Formation of Nitroanthracene and Anthraquinone from the

Jun 20, 2014 - (9,10-AQ) and 9-nitroanthracene (9-NANT), were determined ... In contrast, the rate of formation of 9-NANT across the whole RH range (0...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/est

Formation of Nitroanthracene and Anthraquinone from the Heterogeneous Reaction Between NO2 and Anthracene Adsorbed on NaCl Particles Wenyuan Chen and Tong Zhu* State Key Joint Laboratory of Environmental Simulation and Pollution Control; College of Environmental Sciences and Engineering, Peking University, Beijing 100871, China S Supporting Information *

ABSTRACT: Oxidative derivatives of polycyclic aromatic hydrocarbons (PAHs), that is, nitro-PAHs and quinones, are classed as hazardous semivolatile organic compounds but their formation mechanism from the heterogeneous reactions of PAHs adsorbed on atmospheric particles is not well understood. The heterogeneous reaction of NO2 with anthracene adsorbed on NaCl particles under different relative humidity (RH 0−60%) was investigated under dark conditions at 298 K. The formation of the major products, 9,10-anthraquinone (9,10-AQ) and 9-nitroanthracene (9-NANT), were determined to be second-order reactions with respect to NO2 concentration. The rate of formation of 9,10-AQ under low RH (0−20%) increased as the RH increased but decreased when the RH was further increased in high RH (40−60%). In contrast, the rate of formation of 9-NANT across the whole RH range (0−60%) decreased significantly with increasing RH. Two different reaction pathways are discussed for the formation of 9,10-AQ and 9NANT, respectively, and both are considered to be coupled to the predominant reaction of NO2 with the NaCl substrate. These results suggest that relative humidity, which controls the amount of surface adsorbed water on NaCl particles, plays an important role in the heterogeneous reaction of NO2 with adsorbed PAHs.

■

INTRODUCTION Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous environmental pollutants that mainly originate from natural volatilization (coal tar, crude oil, and creosote), volcanic eruptions, biomass burning, and the incomplete combustion of fossil fuels (petroleum, natural gas, and coal). They have attracted significant attention due to their allergenic, mutagenic, and potentially carcinogenic properties.1−3 Once emitted into the atmosphere, PAHs may react with atmospheric oxidants (O3, OH, and NOx) to form nitrated (NPAHs) or oxygenated (OPAHs) derivatives such as ketones, aldehydes, carboxylates, and quinones, many of which are considered to be not only much more mutagenic and carcinogenic than their parent species but also to be predominantly associated with the particulate phase due to their low volatility.4−7 Field observations have verified that as well as primary emission, atmospheric chemical transformation from their parent PAHs is one of the most important sources of NPAHs and OPAHs.8−11 For these reasons, investigating the reaction of PAHs with atmospheric oxidants, especially for the formation of NPAHs and OPAHs, is necessary and important. A recent review11 has conclude the progress in the research on the homogeneous and heterogeneous reactions of PAHs and especially focused on the product in these atmospheric reaction. Although both homogeneous and heterogeneous reactions in the atmosphere play important roles in the © 2014 American Chemical Society

conversion of PAHs, renewed interest has focused on the potential for the heterogeneous formation of NPAHs and OPAHs, which have been more widely detected on atmospheric particles than in the gas phase.8,12 Heterogeneous reactions of PAHs on particles involve complex processes that have been reported to be affected significantly by a number of factors including PAH molecules, atmospheric oxidants (OH, NOx, and O3 or their mixtures), nature of substrates (e.g., soot, dust, mineral oxides, and sea salt), surface properties of particles (e.g., specific surface area and porosity), and surface coatings (e.g., nitric acid, water or other organic species).11 For example, graphite and silica were chosen as simple models of atmospheric carbonaceous and mineral particles, respectively, to investigate the heterogeneous reactions of O3, NO2, and OH radicals with 11 adsorbed PAHs, which indicated that anthracene (ANT) and benzo[a]pyrene (B[a]P) were the most reactive with NO2, while all of the PAHs studied displayed a similar reactivity with OH, within a given range of uncertainty.13−16 Studies have suggested that the substrate not only controls the reaction kinetics but also alters the reaction pathway for the heterogeneous reaction of Received: Revised: Accepted: Published: 8671

March 28, 2014 June 16, 2014 June 20, 2014 June 20, 2014 dx.doi.org/10.1021/es501543g | Environ. Sci. Technol. 2014, 48, 8671−8678

