Report

Atmospheric New Particle Formation Enhanced by Organic Acids

See allHide authors and affiliations

Science  04 Jun 2004:
Vol. 304, Issue 5676, pp. 1487-1490
DOI: 10.1126/science.1095139

Abstract

Atmospheric aerosols often contain a substantial fraction of organic matter, but the role of organic compounds in new nanometer-sized particle formation is highly uncertain. Laboratory experiments show that nucleation of sulfuric acid is considerably enhanced in the presence of aromatic acids. Theoretical calculations identify the formation of an unusually stable aromatic acid–sulfuric acid complex, which likely leads to a reduced nucleation barrier. The results imply that the interaction between organic and sulfuric acids promotes efficient formation of organic and sulfate aerosols in the polluted atmosphere because of emissions from burning of fossil fuels, which strongly affect human health and global climate.

Aerosols influence the Earth-atmosphere system in several distinct ways (1, 2). Concerns over the human health effects of fine particulate matter constitute the most important element in formulating the national ambient air quality standard (3). Also, aerosols directly or indirectly affect the Earth's radiation budget (4, 5), and light absorption by aerosols causes visibility degradation. Furthermore, modification of clouds and precipitation by aerosols may enhance lightning activity and thus influence tropospheric chemistry (6, 7). The impacts of particulate matter on health, radiation, and cloud microphysics are strongly dependent on the particle sizes.

Several processes determine the aerosol size distribution, including new particle production (as a result of gas-to-particle conversion), growth due to condensation and coagulation, removal rates, and primary emissions. New particle formation or nucleation is the least understood of these steps (8). Much of the previous research has focused on nucleation of sulfuric acid, because sulfate represents an important component of the nucleation mode aerosol (9). It is commonly recognized that binary nucleation of H2O-H2SO4 is not efficient enough to explain atmospheric new particle formation (10). Progress recently has been made in assessing the importance of ternary water–sulfuric acid–ammonia nucleation (11, 12), ion-induced nucleation (13, 14), and nucleation involving iodide species (15, 16).

The role of organic compounds in new particle formation is another potentially important issue (17). Atmospheric measurements reveal that aerosols often contain a considerable amount of organic matter (1821). During photooxidation of volatile organic compounds (VOCs), non- or semivolatile organic products are produced that contribute to secondary organic aerosol (SOA) formation. For example, in the urban atmosphere the aromatic component in gasoline (mainly toluene and xylenes) is responsible for SOA formation caused by oxidation of these compounds (22). Current theory of SOA formation assumes that condensation of low-volatility organic species such as carboxylic or dicarboxylic acids occurs on preexisting particles from primary emissions or formed by homogeneous nucleation, most probably involving sulfuric acid–ammonia–water or ions (8). Alternatively, it is suggested that SOA nucleation may occur through the formation of stable organic heterodimers (23). Currently, few experimental studies have investigated new particle formation from organic acids. Another process, which also influences the chemical composition of organic aerosols, involves absorption of gaseous species onto particulate matter. On the basis of consideration of the thermodynamic equilibrium distribution of a compound between the gas and condensed phases, a gas-particle partitioning model has been proposed (24) and invoked to explain the observed correlation between the SOA yield and the organic aerosol mass concentration (22, 25). More recently, it has been suggested that sulfate aerosols catalyze heterogeneous reactions of carbonyl compounds, leading to a considerably enhanced SOA yield (26). The growth of SOA from both mechanisms depends on preexisting particles, which are linked to new particle formation or primary emissions.

