Atmospheric Secondary Aerosol Formation by Heterogeneous

Particle growth by the heterogeneous reaction of aldehydes was evaluated in 0.5 m3 Teflon film bags under darkness in the presence of background seed ...
13 downloads 0 Views 157KB Size
Download PDF
Environ. Sci. Technol. 2001, 35, 4758-4766

Atmospheric Secondary Aerosol Formation by Heterogeneous Reactions of Aldehydes in the Presence of a Sulfuric Acid Aerosol Catalyst MYOSEON JANG* AND RICHARD M. KAMENS Department of Environmental Sciences and Engineering, CB# 7400, Rosenau Hall, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, 27599

Particle growth by the heterogeneous reaction of aldehydes was evaluated in 0.5 m3 Teflon film bags under darkness in the presence of background seed aerosols. The aldehydes used were as follows: glyoxal, butanal, hexanal, octanal, and decanal. To study acid catalyst effects on aldehyde heterogeneous reactions, one of the Teflon bags was initially filled with seed aerosols composed of ammonium sulfate-aerosol acidified with sulfuric acid. These results were compared to particle growth reactions that contained only ammonium sulfate as a background seed aerosol. The gas-phase aldehydes were then added to the Teflon bags. In selected experiments, 1-decanol was also added to the Teflon bags with aldehydes to clarify particle growth via a heterogeneous hemiacetal/acetal formation in the presence/absence of an acid catalyst. The particle size distribution and growth were measured using a scanning mobility particle sizer (TSI-SMPS), and the results were applied to predicting aerosol growth and size distribution changes by condensation and heterogeneous reactions. Aerosols created from the heterogeneous reactions of aldehydes were collected directly on an ungreased zinc selenide (ZnSe) FTIR disk (25 mm in diameter) by impaction. The ZnSe disks were directly analyzed for product functional groups in the aerosol phase using a Fourier transform infrared (FTIR) spectrometer with a deuterated triglycine sulfate (DTGS) detector. Aerosol growth by heterogeneous aldehyde reactions proceeds via a hydration, polymerization process, and hemiacetal/acetal formation from the reaction of aldehydes with alcohols. These aldehyde heterogeneous reactions were accelerated in the presence of an acid catalyst, H2SO4, and led to higher aerosol yields than when H2SO4 was not present in the seed aerosol. The FTIR spectra obtained from the growing aerosol, also illustrated aldehyde group transformation in the particle phase as a function of the heterogeneous reaction. It was concluded that aldehydes, which can be produced by atmospheric photochemical reactions, can significantly contribute on secondary aerosol formation through heterogeneous reactions in the presence of an acid catalyst.

Introduction The secondary organic aerosol (SOA) formation by gas/ particle (G/P) partitioning of low volatility products from 4758

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001

the photooxidation of reactive organic species has received renewed interest in recent years. This is because of its possible impacts on the earth’s radiative balance associated with climate change (1, 2), visibility degradation (3, 4), and health effects (5). Possible candidates for atmospheric SOA formation involve primarily terpene emissions from terrestrial vegetation (6) and aromatics from anthropogenic sources (7). Atmospheric oxidation reactions of these organic compounds can produce multifunctional oxygenated or nitrated semivolatile organic compounds and result in SOA formation via either a self-nucleation process or G/P partitioning on preexisting particulate matter (8-13). Product compounds that contribute to SOA formation contain a variety of different chemical functional groups: carboxylic acids, aldehydes, alcohols, and nitrates (14-20). It is expected that different chemical functionalities of each of these products will have different chemical affinities and reaction properties and may result in different G/P distributions and secondary aerosol yields. The accommodation properties of secondary organic products and SOA formation yields have been primarily advanced by condensation or partitioning theory (8-12), and these are largely based on vapor pressures and chemical activity coefficients of products in the aerosol phase. However, there may be other possibilities that lead to secondary aerosol formation. For example, some chemical species can be further transformed via heterogeneous reactions between the gas phase and atmospheric particulate matter. The result is an increase in secondary aerosol mass because the new aerosol reaction products either have relatively low vapor pressures or are different from their original parent compounds; this leads to additional partitioning from the gas to the particle phase of the parent compounds. One of the candidates which has the potential for heterogeneous reactions on the particle phase is the chemistry associated with aldehydes in the presence of an acid catalyst such as sulfuric acid. If one considers the vapor pressures of tropospheric aldehydes formed from photochemical reactions of organics, the accommodation probability to aerosol surfaces, for aldehydes with less than eight carbons, will be insignificant. However, deviations from fundamental partitioning behavior have been observed for aldehyde products from the photochemical reaction of toluene (21) and R-pinene (12). The experimental partitioning coefficients of these aldehydes are much higher than predictive partitioning coefficients, calculated from the vapor pressures and activity coefficients (17, 21). This phenomenon can be explained from the following: (1) aldehyde functional groups can further react in the aerosol phase through heterogeneous reactions via hydration, polymerization, and hemiacetal/acetal formation with alcohols; (2) these aldehyde reactions can be radically accelerated by the presence of an atmospheric acid catalyst such as sulfuric acid; (3) without an “end capping reaction”, reverse reactions to the original aldehyde parent structures can occur during sample work up/solvent extraction procedures; (4) this gives the appearance of higher aldehyde concentrations in the particle phase and provides a possible explanation for the highly positive deviation of experimental partitioning coefficients of aldehydes from the predictive ones by the absorptive partitioning theory (12, 21). Understanding the chemical reaction behavior of aldehydes is particularly important, because many secondary * Corresponding author phone: (919)966-3861; fax: (919)966-7911; e-mail: [email protected]. 10.1021/es010790s CCC: $20.00