Environmental Science & Technology

Article

NO2 with ANT adsorbed on SiO2 and MgO.17,18 Heterogeneous reaction chemistry of PAHs is far from being fully understood in aspects of reaction mechanisms, products and the factors that influence the reaction kinetics still lacking. Although many studies have investigated the exposure of adsorbed PAH to atmospheric oxidants, the role of humidity in the heterogeneous reaction of PAHs is not well understood. Furthermore, most research so far has focused on kinetic investigations of the decay of the particulate-PAH concentration over time to assess the reactivity of PAHs rather than on the kinetics and mechanisms of the formation of products that are considered to pose significant risks to human health. Among these oxidative derivatives of PAHs, NPAHs, and quinones, which are believed to contribute to the mutagenicity and toxicity of particulate matter (PM) in ambient air, have attracted the most attention. NPAHs are a class of very potent mutagenic compounds that have been shown to be typically 100 000 times more mutagenic and 10 times more carcinogenic than the corresponding parent PAHs, which require a preliminary enzymatic activation.19,20 Quinones are highly active redox molecules that can undergo a redox cycle with their semiquinone radicals, leading to formation of reactive oxygen species (ROS) that are associated with adverse health effects such as cardiovascular and pulmonary diseases.21−24 Sodium chloride (NaCl) is the major constituent of sea salt particles and NO2 is one of the major pollutants in vehicle exhaust, both of which are ubiquitous in ambient air, especially in coastal cities where severe vehicle exhaust pollution is experienced.25 Many studies have examined the atmospheric heterogeneous reactions of NO2 on NaCl particles, which are used as a model for sea salt.26 Therefore, investigating the chemical behavior of PAHs on sea salt particles in the presence of NO2 is of interest. In this study, we investigated the formation of NANT and anthraquinone (AQ) from the heterogeneous reaction of NO2 with ANT adsorbed on NaCl particles using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and gas chromatography−mass spectrometry (GC−MS). The reaction mechanism and the impact of relative humidity on the reaction kinetics and pathways are discussed. The results will help to understand the chemical behavior of PAHs and the source of NPAHs and quinones in the atmosphere.

Approximately 1.0 g of cleaned NaCl particles was added to 50.0 mL of n-hexane containing 50 mg of ANT. The solvent was homogeneously stirred for 3 h and then slowly evaporated using a rotary evaporator at 303 K. Finally, the particles were dried in a high-purity N2 flow at room temperature for 4 h. The NaCl particles with preloaded ANT were labeled as ANT-NaCl samples. To avoid photodegradation of the adsorbed PAHs, all ANT-NaCl samples were stored in amber glass flasks at 253 K in the dark. Reaction Procedure and Analytical Methods. All experiments were performed under dark conditions at 298 K. Coated particles (20.5 ± 1.5 mg) were evenly deposited in a cylinder-type sample cell made of stainless steel (8.1 mm diameter, 0.5 mm depth), which was compressed using a quartz slice and then placed in the DRIFTS reactor. This procedure is similar to that used in our previous study.26,27 The reactor is a vacuum reaction chamber (model HVC-DR2) surrounding a Harrick Scientific (Pleasantville, NY) diffuse reflectance accessory (model DRA-2CS) located in the sampling compartment of a Nicolet Nexus (Thermo Fisher Scientific Inc.) FTIR spectrometer equipped with a mercury cadmium telluride (MCT) detector. The gas molecules, NO2, diffused freely onto the surface of the sample and reacted with NaCl particles and the adsorbed ANT. Infrared absorption spectra of the reaction products were collected by DRIFTS every 3 min. Data from 128 scans with a 4 cm−1 resolution taken over about 80 s were averaged to produce one spectrum. To analyze the products quantitatively, samples after reaction were extracted three times with 6 mL hexane/CH2Cl2 (1:1, v/v, 2 mL each time). The combined extract was blown down in an N2 flow to volatilize the solvent, followed by reconstituting the residue in 10 mL pure hexane. Deuterated 9-NANT was added as an internal standard for instrumental analysis. Throughout the extraction and analysis procedure, samples were protected from potential photolysis using amber containers or by wrapping the containers with aluminum foil. All of the products were measured by GC-MS (7890A5975C; Agilent, Santa Clara, CA) with an electron-capturenegative ionization (ECNI) detector. The GC injection port was held at 250 °C and an aliquot of 1 μL was injected in the splitless mode. A 15 m Rtx-5MS column (250 μm i.d., 0.25 μm film thickness: Restek, Bellefonte, PA) was used to separate all analytes with a constant flow rate of 1.5 mL·min−1. The GC oven temperature program was as follows: held at 70 °C for 1 min, 6 °C min−1 to 250 °C, 25 °C min−1 to 300 °C, and then held for 2 min. For the identification of target chemicals, the following ion couples were monitored: m/z 208 for 9,10-AQ, m/z 223 for 9-NANT and 2-NANT, m/z 268 for 9,10DNANT, and m/z 232 for deuterated 9-NANT.