To assess the role of low-volatility organic species in new particle formation, we performed laboratory studies of particle nucleation from aromatic acid vapors and their mixtures with H2SO4 (27). Aromatic acids, such as benzoic (C7H6O2), p-toluic (C8H8O2), and m-toluic (C8H8O2) acids, are products from photochemical degradation of aromatic hydrocarbons emitted from automobiles in the urban atmosphere (28) and have been identified in the particle phase (29, 30). Nanometer-sized particles were produced in an aerosol chamber, and the particle concentration and size distribution were monitored with an ultrafine particle counter (model 3025A, TSI Incorporated Particle Instruments, St. Paul, MN) and a nanodifferential mobility analyzer (model 3085, TSI Incorporated Particle Instruments) capable of measuring particle sizes as small as 3 nm (fig. S1). Gas-phase concentrations of the organic and sulfuric acids in the aerosol chamber were monitored with the use of proton-transfer reaction mass spectrometry and chemical ionization mass spectrometry, respectively (31, 32). We initially generated H2SO4 aerosols by introducing gas-phase H2SO4 in a nitrogen carrier gas with a variable relative humidity (RH). For a gaseous H2SO4 concentration in the range of 109 to 1010 molecule cm–3, the particle sizes formed ranged from 3 to 10 nm (Fig. 1), corresponding to the nucleation mode. The observed particle concentration increased when the gaseous H2SO4 concentration or RH was increased. A marked increase in the particle concentration occurred when benzoic acid vapor was added to the aerosol chamber (Fig. 1). With H2SO4 concentrations of 6 × 109 and 8 × 109 molecule cm–3, addition of 0.04 ppb (parts per billion) benzoic acid increased the particle concentration by a factor of 5 (Fig. 1A). For a fixed H2SO4 concentration, higher amounts of benzoic acid resulted in more pronounced particle formation (Fig. 1B). Figure 1 shows that the measured peak diameter of the particle distribution shifted slightly to a larger size with addition of benzoic acid, implying that the presence of benzoic acid both enhanced nucleation and contributed to the growth of the newly nucleated particles. Substantially larger peak diameters (>10 nm) were detected when benzoic acid concentrations were increased by one to two orders of magnitude.

Fig. 1.

Measured particle size distributions of the nucleating aerosols. In (A), the brown and black curves correspond to H2SO4 aerosol formation with a RH of 5% and gaseous H2SO4 of concentrations 6 × 109 and 8 × 109 molecule cm–3, respectively. The green and orange curves are similar to the brown and black curves, respectively, except for addition of 0.04 ppb benzoic acid to the aerosol chamber. In (B), the brown curve corresponds to H2SO4 aerosol formation with a RH of 5% and a gaseous H2SO4 concentration of 7 × 109 molecule cm–3. The green and orange curves are similar to the brown curve, except for the addition of 0.04 and 0.1 ppb benzoic acid (corresponding to 1 × 109 and 2.5 × 109 molecule cm–3), respectively. The experiments were performed at 298 ± 2 K and a total pressure of 760 torr.

The aerosol nucleation rate, J, was estimated on the basis of the ratio of the measured particle concentration to the nucleation time (33). In the absence of organic acids, the nucleation rate is dependent on the gas-phase H2SO4 concentration and RH. Our measured nucleation rate of the H2O-H2SO4 binary system is qualitatively in agreement with previous experimental studies (12). Figure 2 shows that the nucleation rate was considerably increased in the presence of the organic acids. The nucleation rate in the presence of 0.1 ppb benzoic acid is about a factor of 8 to 10 higher than that of the H2O-H2SO4 binary system. Enhanced nucleation rates were also observed for p-toluic and m-toluic acids (Fig. 2, B and C). The nucleation rate was increased by a factor of 5 to 13 in the presence of 0.2 to 0.3 ppb of the two acids. For RH in the range of 4 to 15%, addition of sub-ppb levels of the aromatic acids consistently led to a larger nucleation rate by a factor of 5 or higher than that of the H2O-H2SO4 binary system (fig. S2). The partial pressures of the aromatic acids in those experiments were several orders of magnitude smaller than their corresponding equilibrium vapor pressures; that is, the saturation ratio, S (34), was much smaller than unity. Interestingly, the high nucleation rate was also measured in the absence of water vapor for benzoic acid and p-toluic acid, indicating that binary nucleation of the organic acid–sulfuric acid system is responsible for the enhanced new particle formation (35). Hence, these results suggest a probable interaction between the aromatic acid and sulfuric acid that leads to a reduced (heteromolecular) nucleation barrier. The magnitude of the effect of aromatic acids on H2SO4 nucleation enhancement appears to be comparable to that previously reported for ammonia at similar H2SO4 and ammonia additive concentrations and RH (fig. S3).

