Laboratory Insights into the Diel Cycle of Optical and Chemical Transformations of Biomass Burning Brown Carbon Aerosols.

Transformations of biomass burning brown carbon aerosols (BB-BrC) over their diurnal lifecycle are currently not well studied. In this study, the aging of BB tar proxy aerosols processed by NO3• under dark conditions followed by the photochemical OH• reaction and photolysis were investigated in tandem flow reactors. The results show that O3 oxidation in the dark diminishes light absorption of wood tar aerosols, resulting in higher particle single-scattering albedo (SSA). NO3• reactions augment the mass absorption coefficient (MAC) of the aerosols by a factor of 2-3 by forming secondary chromophores, such as nitroaromatic compounds (NACs) and organonitrates. Subsequent OH• oxidation and direct photolysis both decompose the organic nitrates (ONs, representing bulk functionalities of NACs and organonitrates) in the NO3•-aged wood tar aerosols, thus decreasing particle absorption. Moreover, NACs degrade faster than organonitrates by photochemical aging. The NO3•-aged wood tar aerosols are more susceptible to photolysis than to OH• reactions. The photolysis lifetimes for the ONs and for the absorbance of the NO3•-aged aerosols are on the order of hours under typical solar irradiation, while the absorption and ON lifetimes toward OH• oxidation are substantially longer. Overall, nighttime aging via NO3• reactions increases the light absorption of wood tar aerosols and shortens their absorption lifetime under daytime conditions.

Introduction

With over 3.9 Tg yr–1 emissions, biomass burning (BB) aerosols significantly impact the climate, atmospheric chemistry, and human health.[1−4] Modeling studies estimate a global radiative forcing of +0.03 to +0.57 W m–2 arising mainly from brown carbon components of BB (BB-BrC hereafter).[4−6] The frequency and severity of forest fires are likely to increase globally as a positive feedback to the warmer and drier climate and so are the BB-BrC emissions.[7−9] These scenarios call for systematic research to investigate the global impact of BB-BrC, which is eventually constrained by the understanding of the physical–chemical nature and atmospheric transformations of BB-BrC aerosols.

BB-BrC consist of an intricate organic matrix, with numerous molecular chromophores and inter/intramolecular interactions that make BB-BrC an efficient light absorber, especially in the near UV spectral region.[10−13] Once emitted, BB-BrC undergo atmospheric aging processes with a lifetime of hours to days.[10,14] During daytime, OH• reactions and photolysis dominate the aging of BB-BrC, while the NO3• and O3 oxidation account for the nighttime aging.[15−17] The typical diel pattern will continuously modify BB-BrC and alter their role in atmospheric chemistry and radiative forcing both regionally and globally. However, the temporal continuity in atmospheric aging processes is generally overlooked, and most studies focus on a single reaction pathway in BB-BrC aging or ignore the modifications due to diel atmospheric transformations altogether in climate models.[15,16,18,19]

In previous studies, photochemical aging of BB-BrC was explored and bleaching of the particles as a result of OH• oxidation was observed.[14,15,20] Many fires occur late in the day, suggesting nighttime processing of BB-BrC prior to the photochemical aging.[21−23] Heterogeneous reactions with O3 in the dark degrade the chromophores and decrease the absorption of BB-BrC.[19] In contrast, dark reactions of NO3• with BB-BrC form new chromophores, such as nitrophenols and other nitroaromatics, that enhance the absorption of the particles.[16] These results support previous observations by Lin et al. and Bluvshtein et al. who reported the formation of absorbing nitroaromatic chromophores in BB-BrC following nighttime aging.[24,25] Moreover, Bluvshtein and co-authors also observed a decrease in absorption and further oxidation of BB-BrC the following morning, possibly due to dilution and atmospheric reactions. In short, day and nighttime aging channels can result in different characters of the BB-BrC aging effects. Thus, understanding of the dynamic behavior of BB-BrC and the associated effects may be biased if only a single reaction pathway is considered. Therefore, systematic studies that cover the diurnal cycle of BB-BrC, including NO3• and O3 oxidation reactions that dominate the nighttime aging and OH• reactions that are most important during daytime, are needed to improve the understanding of BB-BrC transformations and to guide modeling developments.

It is challenging to conduct an exact in situ exploration of BB-BrC aging in light of the high complexity of biomass burning emissions and the instantaneous changes in the atmosphere. Laboratory simulations help provide mechanistic understanding necessary to comprehensively interpret field observations and improve modeling. Wood tar aerosols from biomass pyrolysis can be suitable BB-BrC proxies due to their comparable chemical composition and optical properties.[26−28] Such wood tar aerosols have been applied in a number of recent studies to investigate the atmospheric aging of BB-BrC in the laboratory.[16,20,28−30]

Here, the transformation of BB-BrC during the transition from nighttime to daytime oxidation environments is investigated. We use tandem flow reactors that couple the NO3• reaction in the dark and subsequent OH• photo-oxidation to process wood tar aerosols. The optical and chemical transformations are characterized using real-time particle-phase measurements, including a complex refractive index (RI) over the wavelength range of 330–650 nm and bulk aerosol non-refractory composition. In addition, filter samples were collected for functional group analysis and individual particle investigation of the morphology and viscosity. These results elucidate the diurnal behaviors of BB-BrC and help simulate the associated influences on climate and atmospheric chemistry.