■

MATERIALS AND METHODS Chemicals. All chemicals were chromatographic grade and used without further purification. Dichloromethane (CH2Cl2) and n-hexane were obtained from Thermo Fisher Scientific Inc. (Waltham, MA). ANT, 9,10-anthraquinone (9,10-AQ), 9-nitroanthracene (9-NANT), 2-nitroanthracene (2-NANT), and 9,10-dinitroanthracene (9,10-DNANT) were purchased from J&K Scientific Ltd. (Beijing, China). High purity N2 (99.999%; Beijing Haike Yuanchang Gases Inc., Beijing, China) were purged through the reactor before and after the reaction. The experimental gas, which was described in detail in the Supporting Information (SI), was prepared with the required concentration and relative humidity as a mixture of standard gas NO2 (99.99%, 1000 ppm; Wujiang Messer Gases Inc., Wujiang, China), high purity N2, and water vapor. Sample Preparation. Before the adsorption of ANT, NaCl (>99.5%: Sigma-Aldrich Co., St. Louis, MO) particles were ground and then cleaned three times by ultrasonication in dichloromethane, followed by drying at room temperature. The concentration of PAHs, nitro-PAHs, and quinones in the extract of cleaned particles and the pure solvent could not be determined by GC-MS.

■

RESULTS AND DISCUSSION Identification of Reaction Products. As shown in Figure 1, in situ DRIFTS measurements in a flow system were performed for a preliminary analysis of the products of the heterogeneous reaction of ANT-NaCl particles with NO2. The predominant bands could be assigned to the vibration of the nitrate ion, which is a similar result to that reported by Li et al. (2006) for the heterogeneous reaction between NO2 and pure NaCl particles.26,28 The infrared bands in the 1300−1500 cm−1 region and at 1050 and 836 cm−1 were assigned to the ν3 asymmetric stretch, the ν1 symmetric stretch, and the ν2 outof-plane bend, respectively, which are three fundamental vibrations of the surface nitrate ions formed during the 8672

dx.doi.org/10.1021/es501543g | Environ. Sci. Technol. 2014, 48, 8671−8678

Environmental Science & Technology

Article

Figure 1. Typical in situ DRIFTS spectra of ANT-NaCl particles as a function of time (0−360 min) exposed to a flow of 2.7 × 1015 molcules·cm−3 NO2 at 298 K. Absorbance lines from the bottom to the top of the figure represent spectra from reactions at 0, 30, 60···360 min.

reaction. Another band at ∼1778 cm−1 was assigned to ν1+ν2. In the asymmetric stretch ν3 region, the adsorption peak was split due to the presence of the Na+ ion. This splitting behavior for adsorption peak, which is resulted from the strong interaction between negative ion and positive ion in nitrate and the transformation of NO3− ion in this process, has been observed and demonstrated in previous studies.26 The central positions of the adsorption peaks shifted from the initial positions of 1360 and 1425 cm−1 to 1358 and 1416 cm−1, whereas the intensity ratio of these two peaks changed from 1.0 to ∼1.5. This indicates that the NaNO3 crystallite that formed was isolated in the early stage and the chemical environment on the surface of the particles changed with the gradual formation of nitrate. In addition to the vibration of the nitrate ion, several absorption bands of organic matter were observed. The negative peaks at 1618, 956, and 883 cm−1 indicate the consumption and volatilization of ANT.29,30 The strong bands at 1556 and 1518 cm−1 are the characteristic frequencies of 9-NANT.31 The weak bands at 1672, 1304, and 1290 cm−1 can be assigned to 9,10-AQ.32 These results suggest that the heterogeneous reaction between NO2 and ANT adsorbed on NaCl particles leads to the formation of 9-NANT and 9,10-AQ, which are also the products of heterogeneous reactions between NO2 and ANT adsorbed on SiO2 particles, but not MgO particles.17 The difference in the heterogeneous reactivity of NO2 with ANT adsorbed on SiO2 and MgO has been ascribed to the formation of HNO3 on SiO2, which can catalyze the nitration of PAHs by NO2.17 This mechanism can also explain the heterogeneous reactivity of NO2 with ANT adsorbed on NaCl particles because the formation of HNO3 as one of the most important intermediate products in the reaction of NO2 with surface-adsorbed water on NaCl particles is well-known.33,34 GC-MS analysis was performed to unambigously identify specific products of the heterogeneous reaction of surfacebound ANT, which could be extracted by an organic solvent (hexane/CH2Cl2 1:1), eluted from the GC column, and could be detected by MS in the ECNI mode. As shown in Figure 2, four peaks were observed in the total ion chromatogram (TIC) of the products of the heterogeneous reaction between NO2 and ANT adsorbed on NaCl. Every peak was analyzed in detail by MS. By comparing the mass spectra of the products obtained with those from the mass spectra library, these four peaks were identified as 9,10-AQ, 9-NANT, 2-NANT, and 9,10-DNANT, respectively, each of which was confirmed by the

