosol and typical data is shown in (b) for a RH of 43%. Under these conditions, all of the initial nitric acid, approximately 120 mTorr, is consumed as no HNO₃ absorptions are visible in the spectrum. Instead, broad bands due to adsorbed water and new features corresponding to particulate nitrate^58,59 are observed, indicative of the rapid reaction to form deliquesced calcium nitrate product. The appearance of an intense CO₂ absorption is further evidence of the reaction. Relevant absorptions due to nitrate and adsorbed water are listed in Table 1.

Fig. 4 FTIR data from the HAc/HNO₃/CaCO₃ experiments. In the upper panel, (A), we show a reference spectrum for an acetic/nitric acid mixture, (a). Also shown are spectra characterizing the exposure of calcite aerosol to HNO₃ only at 43% RH, (b), and a mixture of acetic and nitric acid at 41% RH, (c). In the lower panel, (B), expanded spectral data for calcite reacting with HNO₃ only at 43% RH, (b), and a mixture of HNO₃ and acetic acid at 41% RH, (c), are shown. A reference spectrum of calcium acetate aerosol is also included, (a). Gas phase water absorptions have been subtracted and spectrum (c) in the lower panel, (B), has been further processed to subtract acetic acid adsorptions. The positions of various bands are indicated.

Spectrum (c) in the upper panel displays the results of exposing calcite aerosol to a HAc/HNO₃ mixture at a RH of 41%. The spectrum is generally similar to that recorded for the nitric acid only experiment, (b), but with the appearance of additional absorptions recognized as arising from HAc. To further characterize the reaction, an expanded and enhanced view of the IR data is shown in the lower panel of Fig. 4 for the HNO₃ only, (b), and mixture, (c), experiments. In addition to subtraction of the gas phase water bands, spectrum (c) has been further processed to remove residual acetic acid features. While increasing the spectral noise, the subtraction process shows that both spectra have most features in common except for the appearance of a new band at ≈1551 cm⁻¹ in the case of the HAc/HNO₃ mixture. To aid in identification of this feature, we recorded a reference spectrum of calcium acetate, Ca(CH₃COO)₂, aerosol, as depicted in (a). Based on this comparison and literature references for acetate IR absorptions,^58,60 we assigned the new feature to the formation of acetate product in the mixture experiments. More frequencies are listed in Table 1.

We analyzed the acetic acid and CO₂ product concentrations as a function of time in the mixed reagent gas experiments using the procedure outlined for the HAc only data. The results are displayed in Fig. 5 for the HAc and CO₂ partial pressures in the upper and lower panels, respectively. The total loss of reactant acetic acid over the course of an experiment is clearly much larger when HNO₃ is present, as can be seen from comparing the data of curves (b) and (c) with the results from an HAc only experiment, (a). Furthermore, the extent of HAc consumption increases when the RH is increased from 21%, (b), to 41%, (c). The fast initial drop in HAc pressure is still observable at 21% RH but is suppressed at the higher RH. Similar to the HAc only data, the decay curves appear to be linear at longer times where the 41% RH data show an exponential approach to linearity. The longer time HAc loss data was fit to either a linear decay, in the case of the 21% RH data, or an exponential plus linear decay function, for the 41% RH curve, and the resultant fits are shown as the solid lines in Fig. 5.

Fig. 5 Partial pressure of reactant acetic acid, (A), and product CO₂, (B), as a function of time for the HAc/HNO₃/CaCO₃ experiments under conditions of 21% RH, (b), and 53% RH, (c). In addition, shown for reference, are data from Fig. 2 for an acetic acid only experiment at 53% RH, (a), and, in the lower panel, (B), results from a nitric acid only experiment at 44% RH, (d). All of the data for times after introduction of the CaCO₃ aerosol, t > 0, have been corrected for wall reactions. The solid lines show fits to the data.

The CO₂ time course data show an immediate jump in pressure upon introduction of the calcite aerosol, similar to the results observed for the HNO₃ only experiment, represented by curve (d). However, unlike the experiment with only nitric acid present where complete consumption of HNO₃ immediately produces a fixed, stoichiometric amount of carbon dioxide, the mixture experiments show continued production of gaseous CO₂ product as the HAc is consumed. The total amount of carbon dioxide produced is larger for the higher RH experiment. In addition, the amount of CO₂ present at longer reaction times is larger than the sum of the contributions from the nitric acid only, curve (d), and the acetic acid only results, curve (a), recorded with similar starting reactant pressures. In other words, the yield of product carbon dioxide is enhanced when nitric acid is present and the enhancement increases with larger RH values. As with the HAc loss data, the 21% RH CO₂ production curve was fit to a straight line and the 41% RH data was fit to an exponential plus linear growth function. The curve fits are shown by the solid lines.