Methods and Experiments

BB-BrC Proxy Preparation

Following a method described previously,[16,20,28] wood tar emulsions were collected from pyrolysis of wood fuel and then the emulsions were concentrated, dissolved in acetonitrile, and filtered through a 0.2 μm PTFE syringe (Acrodisc, Pall). Particles were generated via nebulization of the tar solution with N2 as the carrier gas. A heat shock device and series of activated charcoal denuders were fixed downstream to compact the particles and to remove gaseous volatile components.

BB-BrC Aging Experiments

Diel aging of wood tar aerosol was conducted in tandem flow tube reactors, including a custom-built aerosol flow reactor (AFR, length: 165 cm, I.D.: 10.5 cm), and an aerodyne potential aerosol mass-oxidation flow reactor (PAM-OFR). The experimental setup and parameters are given in the Supporting Information (SI, Text 1, Figure S1, Table S1).

In the nighttime aging simulation, the polydisperse wood tar aerosols passed through a laminar AFR, where they were oxidized in the dark for 300 s. For NO3• oxidation, O3 and NO in a concentration ratio of 5:3 were mixed in a long Teflon tube (length: 650 cm, I.D.: 1.4 cm) for 85 s to produce NO3• and N2O5 prior to introduction into the AFR. To minimize hydration reactions, the experiments were performed under dry conditions in the AFR and Teflon tube (RH < 0.2%). Heterogeneous O3 oxidation of wood tar aerosols was conducted without NO addition to simulate O3-dominating aging in less polluted environments. O3 was generated by flowing oxygen through a low-pressure mercury lamp (78-2046-07, BHK Inc., CA). O3 concentration was monitored at the output of the AFR by an ozone analyzer (model 106-L, 2B Technology). O3 scrubbers and charcoal denuders were connected after the AFR to terminate the reactions and to remove gaseous components, including O3, excess NO, and other volatile species.

The NO3•-aged wood tar aerosols were then transferred into a PAM-OFR, where they underwent OH• aging and photolysis for 190 s at RH ∼35%. 254 nm UV light was applied to photolyze O3 and to produce OH• by reaction of the excited singlet oxygen (O1D) with water vapor. The intensity of the 254 nm UV light in the PAM-OFR was controlled to achieve OH• exposure (OHexp) ranging from 1.3 × 1011 to 9.3 × 1011 molecule cm–3 s, which is equivalent to 1.0–7.2 days of ambient daytime oxidation (EAD) assuming an average OH• concentration of 1.5 × 106 molecule cm–3.[31] The OHexp was deduced from the decay of injected SO2 that was monitored by an SO2 analyzer (model 43i, Thermo) in preliminary experiments. For convenience, the OH• photo-oxidation experiments were labeled with a prefix of EAD, such as EAD-1.0 indicating equivalent to 1.0 day of atmospheric aging and so forth.

To investigate the role of UV irradiation (254 nm) in the transformation of the NO3•-aged wood tar aerosols throughout the PAM-OFR, photolysis tests were conducted separately on the particles under normal experimental conditions in the absence of O3. The UV light intensities were similar to the EAD-1.0 and EAD-7.2 experiments and these two experiments were termed Weak and Strong photolysis, respectively.

Aerosol Characterization

Particle size distributions during the aging experiments were monitored by a scanning mobility particle sizer (SMPS, TSI model 3080). A novel broadband cavity enhanced spectrometer (BBCES) was used to characterize optical evolutions of wood tar aerosols. Detailed operation of BBCES and RI derivation methods have been published elsewhere.[16,32] Briefly, particles were size-selected via an aerosol aerodynamic classifier (AAC; Cambustion, U.K.) and then the aerodynamic monodispersed particles were flow-split and parallel-scanned via another SMPS, a condensation particle counter (CPC, model 3775, TSI), and the BBCES to get their mobility size, number concentration, and broadband light extinction coefficient (αext), respectively. The complex refractive indices, RIs, (RIs = n + ik) were determined between 330 and 650 nm (resolution of 0.5 nm) based on the Mie–Lorenz theory, where the real part n and imaginary part k indicate the intrinsic scattering and absorption strength by the particle material, respectively. Knowing the RIs, the aerosol single-scattering albedo (SSA) was derived. In addition, particle densities were calculated by comparing their aerodynamic and mobility size distributions.

