- Email: [email protected]
Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 3: Carboxylic and dicarboxylic acids
ARTICLE IN PRESS
Atmospheric Environment 40 (2006) 6676–6686 www.elsevier.com/locate/atmosenv
Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 3: Carboxylic and dicarboxylic acids Kelley C. Barsanti, James F. Pankow Department of Environmental and Biomolecular Systems, OGI School of Science & Engineering, Oregon Health & Science University, 20000 NW Walker Road, Beaverton, OR 97006, USA Received 23 December 2005; accepted 12 March 2006
Abstract The term ‘‘accretion reactions’’ has been used to describe the collection of reactions by which organic compounds can react with one another and/or other atmospheric constituents, forming products of higher-molecular weight (MW) and lower volatility, and thus increasing their tendency to condense [Barsanti, K.B., Pankow, J.F., 2004. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 1: Aldehydes and ketones. Atmospheric Environment 38, 4371–4382; Barsanti, K.B., Pankow, J.F., 2005. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 2: Dialdehydes, methylglyoxal, and diketones. Atmospheric Environment 39, 6597–6607]. Studies have shown that a significant fraction of atmospheric organic particulate matter (OPM) may be comprised of high-MW/low-volatility compounds [e.g., Havers, N., Burpa, P., Lambert, J., Klockow, D., 1998. Spectroscopic characterization of humic-like substances in airborne particulate matter. Journal of Atmospheric Chemistry 29, 45–54], which would be consistent with the occurrence of such accretion reactions in the atmosphere, [e.g., Jang, M., Czoschke, N. M., Lee, S., Kamens, R. M., 2002. Heterogeneous atmospheric organic aerosol production by acid-catalyzed particle-phase reactions. Science 298, 814–817]. However, many uncertainties exist regarding accretion reactions as they may occur in the atmosphere, including identification of those reactions most likely to contribute to OPM. Barsanti and Pankow (2004, 2005) have developed and applied a general theoretical approach to evaluate the thermodynamic favorabilities of accretion reactions, including the extents to which they may be relevant for OPM formation in the atmosphere. That approach is applied here in the consideration of OPM formation by reactions of four mono- and dicarboxylic acids (acetic, malic, maleic, and pinic) to form esters and amides. It was concluded that for all of the acids considered, ester and amide formation are thermodynamically favored under the assumed conditions. For malic, maleic, and pinic acids, and likely for similar mono- and dicarboxylic acids, significant OPM formation may occur via ester and amide formation in the atmosphere when kinetically favorable. r 2006 Published by Elsevier Ltd. Keywords: Organic particulate matter; Secondary organic aerosol; SOA; Accretion reactions; Oligomers; Acids; Alcohols; Amides; Esters
Corresponding author. Tel.: +1 503 690 1196.
E-mail address: [email protected] (J.F. Pankow). 1352-2310/$ - see front matter r 2006 Published by Elsevier Ltd. doi:10.1016/j.atmosenv.2006.03.013
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
1. Introduction The molecular level characterization of the organic particulate matter (OPM) found in the atmosphere has proven difficult. Such OPM can be very complex, and at least part of that complexity is a result of the innumerable pathways by which organic compounds can react in the atmosphere. Volatile organic compounds (VOCs) can be oxidized to form multi-functional compounds, and oxidation products of sufficiently low vapor pressure can condense to form OPM. Additionally, oxidation products may undergo ‘‘accretion reactions’’ with one another and/or other atmospheric constituents and thereby add molecular weight (MW), further decrease volatility, and thus increase their tendency to condense (Barsanti and Pankow, 2004, 2005). Significant studies in this context include Haagen-Smit (1952), Tobias and Ziemann (2000), and Jang and Kamens (2001b). It has been suggested that a significant fraction of atmospheric OPM samples can be comprised of high-MW/lowvolatility compounds, loosely referred to as ‘‘oligomers’’ (Havers et al., 1998; Samburova et al., 2005). The presence of such compounds in the atmosphere would be consistent with accretion reactions of a variety of types (Gao et al., 2004a, b; Iinuma et al., 2004; Jang et al., 2002; Kalberer et al., 2004; Limbeck et al., 2003; Tobias and Ziemann, 2000; Tolocka et al., 2004). Chamber studies have focused on accretion reactions as a general mechanism for OPM formation, and on the possibility of a role for acid catalysis in increasing OPM formation by certain such reactions (e.g., Jang and Kamens, 2001b). Some studies have begun to consider the detection and quantification of oligomers per se, and the effects of particle acidity and parent-compound structure. These studies have shown that: (1) atmospherically relevant constituents (though not necessarily at atmospherically relevant levels) can react to form oligomers (Gao et al., 2004a, b; Iinuma et al., 2004; Kalberer et al., 2004; Tolocka et al., 2004); (2) a significant fraction of the total OPM formed in chamber experiments can be comprised of oligomers, in some cases 450% (Gao et al., 2004a; Kalberer et al., 2004); (3) inorganic seed particles are not required for oligomer formation (Kalberer et al., 2004); when inorganic seed particles are present, the extent of oligomer formation may be affected by particle acidity (Gao et al., 2004b; Iinuma et al., 2004; Jang et al., 2002; Tolocka et al.,
6677
2004); and (4) the nature of precursor aerosol seed may affect the type and extent of oligomer formation (Gao et al., 2004a, b; Kalberer et al., 2004). Despite such advances, many uncertainties exist regarding accretion reactions as they may occur with atmospherically relevant compounds. For example, in a chamber study involving the oxidation of a-pinene, a highly relevant biogenic OPM precursor, Tolocka et al. (2004) suggested that one of the several oligomers they detected could be explained by nine different possible combinations of monomers. Barsanti and Pankow (2004, 2005) have developed and applied a general theoretical approach to evaluate the thermodynamic favorabilities of accretion reactions, including the extents to which they may be relevant for OPM formation in the atmosphere. In the consideration of C4–C10 aldehydes, ketones, dialdehydes, diketones, and methylglyoxal, it was concluded that: (1) hydration/oligomerization, hemiacetal/acetal formation, and aldol condensation are not favored for C4–C10 monoand diketones, or for C5 and lower mono- and dialdehydes; (2) aldol condensation of C6 and higher mono- and dialdehydes may contribute to atmospheric OPM formation under some circumstances when kinetically favored; and (3) diol and diol-oligomer formation from glyoxal as well as aldol condensation of methylglyoxal are thermodynamically favored and may contribute significantly to OPM in the atmosphere when kinetically favored. This work applies the method of Barsanti and Pankow (2004) in the consideration of OPM formation by reaction of mono- and dicarboxylic acids to form esters and amides. The study compounds are acetic, malic, maleic, and pinic acid (Fig. 1). Carboxylic acids are of interest because they are produced by oxidation of both anthropogenic and biogenic VOCs (Grosjean et al., 1978; Jang and Kamens, 2001a; Yu et al., 1999); ester and amide formation are among the most important reactions involving carboxylic acids (Loudon, 1995). 2. Overview of theoretical approach 2.1. Thermodynamic framework and mathematical solution process The proposed mechanism by which oxidation products and other atmospheric constituents (e.g.,
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6678
Fig. 1. Parent carboxylic and dicarboxylic acids.