4. Discussion

HAc–CaCO₃ reaction

The results presented illustrate that acetic acid is irreversibly taken up by calcite aerosol and yields a gas phase product, CO₂. The overall reaction can be described in analogy to that of HNO₃ as, The uptake process is distinguished by two mechanisms; a fast component with a relatively large uptake coefficient and a slower, long time process. The fast process is relatively independent of initial HAc pressure and RH. The slower process, however, is sensitive to humidity and exhibits an increased reaction extent as the RH is increased.

A convenient method for characterizing uptake and reaction for both the fast and slow channels in the HAc/CaCO₃ system is to quantify the surface coverage in terms of the amount of HAc lost and the “reaction” coverage calculated from the CO₂ product yield and assuming the 1 ∶ 1 stoichiometry between carbon dioxide and reacted Ca sites as suggested by eqn (2). We further normalize the coverage in terms of the number of monolayers by using a site density for CaCO₃ of n_s= 6.5 × 10¹⁴ sites cm⁻²,^51,61 corresponding to one monolayer. The coverage can then be calculated as,

where ΔP is the pressure of acetic acid lost to calculate the HAc coverage and the pressure of the CO₂ product is used to calculate the reaction coverage. The acetic acid coverage for the fast reaction pathway is determined from the pressure drop immediately upon aerosol introduction while the associated number of reacted layers is calculated from the CO₂ evolved in the first 10 min of reaction. The slower process is characterized from corresponding measurements made after 200 min of reaction time. The RH dependence of the calculated number of HAc or reacted monolayers is shown in Fig. 6 for the fast component, in the upper panel, and the slow component, lower panel. Note that in all cases, the reaction is not HAc limited as much of the initial acetic acid is still present at the end of the experiment.

Fig. 6 Semi log plots of the RH dependence for the acetic acid uptake and the number of reacted layers, calculated from CO₂ production. The upper panel, (A), shows the results for the fast component of the HAc uptake and reaction. The lower panel, (B), shows similar data for the longer time behavior for both the acetic acid only and the HAc/HNO₃ mixture experiments.

As can be discerned in the upper panel of Fig. 6, the amount of HAc adsorbed onto calcitevia the fast reaction pathway is only slightly dependent on initial acetic acid pressure and is relatively independent of RH. While calcite is not hygroscopic, RH values of 50% should correspond to a few monolayers of water coverage on the surface. Apparently, under these conditions, there is minimal interference between coadsorbing water and acetic acid. Since the uptake coefficient for HAc onto dry calcite is likely to be orders of magnitude lower than that for H₂O,⁴⁶ it may be that water and acetic acid do not initially adsorb to the same sites and that water does not completely cover the available surface sites. Alternatively, the acetic acid may efficiently displace adsorbed water.

The lower limit to the uptake coefficient we have estimated for HAc onto dry CaCO₃, γ≥5 × 10⁻⁶, is consistent with previous measurements of the uptake of acetic acid onto various model dust surfaces using a Knudsen cell technique.⁴⁶ Although calcite was not investigated in this study, other dust surfaces yielded uptake coefficients of (2–0.2)× 10⁻³ for HAc under dry conditions. Similarly, a value of γ=(3 ± 1)× 10⁻³ for formic acid uptake by calcite has been observed.⁴⁷ The acetic acid uptake coefficient is probably larger at the higher relative humidity conditions that are more typical of the troposphere. For instance, the uptake of nitric acid onto bulk CaCO₃ powder samples was found to increase with relative humidity from a value of ≈10⁻⁴ measured under dry conditions.⁶² The CaCO₃ surface is also likely to be more reactive towards acetic acid, particularly in the presence of water, and the resultant uptake larger than for the relatively unreactive metal oxide surfaces used in our previous Knudsen cell work. In companion experiments, IR spectroscopy has also been used to identify surface acetate on metal oxides exposed to acetic acid.⁴⁶ Again, this would be in agreement with the overall reaction suggested by eqn (2). Previous measurements have shown an increased amount of surface formate from calcite exposure to formic acid when the RH is raised.⁴⁷