The compositions of non-refractory aerosol constituents were analyzed by a high-resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS) in real-time. The fragments in the mass spectra were classified into five families based on their chemical formulas: hydrocarbon-like (CH+), less-oxygenated (CHO+), more-oxygenated (CHO+), nitrogen-containing (CHON+) organic components, and nitrogen oxides (NO+), where x, y, i, p ≥ 1 and z > 1. Wood tar aerosols were also quantitatively collected on Teflon filters (0.2 μm, Pall). The filter samples were dissolved in methanol via vortex shaking extraction. The extracts were further filtered with a 0.2 μm PTFE syringe (Acrodisc, Pall) before UV–vis absorption (300–800 nm) measurements (Ocean Optics, model USB 650). The extracts were then concentrated by a gentle flow of N2. The concentrated samples were deposited uniformly onto KBr windows (diameter: 13 mm, thickness: 2 mm, Spectra-Tech Inc.) and dried with N2 flow. The infrared absorbance of the deposited samples was recorded over the wavenumber range of 600–4000 cm–1 using a Fourier transform infrared spectrometer (FTIR, Thermo Scientific Nicolet 6700). Samples of individual particles were also impacted onto silicon wafers for topographic morphology analysis by an atomic force microscope (Multimode AFM, Bruker Nano Surfaces).

Solvents (acetonitrile and methanol, Sigma-Aldrich, >99.9%) and gases (O2 and N2, >99.999%; NO and SO2, 50 ppmv diluted in high-purity N2) are in high purity and used as received.

Chemical-Box Modeling and Assessment of Photolysis

To quantify NO3• and O3 oxidation, a chemical-box model accounting for NO-O3-NO3•-N2O5 reactions and their sinks to wood tar particle surface and to the AFR inner wall was developed and is described in the Supporting Information (Text S2, Tables S2 and S3, Figures S2 and S3). The modeling revealed that the reactive uptake of NO3• by wood tar aerosols was about an order of magnitude higher than that of N2O5 and O3, highlighting the dominant oxidation pathway by the NO3• reaction in the AFR. Moreover, the chemical-box modeling suggests an NO3• exposure of 0.31 ppbv h for wood tar aerosols throughout the AFR, which is equivalent to about 15.6 h of ambient aging by NO3• assuming a typical concentration of 20 pptv at night.[22,33] Based on another method that uses the particle surface area-weighted NO3• uptake in the AFR and ambient NO3• sink rate to smoke particles,[16] an equivalent ambient NO3•-aging time (EAN) of about 20.7 h is estimated for wood tar aerosols, confirming that the AFR results are atmospherically relevant. Similarly, we used 370 ppbv h of O3 exposure to wood tar aerosols in separate heterogeneous O3 reaction experiments. Assuming an ambient ozone concentration of 35 ppbv, this O3 exposure corresponds to an atmospheric processing time of about 10.6 h.

Since the PAM-OFR lamps do not replicate the solar spectrum, the irradiation of the particles in the PAM-OFR is converted into an equivalent time under ambient sunlight by considering the effective photolysis; see the method and calculation in the Supporting Information (Text S3, Figure S4 and Tables S4 and S5). In short, nitroaromatic compounds, such as nitrocatechol, nitroguaiacol, etc., that have been measured with high abundance in the NO3•-aged wood tar aerosols and ambient BB-BrC are used as chemical actinometers;[10,11,16,34] their photolysis effect under Weak and Strong photolysis experimental conditions are calculated by multiplying the residence time by their photolysis rates, the latter integrated over the 250–500 nm spectrum in the PAM-OFR. In addition, solar photolysis rates for these actinometers were estimated considering a solar zenith angle (SZA) in the range of 0–80° under cloudless ground level conditions (June 30th, 2019, albedo of 0.19). The ratios of the photolysis effect in the PAM-OFR and the solar photolysis rate yield the equivalent solar photolysis times, which were 0.34 ± 0.29 and 2.49 ± 1.55 h on average for the Weak and Strong photolysis conditions, respectively.

Results and Discussion

Chemical Transformations Inferred from HR-ToF-AMS Measurements

The AMS spectra of wood tar aerosols following various aging pathways are displayed in Figure S5, the features of the chemical evolutions are shown in Figure and summarized in Table S6. Dark reactions with both O3 and NO3• generated more oxygenated organic components (CHO+ and CHO+) and the bulk aerosol O/C ratios increased as a result of aged wood tar aerosols (Figure A,B). Substantial nitrate formation (7.2 wt %) together with an increase of nitrogen-containing ions (CHON+, 3.8–4.3 wt %) were measured for wood tar aerosol after NO3• aging. The low NO2+/NO+ ratio (0.19, smaller than that of 0.45 for standard NH4NO3 particles) in the mass spectrum indicates the organic nature of nitrates.[35,36] In accordance with our previous findings, these organic nitrate (ON) peaks mainly measure the bulk functionalities of organonitrates (-ONO2) and nitroaromatic compounds (-NO2) in NO3•-aged wood tar aerosols.[16] In contrast to O3 oxidation, the 15.6–20.7 h EAN of NO3• reactions increased the size and mass concentration (by over 30%) of the aerosols (Figure S2). These results suggest that heterogeneous chemistry driven by NO3• rather than by O3 is an efficient nighttime source of secondary organic components and ONs.