 2001 American Chemical Society Published on Web 11/10/2001

TABLE 1. Vapor Pressures (mmHg) for Organic Compounds at 295 K compounds

A

B

C

D

E

log pLo at 295 K (regression)a

log pLo at 295 K (group contribution)b

glyoxal butanal hexanal octanal decanal 1-decanol

66.8 -10.9 64.4 82.5 103.0

-3.68E+03 -2.39E+03 -5.02E+03 -6.39E+03 -8.15E+03

-2.26E+01 1.06E+01 -1.94E+01 -2.53E+01 -3.16E+01

1.17E-02 -2.57E-02 6.83E-10 -1.94E-09 -7.23E-10

2.96E-13 1.45E-05 5.54E-06 6.31E-06 6.03E-06

1.98 0.97 -0.03 -1.11 -2.22

1.56 1.83 0.94 0.06 -0.81 -2.05

a The Antoine type equation was used for correlation of vapor pressure as a function of temperature: log po (mmHg) )A + B/T + ClogT + DT L + ET2 where A, B, C, D, and E are regression coefficients for a given compound and T is a temperature (K) (22). b The vapor pressure (pLo) was calculated by (23, 24),

ln P0)

∆Svap(Tb)

R

[ (

(1.8) 1 -

)

( )]

Tb Tb + (0.8) ln T T

(atm)

(A)

where ∆Svap is the entropy of vaporization, R is a gas constant (8.314 J/K/mol), Tb is a boiling point (K), and T is an ambient temperature (K) for a given organic compound. ∆Svap of an organic compound was calculated using modified Trouton’s method developed by Yalkowsky and coworkers, considering parameters related to molecular geometry and association (25, 26). Boiling points (Tb) of organic compounds were calculated by a group contribution method originally developed by Joback and Reid (27) with a modified equation and modified group contribution parameters (28).

organic products from the photochemical reactions of aromatics and terpenes include aldehyde functional groups. This functionality can potentially lead to acid-catalyzed heterogeneous reactions in the aerosol phase resulting in increase of SOA formation yield.

Experimental Section The heterogeneous reactions of aldehydes were carried out in two 0.5 m3 (maximum volume) 2 mil Teflon film bags. All experiments were conducted in the dark to exclude photochemical effects. Organic compound injections to the gasphase atmosphere of the chambers were performed by volatilizing liquid organics in a gently heated manifold with a flowing stream of dry air. Five different aldehydes, which include butanal (98%), hexanal (98%), octanal (99%), decanal (95%), and glyoxal (40 wt % solution in water), were used. All of these aldehydes, with the exception of glyoxal, were easily vaporized into the bags. Glyoxal, however, is readily polymerized at temperatures >150 °C. On average, only 2 wt % of total glyoxal injection, calculated from density (1.265 g per mL) and wt % in a water solution (40%), was injected to the bag. In addition to the above aldehydes, 1-decanol (99%) was used to address the hemiacetal/acetal formation reaction with aldehydes. All of the compounds were purchased from Aldrich (Milwaukee, WI) and are listed in Table 1 along with their calculated vapor pressures (22-28). For a given experiment, two Teflon bags, designated as acid- and nonacid- chambers, were used simultaneously to assess the catalytic effects of sulfuric acid on the heterogeneous reaction of aldehydes. Prior to the addition of hydrocarbons, ammonium sulfate [(NH4)2SO4] seed aerosols were added to the Teflon bags. The aerosol was generated using a commercially available large volume nebulizer (Westmed, Inc., Tucson, AZ) to aspirate the aqueous salt solution into the bag. The aqueous salt concentration of (NH4)2SO4 injected to the nonacidic bag was 0.0067 M. The aqueous solution used to inject seed nuclei into the acid bag was 0.0035 M of (NH4)2SO4 and 0.005 M of sulfuric acid. Tables 2 and 3 show the experimental conditions and aerosol yields from the heterogeneous reaction of aldehydes. The bag temperature during the course of the reactions was 295-296 K. The relative humidity was 44-46% for clean air in Teflon bags and 48-56% after injecting aqueous seed solution into the bags. The particle size distribution of aerosols was monitored with a Scanning Mobility Particle Sizer (SMPS 3936 TSI, Shoreview, MN) and an associated TSI condensation nuclei counter (3025A, TSI). Particle size data from the SMPS were

taken over the size range of 13.8-697 nm, and scanning time was 180 s. Analysis of product functional groups in the aerosol phase was performed using a Fourier transform infrared spectroscopy (FTIR, Nicollet Magma 560, Madison, WI) with a deuterated triglycine sulfate (DTGS) detector. SOAs were collected directly on an ungreased zinc selenide (ZnSe) FTIR disk (25 mm in diameter) by impaction. The sampling rate was 27.2 L/min, and the sampling volume was about 0.3 m3. The scan number for FTIR was 8, and the resolution was 2 cm-1.

Results and Discussion Heterogeneous Reactions of Aldehydes. Atmospheric gasphase aldehydes can migrate to the aerosol phase through either condensation or equilibrium G/P partitioning (8-12, 21) and can further react by heterogeneous reactions. This process can be accelerated in the presence of an acid catalyst. The aerosol formation percent yields by heterogeneous reactions of aldehydes in the Teflon bags were defined as (SOA total volume corrected for seed inorganic volume)/ (volume of an organic injected to the bag) × 100 and are shown in Tables 2 and 3. It is apparent from these results that the heterogeneous reactions of aldehydes in the presence of sulfuric acid, as an acid catalyst, give remarkably higher aerosol yields than in the absence of sulfuric acid. Tables 2 and 3 also show that the adding of an alcohol to acid-catalyzed aldehyde heterogeneous reaction systems resulted in additional aerosol mass. Possible reaction mechanisms for acid-catalyzed aldehyde reactions include hydration, polymerization, and hemiacetal/ acetal formation (29), as shown in Scheme 1. It is known that the equilibrium between the aldehyde and its hydrate, gemdiol, is quickly established and often favors the hydrate form (30). Aldehydes also further react with hydroxy groups of the gem-diol, resulting in higher molecular weight of dimers, trimers and polymers (31, 32). A variety of alcohols originate either directly from vegetation and anthropogenic sources or from secondary organic products via photochemical reactions of primary organic compounds (21). Aldehydes can react with alcohols in the atmosphere forming hemiacetals/ acetals as shown in Scheme 1. In our experiments, 1-decanol was employed to demonstrate the aerosol mass increment by hemiacetal/acetal formation from aldehyde heterogeneous reactions with an alcohol. Hemiacetals, which are often too unstable to be isolated, exist only in equilibrium (3339). In the presence of strong acids, further reaction of hemiacetals leads to acetals which are relatively stable and can be isolated by neutralization (40). Hemiacetals or acetals VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4759