The number of reacted layers in the upper panel of Fig. 6, determined from the amount of CO₂ evolved in the fast reaction, is also independent of the initial acetic acid pressure although it appears that increased RH does increase the production of CO₂. It is likely that the fast reaction pathway is facilitated by surface adsorbed water. The reaction coverage increases by about a factor of 1.5 from 21% to 53% RH, a change in RH that almost doubles the measured water coverage.⁶² Under the driest conditions, barely any CO₂ above background levels is observed so whatever residual surface adsorbed water remains on the calcite is insufficient for appreciable reaction. The measured stoichiometry is also far from that suggested by eqn (2) even when surface water is present. We calculate a ratio of acetic acid reacted to CO₂ produced of 8 ± 2 for the 53% RH experiment while eqn (2) would predict a stoichiometric ratio of 2 ∶ 1. Clearly, much of the acetic acid is adsorbed onto the surface as a reaction intermediate that we cannot detect via our IR probe method. It may be that the reaction pathway is initiated by ligand exchange on the hydroxyl-terminated Ca surface sites,^47,63–67

The carbonic acid would be unstable in the presence of surface water and rapidly decompose to yield CO₂ and H₂O. However, given that four times more adsorbed HAc is present than expected on the basis of eqn (2) and that at the higher RH values multilayer water coverage should be available to facilitate eqn (4), it may be more likely that another, precursor, acetic acid adsorbate species is formed.

Interestingly, in the mixed HAc/HNO₃ studies, evidence for the fast uptake of acetic acid can still be observed in the 21% RH experiment. The extent of the uptake is less than in the HAc only experiment, as demonstrated by the relevant data points in the upper panel of Fig. 6. At 41% RH in the mixed gas experiment, this channel is totally suppressed and there is no indication of a fast, initial, HAc uptake. Despite the extensive processing of the carbonate aerosol by HNO₃ due to calcium nitrate formation and deliquescence, these results suggest that there are still available surface sites onto which the HAc can rapidly adsorb. It may be that the HNO₃ reaction does not lead to a uniform coating of the calcite particle surface but, rather, preferentially reacts at certain sites. The lack of synergistic effects between SO₂ and HNO₃ uptake was previously suggested to be possible evidence for unreacted particulate surface in mixed SO₂/HNO₃ experiments.⁵⁷

In all instances, the calculated coverages for the fast reaction channel are less than unity; about 0.5 in the case of HAc uptake and much lower, 0.06, for the number of reacted layers at 53% RH. Thus, this reaction pathway appears to be surface limited, even at RH values typical of the ambient troposphere. The saturation coverages are on the order of (3.1 ± 0.3)× 10¹⁴ molecules cm⁻² for acetic acid uptake. This value can be compared with the saturation coverage for HAc on less reactive hematite and alumina particles of (6 ± 4)× 10¹³ molecules cm⁻² measured in a Knudsen cell study, albeit at a much lower acetic acid pressure, 6 μTorr.⁴⁶

The time course data in Fig. 2 indicates that the slower pathway leading to HAc uptake and CO₂ evolution is, in contrast to the fast channel, very sensitive to the RH. Similarly, the RH dependence can be observed through the calculated coverages, as summarized by the data in the lower panel of Fig. 6. The HAc uptake increases by approximately a factor of two in going from dry conditions to 53% RH. The reactive coverage increases even more, by greater than an order of magnitude, for the same increase in experimental RH. Correspondingly, the ratio of HAc reacted to CO₂ produced is 5 ± 1 at 53% RH. If only the reaction due to the slower channel is considered, correcting the data for the uptake and reaction from the fast process, the ratio decreases to 3.0 ± 0.2, approaching the value of 2 ∶ 1 expected from the overall reaction of eqn (2).

The observed increase in reaction extent with increase in RH is similar to our previous observations with the SO₂/CaCO₃ system.⁵⁷ In those studies, the reactive uptake of SO₂ onto calcite, quantified by measurement of CO₂ product from the reaction,

was found to increase slightly as the RH was raised to 76% RH. At higher RH values, up to 90% RH, CO₂ production exhibited a proportionally much larger increase. As the RH approaches the deliquescence relative humidity (DRH) for the sulfite product, which is thought to be about 90% DRH, water uptake increases and enhances the reaction yield of eqn (5). In the nitric acid system, the DRH of the corresponding nitrate product from the reaction, is much lower, approximately 13–18% DRH.⁶⁸ Only modest RH values are necessary to manifest extensive reaction and completely react the calcite particles if HNO₃ is in stoichiometric excess in a volume, rather than surface, limited reaction. In the case of the acetate product in eqn (2), the deliquescence point may be as high as 100% DRH.⁶⁹ We were prevented from approaching such high RH in our chamber studies due to wall reactions and spectral interferences. We expect that at RH values above 53%, the reaction extent will continue to increase and will likely manifest a steeper dependence on humidity as the DRH of calcium acetate is approached. In affiliated work with the formic acid reaction with calcite, we have noted a steeply rising product yield at RH values above about 60% with a reaction rate that is an order of magnitude larger than under dry conditions.⁴⁷ At the highest RH we have studied here, 53%, the HAc coverage is about 0.9 and the reaction coverage is approximately 0.2. Thus, the reaction still appears to be limited by particle surface area, rather than bulk. As more of the acetate product is dissolved at higher RH, the reaction will extend into the bulk of the particle and coverages will exceed a monolayer, similar to what we have observed for the SO₂/CaCO₃ reaction.