Figure 1

HR-ToF-AMS measurements of wood tar aerosol aging via various reaction pathways. (A) Normalized fragment compositions of fresh and aged wood tar aerosols following O3 and NO3• reactions, OH• photochemical oxidation (EAD-1.0 to EAD-7.2), and photolysis (WP and SP symbolize Weak and Strong photolysis, respectively). (B) Particulate nitrate (fHR-NO3) formation as a function of the bulk organic O/C ratio, the marker size increases with NO2+/NO+ ratios in the range 0.19–0.44. HR-NO3 measures the organic nitrate (-ONO3) and nitro (-NO2) functionalities only. (C) Van Krevelen diagram shows the aerosol elemental changes with aging (H/C vs O/C). (D) Plot of normalized mass spectrum signals at m/z-44 (CO2+, f44) and m/z-43 (C2H3O+, f43) to trace the oxygenation degree of the organic aerosol. The triangular region (dashed lines) indicates the ambient oxygenated organic aerosol (OOA), which can be further classified into less-oxidized OOA (LO-OOA) and more-oxidized OOA (MO-OOA) with a critical value of 0.12 for f44.[37] (E) Plot of f44 vs normalized fragments at m/z-60 (C2H4O2+, f60). Ambient biomass burning organic aerosols commonly fall inside the triangular region defined by the gray dashed lines.[38] Scheme of the multiple-reaction pathways in a square-chart are added to (D) and (E).

The underlying reaction mechanism by O3 and NO3• aging can be inferred from the trends of H/C ratios and functional group fragments, such as carbonyl and carboxyl/peroxide fractions indicated by f43 (m/z-43, C2H3O+) and f44 (m/z-44, CO2+), respectively. No significant change (p < 0.01) in particle H/C ratio was measured after O3 oxidation (Figure C), while the NO3• aging reduced the hydrogen content of the aerosol, presumably due to hydrogen-abstraction and hydrogen-transfer reactions, which are important channels in the reaction with polyaromatic hydrocarbons (PAHs) and phenols that are abundant in fresh wood tar aerosols to produce nitroaromatic compounds (NACs). In addition, unsaturated carbon bonds are preferentially attacked by these oxidants. Different from the O3 reaction with single double bonds that form Criegee intermediates, NO3• reactions followed by O2 addition generate nitro-oxy-peroxy radicals that can undergo rapid inter/intramolecular reactions and react with NO2, NO3•, and HO2• to generate oxygenated products bearing carbonyl, carboxyl, organic peroxide, or/and nitrate functional groups. In the O3 driven reactions, the Criegee intermediates decompose to form aldehydes and carboxylic acids via further reactions. Accordingly, both f43 and f44 increased after O3 and NO3• oxidation (Figure D). In general, f44 and the O/C ratio correlate with the hygroscopicity and volatility of the organic aerosol[37] but oxidation via NO3• generates organonitrates, which are commonly hydrophobic with low volatility.[39,40] Therefore, the exact reaction pathways should be considered for assessing the physical properties, such as cloud-condensation nucleation activity and volatility, of aged organic aerosols.

Following OH• exposure, the NO3•-aged wood tar aerosol became increasingly more oxidized. More-oxygenated organic compounds (CHO+ and CHO+) were generated while hydrocarbon-like (CH+) and nitrogen-containing (CHON+ and NO+) components decreased in the mass spectra (Figure A). These observations are consistent with the increased O/C ratios and decrease in N/C ratios as shown in Figure C and Table S6, but the H/C ratios increased at an early stage (EAD-1.0) due to OH• addition and then H/C ratios decreased from hydrogen abstraction and fragmentation reactions via further OH• exposure. Although the ion families exhibit similar trends, photochemical aging via photolysis decreased the aerosol H/C, O/C, and N/C ratios to a lesser extent. As a result, photolysis slightly increased the particle carbon oxidation state (O̅S̅C ≈ 2 × O/C-H/C-5 × N/C),[41] while OH• photo-oxidation enhanced the carbon oxidation state of NO3-aged wood tar aerosol remarkably (Table S6).

Decrease in the ON content was detected following both OH• photo-oxidation and photolysis of NO3•-aged wood tar aerosols (Figure B and Table S6). The NO2+/NO+ ratio of the remaining nitrate gradually increased from 0.19 to 0.26, indicating their consistent organic feature but with varying ratios during aging. Meanwhile, the aerosol mass concentration progressively decreased (Figure S6) demonstrating that the photochemical results are a combination of functionalization and fragmentation. Considering the constant low humidity (35% RH) and short residence time (190 s) in the PAM-OFR, hydrolysis is unlikely to be a governing factor of the decrease in particulate nitrates.[42] The decreasing trends of ONs are due to fragmentation or/and evaporation of particle-phase organonitrates and NACs as a result of the OH• reaction and photolysis. Although the two aging channels represent different reaction pathways, the nitrate (-ONO2) and nitro (-NO2) functional groups are commonly cleaved from the carbon-chain and should be released as gaseous reactive nitrogen species, such as HNO3, HONO, or NO2. Besides, other reactions that increase the volatility of organonitrates and NACs, like carbon-chain cleavage and aromatic-ring opening to produce smaller molecules with high volatility that partition into the gas phase, should also contribute to the observed decrease of ONs in aged wood tar aerosols.

