Article pubs.acs.org/est
Modeling the Impact of Iron−Carboxylate Photochemistry on Radical Budget and Carboxylate Degradation in Cloud Droplets and Particles Christian Weller, Andreas Tilgner, Peter Braü er, and Hartmut Herrmann* Leibniz-Institut für Troposphärenforschung, Permoserstraße 15, 04318 Leipzig, Germany S Supporting Information *
ABSTRACT: To quantify the effects of an advanced iron photochemistry scheme, the chemical aqueous-phase radical mechanism (CAPRAM 3.0i) has been updated with several new Fe(III)−carboxylate complex photolysis reactions. Newly introduced ligands are malonate, succinate, tartrate, tartronate, pyruvate, and glyoxalate. Model simulations show that more than 50% of the total Fe(III) is coordinated by oxalate and up to 20% of total Fe(III) is bound in the newly implemented 1:1 complexes with tartronate, malonate, and pyruvate. Up to 20% of the total Fe(III) is found in hydroxo and sulfato complexes. The fraction of [Fe(oxalate)2]− and [Fe(pyruvate)]2+ is significantly higher during nighttime than during daytime, which points toward a strong influence of photochemistry on these species. Fe(III) complex photolysis is an important additional sink for tartronate, pyruvate, and oxalate, with a complex photolysis contribution to overall degradation of 46, 40, and 99%, respectively, compared to all possible sink reactions with atmospheric aqueous-phase radicals, such as •OH, NO3•, and SO4• −. Simulated aerosol particles have a much lower liquid water content than cloud droplets, thus leading to high concentrations of species and, consequently, an enhancement of the photolysis sink reactions in the aerosol particles. The simulations showed that Fe(III) photochemistry should not be neglected when considering the fate of carboxylic acids, which constitute a major part of aqueous secondary organic aerosol (aqSOA) in tropospheric cloud droplets and aqueous particles. Failure to consider this loss pathway has the potential to result in a significant overestimate of aqSOA production.
■
INTRODUCTION The occurrence of iron in tropospheric aerosol particles has been known since the mid-1970s.1 When iron enters the troposphere in particulate form as wind-blown dust or from anthropogenic sources, it is mostly found in different mineral structures, in amorphous hydroxides, such as Fe(OH)3, adsorbed on clay minerals, organic matter, or carbonaceous particles or bound in salts. Iron ions can be released from the original particulate form and be made available for complexation by inorganic or organic substances by a number of processes, e.g., ligand promoted or acidic dissolution.2,3 Iron mainly exists in its oxidation state III or II in natural systems and is converted from one form to the other by photochemical and radical reactions, which are the main driving processes of chemical conversions in the tropospheric aqueous phase. One of these redox cycling reactions is the photoreduction of iron complexes from Fe(III) to Fe(II) oxidation state, which is the focus of the present work. This photoreduction ultimately leads to the decomposition of the compounds bound to iron and the production of reactive radicals, which are suspected to initiate further oxidation processes.3−5 More specifically, it has been postulated that Fe(III)−carboxylate photolysis can be an important source of aqueous HxOy species (HO2, O2−, H2O2, and OH) in atmospheric waters.3,6−12 This qualitative © 2014 American Chemical Society
conclusion has been mainly derived from laboratory experiments involving irradiation of bulk solutions containing very few or only single carboxylate species, and it is questionable whether these findings are readily transferable to tropospheric conditions because real cloud and aerosol particle chemistry is impacted by mass exchange with the gas phase, which plays only a subordinate role in bulk aqueous solution experiments. Additionally, a quantitative understanding of the HxOy source strength from Fe(III) photochemistry is still missing. Because of the complexity of interacting tropospheric physical and chemical processes, which involve many reactants, chemical mechanisms, such as chemical aqueous-phase radical mechanism (CAPRAM),13 coupled with gas-phase mechanisms and microphysics are needed to assess the effects of Fe complex photochemistry. Currently, realistic daytime profiles of oxalic acid and oxalate cannot be produced with atmospheric chemistry mechanisms, such as CAPRAM, because the included Fe3+−oxalato complex photolysis leads to depletion of oxalate. This could arise from an underestimation of oxalate sources and Received: Revised: Accepted: Published: 5652
December March 25, March 28, March 28,
19, 2013 2014 2014 2014
dx.doi.org/10.1021/es4056643 | Environ. Sci. Technol. 2014, 48, 5652−5659
Environmental Science & Technology
Article
Figure 1. Fraction of total Fe(III) bound in different species simulated in CAPRAM: comparison of daytime versus nighttime (top) clouds and (bottom) deliquescent particles in the “Ext Fe” run, which includes new Fe(III) reactions.