The fast component, corresponding to t≤ 10 min, for CO₂ production from the acetic acid reaction appears to depend on the amount of surface adsorbed water on the calcite. Similarly, carbon dioxide evolution from the reaction of SO₂ with calcite exhibits a fast, initial rise that has been ascribed to a process facilitated by surface water. In both systems, a subsequent, slower reaction is observed that eventually becomes linear, suggestive of zero-order kinetics due to saturation of the surface. The measured reaction rate, υ, can be expressed in terms of a surface reaction rate constant, k, as

where S is the BET dust surface area concentration in the chamber and n_s is the CaCO₃ surface site density defined above. Dividing the measured rates by S yields a normalized areal rate, which we have listed for the various experimental conditions in Table 2. The values for HAc decay or CO₂growth are derived either directly from a linear fit or from the slope parameter using an exponential plus linear fitting function, as described above, to the longer time data in Fig. 2 and 5.

Table 2 Areal rates for the linear portions of the HAc loss and CO₂ production curves for the acetic acid only and the HAc/HNO₃ mixture experiments (Fig. 2 and 5) under varying condition of RH

		RH/%
		<1	21	53
HAc only	HAc loss (×10^9 cm^−2 s^−1)		1.7 ± 0.2	3.3 ± 0.4
	CO2 production (×10^8 cm^−2 s^−1)	4.9 ± 0.5	4.1 ± 0.9	56 ± 5

---

		21	41
HAc/HNO3 mix	HAc loss (×10^10 cm^−2 s^−1)	9.8 ± 0.9	11 ± 1
	CO2 production (×10^10 cm^−2 s^−1)	6.5 ± 0.6	6.8 ± 0.6

---

The calculated reaction rates exhibit greater sensitivity to the RH than is apparent from the coverage data. In particular, the rate of CO₂ production increases by an order of magnitude when the RH is approximately doubled from 21 to 53%. Overall, these rates appear to be lower than the corresponding areal rates calculated from our SO₂/CaCO₃ experiments. Extrapolating the SO₂ results to 53% RH suggests a rate of 2.1 × 10¹⁰ cm⁻² s⁻¹ compared to the CO₂ production rate from the HAc reaction of (5.6 ± 0.5)× 10⁹ cm⁻² s⁻¹ at the same RH, about four times smaller. The measured areal rates for the HAc reaction are only a fraction of the bulk phase value for calcite dissolution at neutral pH, as was the case of our previous SO₂ experiments.⁵⁷

HAc–HNO₃–CaCO₃

The results for the mixed reagent gas experiments, with both HAc and HNO₃ present in the reaction chamber, clearly reveal evidence of an enhancement in the HAc reaction with calcite. The reaction with HNO₃ is very rapid, viaeqn (6), converting the calcite particle to nitrate which deliquesces under the RH conditions of these experiments, promoting further reaction by exposing new surface sites from the particle bulk and increasing the amount of surface water, which further facilitates reaction by establishing a quasi-liquid phase. The IR spectra of Fig. 4 show evidence of the aerosol processing in the form of new spectral features due to particulate nitrate and surface adsorbed water. The fast HAc channel can still compete with the HNO₃ reaction at 21% RH, as we have discussed (vide supra), but the HNO₃ reaction is sufficiently extensive at 41% RH that this pathway is suppressed. We have previously determined that, with similar initial HNO₃ pressures, the number of adsorbed water layers increases from about 10–12 to about 16–19 as the RH is raised from 20 to 40%.⁵¹ The dominant loss mechanism for HAc is the slower reaction pathway, as seen by the data in Fig. 5. Thus, the HAc reaction in the mixed gas experiments can be considered to be between gaseous acetic acid and a highly processed calcite particle with a partial, or total, coating of concentrated nitrate solution.