A and B) may contribute to OPM formation is: Ag þ Bg 2Cg Cg 2Cliq
accretion,
(1)
gas=particle ðG=PÞ equilibration of accretion product
ð2Þ
in which C is a relatively high-MW/low-volatility compound. Accretion of A with B may also occur according to the following thermodynamically equivalent schemes: (1) Ag+Bliq2Cliq, then Cliq2Cg; and (2) Aliq+Bliq2Cliq, then Cliq2Cg. The overall conclusions (e.g., predicted extents of OPM formation) will be independent of the actual scheme by which the accretion reaction proceeds, and independent of whether a catalyst is present in the PM phase (Barsanti and Pankow, 2004, 2005). A thermodynamic framework used to predict OPM formation by accretion reactions has been described by Barsanti and Pankow (2004, 2005). Briefly, equilibrium for an accretion reaction (e.g., Eq. (1)) is considered using its equilibrium constant K as Pset by its Gibbs free energy of reaction DG1 ð¼ i ni DGf ;i Þ according to DG1 ¼ RT ln K, where: ui is the stoichiometric coefficient for compound i in the accretion reaction; DGf;i (kJ mol1) is the standard free energy of formation for i; R (kJ mol1 K1) is the ideal gas constant; and T (K) is temperature. For each parent compound and product, the equilibrium constant Kp for gas/ particle (G/P) partitioning (e.g., Eq. (2)) is governed largely by its pure-liquid vapor pressure pL (atm) (Pankow, 1994b). The tendency of any given accretion reaction to form significant OPM depends on its K, the value of Kp for the product, and the mass-concentration driving force for accretion product formation/condensation as provided by the initial amounts of reactants and the atmospheric level of non accretion-related OPM into which the product can condense, referred to here as OPMna (Barsanti and Pankow, 2004). Mass balance considerations lead to the following expression for initial concentration (A0, mg m3) of
the parent compound A (Barsanti and Pankow, 2005): N X nA MWA C i;g A0 ¼ Ag þ Aliq þ nC;i MWC;i i¼1 N X nA MWA þ Ci;liq ð3Þ nC;i MWC;i i¼1 In Eq. (1), Ag and Aliq denote the species A in the gas and liquid phases, respectively. In Eq. (3), we allow that Ag and Aliq can also be used to denote the concentrations (mg m3) of A in the gas and liquid phases. Similarly, in Eq. (3) the parameters Ci,g and Ci,liq represent the gas- and liquid-phase concentrations (mg m3) of the ith C product. N is the number of possible products from A. Each u is an overall stoichiometric coefficient (does not carry sign) for production of Ci from A, and each MW (g mol1) is a molecular weight. The fundamental equations describing formation of any product Ci and the G/P partitioning of A and Ci at equilibrium (Pankow, 1994a), can be used to express Aliq, Ci,g, and Ci,liq in terms of Ag (see Barsanti and Pankow, 2005). Eq. (3) is then easily solved for Ag given the necessary input parameters. Analogous expressions of Eq. (3) exist for reaction schemes different from Eq. (1). In this work, mass balance equations like Eq. (3) were solved using the standard Microsoft Excel Solver, developed by Frontline Systems, Inc., and executed through a subroutine written in Visual Basic. 2.2. Input parameters and assumed initial conditions Needed pL and DG f values were obtained from the literature when available. When not available, pL values were estimated using SPARC (Hilal et al., 1994) and DG f values were estimated using the Benson (1976) method as implemented by the CHETAH (2002) algorithm. For molecular groups not found in the CHETAH database (Table 1), or missing enthalpy of formation ðDH f Þ or entropy
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6679
Table 1 Molecular groups with DH f and S1 values that are not available in the CHETAH (2002) database and so required interpolation in this work DH f (kJ mol1)
Group: Benson notation
Group: physical description
CH–(QC,O,CO)
C with: one single bond to a hydrogen; one single bond to a doubly-bonded carbon; one single bond to an oxygen; and one single bond to a carbonyl C with: one single bond to a hydrogen; one single bond to a nitrogen; one single bond to a carbonyl; and one single bond to a carbon O with: one single bond to a carbon; and one single bond to a carbonyl C with: one double bond to an oxygen; one single bond to a doubly-bonded carbon; one single bond to a nitrogen N with: three single bonds to carbonyls
CH–(N,CO,C) O–(C, CO) CO–(QC,N) N–(3CO)
S (J mol1 K1)
27.5
44.2
20.9
48.7
179.8 130.5
12.9 66.0
127.7
99.6
esterification w/2-methyl-3-buten-1,2-ol O
OH
OH
OH
n
O O
O
O
+ m
HO
+ (n + m -1) H2O
O HO
OH
CH3
CH2
H 3C
O
C2H
amide formation w/diethylamine O
OH
OH O
+ m
HO
O
O H C 3
NH
NC
CH 3
CH3
+ m H 2O
O
OH
CH3
amide formation w/ammonia O
O
OH O
n
+ m NH3
HO OH
OH O
+ (n + m -1) H2O
O N
Fig. 2. Accretion reactions of interest for acetic, malic, maleic, and pinic acids, shown here for malic acid.
(S1) data, their DH f and S1 values were estimated using linear interpolation as described by Frurip et al. (2002). Values of activity coefficients in the OPM phase (z) were set equal to 1 (see Barsanti and Pankow, 2004, 2005). OPM formation by the Fig. 2 accretion reactions was computed for each of the parent acids by reaction with 2-methyl–3-buten–1,2-ol (MBO), diethylamine (DEA), and ammonia (NH3). Individual accretion reactions and products are listed in Table 2. When considering amide formation, simultaneous salt formation between the organic acid and the amine (or ammonia) was not considered, though we note that such formation
could decrease the amount of OPM by amide formation. Also, the effects of the presence of an inorganic acid on amide formation were not considered explicitly, though we note that such an acid would protonate both amines and ammonia, and thus again reduce the amount of OPM by amide formation. In case 1, A0 ranged from 103 to 103 mg m3 with RH ¼ 20%; in case 2, A0 ¼ 1 mg m3 with RH from 5–95%. In both cases, it was assumed that MBO0 and DEA0 ¼ 1 mg m3, NH3 was held constant at 0.07 mg m3 (based on a lower-limit ambient value from Seinfeld and Pandis, 1998), OPMna ¼ 10 mg m3, P ¼ 1 atm, and T ¼ 298 K.
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6680
Table 2 Summary of accretion reactions and products of interest for acetic, malic, maleic, and pinic acids; nomenclature illustrated for malic acid Accretion product
n Parent compound
m Alcohol or amine
Type of linkage
n malic acid+m 2-methyl-3-butene-1,2-ol 2 accretion product+(n+m1) H2O m.1 1 1 ester m.2 1 2 ester m.3 2 1 ester m.4 2 2 ester m.5 2 3 ester m.6 3 2 ester m.7 3 3 ester n malic acid+m diethylamine 2 accretion product+(n+m1) H2O m.8 1 1 amide m.9 1 2 amide n malic acid+m ammonia 2 accretion product+(n+m1) H2O m.10 1 1 amide m.11 1 2 amide m.12 2 1 amide m.13 2 2 amide m.14 2 3 amide m.15 3 1 amide m.16 3 2 amide m.17 3 3 amide m.18 3 4 amide m.19 4 4 amide Corresponding accretion products were considered for acetic acid (a.1, a.3, a.8, a.10, a.12, and a.15), maleic acid (me.1–19), and pinic acid (p.1–19). For maleic acid, formation of the hydrate (me.20) was also considered.
3. Results and discussion 3.1. Predicted additional OPM formation in Case 1 (A0 ¼ 103– 103 mg m3 MBO0 and , DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, and RH ¼ 20%) Fig. 3 illustrates predicted levels of total additional OPM as a function of A0 for acetic, malic, maleic, and pinic acids by the Fig. 2 accretion reactions. Barsanti and Pankow (2005) proposed a ‘‘criterion of importance’’ that a compound would be of interest in an accretion reaction context if it could raise the OPM level by 1% under the assumed conditions. For the conditions of Case 1, when A0 ¼ 1 mg m3, maleic, malic, and pinic acids meet that criterion, with the predicted levels of additional OPM as follows: maleic, 1.2 mg m3; malic, 1.2 mg m3; and pinic, 0.27 mg m3. For each of
Fig. 3. Predicted total additional OPM by accretion reactions of acetic, malic, maleic, and pinic acids as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
these acids, an ester is the dominant accretion product in the OPM phase (me.3, m.6, and p.3, respectively); however, amides (predominately me.15, m.15, and p.15, respectively) also contribute to the additional OPM (see Figs. 4–6). For maleic and malic acids, the linear relationship between predicted additional OPM and A0 in Fig. 3 is a consequence of the essentially complete conversion of the parent compound to condensable products. Extending the lower limit of A0 would eventually result in a lack of complete conversion. While esterification and amide formation are thermodynamically favorable for acetic acid, the pL values of the accretion products are not low enough to allow significant condensation of the products. Thus, even when A0 ¼ 103 mg m3 the , predicted level of additional OPM by the reactions of acetic acid is not significant (E8 105 mg m3). For malic, maleic, and pinic acids, as A0 increases from 103 to 103 mg m3, increases in additional OPM are predicted (Fig. 3) as are changes in OPM composition. For maleic acid, the dominant accretion product shifts from the ester me.3 to the amide me.15 (Fig. 4). For malic acid, this shift is from the ester m.3 to the ester m.6 and then to the amide m.15 (Fig. 5). For pinic acid, the shift is from the amide p.8 to the ester p.3 and then to the amide p.15 (Fig. 6). Recalling that NH3 was held constant at a lower-limit ambient value (0.07 mg m3), when we
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6681
Fig. 4. Fractional contribution of each of the dominant accretion products to total additional OPM as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
Fig. 5. Fractional contribution of each of the dominant accretion products to total additional OPM as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
ARTICLE IN PRESS 6682
K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
Fig. 6. Fractional contribution of each of the dominant accretion products to total total additional OPM as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, RH ¼ 20%).