TABLE 2. Conditions for the Small Bag (0.5 m3) Experiments Used for Heterogeneous Reactions of Aldehydes in the Presence/ Absence of an Acid Catalyst (H2SO4) exp. date mm/dd/yy

aldehyde

alcohol

cat.

gas concn, µL/m3

seed, nm3/cm3

particle, nm3/cm3

yield,a %

seed/particle

11/27/00 11/27/00 11/13/00 11/13/00 11/13/00 11/13/00 11/7/00 11/7/00 11/8/00 11/8/00 11/9/00 11/9/00 11/9/00 11/9/00 11/10/00 11/10/00 11/27/00 11/27/00 11/21/00 11/21/00 11/27/00 11/27/00

butanal butanal glyoxal glyoxal glyoxal glyoxal hexanal hexanal hexanal hexanal hexanal hexanal octanal (C8) octanal octanal octanal octanal octanal C8+glyoxal C8+glyoxal decanal decanal

1-decanol 1-decanol none none 1-decanol 1-decanol none none none none 1-decanol 1-decanol none none 1-decanol 1-decanol 1-decanol 1-decanol 1-decanol 1-decanol 1-decanol 1-decanol

acid non acid non acid non acid non acid non acid non acid non acid non acid non acid non acid non

51.4 51.4 35.5 35.3 40.4 40.4 28.6 28.6 8.57 8.57 17.1 17.1 8.57 8.57 17.1 17.1 17.1 17.1 42.6 42.6 17.1 17.1

1.48E+10 2.53E+9 2.28E+10 7.17E+9 2.04E+10 1.62E+10 5.05E+10 1.30E+10 5.56E+10 3.10E+10 4.72E+10 4.14E+10 6.10E+10 4.09E+10 2.90E+10 1.67E+10 1.80E+10 1.68E+10 8.53E+9 6.60E+10 1.38E+10 1.57E+10

5.03E+10 3.70E+10 3.40E+10 1.78E+10 3.41E+11 9.20E+10 4.04E+11 2.38E+11 6.10E+10 3.41E+10 7.74E+10 3.84E+10 3.90E+11 1.66E+11 6.10E+11 7.57E+10 3.45E+11 2.13E+11 6.00E+11 3.03E+11 2.68E+11 2.14E+11

0.08 0.07 2.33 1.59 2.25 0.55 1.28 0.80 0.21 0.12 0.24 0.04 4.00 1.57 3.43 0.37 1.93 1.17 5.35 2.69 1.50 1.18

0.23 0.05 0.52 0.33 0.05 0.14 0.10 0.04 0.70 0.70 0.47 0.83 0.12 0.19 0.04 0.17 0.04 0.06 0.01 0.02 0.04 0.06

a The aerosol yields were calculated with dividing the organic aerosol volume (aerosol volume - seed aerosol volume × 0.769) by the injected liquid organic volume into the bag air phase. The volume can be used to calculate aerosol yields, assuming that the densities of organic compounds, organic aerosols, and seed aerosols are the same. The correction factor is 0.769 for air volume differences between before and after chemical injection.

TABLE 3. Conditions for the Small Bag (0.5 m3) Experiments Used for Sequential Injections of Aldehydes and 1-Decanol in the Presence/Absence of an Acid Catalysta

date

compd at step 2

gas concn, µL/m3

compd at step 3

11/6/00 11/6/00 11/20/00 11/20/00 11/26/00 11/26/00

1-decanol 1-decanol 1-decanol 1-decanol octanal octanal

8.6 8.6 8.6 8.6 5.7 11.4

octanal octanal octanal octanal 1-decanol 1-decanol

total gas gas concn, concn, µL/m3 µL/m3 7.0 7.0 7.0 7.0 7.0 7.0

13.9 13.9 13.9 13.9 11.6 16.3

cat.

seed nm3/ particle nm3/ particle nm3/ cm3 (step 1) cm3 (step 2) cm3 (step 3)

acid non acid non acid non

5.27E+10 1.10E+10 1.87E+10 7.49E+09 1.83E+10 1.36E+10

8.65E+10 1.20E+10 1.26E+10 6.09E+9 1.87E+11 3.02E+11

3.40E+11 2.20E+11 5.10E+11 1.28E+11 2.89E+11 2.62E+11

step 2 yielda (%) 0.54 0.04 -0.02 0.004 3.03 2.53

step 3 total yieldb yieldc (%) (%) 3.40 2.92 7.01 1.70 1.80 0.14

2.21 1.53 3.58 0.89 2.39 1.56

a Step 1. (NH ) SO or (NH ) SO -H SO seed aerosol injection; step 2. injection of a gas-phase alcohol (aldehyde); and Step 3. injection of a 4 2 4 4 2 4 2 4 gas-phase aldehyde (alcohol). a The aerosol yields for step 2 were calculated with dividing the organic aerosol volume (aerosol volume - seed aerosol volume × 0.769) by the injected organic liquid volume into the bag air phase at step 2. b The aerosol yields for step 3 were calculated with dividing the organic aerosol volume at step 3 only (accumulated aerosol volume - the organic aerosol volume at step 2 × 0.813 - seed aerosol volume at step 1 × 0.625) by the injected liquid organic volume into the bag air phase at step 3. c The total aerosol yields were calculated with dividing the organic aerosol volume at step 3 (accumulated aerosol volume through steps 2 and 3 - seed aerosol volume × 0.625) by the injected total liquid organic volume into the bag air phase at steps 2 and 3. The correction factors are 0.769, 0.813, and 0.625 for air volume differences between chemical injection steps.