The resultant reaction leads to more uptake of HAc and production of CO₂ than is the case with HAc alone. The data reflect the enhanced reactivity as Fig. 5 shows more HAc uptake at long reaction times with the processed aerosol (≈5× larger relative to the 53% RH acetic acid data) and the amount of CO₂ evolved is greater than the sum of carbon dioxide produced from the HNO₃ and HAc only experiment (curves (a) and (d) in the lower panel of Fig. 5) under similar initial conditions. The increased reaction extent is similarly exhibited by the data in the lower panel of Fig. 6 which indicates that the HAc and reaction coverages are larger than the corresponding measurements in the absence of HNO₃. The reaction coverages for the mixed gas experiments were determined by subtracting the amount of CO₂ produced from the HNO₃ reaction, knowing the initial nitric acid pressure and assuming all of the HNO₃ is consumed according to eqn (6). Now, the number of reacted layers resulting from acetic acid chemistry is greater than a monolayer, about 3–4 at 41% RH, indicating the reaction can extend beyond the surface into the bulk of the particle. Finally, the presence of particulate acetate absorption features in the IR spectrum of Fig. 4 shows that the HAc reaction with nitric acid processed calcite aerosol is enhanced.

The derived areal reaction rates for the mixed reagent gas experiments are contained in Table 2. The rates are larger than the HAc only results for both HAc loss and CO₂ production. The rates are still far smaller than bulk dissolution rates measured for calcite. Interestingly, the areal rates are approximately the same for the 21 and 41% RH measurements, implying that the amount of adsorbed water (10–20 layers) is sufficient to at least establish a consistent quasi-liquid like environment on the particle surface for the HAc reaction at either RH. The ratio of the HAc loss rate to the CO₂ production rate is 1.6 ± 0.1 for the 41% RH experiment. The implied stoichiometric ratio is similar to that calculated from quantitative measurements of the HAc and CO₂ pressures but is intermediate between the expected stoichiometric ratio from eqns (2) and (4).

Our results illustrate that in the presence of HNO₃ and sufficient RH, at least greater than the DRH of calcium nitrate, the uptake and reaction of HAc is enhanced, clear evidence of a cooperative, synergistic effect between the two acidic gases. In contrast, our study of the SO₂ reaction with calcite aerosol and HNO₃ manifested no such effect; the production of CO₂ from sulfur dioxide was not enhanced and the SO₂ reaction extent was similar to that with unprocessed carbonate particles. It may be informative to consider the reason for such a contrast. The aqueous phase solubility of HAc and SO₂, both acidic gases, will depend on the relevant effective Henry’s law constant, K^*_H, which, in turn, will be pH dependent. Lower pH conditions will decrease K^*_H for either gas due to their acidic nature, suppressing uptake by the aqueous phase. However, uptake of HAc can also have a contribution from CaAc⁺ since calcium acetate is highly soluble, unlike calcium sulfite, where^44,70

This additional sink for HAc increases the effective Henry’s law constant as the pH is lowered. We have calculated K^*_H for SO₂ and for HAc, both with and without contribution from CaAc⁺, as a function of pH in the upper panel of Fig. 7. Another way to represent the influence of the Henry’s law equilibrium is to plot the mole fraction of each gas in the aqueous phase, X_aq, fromwhere f is the distribution factor.⁴ The distribution factor depends on K^*_H according to, and w_L is the liquid water volume mixing ratio. The exact value of w_L is not critical for the present discussion but we have used a value relevant to the conditions in our reaction chamber based on a procedure outlined previously⁵¹ which combines measurement of the CO₂ pressure increase with the spectroscopically observed drop in gas phase water when the aerosol is introduced. The calculated value is w_L≈ 1.6 × 10⁻⁷ and the resultant plot of eqn (9) is shown in the lower panel of Fig. 7. The pH of any aqueous phase on the particle is expected to be near neutral due to the excess buffering capacity of the carbonate aerosol. Around neutral pH, the aqueous fraction of SO₂ and HAc, when CaAc⁺ is not accounted for, drop off precipitously as acidity increases. However, the CaAc⁺ sink becomes important as the pH decreases and X_aq is more or less independent of pH. To illustrate this behavior, we use a semi log plot to accentuate the small dip around pH = 7. Thus, it may be that the formation of a relatively soluble acetate product species in the case of HAc favors uptake on processed, deliquesced calcite aerosol, unlike the situation with SO₂.

Fig. 7 The calculated effective Henry’s law constant, K_H^*, upper panel, (A), and aqueous phase mole fraction, X_aq, lower panel, (B), for acetic acid, with and without the formation of CaAc⁺, and SO₂.