allow NH3 to be held constant at higher values, up to 10 mg m3 (which is within the range of ambient values discussed by Finlayson-Pitts and Pitts (2000)), the conclusions regarding significance of additional OPM formation by reaction of each acid do not change. For acetic and pinic acids, the levels of additional OPM increase with A0 due to the increased driving force for condensable product formation (Fig. 7). For maleic and malic acids, however, predicted additional OPM decreases slightly with increasing A0 due to the increased favorability of the less-condensable accretion products me.15 and m.15, respectively, over the esters me.3 and m.6, respectively. These results suggest that levels of OPM may not always increase monotonically with increases in total initial concentrations of reactants. 3.2. Predicted OPM formation in Case 2 (A0 ¼ 1 mg m3, MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, and RH ¼ 5% – 95%) The effect of varying RH on the extent of ester and amide formation in the gas phase can be understood by considering that water is a product in both cases. A lower ambient RH value thus
Fig. 7. Predicted levels of total additional OPM as a function of NH3 concentration (held constant), when A0 ¼ 1 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
promotes the formation of both of these types of accretion products, though the relative magnitude of the observed effect will depend on the K value for
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
the reaction. For acetic and pinic acids, the K values for ester and amide formation are not so large that the extent of accretion product formation is not affected by changes in RH. For maleic and malic acid, the K values for some of the accretion reactions are sufficiently large that the reactions go nearly to completion even at relatively high RH. When the accretion reactions are sufficiently favorable, whether significant OPM forms under the conditions considered depends upon the pL values of the accretion products. For acetic acid, even at the lowest RH considered, the level of additional OPM formed is not significant because the pL values for its ester and amide products are relatively large. For pinic acid, the additional OPM formed falls below the criterion of significance when RH rises above 50% (when RH ¼ 50%, additional OPME6 102 mg m3; when RH ¼ 95%, additional OPME2 102 mg m3). If we take into account that predicted DG f values may be uncertain by as much as 3% (Barsanti and Pankow, 2005), then adjusting the K values to favor accretion product formation leads to predicted levels of additional OPM for pinic acid at RH450% that meet the criterion. For maleic and malic acids, conversion to condensable products is nearly complete, and so varying RH has little effect on the predicted levels of additional OPM (large enough K values for some accretion reactions and small enough pL values for some of the corresponding products). The levels and composition of predicted additional OPM in Figs. 3–7 are for T ¼ 298 K. In the ambient atmosphere, temperatures lower than 298 K are common. For all of the accretion reactions considered, DH values are negative so that lowering temperature favors product formation. Assuming DH values are independent of temperature over a small range, for a 10 K decrease in temperature to 288 K, the calculated equilibrium constants increase by 1–3 orders of magnitude depending on the overall number of ester and amide bonds formed (100.9 for products with m ¼ 1, 101.6 for m ¼ 2, 102.6 for m ¼ 3, and 103.4 for m ¼ 4). In addition to increasing favorability for accretion product formation, decreasing temperature leads to lower vapor pressures of the accretion products, which favors partitioning to the OPM. 3.3. Implications for oligomer formation reported in chamber studies The chamber studies of Gao et al. (2004a, b), Iinuma et al. (2004), and Tolocka et al. (2004)
6683
involved oxidation of a-pinene, which is known to produce a number of cyclic compounds including pinic acid, hydroxy pinonaldehyde, and pinonaldehyde (Yu et al., 1999). It has been suggested that cyclic compounds can undergo ring opening in the presence of a weak acid (Chakraborty et al., 2003; Shi and Xu, 2002; Tolocka et al., 2004). For hydroxy pinonaldehyde and pinonaldehyde there are potentially eight products of ring opening (Fig. 8); addition of water across the unsaturated bond of each of those products can form a total of four mono-alcohols and two diols, with which carboxylic acids may react to form esters. To determine whether esterification might have been important in the a -pinene/O3 chamber studies of Gao et al. (2004a, b), Iinuma et al. (2004), and Tolocka et al. (2004), OPM formation was evaluated for esterification of pinic acid with the hydrated products from ring opening of hydroxy pinonaldehyde and pinonaldehyde (see Fig. 8 and Table 3). Initial concentrations of pinic acid were calculated by combining reacted a-pinene (DHC) values ranging from 6 to 19,000 mg m3 with an average molar yield value from chamber experiments of Yu et al. (1999). For the hydrated products of ring opening, hydroxy pinonaldehyde and pinonaldehyde were first assumed to form in amounts corresponding to the amount of pinic acid, then undergo ring opening by schemes I and II in equal proportions (100% complete), then hydration (to equilibrium), and finally, reaction with the pinic acid (to equilibrium). The average RH for the chamber experiments, 50%, was assumed. Other assumptions were: OPMna ¼ 0 mg m3 (i.e., no initial organic aerosol), P ¼ 1 atm, and T ¼ 298 K. The results are summarized in Table 4. The predicted levels of OPM in Table 4 are nonnegligible for all DHC values considered. As DHC increases, the dominant accretion product does not change, but other accretion products contribute with increasing proportions to the OPM formed. The molecular weights of the dominant accretion products (256–695 g mol1) are within the MW range of the ‘‘oligomers’’ reported in the chamber experiments of Gao et al. (2004a, b) and Iinuma et al. (2004); the functional groups on the dominant accretion products are also consistent with those suggested for the chamber experiments of interest. Based on the estimation method utilized here to obtain DG f values, the addition of water across an unsaturated bond is thermodynamically favorable. If kinetically favorable, other cyclic oxidation
ARTICLE IN PRESS 6684
K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
Fig. 8. Potential schemes (I and II) and products of ring opening from hydroxy pinonaldehyde and pinonaldehyde. A product of ring opening (hpa/pa) is shown reacting with water to form an alcohol (hpa.20/pa.20) and then the alcohol reacting with pinic acid to form an ester (hpa.21/pa.21). Table 3 Summary of accretion reactions and products for pinic acid reacting with the products and hydrated products of ring opening from hydroxy pinonaldehyde and pinonaldehyde; nomenclature illustrated for the product hpb m H2O or alcohol
Type of linkage
n pinic acid+m H2O2accretion product (alcohol) hpb.20 1
1
alcohol
n pinic acid+m alcohol/diol2accretion product+(n1) H2O hpb.21 1 hpb.22 1 hpb.23 2 hpb.24 2 hpb.25 2 hpb.26 3 hpb.27 3
1 2 1 2 3 2 3
ester ester ester ester ester ester ester
Accretion product
n Parent compound
Corresponding accretion reactions and products were considered for pinic acid with the additional products and hydrated products of ring opening from hydroxy pinonaldehyde (accretion product: hpa.20-hpa.22, hpc.20-hpc.22, and hpd.20-hpd.27) and from pinonaldehyde (accretion product: pa.20-pa.22, pb.20-pb.22, pc.20-pc.22, and pd.20-pd.22).
products and unsaturated aliphatic oxidation products may form alcohols. Given the numerous generation possibilities for reactants, esterification
reactions may be important in chamber experiments involving oxidation of VOCs other than a-pinene, including those of Gao et al. (2004a, b).