The contribution of the OH• reaction and photolysis to ON depletion can be separated from the two experiments. In Weak and Strong photolysis scenarios that equal to 0.3 and 2.5 h solar irradiation (method and SI, Text S3), the particle-phase nitrates decreased by 13 and 35%, respectively. The particulate nitrate mass concentration as a function of irradiation time was fit assuming a time-dependent exponential degradation. As such, an ambient solar photolysis lifetime of 5.9 ± 2.6 h was estimated for the ONs in the NO3•-aged wood tar aerosols (for details on the lifetime estimation please refer to SI, Figure S7). We assume no synergetic effect between OH• reactions and photolysis in aging of wood tar aerosol. Subtracting the photolysis contribution, the lifetime of the particulate nitrates against OH• reactions is around 32.3 ± 14.6 days, assuming an ambient average OH• concentration of 1.5 × 106 molecule cm–3. Therefore, the particulate ONs are more susceptible to photolysis than to OH• reactions. This observation is consistent with previous studies on the chemical transformations of NACs and organonitrates, where faster direct photolysis than OH• reactions in their atmospheric losses was highlighted.[43−47] In short, photolysis lifetimes of minutes to hours have been estimated for many nitrophenols and organonitrates, while their lifetimes with respect to OH• reactions range from days to weeks during atmospheric transport.[43−45,48−50] Moreover, based on the lifetime, the second-order degradation kinetic rate for the reaction of OH• with ONs in NO3•-aged wood tar aerosols was estimated to be (2.4 ± 1.1) × 10–13 cm3 molecule–1 s–1, which is comparable with the kinetics for most atmospheric relevant nitrophenols and organonitrates in the range of (2.6–67.2) × 10–13 cm3 molecule–1 s–1.[44,46,51−53]

A plot of f44 vs f43 (Figure D) demonstrates that the photochemical aging of NO3•-aged wood tar aerosols generated more carboxyl or/and peroxide-containing products at the expense of carbonyl compounds and this is more prominent for OH• photo-oxidation. In addition, a rapid and consistent decay of the anhydrous sugar marker, m/z-60 for C2H4O2+, was observed with respect to all of the aging processes studied here (Figure E); the result again suggests that biomass burning aerosols lose some of their specific chemical tracers during atmospheric transport.[38,54]

Chemical Transformations Deduced from FTIR Analysis

Based on the complementary FTIR spectra shown in Figure A, a large contribution of carboxyl, carbonyl, hydroxy, alkyl, and carbon double-bond containing components can be identified in the unprocessed wood tar aerosols, which are supported by previous molecular characterization where organic acids, aldehydes, PAHs, and phenols were identified.[16,20] As expected, the strength of the C=C (1599, 1515, and 1458 cm–1) vibrations and C-H stretch (3060–2800 cm–1) decreased following O3 oxidation in the dark, while the signal of oxygenated bands (-OH, C=O, C-O stretch) increased in aged wood tar aerosol, consistent with O3 reactions with double bonds to form aldehydes, ketones, and carboxylic acids. The striking differences in the IR spectra of NO3•-aged wood tar are observed in the fingerprint region that can be attributed to the stretching of nitrate (-ONO3 at 1645, 1280, and 854 cm–1) and nitro (-NO2 at 1359, 1386, 1528, and 1554 cm–1) functional groups,[55,56] further supporting the formation of organonitrates and NACs in the NO3•-aged wood tar aerosol.[16] In addition, alkyl (C-H), unsaturated carbon bonds (C=C) and hydroxyl C=C-OH vibrations decreased. Yet, the C=O vibration strengthened and a blue shift of this band from the initial peak at 1710 cm–1 for fresh wood tar particles to 1730 cm–1 is observed following the NO3• reactions, indicating more aldehyde/ketone formation than carboxyl as the aldehyde/ketone associated C=O vibration commonly appears at higher wavenumbers.[57,58]

Figure 2

FTIR spectra of fresh and processed wood tar aerosols. (A) Chemical transformations of wood tar aerosols via O3 and NO3• reactions. (B) NO3•-aged wood tar aerosol following photochemical oxidation (EAD-7.2, OHexp: 9.3 × 1011 molecule cm–3 s, and equivalent to 7.2 days of atmospheric OH• aging) and Strong photolysis (254 nm UV light exposure: 9.3 × 1016 photon cm–2 and equivalent to about 2.5 h of ambient solar photolysis). The signals were normalized to an integrated area of the entire spectrum. Typical absorption peaks corresponding to specific chemical bonds were identified and marked. Absorption peaks for RONO2 and Ar-NO2 were convoluted on the basis of fresh wood tar aerosol spectra (right axis).