Atmospheric implications

There are several conclusions from the current work that may bear on the role of heterogeneous processes in atmospheric cycles of acetic acid. It is also likely that these findings are equally applicable to formic acid, another relatively abundant organic acid that is highly correlated with acetic acid in field measurements.^1,6 The loss of atmospheric acetic acid due to reactive uptake on available calcite aerosol surfaces, and other mineral dust particles, may be competitive with homogeneous, loss pathways, especially in dusty urban and desertified environments. Based on the current work, and the previous Knudsen cell studies for associated metal oxide components of mineral dust, the acetic acid uptake coefficient for the fast reaction pathway is sufficiently large that dust may be a significant sink compared to photolytic pathways.⁴⁶ The uptake coefficient can be expected to be larger under higher RH conditions typical of the ambient troposphere. However, the fast reaction channel saturates at sub-monolayer coverages and results in relatively little processing of the calcite aerosol, corresponding to only about one-fifth of a monolayer at 53% RH. The acetic acid saturation coverages we have determined would be reached in approximately 0.25–3 h for an acetic acid concentrations of 10 ppb and assuming an uptake coefficient of γ≈ 10⁻⁵–10⁻⁴.

Perhaps of greater significance is the slower reaction pathway we have identified which also saturates the surface but leads to continued acetic acid consumption and CO₂ production. While RH values in the troposphere will usually be below the DRH of the calcium acetate product, the slower reaction is sensitive to RH and will be enhanced under typical tropospheric conditions. Continued processing of the aerosol would extend beyond the initially exposed surface area into the bulk of the particle, suggesting that dust aerosol mass may be a more critical consideration that surface area in atmospheric chemistry models that seek to incorporate heterogeneous dust interactions. In addition, it should be noted that the heterogeneous process converts acetic acid to CO₂, an important greenhouse gas, although it is unlikely to constitute a major source.

The reaction between acetic acid and calcite aerosol that has been processed in the atmosphere to increase particulate hygroscopicity, for instance via HNO₃ chemistry, will result in enhanced HAc uptake and reaction. The buffering capacity of the carbonate and the formation of soluble acetate product will provide a thermodynamic driving force for acetic acid partitioning to the particle phase. Subsequent reaction can be extensive and extend beyond the surface into the bulk of the product. However, the calcite dissolution rates appear to be orders of magnitude less than the bulk rate, at least under our experimental conditions. Finally, there is a clear synergistic effect between HNO₃ and acetic acid when reacting with calcite aerosol. The increased affinity of calcium carbonate for acetic acid may partially rationalize the correlation between particulate nitrate and organic acids observed in field measurements.^28,48

5. Conclusion

The reaction of acetic acid with calcite aerosol under varying conditions of relative humidity has been examined. The uptake of acetic acid by calcite and the production of gas phase CO₂ is observed to occur under both dry and wet conditions and exhibits distinct fast and slow reaction pathways. The fast pathway is characterized by an uptake coefficient of γ≥ 5 × 10⁻⁶ and saturates the surface at sub-monolayer coverages. The fast uptake is relatively independent of initial pressure and RH. The corresponding production of CO₂ is facilitated by surface adsorbed water which increases with RH. The slower reaction pathway also eventually saturates the surface but exhibits continued uptake and CO₂ evolution. The reaction yield increases with RH but extensive processing of the calcite aerosol is limited by the very high DRH of the calcium acetate product. Experiments exposing calcite aerosol to mixtures of HAc and HNO₃ reveal a strong synergistic effect leading to enhanced consumption of acetic acid and concomitant processing of the particle bulk. These results suggest that heterogeneous chemistry involving mineral dust may be a significant sink for atmospheric organic acids and may help explain observed correlations between particulate nitrate and organics.

Acknowledgements

This material is based upon work supported by the National Science Foundation under Grant No. CHE-0503854. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