Figure 2. Total ion chromatogram (TIC) of the products of the heterogeneous reaction of anthracene adsorbed on NaCl with NO2.

spectra of its standard samples. Because the peak area of 9,10AQ and 9-NANT was about 2 orders of magnitude larger than that of 2-NANT and 9,10-DNANT, the production of 2-NANT and 9,10-DNANT was negligible, and 9,10-AQ and 9-NANT were the only two species targeted in our follow-up analysis. Kinetics of the Reaction Products. The kinetics of the heterogeneous reaction between NO2 and ANT adsorbed on NaCl particles were determined by monitoring the growth of 9,10-AQ and 9-NANT as a function of NO2 exposure time at 298 K. When exposed to a flow of 2.7 × 1015 molecules·cm−3 NO2 at 298 K, the increase of 9,10-AQ and 9-NANT concentration on sample particles (ng product/mg particle) displayed a near-linear pattern within a 360 min period (Figure 3), which suggests that the reactions can be reasonably described by linear kinetics. Therefore, the general rate law for the formation of 9,10-AQ and 9-NANT can be provided by eq I: [product] = k 0·[ANT(ads)]m ·[H 2O]n ·[NO2 ]l ·t = k·t

(I)

where m, n, and l are the reaction orders for the adsorbed ANT, water vapor, and NO2 gas, respectively; k0 and t are the reaction rate constant and the reaction time (min), respectively; and k is the apparent reaction rate constant (ng·mg−1·min−1). [Product] is the surface concentration of 9,10-AQ or 9-NANT on sample particles, which was determined by GC-MS; [Ant(ads)] is the concentration of ANT on the surface of the sample particles, which was determined by sample preparation; and [H2O] is the concentration of water vapor, which was determined from the relative humidity. 8673

dx.doi.org/10.1021/es501543g | Environ. Sci. Technol. 2014, 48, 8671−8678

Environmental Science & Technology

Article

considered the final concentration of 9,10-AQ and 9-NANT on particles exposed to different concentrations of NO2 (50, 100, 150, and 200 ppm) for the same time (180 min) under certain relative humidity conditions (RH 0, 10, 20, 40, and 60%). As shown in Figure 4, the production of 9,10-AQ increased dramatically with an increase of relative humidity in the range of RH 0−20% and decreased remarkably with a further increase of relative humidity from RH 40% to RH 60%, which indicates that the formation of 9,10-AQ from the heterogeneous reaction of adsorbed ANT on NaCl particles with NO2 can be promoted by the provision of additional water vapor under low relative humidity conditions (RH 0−20%) and can be suppressed by the provision of additional water vapor under high relative humidity conditions (RH 40−60%). Over the whole range considered in this study (RH 0−60%), the yield of 9-NANT decreased significantly with an increase of relative humidity, which indicated that water vapor can slow down the formation of 9-NANT. Because each curve in Figure 4 was a monotonic increasing function of the NO2 concentration (0−200 ppm), the formation of both 9,10-AQ and 9-NANT under certain relative humidity conditions can be enhanced significantly by the provision of additional NO2. As shown in Figure 5, the logarithm of both the 9,10-AQ and 9-NANT concentration, under a specific relative humidity, increased near-linearly with an increasing NO2 concentration. At a specific relative humidity, the [H2O] value is a constant that is a function of relative humidity. Therefore, the product concentration after the same reaction time (t0 = 180 min) at a specific relative humidity was only determined by the NO2 concentration by as shown by eq II:

Figure 3. Formation of 9,10-anthraquinone (circles) and 9-nitroanthracene (squares) in a flow of 2.7 × 1015 molcules·cm−3 NO2 at 298 K under (a) dry condition (RH 0%) and (b) wet condition (RH 10%).