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6685
Table 4 Total levels and composition of predicted OPM formed by esterification of pinic acid with the hydrated products of ring opening from hydroxy pinonaldehyde and pinonaldehyde DHC (mg m3)
Initial concentration of pinic acid (mg m3)
Total OPM (mg m3)
6 56 280 560
0.41 3.8 19 38
0.15 1.7 8.9 18
840
57
27
19000
1300
650
Dominant accretion products (40.1% of total OPM) in order of decreasing contribution ester ester ester ester ester ester ester ester ester
(hpd.23), ester (hpb.23), diester (hpb.26) (hpd.23), ester (hpb.23), diester (hpb.26) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21), (pd.21) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21), (pd.21) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21), (pd.21), ester (hpa./pa.21), hydrate (pb.20), hydrate (hpc./pc.20)
P ¼ 1 atm, T ¼ 298 K, RH ¼ 50%.
4. Conclusions The approach used herein, including the method used to predict DG f values and therefore K values for accretion reactions, indicates that for all acids considered, ester and amide formation are thermodynamically favored under the assumed conditions. In the case of malic, maleic, and pinic acids, the accretion products have sufficiently low pL values that they are predicted to reside predominately in the OPM phase, contributing to total OPM. Thus, for malic, maleic, and pinic acids, and likely for similar carboxylic/dicarboxylic acids, significant OPM formation may occur via esterification and amide formation in the atmosphere when kinetically favorable. Acknowledgment This work was supported by the Electric Power Research Institute. References Barsanti, K.B., Pankow, J.F., 2004. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 1: Aldehydes and ketones. Atmospheric Environment 38, 4371–4382. Barsanti, K.B., Pankow, J.F., 2005. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 2: Dialdehydes, methylglyoxal, and diketones. Atmospheric Environment 39, 6597–6607. Benson, S.W., 1976. Thermochemical Kinetics. Wiley, New York, pp. 336. Chakraborty, D., Rodriguez, A., Chen, E.Y.X., 2003. Catalytic ring-opening polymerization of propylene oxide by organo-
borane and aluminum Lewis acids. Macromolecules 26, 5470–5481. CHETAH Version 7.3: The ASTM Computer Program for Chemical Thermodynamic and Energy Release Evaluation (CHETAH), 2002. ASTM International, Pennsylvania. Finlayson-Pitts, B.J., Pitts, J.N., 2000. Chemistry of the Upper and Lower Atmosphere. Academic Press, California, p. 969. Frurip, D., Downey, J.R., LaBarge, M., 2002. Estimation of Missing Group Data of Organic Species for Use in the CHETAH Program. ASTM International, Pennsylvania. Gao, S., Keywood, M., Ng, N.L., Surratt, J., Varutbangkul, R.B., Bahreini, R., Flagan, R.C., Seinfeld, J.H., 2004a. Lowmolecular-weight and oligomeric components in secondary organic aerosol from the ozonolysis of cycloalkenes and apinene. Journal of Physical Chemistry A 108, 10147–10164. Gao, S., Ng, N.L., Keywood, M., Varutbangkul, R.B., Bahreini, R., Nenes, A., He, J., Yoo, K.Y., Beauchamp, J.L., Hodyss, R.P., Flagan, R.C., Seinfeld, J.H., 2004b. Particle phase acidity and oligomer formation in secondary organic aerosol. Environmental Science and Technology 38, 6582–6589. Grosjean, D., Van Cauwenberghe, K., Schmid, J.P., Kelley, P.E., Pitts, J.N., 1978. Identification of C3–C10 aliphatic dicarboxylic acids in airborne particulate matter. Environmental Science and Technology 12, 313–317. Haagen-Smit, A.J., 1952. Chemistry and physiology of Los Angeles smog. Industrial Engineering and Chemistry 44, 1342–1346. Havers, N., Burpa, P., Lambert, J., Klockow, D., 1998. Spectroscopic characterization of humic-like substances in airborne particulate matter. Journal of Atmospheric Chemistry 29, 45–54. Hilal, S.H., Carreira, L.A., Karickhoff, S.W., 1994. Estimation of chemical reactivity parameters and physical properties of organic molecules using SPARC. In: Murray, J.S. (Ed.), Quantitative Treatments of Solute/Solvent Interactions. Elsevier, Amsterdam (http://ibmlc2.chem.uga.edu/sparc/index.cfm). Iinuma, Y., Bo¨ge, O., Gnauk, T., Hermann, H., 2004. Aerosolchamber study of the a-pinene/O3 reaction: influence of particle acidity on aerosol yields and products. Atmospheric Environment 38, 761–773.
ARTICLE IN PRESS 6686
K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
Jang, M., Kamens, R.M., 2001a. Characterization of secondary aerosol from the photooxidation of toluene in the presence of NOx and 1-propene. Environmental Science and Technology 35, 3626–3639. Jang, M., Kamens, R.M., 2001b. Atmospheric secondary aerosol formation by heterogeneous reactions of aldehydes in the presence of a sulfuric acid aerosol catalyst. Environmental Science and Technology 35, 4758–4766. Jang, M., Czoschke, N.M., Lee, S., Kamens, R.M., 2002. Heterogeneous atmospheric organic aerosol production by acid-catalyzed particle-phase reactions. Science 298, 814–817. Kalberer, M., Paulsen, D., Sax, M., Steinbacher, M., Dommen, J., Prevot, A.S.H., Fisseha, R., Weingartner, E., Frankevich, V., Zenobi, R., Baltensperger, U., 2004. Identification of polymers as major components of atmospheric organic aerosols. Science 12, 1659–1662. Loudon, M.G., 1995. Organic Chemistry, 3rd ed. Benjamin, California, p. 1390. Limbeck, A., Kulmala, M., Puxbaum, H., 2003. Secondary organic aerosol formation in the atmosphere via heterogeneous reaction of gaseous isoprene on acidic particles. Geophysical Research Letters 30, doi:10.1029/2003/ GL017738. Pankow, J.F., 1994a. An absorption model of gas/particle partitioning of organic compounds in the atmosphere. Atmospheric Environment 28, 185–188.
Pankow, J.F., 1994b. An absorption model of the gas/particle partitioning involved in the formation of secondary organic aerosol. Atmospheric Environment 28, 189–193. Samburova, V., Kalberer, M., Zenobi, R., 2005. Characterization of high molecular weight compounds in urban atmospheric particles. Atmospheric Chemistry and Physics Discussions 5, 437–454. Seinfeld, J.H., Pandis, S.N., 1998. From Air Pollution to Climate Change. Wiley, New York, p. 1325. Shi, M., Xu., B., 2002. Lewis acid-catalyzed ring-opening reactions of methylenecyclopropanes with alcoholic or acidic nucleophiles. Organic Letters 13, 2145–2148. Tobias, H.J., Ziemann, P.J., 2000. Thermal desorption mass spectrometric analysis of organic aerosol formed from reactions of 1-tetradecene and O3 in the presence of alcohols and carboxylic acids. Environmental Science and Technology 34, 2105–2115. Tolocka, M.P., Jang, M., Ginter, J.M., Cox, F.J., Kamens, R.M., Johnston, M.V., 2004. Formation of oligomers in secondary organic aerosol. Environmental Science and Technology 38, 1428–1434. Yu, J., Cocker, D.R., Griffin, R.J., Flagan, R.C., Seinfeld, J.H., 1999. Gas-phase oxidation products of monoterpenes: gaseous and particulate products. Journal of Atmospheric Chemistry 34, 207–258.