Normalized FTIR spectra of NO3•-aged wood tar aerosol in photochemical aging are shown in Figure B. Consistent with the AMS measurements, all signals of the nitrate and nitro group associated bands decreased following photochemical aging. It is noteworthy that the nitrate bands at 1645, 1280, and 854 cm–1 decreased to a lesser extent compared to the loss of nitro stretching at 1554, 1528, and 1359 cm–1 in both OH• photo-oxidation and photolysis. To be more specific, the deconvoluted peak area ratios of integrated nitrate to nitro bands (-ONO2/-NO2) increased with photochemical aging of NO3•-aged wood tar aerosols (Figure S8). These results demonstrate the higher decomposition rate of NACs than organonitrates in NO3•-aged wood tar aerosols. It is reasonable that most NACs, identified as nitrophenols, are labile to photochemistry. The variation of the NO2+/NO+ ratio for the remaining nitrates from the HR-ToF-AMS measurements follow the decomposition preference of organonitrates and NACs in photochemistry. Apart from the obvious depletion of nitrate and nitro bands, photochemical aging increased the -OH stretch peak. In OH• photo-oxidation, the -OH stretch, C=O vibration, and unsaturated hydroxyl C=C-OH vibration increased at the expense of sp2-aromatic and aliphatic C-H stretching. Thus, considerable carboxyl, alcohol, or/and phenol hydroxy formation in the aging can be expected. Upon photolysis, a stronger -OH stretch and unsaturated hydroxyl C=C-OH vibration were observed in the aerosol, while the C=O band slightly decreased. These results suggest that carbonyl or/and carboxyl decreased, while phenols and possibly other alcohols were generated in the photodissociation of organonitrates and NACs. Based on the measured chemical transformations, related references and the MCM platform (http://mcm.leeds.ac.uk/MCM), mechanisms for degradation of typical NACs and organonitrates with respect to photolysis and OH• reactions are summarized and proposed in Figure S9.

Evolution of Optical Properties

RI retrieval based on the Mie–Lorenz theory assumes a spherical morphology and uniform composition of the particles. According to the AFM measurements (Figure S10), the wood tar aerosols generated in this study are spherical semiliquid particles with a relatively low viscosity. Meanwhile, the uniform mixing states of aged wood tar aerosols are inferred from the size-resolved bulk chemical composition via HR-ToF-AMS measurements (Figure S11).

Figure shows the evolution of the wavelength-dependent RIs for wood tar aerosol in the different aging schemes. Note that the optical results for photolysis experiments are presented in Figure S12. The exemplary optical parameters are summarized in Table S7 and detailed RI distributions are displayed in Figure S13. The imaginary RI component derived from the mass absorption coefficients (MAC) of methanol-extracted solutions exhibit almost identical distribution and temporal evolution trends as those derived from direct particle light extinction measurements (Figure S13), thus justifying the precision and reliability of the BBCES measurements.

Figure 3

Optical evolution of wood tar aerosols through various atmospheric aging processes, including O3 and NO3• reactions in the dark (gray area: nighttime aging simulation) and photochemical transformation of NO3•-aged wood tar aerosols (yellow area: daytime photochemical simulation). (A) Continuous wavelength-dependent complex real refractive index (RI) as a function of atmospheric transformation. (B) Continuous wavelength-dependent complex imaginary RI as a function of atmospheric evolution (left y-axis). Absorption Ångström exponent (AAE) was derived to indicate the wavelength dependence of light absorption (right y-axis). (C) Wavelength-dependent single-scattering albedo (SSA) was estimated based on retrieved RIs for wood tar particles at 200 nm. (D) Changes of the wavelength-weighted mean mass absorption coefficient (MAC, m2 g–1) for wood tar aerosols via various atmospheric aging processes. Two categories are shown: near UV absorption (330–400 nm) and visible absorption (400–550 nm). Uncertainties for each RI and SSA distributions (±0.005 for the real part, ±0.006 for the imaginary part, and ±0.013 for the SSA on average from 330 to 550 nm) are not presented for clarity.