can also be easily hydrolyzed or return to their original aldehydes and alcohols during typical sampling workup procedures that involve solvent extractions and derivatization procedures (41, 42). Compared to aldehydes, ketones convert less easily to hemiketal/ketals. However, the rate of hydrolysis reactions enormously increases with the following series: formals < acetals < ketals (43). In this study, sulfuric acid was employed as an acid catalyst for the heterogeneous reaction of aldehydes because it is ubiquitous in atmospheric environments and provides an acidity strong enough to catalyze aldehyde reactions. Hammett and Deyrup (44) introduced an acidity measurement of the proton, which can be applied to moderately concentrated acid media such as the sulfuric acid-water system (45, 46). The treatment begins with the protonation equilibrium of simple a base (B) as described by

B + H+ a BH+

(1)

The equilibrium constant for this reaction is commonly 4760

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001

given by

KBH + )

[B][H +] fBfH + [BH +] fBH +

(2)

where fB, fBH+, and fH+ are the activity coefficients of each species. Hammett proposed that the acidity, h0, of the medium should be defined by h0 t [H +]fBfH +/fBH +. The negative logarithm of the acidity of the system is called as an acidity function, H0 described by

H0 ) - log h0 ) pKBH + - log

[BH +] [B]

(3)

Table 4 shows the acidity scale (pKa) of selected compounds in diluted aqueous solutions along with the pKBH+ values. The acidity functions indicate the ability of the acid medium to donate a proton to a base. Figure 1A illustrates H0 values for various sulfuric acid-water compositions. Liu

SCHEME 1. Acid-Catalyzed Aldehyde Reaction Mechanisms for Hydration, Polymerization, and Hemiacetal/Acetal Formation

TABLE 4. Scales of Acidity for Organic Compounds Interesting in Aldehyde Reactions compounds

p Ka

H2O phenol CH3COOH HNO3 H3O+ CH3CH2OH2+ (CH3)3COH2+ (CH3O+H)(CH3O)CH2 H2SO4 HCl (CH3)2C(dO+H) C6H5C(dO+H)

15.7 10 4.8 -1.4

pKBH+

-1.7 -2.4 -3.8 -4.6 -5.2 -7.0 -7.2 -7.6

comments

protonation of ethanol protonation of acetal protonation of acetone protonation of benzaldehyde

and Levi (47, 48) illustrated the relative humidity (RH) to sulfuric acid weight % of different aqueous aerosol solutions (Figure 1B). The RH values can be calculated from the water partial pressure over aqueous sulfuric acid solution at a given temperature (48). These relationships suggest that sulfuric acid is very hygroscopic and absorbs significant amounts of water under typical atmospheric conditions. The necessary acidity to catalyze heterogeneous aldehyde reactions can be estimated from the acidity function at a given humidity (Figure 1C). The first step of heterogeneous aldehyde reactions, such as formation of aldehyde hydrates or hemiacetal, is closely controlled by the nature of acid catalysts and determined by acid strength, which is expressed by pKa and pKBH+. Simple protonation studies of aliphatic aldehydes were not entirely successful because aliphatic aldehydes are readily hydrated

FIGURE 1. (A) H0 as a function of weight % of sulfuric acid in aqueous solution at 298 K. The reported values cover 298 ( 10 K. (B) RH as a function of weight % of sulfuric acid in aqueous solution at 295 K. (C) RH as a function of the H0 values in aqueous solution at 295 K. The regression line using a polynomial equation is included for each figure. in diluted aqueous acid solutions (43, 49). The protonation of an aldehyde carbonyl in the reaction of water and alcohols is close to a general acid catalysis process (43), which can occur as a result of hydrogen bonding between the reactant and a proton donor. General acid catalysis is complex because of the varied contribution from each potential hydrogen bonding species to the rate. Acetals, which are weak oxygen bases, either hydrolyze to aldehyde hydrates or reverse react to their original aldehydes in the presence of acid catalysts (33, 41). Both the hydrolysis of acetals and the reaction of hemiacetals to acetals by addition of a second mole of an alcohol is a specific acidcatalyzed reaction (43). A specific acid-catalyzed reaction is dependent on the equilibrium for protonation of the reactants and corresponds to H0. For our experimental conditions, the relative humidity was about 45%. At the given amounts of sulfuric acid in seed aerosols (47), the estimated H0 value is approximately -3.0 (Figure 1C). When the pKBH+ values, which range from -2 to -5 for different hydroxyl groups (Table 4), are applied to eq 3, one can predict that some hydroxyl groups would possibly be protonated under our experimental conditions. For example, formaldehyde VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4761

FIGURE 2. SOA yields obtained from sequential injections of aldehydes and 1-decanol in the presence of an acid catalyst (A) and in the absence of an acid catalyst (N): step 1. (NH4)2SO4 or (NH4)2SO4-H2SO4 seed aerosol injection, step 2. injection of a gasphase alcohol (aldehyde), and step 3. injection of a gas-phase aldehyde (alcohol). The “total” in the x-axis denotes the total aerosol yields calculated by dividing the accumulated organic aerosol volume by the injected total organic volume into the bag air phase at step 2 and 3 (see the captions below Table 3). dimethyl acetate (the pKBH+ value of the conjugate acid ) -4.6) would be 2.7% protonated, and an alcohol like ethanol (pKBH+ ) -2.4) would be more than 75% protonated under these conditions. Of interest in Table 2 is the aerosol yield as it relates to the carbon number of the aldehyde. If one considers the volatility of organic compounds, aldehydes with less than eight carbons partition to only a minor extent to the aerosol phase. In these experiments, for example, butanal produced only trivial aerosol yields, because of its high vapor pressure. Hence, the resulting acid-catalyst effects on the heterogeneous reaction were negligible. However, hexanal, a six carbon aldehyde, with a lower vapor pressure in comparison to butanal, had an observable aerosol yield (0.24%). In addition, the acid catalyst effect on the aerosol yield from hexanal was significant. Octanal aerosol yields were greater than hexanal, but decanal was less than octanal. A possible explanation is that the reactivity of an aldehyde may decrease with increasing carbon number due to steric effects of long chain alkyl groups. This trend is supported by observation, in that good linearity was obtained for a plot of log Ke/Ko (Ke is an equilibrium constant for acetal or ketal formation normalized by Ko, the equilibrium constant for a standard aldehyde or ketone) vs the Taft parameter describing steric effects on acetal and ketal formation (34). A similar relationship for the hydrolysis rates of dialkyl acetals has been reported (42). The heterogeneous reaction for glyoxal as a dialdehyde is more interesting than other monoaldehydes because it has an extremely high volatility (boiling point, 50.4 °C) which is similar to butanal as shown in Table 1 but appears to be associated with very significant aerosol yields as shown in Table 2. Due to the dialdehyde group, glyoxal is much more reactive with respect to hydration, polymerization, and hemiacetal formation then any other monoalkylaldehydes. It is well-known that both glyoxal and glyoxal dihydrate [(OCHCHO)3‚2H2O] are quickly polymerized by direct exposure to the humidity in air or direct contact with water (50). When glyoxal was injected with 1-decanol and octanal, the aerosol yield (6.0%) dramatically increased in the presence of H2SO4, as shown in Table 3. This result is extremely important because photochemical reactions of organic aromatic compounds produce dicarbonyls (e.g., glyoxal, methyl glyoxal, and 3-hydroxy-2-oxo-propanal), tricarbonyls 4762