[product]t0 = k 0·[ANT(ads)]m ·[H 2O]n ·t0·[NO2 ]l = k′·[NO2 ]l

13,14,16,17

Several studies have reported that the degradation of ANT can be described by pseudo-first-order kinetics and that a plateau is observed for the reaction of NO2 with ANT adsorbed on mineral substrates and graphite particles. However, the [Ant(ads)] value in this study could be regarded as a constant because the sample was prepared in a unified manner and the concentration of ANT preloaded onto the particle was in excess. Therefore, the apparent reaction rate constant (k) was only determined by the NO2 concentration and relative humidity. From dry (RH 0%) to wet (RH 10%) conditions, the k value, obtained by the slope of linear fitting (Figure 3), for the formation of 9,10-AQ increased from 0.056 ng·mg−1·min−1 to 0.320 ng·mg−1·min−1, while the k value for the formation of 9-NANT decreased from 0.541 ng·mg−1·min−1 to 0.194 ng·mg−1·min−1. When exposed to the same concentration of NO2 (2.7 × 1015 molecules·cm−3), the k value ratio of 9,10-AQ to 9-NANT rose from 0.1 at RH 0% to 1.6% at RH 10%, which indicates that dry conditions contribute to the formation of 9-NANT, while wet conditions contribute to the formation of 9,10-AQ. These preliminary results suggest that two parallel reaction pathways may be related to the formation of 9,10-AQ and 9-NANT, respectively, which may be significantly affected by the relative humidity. Effect of Relative Humidity. To investigate in detail the effects of the relative humidity on the reaction kinetics and reaction pathways for the heterogeneous reactions of ANT adsorbed on NaCl particles with NO2, a series of experiments were performed in a flow system. This experiment only

(II)

where k′ is a constant related to relative humidity and l represents the reaction orders of the heterogeneous reactions for NO2 concentration, which can be obtained from the slope of the bilogarithmic plot (Figure 5). Table 1 shows that when the relative humidity changed from RH 0% to RH 60%, the slope for 9,10-AQ ranged from 1.61 to 2.51, while that for 9-NANT ranged from 1.79 to 2.83. The mean slopes for 9,10-AQ and 9-NANT were 1.96 ± 0.34 and 1.96 ± 0.34, respectively, both of which approximated to 2. Therefore, the formation of both of 9,10-AQ and 9-NANT were determined to be second-order reactions with respect to the NO2 concentration. Discussion of the Reaction Mechanism. According to the kinetic analysis, reaction orders, and effect of relative humidity, two different mechanisms are proposed for the formation of 9,10-AQ and 9-NANT from the heterogeneous reaction of adsorbed ANT with NO2, each of which is coupled with the heterogeneous reaction of the NaCl substrate with NO2, as shown in Figure 6. Previous studies25,26,28,34 have demonstrated that two pathways are associated with the formation of sodium nitrate, shown in in situ DRIFTS spectra (Figure 1), from the heterogeneous reaction of an NaCl substrate:

8674

2NO2(g) ⇌ N2O4(g)

(1)

NaCl(s) + N2O4(g) → NaNO3(s) + ClNO(g)

(2)

N2O4(g) ⇌ N2O4(ads)

(3)

dx.doi.org/10.1021/es501543g | Environ. Sci. Technol. 2014, 48, 8671−8678

Environmental Science & Technology

Article

Figure 4. Concentration of (a) (c) 9,10-anthraquinone and (b) (d) 9-nitroanthracene on the particles after ANT-NaCl samples were exposed to different concentrations of NO2 (50, 100, 150, 200 ppm) for the same reaction time (180 min) under certain relative humidity conditions (RH 0, 10, 20, 40, 60%).

