Heterogeneous and Photochemical Reactions of Solid Benzophenone

Nov 8, 2008 - Brian R. Nichols, Christopher Rapa, Vincent Costa, and Ryan Z. Hinrichs*. Department of Chemistry, Drew UniVersity, Madison, New Jersey ...
2 downloads 0 Views 359KB Size
J. Phys. Chem. C 2009, 113, 2111–2119

2111

Heterogeneous and Photochemical Reactions of Solid Benzophenone-Catechol Films with NO2† Brian R. Nichols, Christopher Rapa, Vincent Costa, and Ryan Z. Hinrichs* Department of Chemistry, Drew UniVersity, Madison, New Jersey 07940 ReceiVed: July 23, 2008; ReVised Manuscript ReceiVed: September 22, 2008

Solid-air interfaces are ubiquitous to the atmosphere, and heterogeneous reactions between gaseous oxidants and surface adsorbed organics on these interfaces can impact tropospheric chemistry. Solid benzophenone-catechol films serve as model photosensitizer-polyphenol compounds that we reacted with NO2 in the parts-perbillion range under dark and light conditions at 300 K and 20% relative humidity. Attenuated total reflectance infrared spectroscopy (ATR-FTIR) monitored chemical changes in the organic film during these reactions to directly identify condensed-phase products. Catechol, when mixed with benzophenone or dicyclohexyl ketone, reacted with NO2 under dark conditions, forming 4-nitrocatechol as the exclusive condensed-phase product; pure catechol films did not react. Kinetic isotope experiments found rate[C6H4(OH)2]/rate[C6H4(OD)2] ) 3.3 ( 0.5, indicating that breaking an O-H bond was critical to the rate-determining step. A mechanism involving the ortho-semiquinone radical, possibly stabilized by hydrogen bonding to the coadsorbed ketone, is discussed. The reaction was also found to be second order (2.09 ( 0.18) with respect to NO2, suggesting a possible pre-equilibrium with N2O4. Although benzophenone is a well-known photosensitizer, the rate of 4-nitrocatechol formation was not enhanced by UV-A/visible radiation. This observation eliminates this pathway as a possible photoenhanced daytime source of HONO. However, ATR-FTIR detected additional photochemical products resulting from a photoinitiated reaction between benzophenone and 4-nitrocatechol. These results highlight the potential for heterogeneous chemistry involving surface adsorbed organics to form nitroaromatic compounds, which are of interest due to their phytotoxic and UV absorbing properties. I. Introduction Airborne particulate and ground level surfaces (e.g., soils, buildings) provide unique solid-air interfaces that can support heterogeneous chemistry and impact tropospheric photochemical cycles.1 A complete understanding of these solid-air interfaces, however, requires the inclusion of heterogeneous and photochemical reactivity of surface adsorbed organics.2 For instance, sea salt3 and mineral4,5 aerosols acquire organic coatings during formation processes and from adsorption of semivolatile compounds during transport. Both biogenic and anthropogenic emissions (from sources such as vegetation and combustion) generate organic aerosols and contribute to the organic content of inorganic aerosols.6-9 Urban surfaces, such as buildings and windows, also contain surface films composed of complex mixtures of organic and inorganic compounds,10 whereas rural surfaces such as soils are rich in degraded biomolecules.11 These surface adsorbed organics may alter the physical and chemical properties of these atmospherically relevant interfaces.12-14 Interest in adsorbed organics initially focused on water uptake since organic films might impact an aerosol’s ability to act as cloud condensing nuclei, thereby altering the earth’s radiative balance.15,16 Saxena et al. found that organics can increase water absorption by water-soluble inorganic aerosols in rural environments, but an opposite effect was noted for organics in urban air masses.17 The difference in these environments was attributed to the hydrophobic versus hydrophilic nature of the organic films. Chemical processing of organic films, through heteroge† Part of the special section “Physical Chemistry of Environmental Interfaces”. * To whom correspondence should be addressed. Phone: 973.408.3853. Fax: 973.408.3572. E-mail: [email protected].