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001

FIGURE 3. FTIR spectra for pure 1-decanol, pure octanal, and the octanal/1-decanol/(NH4)2SO4-H2SO4 aerosols impacted on the ungreased ZnSe FTIR disk.

TABLE 5. FTIR Peak Assignments functional group OH stretching hydrogen bonding free hydrogen bonded CH stretching in -CH2-, and -CH3, groups CH stretching of the aldehyde group carbonyl stretching hydrogen bonding free hydrogen bonded deformation of C-O-H, -CH2-, and -CH3 groups C-O-C stretching C-O stretching for primary alcohols NH stretching in (NH4)2SO4 SdO stretching in (NH4)2SO4

wavelength (cm-1) 3100-3600 3500-3650 3300 2830-2970 2715, 2815 1670-1750 1715-1726 1680-1690 1280-1450 1182 1030-1080 3150-3250 1100

(e.g., trioxopropane and 2,3-dioxobutanal), and multifunctional carbonyls that conjugate with other carbonyl and carbon double bonds (21, 51, 52). To investigate the effect of hemiacetal/acetal formation on aerosol yields, aldehydes and alcohols were injected in three sequential steps as shown in Table 3. The aerosol yields for each step [step 1: (NH4)2SO4 or (NH4)2SO4-H2SO4 seed aerosol injection, step 2: injection of a gas phase alcohol (aldehyde), step 3: injection of a gas phase aldehyde (alcohol)] are reported in Figure 2. When only 1-decanol was added to the chamber at step 2 of the experiment on 11/20/00, no aerosol yield was observed. When octanal was added to the Teflon bag in step 3, a remarkable aerosol yield of 7.01% was obtained in the presence of an acid-catalyzed organic aerosol but 1.7% in the absence of an acid catalyst. When the order was reversed (addition of octanal in step 2 and 1-decanol in step 3), significant aerosol yields were obtained from both step 2 and step 3 in the presence of the acid catalyst (Figure 2). Without the acid catalyst, aerosol yields in the third step were negligible (0.14%). Again the importance of an acid catalyst is highlighted in the aerosol mass accumulation process. FTIR Analysis of Heterogeneous Reactions. FTIR spectroscopy was used to directly observe the reaction and transformation of chemical functional groups as the heterogeneous reactions proceeded. The organic aerosol samples for the FTIR spectra were collected on an ungreased ZnSe FTIR disk by impaction. Figure 3 shows typical FTIR spectra of the SOA formed from the heterogeneous reaction of octanal with 1-decanol, in the presence of the H2SO4 acid catalyst on

TABLE 6. Curve Fitting Results for Aldehydic Carbonyl Stretching in the Octanal System, Which Was Reacted Directly on a ZnSe FTIR Disk by Adding Small Amounts of Acid Catalyst Aqueous Solutiona rex. rex. free free NO. t (min) ν (cm-1) HWHH 1 2 3 4 5 6 7

0 5 10 25 45 70 360

1726 1716 1717 1717 1717 1717 1718

16 18 18 18 18 18 18

AF 0.7 0.145 0.09 0.08 0.055 0.033 0.013

H-bonded H-bonded ν (cm-1) HWHH 1685 1686 1686 1686 1686 1686 1686

19 18 18 18 18 18 18

AH

AH/AF

0.045 0.01 0.01 0.01 0.009 0.007 0.008

0.064 0.069 0.111 0.125 0.164 0.212 0.615

a ν: the center frequency. HWHH: the half width at half-height. A : H the peak absorbance for the hydrogen bonded carbonyl stretching. AF: the peak absorbance for the free carbonyl stretching.

FIGURE 4. FTIR spectra of the SOAs impacted on the ungreased ZnSe FTIR disk for the octanal/(NH4)2SO4-H2SO4, octanal/1-decanol, and glyoxal/1-decanol aerosol systems in the presence/absence of an acid catalyst.

FIGURE 5. Time series FTIR spectra for the octanal/acid-catalyst system, which was reacted directly on a ZnSe FTIR disk by adding small amounts of acid catalyst aqueous solution.

FIGURE 7. Curve-fitting for the CdO (7A) and O-H (7B) stretching regions in the FTIR spectra of SOAs impacted on the ungreased ZnSe FTIR disks for the octanal/1-decanol/acid catalyst system.

FIGURE 6. Curve-fitting for the CdO stretching regions (at reaction time 10 min and 6 h) of the FTIR spectra from the octanal/acid catalyst system used in Figure 5. the particle phase. The O-H stretching of the hydrogenbonded alcoholic hydroxyl group is seen at 3100-3600 cm-1, and the CdO stretching of aldehydes is shown at 1640-1780 cm-1. The detailed assignment of spectra bands (53) for SOAs formed from the heterogeneous reaction of octanal/1decanol/acid-catalyst system is summarized in Table 5. In particular, the transformation of the CdO to the C-O-C group (stretching at 1182 cm-1) is characteristic for heterogeneous reactions of aldehydes.