N2O4(ads) + H 2O(ads) ⇌ HNO2(ads) + HNO3(ads)

(4)

NaCl(s) + HNO3(ads) → NaNO3(s) + HCl(g)

(5)

HNO2(ads) ⇌ HONO(g)

(6)

surface. Therefore, one can reasonably assume that surfaceadsorbed water that depends on the relative humidity will also affect the formation of 9,10-AQ and 9-NANT from the heterogeneous reaction between NO2 and ANT adsorbed on NaCl particles. As described above, the rate of formation of 9,10-AQ, which is a second-order reaction, initially increases but then decreases with the increase of relative humidity. Because the 9 and 10 positions of ANT are highly reactive sites, the addition reaction of ANT with adsorbed dimers of nitrogen dioxide, N2O4 (ads), at the 9 and 10 positions, which leads to a decrease in the resonance energy of the aromatic ring from 351 kJ·kmol−1 to 300 kJ·kmol−1, is generally considered to be one of the most important initial reactions for the formation of 9,10-AQ:40

One mechanism, shown by reactions 1 and 2, is that the gaseous dimer of nitrogen dioxide, N2O4 (g), reacts directly with NaCl particles to generate NaNO3 and gaseous ClNO.26 Another mechanism, shown by reactions 3−6, is related to the heterogeneous hydrolysis of NO2 on the surface of the NaCl particles with the adsorbed dimer of nitrogen dioxide, N2O4 (ads), as an important precursor surface species in the reaction. Finlayson-Pitts et al.34 proposed that the symmetric form of the NO2 dimer, N2O4, is taken up on the surface and isomerizes to the asymmetric form, ONONO2, followed by autoionizing to NO+NO3−. This intermediate can react with water to generate gaseous HONO and surface-adsorbed HNO3. The HNO3 then reacts with the NaCl substrate to generate NaNO3 and gaseous HCl. Studies have reported that surface-adsorbed water has a significant influence on the heterogeneous reaction of NO2 on the surface of NaCl particles.35−37 This is due to the surface structure modification of NaCl particles when they are exposed to water vapor, which has been demonstrated by both atomic force microscopy (AFM)38 and infrared (IR) spectroscopy.39 At RH < 30%, water adsorbs primarily at the step edges and water-surface bonds are favored over water−water hydrogen bonds. At RH > 35%, a uniform layer of water is formed and the surface steps are observed to evolve slowly. At about RH 73%, the step structure becomes unstable and disappears abruptly due to the dissolution (deliquescence) of the salt

Dinitro-anthracene (marked as P), the intermediate generated from the addition reaction of ANT, reacts with water adsorbed on the surface of NaCl particles to generate another intermediate, dihydroxy-ANT (marked as Q), which is readily oxidized to 9,10-AQ by oxidants such as NO2 and O2. Because dihydroxy-ANT is extremely unstable, the reaction 9 is very fast while reaction 8 is the rate-determining steps in the whole 8675

dx.doi.org/10.1021/es501543g | Environ. Sci. Technol. 2014, 48, 8671−8678

Environmental Science & Technology

Article

where [H2O(ads)] is the surface concentration of adsorbed water; and K1, K3 and K7 are the equilibrium constants of reactions 1, 3, and 7, respectively. The K1 and K7 value are almost the constant. The K3 value is determined by the number of active surface sites for N2O4 adsorption, which is influenced by amount of waters adsorbed on the surface. When the relative humidity is raised, the [H2O(ads)] value will increase while the K value will decrease. When RH < ∼35%, the adsorbed water comprises less than a monolayer and sufficient active sites are available for N2O4 adsorption on the surface of NaCl particles. Therefore, the variation of the K value is negligible compared with the [H2O(ads)] value in proportion to the coverage of adsorbed water, which leads to a higher yield of 9,10-AQ with an increase of relative humidity. When RH > ∼35%, the surface of the sample particles is completely covered by adsorbed water. Although the adsorbed water layer becomes thicker, the [H2O(ads)] value primarily related to the coverage of adsorbed water is approximately constant. A change in the K value becomes dominant, which leads to a lower yield of 9,10-AQ with an increase in relative humidity. Over the whole range considered in this study (RH 0−60%), the rate of formation of 9-NANT, which is a second-order reaction, decreased significantly with an increase in relative humidity. Because the 9 and 10 positions of ANT have the highest electron density, one can reasonably assume that the electrophilic substitution at the 9 or 10 position is one of the most important reaction pathways for the formation of 9-NANT. Although NO+ has been proven to be a strong oxidant toward PAH with a redox potential (E0) of