Fig. 2.

Estimated nucleation rate (J) as a function of gaseous H2SO4 concentration. The solid triangles correspond to H2SO4 aerosol formation with a RH of 5%, and the solid circles correspond to particle formation with 5% RH and in the presence of 0.1 ppb benzoic acid (A), 0.2 ppb p-toluic acid (B), or 0.3 ppb m-toluic acid (C). The curves are fit to the experimental data. The experiments were performed at 298 ± 2 K and a total pressure of 760 torr.

We also examined (homomolecular) nucleation of the aromatic acids in the absence of sulfuric acid and water. New particle formation was only detected when a substantial saturation ratio was established in the aerosol chamber. The minimum S required to produce detectable new particles was about 45 for benzoic acid and even higher for p-toluic and m-toluic acids. Similarly, water was observed to have a negligible influence on the organic particle formation for benzoic and p-toluic acids, because the two organic acids are insoluble in water and the organic aerosols formed are hydrophobic. In general, particle formation can be qualitatively predicted in terms of fundamental thermodynamic and kinetic principles (36). The spontaneous gas-to-particle conversion process corresponds to a decreased free energy and is thermodynamically favorable but kinetically hindered. During nucleation, a thermodynamically stable cluster or critical embryo is generated before condensation growth of the particle, and this embryo formation involves an energy barrier. Also, condensation growth of nucleated critical embryos will be retarded because of increased activity due to the Kelvin barrier. Hence, particle nucleation and subsequent growth in a single-component system occur only if the system is supersaturated (S > 1). It is conceivable that large barriers generally exist for other carboxylic or dicarboxylic acids, as shown in our experiments for the aromatic acids. The atmospheric concentrations of the low-volatility organic compounds are typically at the ppb level or less, even under polluted conditions (1, 2). Although certain dicarboxylic acids do reach their saturation points in the atmosphere (18), the high supersaturation required for homomolecular nucleation likely renders new particle formation from those compounds implausible. Previous smog chamber studies reported homogeneous nucleation from low-volatility organic compounds, but those experiments were carried out with the use of hydrocarbon concentrations that were several orders of magnitude higher than those found under the ambient conditions (25).

To gain an insight into the nature of the interaction between aromatic and sulfuric acids at the molecular level, we performed quantum chemical calculations that show the formation of surprisingly stable aromatic acid–sulfuric acid complexes (fig. S4). The equilibrium aromatic acid–sulfuric acid structure exhibits a nearly planar eight-membered ring: There are two hydrogen bonds, with the organic acid molecule acting as both a hydrogen bond donor and acceptor. The strength of the hydrogen bonding is reflected by the calculated bond lengths. For the benzoic–sulfuric acid complex, for example, the hydrogen bond is 1.503 Å for C=O···HOS and 1.705 Å for COH···O=S, nearly comparable to weak covalent bonds. The energetics of the complexes was quantified with a series of quantum chemical calculations (Table 1). The bonding energies of the complexes are about 20 kcal mol–1 for benzoic and p-toluic acids and are about 4 kcal mol–1 higher for m-toluic acid. For comparison, the bonding energy is about 10 kcal mol–1 for the H2O-H2SO4 complex (37) and 25 kcal mol–1 for the H2O-H2SO4-NH3 system (38). The large stability of the organic acid–sulfuric acid complex implies that the aromatic acid molecule bonds irreversibly to H2SO4 under atmospheric conditions. The complex formation between aromatic and sulfuric acids most likely reduces the barrier in heteromolecular nucleation and helps condensation growth of the nucleated critical embryo by overcoming the Kelvin effect, explaining the enhanced new particle formation observed in our experiments. Additional calculations were performed for glutaric acid, indicating that stable complex formation with sulfuric acid represents a general feature for organic compounds with the carboxylic or dicarboxylic functional group. It is likely, though, that for smaller organic acids the effect on H2SO4 nucleation may be less important than that observed for the aromatic acids (39).