By tracking the CdO stretching peak of octanal, one can see a dramatic decrease in the intensity of this peak over the course of the heterogeneous reaction of aldehyde with 1-decanol in the presence of an acid catalyst (Figure 3). Peak broadening in aerosol samples, in comparison with pure components, also occurred due to formation of large molecules as the heterogeneous reaction proceeded (Figure 3). Figure 4 shows the aerosol FTIR spectra for different aerosol systems: the octanal/acid-catalyst system and the octanal/1-decanol and glyoxal/1-decanol systems in the presence or absence of an acid catalyst. The spectra for the aerosol formed in the presence of acid-catalyst have similar absorption peaks with differences in the relative intensity of functional groups. Without the acid catalyst, only the characteristic peaks for (NH4)2SO4 seed aerosols appear in both the octanal/1-decanol and glyoxal/1-decanol spectra. VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4763

TABLE 7. Population Parameters (D h pg and σg) for Seed and Grown Particles and Optimized Constants (r and β) for SOAs from the Heterogeneous Reaction of Octanal/1-Decanol and Glyoxal Systems in the Presence/Absence of an Acid Catalyst reaction system hexanal/1-decanol hexanal/1-decanol glyoxal glyoxal

catalyst acid non acid non

D h pg (seed)

σg (seed)

N0 (seed)

D h pg (grown)

σg (grown)

N0 (grown)

r

β

58.7 58.3 57.8 41.3

2.02 1.96 1.94 1.76

387600 389479 221899 157819

71.9 66.8 80.9 60.3

1.88 1.89 1.74 1.83

418945 401901 269261 184124

39.1 35.2 36.8 39.8

0.019 0.003 0.053 0.034

FIGURE 8. Particle size distributions of seeds and SOAs for hexanal/1-decanol systems in the presence/absence acid catalyst. To demonstrate the acid catalyzed aldehyde reaction, octanal was directly reacted on a ZnSe FTIR window by adding small amounts of aqueous H2SO4 acid catalyst solution (0.005 M). The spectra of the octanal/acid-catalyst system changed progressively as a function of time as shown in Figure 5. The aldehydic C-H stretching at 2715 cm-1 immediately disappeared, the CdO stretching band at 1726 cm-1 gradually decreased, and the OH stretching at 3100-3600 cm-1 increased as hydrates formed. The carbonyl stretching region (1640-1780 cm-1) suggests the presence of at least two bands and indicates the degree of hydrogen bonding. The procedure of least-squares curve fitting was implemented for the multicomponent analysis. The fitting parameters are the center frequency (ν), the peak absorbance (A), and the half width at half-height (HWHH). Most band shapes in the FTIR absorbance spectra were approximated by combinations of Lorentzian and Gaussian functions (54). This treatment was used in Figure 6 for the curve-fitting results for the carbonyl stretching of the octanal/ acid-catalyst system shown at Figure 5. Table 6 summarizes the curve-fitting results in this region. The free CdO stretching band was fixed at 1717 cm-1, and the hydrogen bonded CdO at 1686 cm-1. In this study a Gaussian to the Lorenzian fraction of 0.6 resulted in optimal results. Table 6 shows that the degree of hydrogen bonding is associated with the progress of the aldehyde reaction in the presence of the H2SO4 acid catalyst. Least-squares curve-fitting was also performed for carbonyl and hydroxyl stretching bands for the aerosol sample from the octanal/1-decanol/acid catalyst reaction system (Figure 7). Figure 7A shows the ratio of AH (the peak absorbance for hydrogen bonding) to AF (the peak absorbance for free carbonyl stretching) is much higher (2.3) than in the octanal/acid-catalyst system (0.06-0.62 in Table 5). It was assumed that the OH stretching region (3100-3600 cm-1) of octanal/1-decanol/acid-catalyzed aerosols comprise three 4764

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001

stretching bands attributed to free O-H (ν ) 3505 cm-1, HWHH ) 80 cm-1, A ) 0.009), hydrogen bonded O-H (ν ) 3370 cm-1, HWHH ) 110 cm-1, A ) 0.01), and N-H stretching from (NH4)2SO4 seed aerosols. The Gaussian fraction of the band shape was 0.9 for O-H stretching. Unlike 1-decanol, which exhibits mostly hydrogen bonded hydroxyl stretching (ν ) 3370 cm-1), the aerosols created from the heterogeneous reaction of octanal with 1-decanol in the presence of acid catalyst contains about 47% free hydroxyl stretching (AH/AF ) 1.11). One possible explanation is that the tertiary hydroxyl group of acetals formed through the heterogeneous reaction of octanal is surrounded by branched long chain alkyl groups and partially isolates the hydroxyl groups from the surrounding environment. All of the above observations and interpretations support the idea that the aerosol generated by the heterogeneous reaction of aldehyde includes polymers, hemiacetals/acetals, and aldehyde hydrates (gem-diols) and that hydrogen bonded hydroxyl groups are very important linkages observed in these systems. Aerosol Growth and Size Distributions. Of interest is the change of particle size distributions as a function of secondary aerosols formation. The aerosol growth and population changes in the Teflon bags were monitored with a differential mobility analyzer as the heterogeneous reaction of aldehydes progressed with time. The initial seed particles of (NH4)2SO4 or (NH4)2SO4/H2SO4 were log-normally distributed with geometric mean diameters (D h pg) and standard geometric deviations (σg) (55). Table 7 shows D h pg and σg values for hexanal/1-decanol and glyoxal aerosol systems. For the hexanal/1-decanol system, D h pg increased from 58.3 nm (seed) to 66.8 nm (SOA) without an acid catalysis and from 58.7 nm (seed) to 71.9 nm (SOA) with an acid catalysis. For glyoxal D h pg increased from 41.3 to 60.3 nm without an acid catalyst and 57.8 to 80.9 nm with an acid catalyst. The log-normal fits using D h pg and σg are represented for both seed aerosols and aerosol growth from heterogeneous reactions, as shown

in Figure 8 and Figure 9 (Supporting Information). As the reaction proceeded, the organic aerosol mass increased due to the heterogeneous reactions of aldehydes (55, 56). In turn, this increment accelerated the condensation of gas-phase organic compounds on the aerosol phase as predicted by gas-particle partitioning theory (8-12, 57). It is possible to develop (see Supporting Information) a predictive equation for the size distribution as a function of the initial size distribution parameters of the seed aerosols (D h pg, σg, and N0) and constants for condensation and heterogeneous reaction rates, R and β (see eq S10):

n(Dp,t) )