Atmospheric Environment 40 (2006) 6676–6686 www.elsevier.com/locate/atmosenv
Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 3: Carboxylic and dicarboxylic acids Kelley C. Barsanti, James F. Pankow Department of Environmental and Biomolecular Systems, OGI School of Science & Engineering, Oregon Health & Science University, 20000 NW Walker Road, Beaverton, OR 97006, USA Received 23 December 2005; accepted 12 March 2006
Abstract The term ‘‘accretion reactions’’ has been used to describe the collection of reactions by which organic compounds can react with one another and/or other atmospheric constituents, forming products of higher-molecular weight (MW) and lower volatility, and thus increasing their tendency to condense [Barsanti, K.B., Pankow, J.F., 2004. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 1: Aldehydes and ketones. Atmospheric Environment 38, 4371–4382; Barsanti, K.B., Pankow, J.F., 2005. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 2: Dialdehydes, methylglyoxal, and diketones. Atmospheric Environment 39, 6597–6607]. Studies have shown that a significant fraction of atmospheric organic particulate matter (OPM) may be comprised of high-MW/low-volatility compounds [e.g., Havers, N., Burpa, P., Lambert, J., Klockow, D., 1998. Spectroscopic characterization of humic-like substances in airborne particulate matter. Journal of Atmospheric Chemistry 29, 45–54], which would be consistent with the occurrence of such accretion reactions in the atmosphere, [e.g., Jang, M., Czoschke, N. M., Lee, S., Kamens, R. M., 2002. Heterogeneous atmospheric organic aerosol production by acid-catalyzed particle-phase reactions. Science 298, 814–817]. However, many uncertainties exist regarding accretion reactions as they may occur in the atmosphere, including identification of those reactions most likely to contribute to OPM. Barsanti and Pankow (2004, 2005) have developed and applied a general theoretical approach to evaluate the thermodynamic favorabilities of accretion reactions, including the extents to which they may be relevant for OPM formation in the atmosphere. That approach is applied here in the consideration of OPM formation by reactions of four mono- and dicarboxylic acids (acetic, malic, maleic, and pinic) to form esters and amides. It was concluded that for all of the acids considered, ester and amide formation are thermodynamically favored under the assumed conditions. For malic, maleic, and pinic acids, and likely for similar mono- and dicarboxylic acids, significant OPM formation may occur via ester and amide formation in the atmosphere when kinetically favorable. r 2006 Published by Elsevier Ltd. Keywords: Organic particulate matter; Secondary organic aerosol; SOA; Accretion reactions; Oligomers; Acids; Alcohols; Amides; Esters
Corresponding author. Tel.: +1 503 690 1196.
E-mail address: [email protected] (J.F. Pankow). 1352-2310/$ - see front matter r 2006 Published by Elsevier Ltd. doi:10.1016/j.atmosenv.2006.03.013
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
1. Introduction The molecular level characterization of the organic particulate matter (OPM) found in the atmosphere has proven difficult. Such OPM can be very complex, and at least part of that complexity is a result of the innumerable pathways by which organic compounds can react in the atmosphere. Volatile organic compounds (VOCs) can be oxidized to form multi-functional compounds, and oxidation products of sufficiently low vapor pressure can condense to form OPM. Additionally, oxidation products may undergo ‘‘accretion reactions’’ with one another and/or other atmospheric constituents and thereby add molecular weight (MW), further decrease volatility, and thus increase their tendency to condense (Barsanti and Pankow, 2004, 2005). Significant studies in this context include Haagen-Smit (1952), Tobias and Ziemann (2000), and Jang and Kamens (2001b). It has been suggested that a significant fraction of atmospheric OPM samples can be comprised of high-MW/lowvolatility compounds, loosely referred to as ‘‘oligomers’’ (Havers et al., 1998; Samburova et al., 2005). The presence of such compounds in the atmosphere would be consistent with accretion reactions of a variety of types (Gao et al., 2004a, b; Iinuma et al., 2004; Jang et al., 2002; Kalberer et al., 2004; Limbeck et al., 2003; Tobias and Ziemann, 2000; Tolocka et al., 2004). Chamber studies have focused on accretion reactions as a general mechanism for OPM formation, and on the possibility of a role for acid catalysis in increasing OPM formation by certain such reactions (e.g., Jang and Kamens, 2001b). Some studies have begun to consider the detection and quantification of oligomers per se, and the effects of particle acidity and parent-compound structure. These studies have shown that: (1) atmospherically relevant constituents (though not necessarily at atmospherically relevant levels) can react to form oligomers (Gao et al., 2004a, b; Iinuma et al., 2004; Kalberer et al., 2004; Tolocka et al., 2004); (2) a significant fraction of the total OPM formed in chamber experiments can be comprised of oligomers, in some cases 450% (Gao et al., 2004a; Kalberer et al., 2004); (3) inorganic seed particles are not required for oligomer formation (Kalberer et al., 2004); when inorganic seed particles are present, the extent of oligomer formation may be affected by particle acidity (Gao et al., 2004b; Iinuma et al., 2004; Jang et al., 2002; Tolocka et al.,
6677
2004); and (4) the nature of precursor aerosol seed may affect the type and extent of oligomer formation (Gao et al., 2004a, b; Kalberer et al., 2004). Despite such advances, many uncertainties exist regarding accretion reactions as they may occur with atmospherically relevant compounds. For example, in a chamber study involving the oxidation of a-pinene, a highly relevant biogenic OPM precursor, Tolocka et al. (2004) suggested that one of the several oligomers they detected could be explained by nine different possible combinations of monomers. Barsanti and Pankow (2004, 2005) have developed and applied a general theoretical approach to evaluate the thermodynamic favorabilities of accretion reactions, including the extents to which they may be relevant for OPM formation in the atmosphere. In the consideration of C4–C10 aldehydes, ketones, dialdehydes, diketones, and methylglyoxal, it was concluded that: (1) hydration/oligomerization, hemiacetal/acetal formation, and aldol condensation are not favored for C4–C10 monoand diketones, or for C5 and lower mono- and dialdehydes; (2) aldol condensation of C6 and higher mono- and dialdehydes may contribute to atmospheric OPM formation under some circumstances when kinetically favored; and (3) diol and diol-oligomer formation from glyoxal as well as aldol condensation of methylglyoxal are thermodynamically favored and may contribute significantly to OPM in the atmosphere when kinetically favored. This work applies the method of Barsanti and Pankow (2004) in the consideration of OPM formation by reaction of mono- and dicarboxylic acids to form esters and amides. The study compounds are acetic, malic, maleic, and pinic acid (Fig. 1). Carboxylic acids are of interest because they are produced by oxidation of both anthropogenic and biogenic VOCs (Grosjean et al., 1978; Jang and Kamens, 2001a; Yu et al., 1999); ester and amide formation are among the most important reactions involving carboxylic acids (Loudon, 1995). 2. Overview of theoretical approach 2.1. Thermodynamic framework and mathematical solution process The proposed mechanism by which oxidation products and other atmospheric constituents (e.g.,
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6678
Fig. 1. Parent carboxylic and dicarboxylic acids.