The absorption properties and wavelength dependence (characterized by Absorption Ångström exponent, AAE) of the fresh wood tar aerosols are consistent with our previous studies[16,20] and consistent with the values found in previous field and laboratory studies of BB-BrC.[59−63] Nighttime aging by O3 or NO3• reactions lead to distinctly different optical results for wood tar aerosols. O3 oxidation decreased the light absorption of wood tar aerosol (commonly termed “bleaching”), especially at longer wavelengths (Figure B). For example, the imaginary RIs decrease from (0.017 ± 0.005)i to (0.011 ± 0.003)i at 405 nm but the value difference at 365 nm is within the detection precision of the BBCES system. Clearly, the faster decay of absorption at longer wavelengths leads to a higher AAE (10.9) of the aged wood tar aerosols. In contrast, the real part distribution increased by 0.018 (Figure A), resulting in an increase of the particle single-scattering albedo (SSA). Unless otherwise stated, the SSA values in this study were calculated for 200 nm particles. After O3 oxidation in the dark, the SSA for wood tar aerosols changed insignificantly from 0.85 to 0.86 at 365 nm and increased weakly from 0.92 to 0.95 at 405 nm (Figure C and Table S7). The wavelength-weighted mean MAC () can be used to quantify the absorption changes following atmospheric transformations:We divided the particle absorption into UV (330–400 nm) and visible (400–550 nm) ranges, namely MAC_UV and MAC_Vis, respectively. In Figure D, the MAC_UV decreased by less than 5% and the MAC_Vis decreased by more than 25% for the wood tar aerosol after atmospheric exposure equivalent to 35 ppbv O3 for 10.6 h in the dark. Similarly, Browne et al. observed rapid bleaching of BB-BrC by O3 oxidation, but to a lesser extent compared to our study, while the AAE increased from an initial 7.5 to a final 9.5.[19] In a similar manner, bleaching by O3 oxidation was also observed for tar ball aerosols.[20] It was suggested that some chromophores or/and intramolecular interactions, such as charge-transfer-conjugations among oxygenated functional groups (carbonyl), which can contribute to absorption at longer wavelengths, were preferentially depleted by O3 reactions, thereby resulting in wavelength-dependent bleaching and a higher AAE. In the Browne et al. study, after initial bleaching some residual absorption by BB-BrC was still observed following additional O3 oxidation, suggesting that a fraction of the chromophores and/or intramolecular interactions were substantially resistant to bleaching or that formation of oxygenated moieties and some secondary chromophores offset the absorption reduction.[19] Thus, it is suggested that BB-BrC have a potentially long absorption lifetime with respect to O3 reactions during atmospheric transport.

Consistent with our recent report,[16] the results here show that NO3• reactions in the dark enhance the light absorption of BB-BrC. After an EAN of 15.6–20.7 h, the light absorption in the visible range increased approximately 3 times and doubled in the UV region for wood tar aerosols. As a result, the SSA of aged wood tar particles reduced to 0.70 at 365 nm and 0.80 at 405 nm, respectively (SI, Table S7). The formation of absorbing nitroaromatics, especially the nitrophenols, contributed most of the absorption enhancement, since addition of the nitrogroup leads to a significant red shift of phenol absorption.[64,65] Moreover, unidentified chromophores and products with a high molecular weight, synergistically with possible intramolecular interactions, can also contribute to the enhanced absorption.[16,24]

Wavelength-dependent bleaching was observed during photochemical aging of NO3•-aged wood tar aerosols. As shown in Figure and Table S7, more rapid bleaching at the longer wavelength increased the AAE gradually during the course of OH• photo-oxidation. Apart from the dominant loss of absorption, the real part (n) distributions shifted to lower values within approximately 4 days of equivalent atmospheric aging (EAD-4.4) and then remained stable with further aging (Figure A). Overall, subsequent OH• photochemical reactions lead to higher SSA values for NO3•-aged wood tar aerosols.

Consistent with the loss of nitrated groups seen in the FTIR and AMS results, photolysis led to bleaching of NO3•-aged wood tar aerosols but to a lesser extent than OH• photo-oxidation (Figures S12 and S13 and Table S7), demonstrating that both photolysis and OH• uptake channels can diminish light absorption. Similarly, photolysis resulted in the faster decay of absorption at the longer wavelength, thus increasing the AAE of the aged wood tar aerosols. An equivalent 0.3 h of solar irradiation (Weak photolysis) decreased the MAC_UV by 8.6%, while it degraded the MAC_Vis by 14.6%. The Strong photolysis by an equivalent 2.5 h of solar irradiation resulted in a decrease of 22.9 and 31.3% for MAC_UV and MAC_Vis, respectively. Assuming an exponential decay of the integrated absorption with aging time, the light absorbing lifetime for nighttime NO3•-aged wood tar aerosols due to photolysis by sunlight was estimated to be 9.9 ± 5.1 and 6.9 ± 2.9 h at the UV and visible regions, respectively. Excluding the photolysis induced bleaching, the absorption lifetime attributed to OH• heterogeneous reactions for NO3•-aged wood tar aerosols was 31.3 ± 16.6 and 27.8 ± 15.4 days for UV and visible ranges, respectively. It is worthy to note that the absorption lifetimes are fairly consistent with those of ONs against OH• reactions (32.3 ± 14.6 days) and those due to photolysis by solar radiation (5.9 ± 2.6 h), further suggesting that the organonitrates and NACs are the main contributors to the absorption of NO3•-aged wood tar aerosol. Note that the absorption lifetimes were estimated based on the aging time before the wood tar aerosol reached the final light-absorbance state. Several investigations have reported that the decay of BB-BrC absorption following photochemical aging leveled off due to formation of persistent chromophores with a relatively high molecular weight. Approximately 70% of the initial BB-BrC absorption was lost due to rapid reactions.[10,14,63,66,67] Assuming an exponential decay of the absorption to a final state with the remaining 30% absorbance in both photolysis and OH• reactions, the lifetimes for the NO3-aged wood tar aerosols were estimated to be 11.9 ± 6.1 and 8.4 ± 3.4 h for the UV and visible channels, respectively, with respect to photolysis. The absorption lifetimes against OH• oxidation are 37.6 ± 20.0 and 33.4 ± 18.6 days for UV and visible ranges, respectively.