βDpe - 2βt

(

2

(R + βDp )e exp -

ln2

- 2βt

((

N0 - R x2πlnσg

)

(R + βDp2)e - 2βt - R β 2ln2σg

1/2

/D h pg

)

)

Hill. The authors thank Sangdon Lee for assisting with measuring particle distribution using the TSI-SMPS. Last we wish to thank the reviewers for their very helpful comments and insights.

Supporting Information Available A predictive equation for the size distribution as a function of the initial size distribution parameters of the seed aerosols (D pg, σg, and N0) and constants for condensation and heterogeneous reaction rates (R and β); Figure 9 shows particle size distributions of seeds and SOAs for glyoxal systems in the presence/absence acid catalyst. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (4)

The R and β values are characteristic constants for condensation and volume reaction, and these constants contain a number of parameters including Di (the diffusion coefficient for species i in air), Mi [the molecular weight of species i, (g/mol)], T [the temperature (K)], R (gas constant), ip (the vapor pressure in the gas phase), ip0 (the equilibL Li rium vapor pressure the product vapor pressure), kυ (rate constant for the product formation), Hυ (adsorption equilibrium constant), σs (surface tension of a particle), Fp (liquidphase density), m (mass of a particle), and υ j (volume occupied by a molecule in the liquid phase) (see Supporting Information). The constants R and β, which were obtained by optimizing the model fit (eq 4) to the aerosol distribution experimentally observed for reaction systems, are summarized in Table 7. The contribution of heterogeneous reactions of aldehydes on aerosol growth can be successfully demonstrated by comparing R and β constants in the presence of an acid catalyst to those in the absence of an acid catalyst. R and β are evaluated at 39.1 and 0.019 with the H2SO4 acid catalysis, and at 35.2 and 0.003 without the acid catalysis for the hexanal/1-decanol system, and 36.8 and 0.053 with H2SO4 and 39.8 and 0.034 without H2SO4 for the glyoxal system (Table 7). The particle distributions observed from acidcatalyzed heterogeneous reaction systems were shifted to larger particle sizes than the nonacid-catalyzed systems, as shown in Table 7, Figure 8, and Figure 9 (Supporting Information). It is also apparent that the model parameter β for the contribution of heterogeneous reaction to aerosol growth was higher with an acid catalysis than without an acid catalysis as shown in Table 7. Figures 8 and 9 show the size distributions produced by eq 4 for hexanal/1-decanol and glyoxal systems after 3 min of reaction, both with and without an acid catalysis. These distributions are compared to experimental data measured by a TSI-SMPS in the bag and show that model predictions for aerosol population can closely simulate our experimental size distribution data. The model fitting parameters (R and β), which were observed in the controlled reaction systems for different aldehydes and alcohols, RH, T, and acidity, will permit us in the future to apply these parameters to predicting accommodation of gasphase carbonyl products. This, in turn, makes it possible to predict particle size distributions of growing organic aerosols using either partitioning or the heterogeneous reaction chemistry of gas-phase organic products.

Acknowledgments This work was supported by a Grant from National Science Foundation (ATM 9708533) and STAR grants from U.S. EPA (R-82817601-0) to the University of North Carolina at Chapel