A and B) may contribute to OPM formation is: Ag þ Bg 2Cg Cg 2Cliq
accretion,
(1)
gas=particle ðG=PÞ equilibration of accretion product
ð2Þ
in which C is a relatively high-MW/low-volatility compound. Accretion of A with B may also occur according to the following thermodynamically equivalent schemes: (1) Ag+Bliq2Cliq, then Cliq2Cg; and (2) Aliq+Bliq2Cliq, then Cliq2Cg. The overall conclusions (e.g., predicted extents of OPM formation) will be independent of the actual scheme by which the accretion reaction proceeds, and independent of whether a catalyst is present in the PM phase (Barsanti and Pankow, 2004, 2005). A thermodynamic framework used to predict OPM formation by accretion reactions has been described by Barsanti and Pankow (2004, 2005). Briefly, equilibrium for an accretion reaction (e.g., Eq. (1)) is considered using its equilibrium constant K as Pset by its Gibbs free energy of reaction DG1 ð¼ i ni DGf ;i Þ according to DG1 ¼ RT ln K, where: ui is the stoichiometric coefficient for compound i in the accretion reaction; DGf;i (kJ mol1) is the standard free energy of formation for i; R (kJ mol1 K1) is the ideal gas constant; and T (K) is temperature. For each parent compound and product, the equilibrium constant Kp for gas/ particle (G/P) partitioning (e.g., Eq. (2)) is governed largely by its pure-liquid vapor pressure pL (atm) (Pankow, 1994b). The tendency of any given accretion reaction to form significant OPM depends on its K, the value of Kp for the product, and the mass-concentration driving force for accretion product formation/condensation as provided by the initial amounts of reactants and the atmospheric level of non accretion-related OPM into which the product can condense, referred to here as OPMna (Barsanti and Pankow, 2004). Mass balance considerations lead to the following expression for initial concentration (A0, mg m3) of
the parent compound A (Barsanti and Pankow, 2005): N X nA MWA C i;g A0 ¼ Ag þ Aliq þ nC;i MWC;i i¼1 N X nA MWA þ Ci;liq ð3Þ nC;i MWC;i i¼1 In Eq. (1), Ag and Aliq denote the species A in the gas and liquid phases, respectively. In Eq. (3), we allow that Ag and Aliq can also be used to denote the concentrations (mg m3) of A in the gas and liquid phases. Similarly, in Eq. (3) the parameters Ci,g and Ci,liq represent the gas- and liquid-phase concentrations (mg m3) of the ith C product. N is the number of possible products from A. Each u is an overall stoichiometric coefficient (does not carry sign) for production of Ci from A, and each MW (g mol1) is a molecular weight. The fundamental equations describing formation of any product Ci and the G/P partitioning of A and Ci at equilibrium (Pankow, 1994a), can be used to express Aliq, Ci,g, and Ci,liq in terms of Ag (see Barsanti and Pankow, 2005). Eq. (3) is then easily solved for Ag given the necessary input parameters. Analogous expressions of Eq. (3) exist for reaction schemes different from Eq. (1). In this work, mass balance equations like Eq. (3) were solved using the standard Microsoft Excel Solver, developed by Frontline Systems, Inc., and executed through a subroutine written in Visual Basic. 2.2. Input parameters and assumed initial conditions Needed pL and DG f values were obtained from the literature when available. When not available, pL values were estimated using SPARC (Hilal et al., 1994) and DG f values were estimated using the Benson (1976) method as implemented by the CHETAH (2002) algorithm. For molecular groups not found in the CHETAH database (Table 1), or missing enthalpy of formation ðDH f Þ or entropy
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6679
Table 1 Molecular groups with DH f and S1 values that are not available in the CHETAH (2002) database and so required interpolation in this work DH f (kJ mol1)
Group: Benson notation
Group: physical description
CH–(QC,O,CO)
C with: one single bond to a hydrogen; one single bond to a doubly-bonded carbon; one single bond to an oxygen; and one single bond to a carbonyl C with: one single bond to a hydrogen; one single bond to a nitrogen; one single bond to a carbonyl; and one single bond to a carbon O with: one single bond to a carbon; and one single bond to a carbonyl C with: one double bond to an oxygen; one single bond to a doubly-bonded carbon; one single bond to a nitrogen N with: three single bonds to carbonyls
CH–(N,CO,C) O–(C, CO) CO–(QC,N) N–(3CO)
S (J mol1 K1)
27.5
44.2
20.9
48.7
179.8 130.5
12.9 66.0
127.7
99.6
esterification w/2-methyl-3-buten-1,2-ol O
OH
OH
OH
n
O O
O
O
+ m
HO
+ (n + m -1) H2O
O HO
OH
CH3
CH2
H 3C
O
C2H
amide formation w/diethylamine O
OH
OH O
+ m
HO
O
O H C 3
NH
NC
CH 3
CH3
+ m H 2O
O
OH
CH3
amide formation w/ammonia O
O
OH O
n
+ m NH3
HO OH
OH O
+ (n + m -1) H2O
O N
Fig. 2. Accretion reactions of interest for acetic, malic, maleic, and pinic acids, shown here for malic acid.
(S1) data, their DH f and S1 values were estimated using linear interpolation as described by Frurip et al. (2002). Values of activity coefficients in the OPM phase (z) were set equal to 1 (see Barsanti and Pankow, 2004, 2005). OPM formation by the Fig. 2 accretion reactions was computed for each of the parent acids by reaction with 2-methyl–3-buten–1,2-ol (MBO), diethylamine (DEA), and ammonia (NH3). Individual accretion reactions and products are listed in Table 2. When considering amide formation, simultaneous salt formation between the organic acid and the amine (or ammonia) was not considered, though we note that such formation
could decrease the amount of OPM by amide formation. Also, the effects of the presence of an inorganic acid on amide formation were not considered explicitly, though we note that such an acid would protonate both amines and ammonia, and thus again reduce the amount of OPM by amide formation. In case 1, A0 ranged from 103 to 103 mg m3 with RH ¼ 20%; in case 2, A0 ¼ 1 mg m3 with RH from 5–95%. In both cases, it was assumed that MBO0 and DEA0 ¼ 1 mg m3, NH3 was held constant at 0.07 mg m3 (based on a lower-limit ambient value from Seinfeld and Pandis, 1998), OPMna ¼ 10 mg m3, P ¼ 1 atm, and T ¼ 298 K.
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6680
Table 2 Summary of accretion reactions and products of interest for acetic, malic, maleic, and pinic acids; nomenclature illustrated for malic acid Accretion product
n Parent compound
m Alcohol or amine
Type of linkage
n malic acid+m 2-methyl-3-butene-1,2-ol 2 accretion product+(n+m1) H2O m.1 1 1 ester m.2 1 2 ester m.3 2 1 ester m.4 2 2 ester m.5 2 3 ester m.6 3 2 ester m.7 3 3 ester n malic acid+m diethylamine 2 accretion product+(n+m1) H2O m.8 1 1 amide m.9 1 2 amide n malic acid+m ammonia 2 accretion product+(n+m1) H2O m.10 1 1 amide m.11 1 2 amide m.12 2 1 amide m.13 2 2 amide m.14 2 3 amide m.15 3 1 amide m.16 3 2 amide m.17 3 3 amide m.18 3 4 amide m.19 4 4 amide Corresponding accretion products were considered for acetic acid (a.1, a.3, a.8, a.10, a.12, and a.15), maleic acid (me.1–19), and pinic acid (p.1–19). For maleic acid, formation of the hydrate (me.20) was also considered.
3. Results and discussion 3.1. Predicted additional OPM formation in Case 1 (A0 ¼ 103– 103 mg m3 MBO0 and , DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, and RH ¼ 20%) Fig. 3 illustrates predicted levels of total additional OPM as a function of A0 for acetic, malic, maleic, and pinic acids by the Fig. 2 accretion reactions. Barsanti and Pankow (2005) proposed a ‘‘criterion of importance’’ that a compound would be of interest in an accretion reaction context if it could raise the OPM level by 1% under the assumed conditions. For the conditions of Case 1, when A0 ¼ 1 mg m3, maleic, malic, and pinic acids meet that criterion, with the predicted levels of additional OPM as follows: maleic, 1.2 mg m3; malic, 1.2 mg m3; and pinic, 0.27 mg m3. For each of
Fig. 3. Predicted total additional OPM by accretion reactions of acetic, malic, maleic, and pinic acids as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
these acids, an ester is the dominant accretion product in the OPM phase (me.3, m.6, and p.3, respectively); however, amides (predominately me.15, m.15, and p.15, respectively) also contribute to the additional OPM (see Figs. 4–6). For maleic and malic acids, the linear relationship between predicted additional OPM and A0 in Fig. 3 is a consequence of the essentially complete conversion of the parent compound to condensable products. Extending the lower limit of A0 would eventually result in a lack of complete conversion. While esterification and amide formation are thermodynamically favorable for acetic acid, the pL values of the accretion products are not low enough to allow significant condensation of the products. Thus, even when A0 ¼ 103 mg m3 the , predicted level of additional OPM by the reactions of acetic acid is not significant (E8 105 mg m3). For malic, maleic, and pinic acids, as A0 increases from 103 to 103 mg m3, increases in additional OPM are predicted (Fig. 3) as are changes in OPM composition. For maleic acid, the dominant accretion product shifts from the ester me.3 to the amide me.15 (Fig. 4). For malic acid, this shift is from the ester m.3 to the ester m.6 and then to the amide m.15 (Fig. 5). For pinic acid, the shift is from the amide p.8 to the ester p.3 and then to the amide p.15 (Fig. 6). Recalling that NH3 was held constant at a lower-limit ambient value (0.07 mg m3), when we
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6681
Fig. 4. Fractional contribution of each of the dominant accretion products to total additional OPM as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
Fig. 5. Fractional contribution of each of the dominant accretion products to total additional OPM as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
ARTICLE IN PRESS 6682
K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
Fig. 6. Fractional contribution of each of the dominant accretion products to total total additional OPM as a function of A0, when A0 ¼ 103–103 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, OPMna ¼ 10 mg m3, RH ¼ 20%).