neous reactions with atmospheric radicals, may alter their hydrophobic nature through oxidation.18 For instance, the oxidation of fatty acids, predominately on oleic acid, which is ubiquitous to the troposphere, serves as a principal laboratory model for understanding the chemical processing of organic coatings (e.g., see refs 13 and 19 and references therein). Characterization of atmospheric aerosols by Russel and coworkers using soft X-ray spectromicroscopy identified shorterchain, more oxygenated organics on the surface relative to the interior, providing experimental support for such chemical processing in the troposphere.20 More recently, laboratory studies have demonstrated that humic acid films and aerosols photochemically reduce nitrogen dioxide to nitrous acid, HONO, providing a potential pathway to explain a missing daytime source of HONO.21,22 This is of particular interest given the importance of humic and fulvic acids in soils11 and the extent of humic-like substances (HULIS) in the troposphere.23 Additional flow-tube experiments also observed the photochemical reduction of NO2 to HONO on small organic compounds.24 Mixtures containing a photosensitizer, such as aromatic ketones, and an electron donor, such as phenols, were especially effective at generating HONO under photochemical conditions. These experiments, however, focused exclusively on the gas phase reactants and products. The investigations described below complement these prior studies using attenuated total reflectance infrared spectroscopy to monitor the organic films under similar dark and photochemical conditions. Attenuated total reflectance infrared spectroscopy (ATRFTIR) provides a technique for monitoring condensed-phase species involved in heterogeneous chemistry with gaseous

10.1021/jp806525n CCC: $40.75  2009 American Chemical Society Published on Web 11/08/2008

2112 J. Phys. Chem. C, Vol. 113, No. 6, 2009 oxidants. The infrared beam of a Thermo Nicolet 6700 is reflected through a 45 degree germanium multipass ATR crystal. An evanescent wave penetrates ∼2 µm beyond the ATR crystal surface with each total reflection, allowing condensed-phase materials in contact with the crystal to attenuate the IR radiation via absorption. By monitoring the intensity of the IR beam using a liquid-nitrogen-cooled MCT detector, vibrational spectra for these condensed-phase materials can be recorded. Gas-phase species in the region of the evanescent wave may also absorb radiation; however, due to the path length of only a few micrometers, they fall below the detection limit of our instrument and therefore do not contribute to the IR spectra. The specific system studied herein is a thin film mixture of benzophenone and catechol exposed to nitrogen dioxide concentrations in the parts-per-billion (ppb) range, which overlap values found in highly polluted regions of the boundary layer. This system was chosen as a model photosensitizer-electron donor mixture, which has been previously studied by George and co-workers.24 A similar system, containing benzophenonephenol films, was recently shown to exhibit photosensitized chemistry with ozone.25 Catechol is also of direct atmospheric relevance, as it is a major emission from the pyrolysis of hardwoods and has been detected it in biomass burning plumes and residential combustion smoke.26,27 Furthermore, it serves as the simplest prototype for polyphenolic compounds, which constitute a significant fraction of the water-soluble organic matter in soils.28 Benzophenone has been extensively studied as a photosensitizer and therefore serves as a prototype aromatic ketone. An understanding of the heterogeneous reactivity of a thin film mixture of benzophenone and catechol serves, to a first approximation, as a model for understanding the tropospheric chemistry of surface adsorbed polyphenols and aromatic ketones. II. Experimental Methods Thin films of catechol, benzophenone, and other organic mixtures were coated on the surface of the Ge ATR crystal by dissolving up to 10 mg of each compound in less than 0.5 mL of acetone or ether (Aldrich, spectrophotometic grade), which was subsequently allowed to evaporate. Given the surface area of the ATR crystal (430 mm2) and assuming a solid-phase density of 1 g cm-3, the thickness of the solid layer was estimated to be on the order of tens of micrometers. All organic compounds were used without further purification: catechol (Aldrich 99%), catechol-d6 (CDN Isotopes 99.4 atom % D), benzophenone (Aldrich sublimed 99+%), dicyclohexyl ketone (Aldrich 98%), acenaphthene (Aldrich 99%), 3-nitrocatechol (Sunshine Chemlab), 4-nitrocatechol (Aldrich 99%), and 3,5dinitrocatechol (Aldrich 99%). A Pike Technologies 500 µL flow cell, which was modified to allow for photochemical studies, housed the coated Ge crystal. The continuous flow manifold depicted in Figure 1 introduced reactant gas mixtures to the ATR flow cell in a nitrogen (UHP) carrier gas. Divergent paths of “dry” and “wet” flow allowed the relative humidity (RH) to be varied. Both paths contain needle valves and flow meters; the “wet” flow also passed through a fritted bubbler containing deionized water (>18 MΩ), which was held at constant temperature using a thermostatted water bath. A relative humidity gauge (Vasalia) measured the water vapor content downstream of the ATR flow cell. All experiments described in this paper were conducted at 20% relative humidity and 300 K, corresponding to a water vapor concentration of 1.72 × 1017 molecules cm-3. Dilute mixtures of NO2 (Aldrich g99.5%) in UHP-N2 were premixed in a 5-L glass bulb with mole fractions ranging from

Nichols et al.