(1) Twomey, S. Atmos. Environ. 1991, 25A, 2435-2442. (2) Pilinis, C.; Pandis, S.; Seinfeld, J. H. J. Geophys. Res. 1995, 100, 18739-18754. (3) Eldering, A.; Larson, S. M.; Hall, J. R.; Hussey, K. J.; Cass, G. R. Environ. Sci. Technol. 1993, 27, 626-635. (4) Eldering, A.; Cass, G. R. J. Geophy. Res. 1996, 101, 19343-19369. (5) EPA, Air Quality Criteria for Particulate Matter; EPA/600/P95/001cF; Environmental Protection Agency; Washington, DC, 1996. (6) Altshuller, A. P. Atmos. Environ. 1983, 17, 2131-2165. (7) Jeffrie, H. E. Photochemical air pollution. In Composition, chemistry, and climate of the atmosphere; Singh, H. B., Ed.; Van Nostrand Reinhold: New York, 1995. (8) Odum, J. R.; Hoffmann, T.; Bowman, F.; Collins, D.; Flagan, R. C.; Seinfeld, J. H. Environ. Sci. Technol. 1996, 30, 2580-2585. (9) Odum, J. R.; Jungkamp, T. P. W.; Seinfeld, J. H. Science 1997, 276, 96-99. (10) Hoffmann, T.; Odum, J. R.; Bowman, F.; Collins, D.; Klockow, D.; Flagan, R. C.; Seinfeld, J. H. Atmos. Environ. 1997, 26, 189222. (11) Kamens, R. M.; Jang, M.; Chien, C. J.; Leach, K. Environ. Sci. Technol. 1999, 33, 1430-1438. (12) Kamens, R. M.; Jaoui, M. Environ. Sci. Technol. 2001, 35, 13941405. (13) Yu, J.; Cocker, D. R.; Griffin, R. J.; Flagan, R. C.; Seinfeld, J. H. J. Atmos. Chem. 1999, 34, 207-258. (14) Noziere, B.; Barnes, I. J. Geophys. Res. 1998, 103, 25587-25597. (15) Noziere, B.; Barnes, I.; Becker, K. H. J. Geophys. Res. 1999, 104, 23645-23656. (16) Glasius, M.; Lahaniati, M.; Calogirou, A.; Bella, D. D.; Jensen. N. R.; Hjorth, J.; Kotzias, D.; Larsen, B. R. Environ. Sci. Technol. 2000, 34, 1001-1010. (17) Jang, M.; Kamens, R. M. Atmos. Environ. 1999, 33, 459-474. (18) Jaoui, M.; Kamens, R. M. J. Geophys. Res. 2001, 106, 1254112558. (19) Hull, L. A. Terpene Ozonalysis Products. In Atmospheric Biogenic Hydrocarbons; Bufalini, J. J., Arnts, R. R., Eds.; Ann Arbor Science: Michigan, 1981; Vol. 2, pp 161-186. (20) Wangberg, I.; Barnes, I.; Becker, K. H. Environ. Sci. Technol. 1997, 31, 2130-2135. (21) Jang, M.; Kamens, R. M. Environ. Sci. Technol. 2001, 35, 36263639. (22) Yaws, C. L. Chemical properties handbook: physical, thermodynamic, environmental transport, safety and health related propertied for organic and inorganic chemicals; McGraw-Hill: New York, 1999. (23) Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry; John Wiley & Sons: New York, 1993. (24) Mackay, D.; Bobra, A.; Chan, D. W.; Shiu, W. Y. Environ. Sci. Technol. 1982, 16, 645-649. (25) Zhao, L.; Li, P.; Yalkowsky, S. H. J. Chem. Inf. Comput. Sci. 1999, 39, 1112-1116. (26) Zhao, L.; Ni, N.; Yalkowsky, S. H. Ind. Eng. Chem. 1999, 39, 324-327. (27) Joback, K. G.; Reid, R. C. Chem. Eng. Commun. 1987, 57, 233. (28) Stein, S. E.; Brown, R. L. J. Chem. Inf. Comput. Sci. 1994, 34, 581-587. (29) Aldehyde and Ketones. In Comprehensive Organic Chemistry: The synthesis and reactions of organic compounds; Barton, D., Ollis, W. D., Eds.; Pergamon Press: New York, 1979; pp 9601013. VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4765

(30) Klass, D. L.; Jensen, W. N.; Blair, J. S.; Martinek, T. W. J. Org. Chem. 1963, 28, 3029-3034. (31) Walker, F. Formaldehyde, 3rd ed.; Reinhold Publishing: New York, 1964; 140-205. (32) Iraci, L. T.; Tolbert, M. A. J. Geophys. Res. 1997, 102, 1609916107. (33) Guthrie, J. P. Can. J. Chem. 1975, 53, 898-906. (34) Bell, J. M.; Kubler, D. G.; Sartwell, P.; Zepp, R. G. J. Org. Chem. 1965, 30, 4284-4290. (35) Chastrette, F.; Chastrette, M.; Bracoud, C. Bull. Soc. Chim. Fr. 1986, 5, 822-836. (36) Berger, P. A.; Garbow, J. R.; DasGupta, A. M.; Remsen, E. E. Macromolecules 1997, 30, 5178-5180. (37) Meadows, G. W.; Darwent, B. B. Can. J. Chem. 1952, 30, 10151023. (38) Tobias, H. J.; Ziemann, P. J. Anal. Chem. 1999, 71, 3428-3435. (39) Tobias, H. J.; Ziemann, P. J. Environ. Sci. Technol. 2000, 34, 2105-2115. (40) Deyrup, A. J. J. Am. Chem. Soc. 1934, 56, 442-444. (41) Cordes, E. H.; Bull, H. G. Chem. Rev. 1974, 74, 581-603. (42) Maurice, M.; Kreevoy, M. M.; Taft, R. W. J. Am. Chem. Soc. 1955, 77, 5590-5597. (43) Carey, F. A.; Sundberg, R. J. Advanced organic chemistry: part A Structure and Mechanisms, 4th ed.; Plenum Press: 2000; New York. (44) Hammett, L. P.; Deyrup, A. J. J. Am. Chem. Soc. 1932, 54, 2721. (45) Liler, M. Reaction Mechanisms in Sulphuric Acid; Academic Press: New York, 1971. (46) Rochester, C. H. Acidity Function; Academic Press: New York, 1970.

4766

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001

(47) Generation of submicron sulfuric acid aerosol by vaporization and condensation. In Generation of aerosols and facilities for exposure experiments; Willeke, K., Ed.; Ann Arbor Science Publishers Inc.: Ann Arbor, 1980; pp 317-336. (48) Perry’s Chemical Engineering Handbook; Perry, R. H., Ed.; Mcgraw-Hill Cor.: New York, 1984. (49) Jones, R. A. Y. Physical and Mechanistic Organic Chemistry; Cambridge University Press: New York, 1979. (50) Calcateria, L. T.; Gibson, H. W. United States Patent 4717544, 1988. (51) Yu, J.; Jeffries, H. E.; Sexton, K. G. Atmos. Environ. 1997, 31, 2261-2280. (52) Liu, X.; Jeffries, H. E.; Sexton, K. G. Environ. Sci. Technol. 1999, 33, 4212-4220. (53) Dolphin, D.; Wick, A. Tabulation of Infrared Spectral Data; John Wiley & Sons: 1977; New York. (54) Fourier Transform Infrared Spectrometry; Elving, P. J., Winefordner, J. D., Eds.; John Wiley & Sons: New York, 1986; in Chemical Analysis vol. 83. (55) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics; John Wiley & Sons. Inc.: New York, 1998. (56) McMurry, P. H.; Wilson, J. C. Atmos. Environ. 1982, 16, 121-134. (57) Cruz, C. N.; Pandis, S. N. Aerosol Sci. Technol. 1999, 31, 392407.

Received for review March 27, 2001. Revised manuscript received September 19, 2001. Accepted September 25, 2001. ES010790S