allow NH3 to be held constant at higher values, up to 10 mg m3 (which is within the range of ambient values discussed by Finlayson-Pitts and Pitts (2000)), the conclusions regarding significance of additional OPM formation by reaction of each acid do not change. For acetic and pinic acids, the levels of additional OPM increase with A0 due to the increased driving force for condensable product formation (Fig. 7). For maleic and malic acids, however, predicted additional OPM decreases slightly with increasing A0 due to the increased favorability of the less-condensable accretion products me.15 and m.15, respectively, over the esters me.3 and m.6, respectively. These results suggest that levels of OPM may not always increase monotonically with increases in total initial concentrations of reactants. 3.2. Predicted OPM formation in Case 2 (A0 ¼ 1 mg m3, MBO0 and DEA0 ¼ 1 mg m3, NH3 ¼ 0.07 mg m3, and RH ¼ 5% – 95%) The effect of varying RH on the extent of ester and amide formation in the gas phase can be understood by considering that water is a product in both cases. A lower ambient RH value thus
Fig. 7. Predicted levels of total additional OPM as a function of NH3 concentration (held constant), when A0 ¼ 1 mg m3 (with MBO0 and DEA0 ¼ 1 mg m3, OPMna ¼ 10 mg m3, and RH ¼ 20%).
promotes the formation of both of these types of accretion products, though the relative magnitude of the observed effect will depend on the K value for
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
the reaction. For acetic and pinic acids, the K values for ester and amide formation are not so large that the extent of accretion product formation is not affected by changes in RH. For maleic and malic acid, the K values for some of the accretion reactions are sufficiently large that the reactions go nearly to completion even at relatively high RH. When the accretion reactions are sufficiently favorable, whether significant OPM forms under the conditions considered depends upon the pL values of the accretion products. For acetic acid, even at the lowest RH considered, the level of additional OPM formed is not significant because the pL values for its ester and amide products are relatively large. For pinic acid, the additional OPM formed falls below the criterion of significance when RH rises above 50% (when RH ¼ 50%, additional OPME6 102 mg m3; when RH ¼ 95%, additional OPME2 102 mg m3). If we take into account that predicted DG f values may be uncertain by as much as 3% (Barsanti and Pankow, 2005), then adjusting the K values to favor accretion product formation leads to predicted levels of additional OPM for pinic acid at RH450% that meet the criterion. For maleic and malic acids, conversion to condensable products is nearly complete, and so varying RH has little effect on the predicted levels of additional OPM (large enough K values for some accretion reactions and small enough pL values for some of the corresponding products). The levels and composition of predicted additional OPM in Figs. 3–7 are for T ¼ 298 K. In the ambient atmosphere, temperatures lower than 298 K are common. For all of the accretion reactions considered, DH values are negative so that lowering temperature favors product formation. Assuming DH values are independent of temperature over a small range, for a 10 K decrease in temperature to 288 K, the calculated equilibrium constants increase by 1–3 orders of magnitude depending on the overall number of ester and amide bonds formed (100.9 for products with m ¼ 1, 101.6 for m ¼ 2, 102.6 for m ¼ 3, and 103.4 for m ¼ 4). In addition to increasing favorability for accretion product formation, decreasing temperature leads to lower vapor pressures of the accretion products, which favors partitioning to the OPM. 3.3. Implications for oligomer formation reported in chamber studies The chamber studies of Gao et al. (2004a, b), Iinuma et al. (2004), and Tolocka et al. (2004)
6683
involved oxidation of a-pinene, which is known to produce a number of cyclic compounds including pinic acid, hydroxy pinonaldehyde, and pinonaldehyde (Yu et al., 1999). It has been suggested that cyclic compounds can undergo ring opening in the presence of a weak acid (Chakraborty et al., 2003; Shi and Xu, 2002; Tolocka et al., 2004). For hydroxy pinonaldehyde and pinonaldehyde there are potentially eight products of ring opening (Fig. 8); addition of water across the unsaturated bond of each of those products can form a total of four mono-alcohols and two diols, with which carboxylic acids may react to form esters. To determine whether esterification might have been important in the a -pinene/O3 chamber studies of Gao et al. (2004a, b), Iinuma et al. (2004), and Tolocka et al. (2004), OPM formation was evaluated for esterification of pinic acid with the hydrated products from ring opening of hydroxy pinonaldehyde and pinonaldehyde (see Fig. 8 and Table 3). Initial concentrations of pinic acid were calculated by combining reacted a-pinene (DHC) values ranging from 6 to 19,000 mg m3 with an average molar yield value from chamber experiments of Yu et al. (1999). For the hydrated products of ring opening, hydroxy pinonaldehyde and pinonaldehyde were first assumed to form in amounts corresponding to the amount of pinic acid, then undergo ring opening by schemes I and II in equal proportions (100% complete), then hydration (to equilibrium), and finally, reaction with the pinic acid (to equilibrium). The average RH for the chamber experiments, 50%, was assumed. Other assumptions were: OPMna ¼ 0 mg m3 (i.e., no initial organic aerosol), P ¼ 1 atm, and T ¼ 298 K. The results are summarized in Table 4. The predicted levels of OPM in Table 4 are nonnegligible for all DHC values considered. As DHC increases, the dominant accretion product does not change, but other accretion products contribute with increasing proportions to the OPM formed. The molecular weights of the dominant accretion products (256–695 g mol1) are within the MW range of the ‘‘oligomers’’ reported in the chamber experiments of Gao et al. (2004a, b) and Iinuma et al. (2004); the functional groups on the dominant accretion products are also consistent with those suggested for the chamber experiments of interest. Based on the estimation method utilized here to obtain DG f values, the addition of water across an unsaturated bond is thermodynamically favorable. If kinetically favorable, other cyclic oxidation
ARTICLE IN PRESS 6684
K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
Fig. 8. Potential schemes (I and II) and products of ring opening from hydroxy pinonaldehyde and pinonaldehyde. A product of ring opening (hpa/pa) is shown reacting with water to form an alcohol (hpa.20/pa.20) and then the alcohol reacting with pinic acid to form an ester (hpa.21/pa.21). Table 3 Summary of accretion reactions and products for pinic acid reacting with the products and hydrated products of ring opening from hydroxy pinonaldehyde and pinonaldehyde; nomenclature illustrated for the product hpb m H2O or alcohol
Type of linkage
n pinic acid+m H2O2accretion product (alcohol) hpb.20 1
1
alcohol
n pinic acid+m alcohol/diol2accretion product+(n1) H2O hpb.21 1 hpb.22 1 hpb.23 2 hpb.24 2 hpb.25 2 hpb.26 3 hpb.27 3
1 2 1 2 3 2 3
ester ester ester ester ester ester ester
Accretion product
n Parent compound
Corresponding accretion reactions and products were considered for pinic acid with the additional products and hydrated products of ring opening from hydroxy pinonaldehyde (accretion product: hpa.20-hpa.22, hpc.20-hpc.22, and hpd.20-hpd.27) and from pinonaldehyde (accretion product: pa.20-pa.22, pb.20-pb.22, pc.20-pc.22, and pd.20-pd.22).
products and unsaturated aliphatic oxidation products may form alcohols. Given the numerous generation possibilities for reactants, esterification
reactions may be important in chamber experiments involving oxidation of VOCs other than a-pinene, including those of Gao et al. (2004a, b).