Figure 1. Schematic of continuous flow manifold attached to modified ATR-FTIR flow cell, where organic films were exposed to NO2. F, flow meters; B, fritted water bubbler; P, capacitance manometers; and RH, relative humidity gauge.

5 × 10-4 to 3 × 10-3. A third needle valve controlled the addition of the dilute NO2 mixture to the humidified carrier gas flow. A capacitance monometer monitored the pressure decrease in the 5-L glass bulb, allowing the flow rate to be calculated. Nitrogen dioxide concentrations in the ATR flow cell were calculated using the calibrated flow rates, NO2 mole fraction in the premixed bulb, and the pressure in the ATR flow cell measured by a second capacitance monometer. Calculated NO2 concentrations were estimated to be accurate within 13%. A needle valve downstream of the flow cell assisted in controlling the total flow rate through the system, and a liquid nitrogen trap condensed unreacted NO2 upstream of the vacuum pump. All experiments were conducted at 40 kPa with a residence time of ∼8 s in the ATR flow cell. Isotopic studies using 15N-labeled NO2 (Icon Services 99%) were conducted in a similar manner. The spectral output of a xenon arc lamp (Oriel 150 W Solar Simulator), directed through a liquid light guide, could be introduced to the flow cell through a borosilicate window, which absorbed all radiation with wavelengths λ < 300 nm. A cooled water filter removed infrared radiation from the xenon lamp output to minimize sample heating. Temperature changes for the ATR flow cell, measured by a thermocouple, were typically 1 K. A radiant power meter (Newport) measured the irradiance of the spectral output at the ATR crystal surface. Total irradiance between 300 and 800 nm was between 100 and 150 W m2 for all photochemical experiments. III. Results 3.1. Characterization of Solid Benzophenone-Catechol Mixtures. Dominant spectral features for pure benzophenone and catechol are shown in Figure 2 and assigned in Table 1. The major vibrational absorption bands for catechol correspond to an O-H stretch at 3450 cm-1 and a slightly broader O-H stretch at 3325 cm-1, which was red-shifted due to intramolecular hydrogen bonding.29 The main feature for benzophenone was assigned to the carbonyl stretching vibration at 1650 cm-1.30 Several C-H and aromatic vibrations were also observed for both catechol and benzophenone. Also included in Figure 2 are spectra for benzophenone-catechol mixtures ranging from 5:1 (bottom) to 5:5 (top) mass ratios. The spectra for these solid-phase mixtures indicate a strong hydrogen bonding interaction between catechol and benzophenone. Specifically, catechol O-H stretching vibrations appear primarily as a broad peak between 3115 and 3500 cm-1, whereas the benzophenone carbonyl absorbance splits into two peaks at 1643 and 1659 cm-1. The red-shifted CdO peak at 1643 cm-1

Reactions of Benzophenone-Catechol Films with NO2

Figure 2. ATR-FTIR spectra of pure solid-phase catechol (top) and benzophenone (bottom). Spectra of benzophenone-catechol mixtures shown between for 5:1, 5:3, and 5:5 mass ratios.

is assigned to carbonyls accepting hydrogen bonds from catechol. The blue-shifted CdO absorbance at 1659 cm-1 dominates at low catechol mixing ratios and is therefore consistent with disruption of the pure benzophenone crystalline structure. The benzophenone carbonyl is conjugated with the aromatic rings via resonance structures, which are stabilized in the pure benzophenone solid by π-stacking. Disrupting these interactions in the solid mixture strengthens the carbonyl bond, resulting in a higher vibrational frequency. Consistent with this explanation is the observation that the CdO stretch for benzophenone dissolved in CCl4 is blue-shifted relative to solidphase benzophenone.30 Benzophenone adsorbed on rutile produced a similar blue-shifted peak, which was attributed to a disordered glassy state.31 In a similar fashion, a minor shoulder at 3550 cm-1 is attributed to free O-H stretching vibrations for a disordered catechol state in which intermolecular hydrogen bonding interactions are disrupted. The C-H and aromatic vibrations noted for pure benzophenone and catechol showed, at most, very minor shifts (