Table 1.

Bonding energies, D0 (in kcal mol-1), of the aromatic acid—sulfuric acid complexes. BA-SA denotes benzoic acid—sulfuric acid complex; PTA-SA, p-toluic acid-sulfuric acid complex; and MTA-SA, m-toluic acid—sulfuric acid complex. All energies are corrected with the zero-point energies (ZPE). The quantum chemical methods used in the present study are similar to those used by Suh et al. (28).

Complex D0
BA-SA 19.85View inline
17.62View inline
18.63View inline
17.84View inline
PTA-SA 19.99View inline
MTA-SA 23.72View inline
  • View inline* Determined with B3LYP/6-31G(d,p)//B3LYP/6-31G(d,p).

  • View inline Determined with CCSD(T)/6-31G(d) + CF//B3LYP/6-31G(d,p).

  • View inline Determined with QCISD(T)/6-31G(d)//MP2(full)/6-31G(d).

  • View inline§ Determined with G2(MP2, SVP).

  • Organic acids have been widely identified as common components in atmospheric particulate matter (1821). Our experimental study shows that homomolecular nucleation of aromatic acids is unlikely to occur under atmospheric conditions, but that the interaction between aromatic acids and sulfuric acid promotes efficient heteromolecular nucleation. The gas-phase concentration of organic acids is substantially higher than that of gaseous H2SO4 in the atmosphere (18); thus, organic acids can also contribute considerably to the initial growth of the newly nucleated embryos, which is important for subsequent particle growth by adsorption or heterogeneous reactions of other organic vapors. The particle formation mechanism proposed in this study can have major implications for SOA and sulfate aerosol formation in polluted areas, because both organic and sulfuric acids are photochemical degradation products linked to the emissions from the burning of fossil fuels (1, 2). In particular, new particle formation can occur efficiently over a large portion of northern America, eastern Asia, and some parts of central Europe because of the concurrent anthropogenic VOC and SO2 emissions in those regions (fig. S5) (40). Our results suggest an alternative cause for efficient aerosol nucleation frequently observed in the polluted atmosphere, in addition to the available theories of water–sulfuric acid–ammonia ternary nucleation and ion-induced nucleation. For example, enhanced new particle formation (with a particle size of 3 to 4 nm) was observed in anthropogenic plumes advecting from Asia, which were identified by elevated CO and SO2 concentrations (41). The high CO amount was indicative of the abundance of VOCs inside those plumes. It was speculated that a high SO2 concentration, in conjunction with other unidentified, possibly co-emitted species, was responsible for nucleation. In contrast, the same study revealed that few 3- to 4-nm particles were detected in the clean background and even within a volcanic plume that had a high H2SO4 but low CO concentration. Also, measurements of aerosol hygroscopicity during the 1999 Houston Supersite Project indicated a dominance of the organic matter in the fine particle mode, which could not be explained by the formation of ammonium sulfate (21). Those measurements likely can be explained by the importance of organic acids in particle nucleation and growth in the presence of sulfuric acid, because of the large abundance of both types of acids in urban environments and in the tropospheric boundary layer influenced by anthropogenic pollution.

    Supporting Online Material

    www.sciencemag.org/cgi/content/full/304/5676/1487/DC1

    Figs. S1 to S5 References

    References and Notes

    View Abstract

    Navigate This Article