an overestimation of Fe3+−oxalato complex photolysis or an incomplete representation of the iron chemistry. Other models,14 which do not include Fe3+−oxalato complex photolysis, overestimate oxalate with respect to concentrations found in ambient particle and cloudwater samples. This overestimation may lead to false conclusions regarding the connected chemical subsystems and an overestimation of the production of aqueous secondary organic aerosol, sometimes addressed as “aqSOA”, because carboxylic acids constitute a considerable fraction of secondary organic aerosol (SOA). Measurements of carboxylic acids (including oxalate) and metal concentrations in stratocumulus cloudwater samples have shown an inverse relation between metals, particularly, Fe, and oxalate.15 Box model calculations have suggested that the photolysis of Fe−oxalato complexes is a significant oxalate sink.15 Despite the recent inclusion of Fe(III)−oxalato complex photolysis in their model, Long et al. concluded that HxOy and Fe cycling is still not well understood with respect to the role of oxalate complexes.16 Furthermore, other relevant ligands for Fe(III) have not yet been considered in the state of the art tropospheric condensed-phase chemistry models. Data from laboratory investigations on absorption cross-sections, quantum yields, and transient and stable species are essential to an understanding of the fundamentals of Fe complex photochemistry and represent invaluable input for modeling studies. Such data have been used in atmospheric chemistry box
modeling to characterize the influence of an extended iron photochemistry on important atmospheric species, which is reported in the following.
■
MODEL AND MECHANISM DESCRIPTION Overview. New experimentally obtained absorption crosssections and quantum yields have been implemented17,18 into the chemical aqueous-phase radical mechanism (CAPRAM 3.0i,13,19), along with a reaction scheme that connects to the already existing reactions in the mechanism. Details concerning the mechanism CAPRAM 3.0i, including mechanism tables, can be found elsewhere.20 Simulations have been performed using CAPRAM as part of the spectral aerosol cloud chemistry interaction model (SPACCIM21). Fe(III) complexes, whose carboxylate ligands were previously part of the mechanism, namely, oxalate, malonate, glyoxalate, pyruvate, succinate, tartronate , and tartrate, were chosen for implementation. The impact of the new implemented Fe(III) photochemistry has been analyzed through comparison of a model “reference run without extended Fe(III) photochemistry” (here abbreviated as “Ref wo”) and a run “with extended Fe(III) photochemistry” (here abbreviated as “Ext Fe”). Formation of the inorganic complexes [FeOH]2+ and [Fe(SO4)]+ (photolysis included), [Fe(OH)2]+, [Fe(Cl)]2+, and [Fe(OH)2Fe]4+, and oxalate complexes [Fe(oxalate)]+, [Fe(oxalate)2]−, and [Fe(oxalate)3]3−, including the photolysis of the two latter species, 5653
dx.doi.org/10.1021/es4056643 | Environ. Sci. Technol. 2014, 48, 5652−5659
Environmental Science & Technology
Article
Table 1. Fraction of Main Sink Reactions of the Ligands (L) with Respect to the Sum of All Sink Reactions for Total Simulation (∑108 h) of Ext Fea sink reaction FeL + hν FeL2 + hν FeL3 + hν LH/LH2 + •OH L−/LH− + •OH LH/LH2 + NO3• L−/LH− + NO3• LH/LH2 + SO4• −
pyruvate (%) 40
glyoxalate (%)
malonate (%) 5.6
28 18 0.8 14
oxalate (%)
2.3
93 3.9 0.1