References

1. P. Khare, N. Kumar, K. M. Kumari and S. S. Srivastava, Rev. Geophys., 1999, 37, 227–248.
2. R. J. Kieber, B. Peake, J. D. Willey and G. B. Avery, Atmos. Environ., 2002, 36, 3557–3563.
3. T. Reiner, O. Mohler and F. Arnold, J. Geophys. Res., [Atmos.], 1999, 104, 13943–13952.
4. J. H. Seinfeld and S. N. Pandis, Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, John Wiley & Sons, New York, 1998.
5. R. W. Talbot, B. W. Mosher, B. G. Heikes, D. J. Jacob, J. W. Munger, B. C. Daube, W. C. Keene, J. R. Maben and R. S. Artz, J. Geophys. Res., [Atmos.], 1995, 100, 9335–9343.
6. A. Chebbi and P. Carlier, Atmos. Environ., 1996, 30, 4233–4249.
7. K. Kawamura, L. L. Ng and I. R. Kaplan, Environ. Sci. Technol., 1985, 19, 1082–1086.
8. R. W. B. Talbot, K. M. Harriss, R. C. Cofer III and W. R., J. Geophys. Res., [Atmos.], 1988, 93, 1638–1652.
9. J. P. Lacaux, J. Loemba-Ndembi, B. Lefeivre, B. Cros and R. Delmas, Atmos. Environ., 1992, 26A, 541–551.
10. R. W. Talbot, M. O. Andreae, H. Berresheim, P. Artraxo, M. Garstang, R. C. Harriss, K. M. Beecher and S. M. Li, J. Geophys. Res., [Atmos.], 1990, 95, 16955–16969.
11. G. Enders, R. Dlugi, R. Steinbrecher, B. Clement, R. Daiber, J. v. Eijk, S. Gab, M. Haziza, G. Helas, U. Herrmann, J. Kesselmeier, D. Kotzias, K. Kourtidis, H.-H. Kurth, R. T. McMillen, G. Roider, W. Schurmann, U. Teichmann and L. Torres, Atmos. Environ., 1992, 26, 171–189.
12. W. C. Keene and J. N. Galloway, Tellus B, 1988, 40B, 322–334.
13. E. Sanhueza and M. O. Andreae, Geophys. Res. Lett., 1991, 18, 1707–1710.
14. S. Madronich, R. B. Chatfield, J. G. Calvert, G. K. Moortgat, B. Veyret and R. Lesclaux, Geophys. Res. Lett., 1990, 17, 2361–2364.
15. S. Madronich and J. G. Calvert, J. Geophys. Res., [Atmos.], 1990, 95, 5697–5715.
16. D. J. Jacob and S. C. Wofsy, J. Geophys. Res., [Atmos.], 1988, 93, 1477–1486.
17. H. A. Khwaja, Atmos. Environ., 1995, 29, 127–139.
18. D. Grosjean, E. Grosjean and E. L. Williams, II, Environ. Sci. Technol., 1994, 28, 186–196.
19. M. O. Andreae, R. W. Talbot, T. W. Andreae and R. C. Harriss, J. Geophys. Res., [Atmos.], 1988, 93, 1616–1624.
20. F. Maupetit and R. J. Delmas, J. Geophys. Res., [Atmos.], 1994, 99, 16491–16500.
21. K. C. Weathers, G. E. Likens, F. H. Bormann, S. H. Bicknell, B. T. Bormann, B. C. Daube, Jr, J. S. Eaton, J. N. Galloway, W. C. Keene, K. D. Kimball, W. H. McDowell, T. G. Siccama, D. Smiley and R. A. Tarrant, Environ. Sci. Technol., 1988, 22, 1018–1026.
22. M. O. Andreae, R. W. Talbot, H. Berresheim and K. M. Beecher, J. Geophys. Res., [Atmos.], 1990, 95, 16987–16999.
23. J. P. Lacaux, R. Delmas, G. Kouadio, B. Cros and M. O. Andreae, J. Geophys. Res., [Atmos.], 1992, 97, 6195–6206.
24. W. C. Keene and J. N. Galloway, Atmos. Environ., 1984, 18, 2491–2497.
25. R. W. Talbot, M. O. Andreae, H. Berresheim, D. J. Jacob and K. M. Beecher, J. Geophys. Res., [Atmos.], 1990, 95, 16799–16811.
26. W. C. Keene, J. N. Galloway and J. D. Holden, Jr, J. Geophys. Res., [Atmos.], 1983, 88, 5122–5130.
27. D. Grosjean, Environ. Sci. Technol., 1989, 23, 1506–1514.
28. S.-H. Lee, D. M. Murphy, D. S. Thomson and A. M. Middlebrook, J. Geophys. Res., [Atmos.], 2002, 107 DOI:10.1029/2000JD000011.
29. C. Sabbioni, N. Ghedini and A. Bonazza, Atmos. Environ., 2003, 37, 1261–1269.
30. A. R. Garcia, J. L. Da Silva and L. M. Ilharco, Surf. Sci., 1998, 415, 183–193.
31. J. C. Hemminger, Int. Rev. Phys. Chem., 1999, 18, 387–417.