In conclusion, direct photolysis by sunlight is more important than OH• reactions in determining the fate of NO3•-aged wood tar aerosols. The overall absorption lifetime coupling diurnal photolysis and OH• reactions for the nighttime aged aerosols should be about 10 h on average, which is a relevant timescale for regional-scale atmospheric transport. The lifetime reported here is comparable with results from many studies and observations, where a decay time of hours to several days was reported for BB-BrC absorption under typical atmospheric conditions[15,23,34,68,69] Fleming et al. found a relatively slow bleaching rate for fresh BB-BrC by photolysis. The photolysis lifetimes range from 10 to 41 days, depending on the fuel types and burning conditions.[10] They suggested that heterogeneous oxidation by OH• is a more important pathway in fresh BB-BrC degradation, which is different from our results for NO3•-aged BB-BrC. This discrepancy can be explained by the chemical difference of the BB-BrC, or more specifically, explained by the continuous diel transformation. According to our previous study on photochemical aging of fresh tar aerosol, we found lower sensitivity to photolysis for the fresh aerosols, while a substantial decrease of their absorption was observed following OH• oxidation with a lifetime of about 4.3 days (SI, Figure S14).[20] The nighttime NO3• aging augments the absorption of wood tar aerosols via generation of nitroaromatic and organonitrate chromophores, while these secondary chromophores are susceptible to photodegradation in the diel transition. In fact, according to the study by Fleming et al., the nitrogen-containing compounds, mostly NACs, have shorter photolysis lifetimes (within 12 h) than those of PAHs and lignin-derived compounds.[10] Therefore, we suggest that NO3• reactions convert chromophores with relatively long lifetimes, such as phenols and PAHs, into short-lived nitroaromatic compounds with respect to photobleaching. More work on this topic is hence recommended.

Atmospheric Implications

O3 and NO3• are ubiquitous oxidants in nighttime biomass burning plumes but they cause different effects in the transformation of BB-BrC. The former oxidizes BB-BrC concurrent with chromophore depletion, thereby, decreasing particle light absorption. By contrast, the NO3• reaction enhances the light absorption of BB-BrC by generating substantial chromophores with nitrate and nitro functional groups, such as organonitrates and NACs.[16] During the transition from the night to daytime photochemical environment, NO3•-aged BB-BrC lose their absorption through OH• driven reactions and photolysis. The decomposition of organonitrates and NACs, represented by bulk ONs in this study, is consistent with the bleaching of the NO3•-aged BB-BrC in photochemical aging. NACs decompose faster than organonitrates, presumably due to the strong absorption of NACs in the near UV region of the solar spectra and their chemical structures with the electron-donating nitro-group and benzene ring that make NACs prefer to react in photochemistry.

OH• reactions and photolysis both make the NO3•-aged BB-BrC scatter and absorb sunlight less efficiently by decreasing their RIs and particle size, eventually resulting in higher SSA for the particles. Moreover, photolysis plays a more important role than the OH• reaction in determining the fate of nighttime processed aerosols. Specifically, the particle-phase ONs in the NO3•-aged BB-BrC have a photolysis lifetime on the order of hours (5.9 ± 2.6 h), while their lifetime against OH• reactions is on the order of weeks (32.3 ± 14.6 days). Accordingly, ONs gained efficiently at night via NO3• reactions decompose rapidly during the following day, demonstrating that the NO3•-aged BB-BrC serve as the temporary NO sink. These reactive nitrogen species, such as HNO3, HONO, and NO2, released from the NO3•-aged BB-BrC during photochemical aging are expected to impact NO and O3 budgets in the troposphere.

The absorption of NO3•-aged BB-BrC decays with a lifetime of 31.3 ± 16.6 and 27.8 ± 15.4 days at the UV and visible range, respectively, due to OH• reactions. Solar photolysis shortens the absorption lifetimes to 9.9 ± 5.1 and 6.9 ± 2.9 h at the UV and visible regions, respectively. The short lifetime suggests the rapid transformation in absorption of the aerosols in the atmosphere. Compared with the previous study in which bleaching via the OH• reaction was highlighted in aging of fresh BB-BrC,[10,20] the current work shows that NO3• oxidation at night increases the susceptibility of BB-BrC to solar photolysis. Overall, these findings demonstrate that the entire diel oxidation cycle of BB-BrC must be considered in assessment of their atmospheric lifetime for accurate incorporation into models of climate and atmospheric chemistry.

This study shows that the transition from night to daytime oxidative environments impacts the fate of BB-BrC. The bulk aerosol chemical transformations, although limited in nature in the present study, are consistent with the optical evolution of BB-BrC on the basis of our previous findings. We suggest that the detailed analysis and identification of molecular chromophore identifications are crucial for explaining the chemical and optical results for BB-BrC following complicated aging processes. Such detailed chemical and optical property measurements provide an important direction for future research.