ARTICLE IN PRESS K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
6685
Table 4 Total levels and composition of predicted OPM formed by esterification of pinic acid with the hydrated products of ring opening from hydroxy pinonaldehyde and pinonaldehyde DHC (mg m3)
Initial concentration of pinic acid (mg m3)
Total OPM (mg m3)
6 56 280 560
0.41 3.8 19 38
0.15 1.7 8.9 18
840
57
27
19000
1300
650
Dominant accretion products (40.1% of total OPM) in order of decreasing contribution ester ester ester ester ester ester ester ester ester
(hpd.23), ester (hpb.23), diester (hpb.26) (hpd.23), ester (hpb.23), diester (hpb.26) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21), (pd.21) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21), (pd.21) (hpd.23), ester (hpb.23), diester (hpb.26), ester (hpc./pc.21), ester (pb.21), (pd.21), ester (hpa./pa.21), hydrate (pb.20), hydrate (hpc./pc.20)
P ¼ 1 atm, T ¼ 298 K, RH ¼ 50%.
4. Conclusions The approach used herein, including the method used to predict DG f values and therefore K values for accretion reactions, indicates that for all acids considered, ester and amide formation are thermodynamically favored under the assumed conditions. In the case of malic, maleic, and pinic acids, the accretion products have sufficiently low pL values that they are predicted to reside predominately in the OPM phase, contributing to total OPM. Thus, for malic, maleic, and pinic acids, and likely for similar carboxylic/dicarboxylic acids, significant OPM formation may occur via esterification and amide formation in the atmosphere when kinetically favorable. Acknowledgment This work was supported by the Electric Power Research Institute. References Barsanti, K.B., Pankow, J.F., 2004. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 1: Aldehydes and ketones. Atmospheric Environment 38, 4371–4382. Barsanti, K.B., Pankow, J.F., 2005. Thermodynamics of the formation of atmospheric organic particulate matter by accretion reactions—Part 2: Dialdehydes, methylglyoxal, and diketones. Atmospheric Environment 39, 6597–6607. Benson, S.W., 1976. Thermochemical Kinetics. Wiley, New York, pp. 336. Chakraborty, D., Rodriguez, A., Chen, E.Y.X., 2003. Catalytic ring-opening polymerization of propylene oxide by organo-
borane and aluminum Lewis acids. Macromolecules 26, 5470–5481. CHETAH Version 7.3: The ASTM Computer Program for Chemical Thermodynamic and Energy Release Evaluation (CHETAH), 2002. ASTM International, Pennsylvania. Finlayson-Pitts, B.J., Pitts, J.N., 2000. Chemistry of the Upper and Lower Atmosphere. Academic Press, California, p. 969. Frurip, D., Downey, J.R., LaBarge, M., 2002. Estimation of Missing Group Data of Organic Species for Use in the CHETAH Program. ASTM International, Pennsylvania. Gao, S., Keywood, M., Ng, N.L., Surratt, J., Varutbangkul, R.B., Bahreini, R., Flagan, R.C., Seinfeld, J.H., 2004a. Lowmolecular-weight and oligomeric components in secondary organic aerosol from the ozonolysis of cycloalkenes and apinene. Journal of Physical Chemistry A 108, 10147–10164. Gao, S., Ng, N.L., Keywood, M., Varutbangkul, R.B., Bahreini, R., Nenes, A., He, J., Yoo, K.Y., Beauchamp, J.L., Hodyss, R.P., Flagan, R.C., Seinfeld, J.H., 2004b. Particle phase acidity and oligomer formation in secondary organic aerosol. Environmental Science and Technology 38, 6582–6589. Grosjean, D., Van Cauwenberghe, K., Schmid, J.P., Kelley, P.E., Pitts, J.N., 1978. Identification of C3–C10 aliphatic dicarboxylic acids in airborne particulate matter. Environmental Science and Technology 12, 313–317. Haagen-Smit, A.J., 1952. Chemistry and physiology of Los Angeles smog. Industrial Engineering and Chemistry 44, 1342–1346. Havers, N., Burpa, P., Lambert, J., Klockow, D., 1998. Spectroscopic characterization of humic-like substances in airborne particulate matter. Journal of Atmospheric Chemistry 29, 45–54. Hilal, S.H., Carreira, L.A., Karickhoff, S.W., 1994. Estimation of chemical reactivity parameters and physical properties of organic molecules using SPARC. In: Murray, J.S. (Ed.), Quantitative Treatments of Solute/Solvent Interactions. Elsevier, Amsterdam (http://ibmlc2.chem.uga.edu/sparc/index.cfm). Iinuma, Y., Bo¨ge, O., Gnauk, T., Hermann, H., 2004. Aerosolchamber study of the a-pinene/O3 reaction: influence of particle acidity on aerosol yields and products. Atmospheric Environment 38, 761–773.
ARTICLE IN PRESS 6686
K.C. Barsanti, J.F. Pankow / Atmospheric Environment 40 (2006) 6676–6686
Jang, M., Kamens, R.M., 2001a. Characterization of secondary aerosol from the photooxidation of toluene in the presence of NOx and 1-propene. Environmental Science and Technology 35, 3626–3639. Jang, M., Kamens, R.M., 2001b. Atmospheric secondary aerosol formation by heterogeneous reactions of aldehydes in the presence of a sulfuric acid aerosol catalyst. Environmental Science and Technology 35, 4758–4766. Jang, M., Czoschke, N.M., Lee, S., Kamens, R.M., 2002. Heterogeneous atmospheric organic aerosol production by acid-catalyzed particle-phase reactions. Science 298, 814–817. Kalberer, M., Paulsen, D., Sax, M., Steinbacher, M., Dommen, J., Prevot, A.S.H., Fisseha, R., Weingartner, E., Frankevich, V., Zenobi, R., Baltensperger, U., 2004. Identification of polymers as major components of atmospheric organic aerosols. Science 12, 1659–1662. Loudon, M.G., 1995. Organic Chemistry, 3rd ed. Benjamin, California, p. 1390. Limbeck, A., Kulmala, M., Puxbaum, H., 2003. Secondary organic aerosol formation in the atmosphere via heterogeneous reaction of gaseous isoprene on acidic particles. Geophysical Research Letters 30, doi:10.1029/2003/ GL017738. Pankow, J.F., 1994a. An absorption model of gas/particle partitioning of organic compounds in the atmosphere. Atmospheric Environment 28, 185–188.
Pankow, J.F., 1994b. An absorption model of the gas/particle partitioning involved in the formation of secondary organic aerosol. Atmospheric Environment 28, 189–193. Samburova, V., Kalberer, M., Zenobi, R., 2005. Characterization of high molecular weight compounds in urban atmospheric particles. Atmospheric Chemistry and Physics Discussions 5, 437–454. Seinfeld, J.H., Pandis, S.N., 1998. From Air Pollution to Climate Change. Wiley, New York, p. 1325. Shi, M., Xu., B., 2002. Lewis acid-catalyzed ring-opening reactions of methylenecyclopropanes with alcoholic or acidic nucleophiles. Organic Letters 13, 2145–2148. Tobias, H.J., Ziemann, P.J., 2000. Thermal desorption mass spectrometric analysis of organic aerosol formed from reactions of 1-tetradecene and O3 in the presence of alcohols and carboxylic acids. Environmental Science and Technology 34, 2105–2115. Tolocka, M.P., Jang, M., Ginter, J.M., Cox, F.J., Kamens, R.M., Johnston, M.V., 2004. Formation of oligomers in secondary organic aerosol. Environmental Science and Technology 38, 1428–1434. Yu, J., Cocker, D.R., Griffin, R.J., Flagan, R.C., Seinfeld, J.H., 1999. Gas-phase oxidation products of monoterpenes: gaseous and particulate products. Journal of Atmospheric Chemistry 34, 207–258.