32. P. Baumann, H. P. Bonzel, G. Pirug and J. Werner, Chem. Phys. Lett., 1996, 260, 215–222.
33. Q. Gao and J. C. Hemminger, J. Electron Spectrosc. Relat. Phenom., 1990, 54–55, 667–676.
34. B. A. Sexton, Chem. Phys. Lett., 1979, 65, 469–471.
35. R. J. Madix, J. L. Falconer and A. M. Suszko, Surf. Sci., 1976, 54, 6–20.
36. K. U. Goss, Environ. Sci. Technol., 1993, 27, 2127–2132.
37. K.-U. Goss and R. P. Schwarzenbach, J. Colloid Interface Sci., 2002, 252, 31–41.
38. C. M. Roth, K.-U. Goss and R. P. Schwarzenbach, J. Colloid Interface Sci., 2002, 252, 21–30.
39. K. U. Goss, Environ. Sci. Technol., 1992, 26, 2287–2294.
40. J. D. Kubicki, L. M. Schroeter, M. J. Itoh, B. N. Nguyen and S. E. Apitz, Geochim. Cosmochim. Acta, 1999, 63, 2709–2725.
41. S. J. Hug and B. Sulzberger, Langmuir, 1994, 10, 3587–3597.
42. E. C. Yost, M. I. Tejedor-Tejedor and M. A. Anderson, Environ. Sci. Technol., 1990, 24, 822–828.
43. M. V. Biber and W. Stumm, Environ. Sci. Technol., 1994, 28, 763–768.
44. C. N. Fredd and H. S. Fogler, Chem. Eng. Sci., 1998, 53, 3863–3874.
45. D. E. Nierode and B. B. Williams, Soc. Petrol. Eng. J., 1971, 11, 406–418.
46. S. Carlos-Cuellar, P. Li, A. P. Christensen, B. J. Krueger, C. Burrichter and V. H. Grassian, J. Phys. Chem. A, 2003, 107, 4250–4261.
47. H. A. Al-Hosney, S. Carlos-Cuellar, J. Baltrusaitis and V. H. Grassian, Phys. Chem. Chem. Phys., 2005, 7, 3587–3595.
48. S.-H. Lee, D. M. Murphy, D. S. Thomson and A. M. Middlebrook, J. Geophys. Res., [Atmos.], 2003, 108, 8417–8424.
49. L. M. Russell, S. F. Maria and S. C. B. Myneni, Geophys. Res. Lett., 2002, 29(16), 1779.
50. A. P. Prince, PhD thesis, University of Iowa, 2003.
51. A. P. Prince, V. H. Grassian, P. Kleiber and M. A. Young, Phys. Chem. Chem. Phys., 2006, 9, 622–634.
52. A. P. Prince, J. Wade, V. H. Grassian, P. Kleiber and M. A. Young, Atmos. Environ., 2002, 36, 5729–5740.
53. M. Haurie and A. Novak, J. Chim. Phys. Phys.-Chim. Biol., 1966, 63, 1584–1586.
54. A. Burneau, F. Genin and F. Quiles, Phys. Chem. Chem. Phys., 2000, 2, 5020–5029.
55. J. Chocholousova, J. Vacek and P. Hobza, J. Phys. Chem. A, 2003, 107, 3086–3092.
56. M. Haurie and A. Novak, J. Chim. Phys. Phys.-Chim. Biol., 1965, 62, 146–157.
57. A. P. Prince, P. Kleiber, V. H. Grassian and M. A. Young, Phys. Chem. Chem. Phys., 2007, 9, 3432.
58. K. Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds, John Wiley and Sons, New York, 1963.
59. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, John Wiley and Sons, New York, 1986.
60. F. Quiles and A. Burneau, Vib. Spectrosc., 1998, 16, 105–117.
61. G. M. Underwood, P. Li, H. Al-Abadleh and V. H. Grassian, J. Phys. Chem. A, 2001, 105, 6609–6620.
62. A. L. Goodman, G. M. Underwood and V. H. Grassian, J. Geophys. Res., [Atmos.], 2000, 105, 29053–29064.
63. H. A. Al-Hosney and V. H. Grassian, J. Am. Chem. Soc., 2004, 126, 8068–8069.
64. S. I. Kuriyavar, R. Vetrivel, S. G. Hegde, A. V. Ramaswamy, D. Chakrabarty and S. Mahapatra, J. Mater. Chem., 2000, 10, 1835–1840.
65. S. L. Stipp and J. M. F. Hochella, Geochim. Cosmochim. Acta, 1991, 55, 1723–1736.
66. S. L. Stipp, Geochim. Cosmochim. Acta, 1999, 63, 3121–3131.
67. H. A. Al-Abadleh, H. A. Al-Hosney and V. H. Grassian, J. Mol. Catal. A: Chem., 2005, 228, 47–54.
68. I. N. Tang and K. H. Fung, J. Chem. Phys., 1997, 106, 1653–1660.
69. J. Schonherr and M. Luber, Plant Soil, 2001, 236, 117–122.
70. C. A. Colman-Porter and C. B. Monk, J. Chem. Soc., 1952, 4363–4368.

---