9권2호_배귀남(97-112).pdf506.7KB

1. Introduction

Diseases associated with mosquitoes affect millions of people worldwide every year. Some mosquitoes transmit diseases such as malaria, yellow fever, and various other infections to humans and animals. The global malaria program of the World Health Organization (WHO) has helped more than 100 countries to prevent malaria (WHO, 2010). Many localities and houses have established mosquito control and prevention programs employing methods ranging from simple to complex, such as the use of mosquito nets and the application of chemical and biological insecticides. To protect human health and the environment and to minimize other risks, insecticide labels are required to identify their active chemicals and to provide guidance for safe usage. Normally, people are not required to relocate when they use insecticides in their homes to eliminate or prevent mosquitoes. However, no insecticide is completely safe and care must be exercised in the use of such products because of their potential toxicity (Casida, 1980; Zaim et al., 2000; Schummer et al., 2010).

Domestic insecticides are applied using sprayers and electrically heated evaporators. Natural or synthetic pyrethroid insecticides are commonly used against mosquitoes, houseflies, and cockroaches (Lines, 1996; Sirak-Wizeman et al., 2008; Mongkalangoon et al., 2009). In addition to the pyrethroids used for insect minimization or elimination, such as prallethrin, geraniol, permethrin, and deltamethrin, domestic insecticides also incorporate volatile organic compounds (VOCs), which are associated with fragrances to ensure comfortable conditions after their application. Most of these fragrant home insecticides are extracted from natural ingredients. The fragrances are mostly VOCs, which can react strongly with ozone to produce secondary aerosols. Previous studies reported that VOCs, particularly terpenes, emitted from air fresheners react with ozone to form secondary aerosols (Nazaroff and Weschler, 2004; Singer et al., 2006; Coleman et al., 2008; Vu et al., 2011b). Thus, the contribution of secondary aerosols produced from the reaction of ozone with VOCs can contribute to indoor air pollution, together with other outdoor and indoor primary sources. This contribution might explain the findings of previous studies that organic content accounts for 5–40% of the total mass of indoor particles (Odum et al., 1996; Hoffmann et al., 1997; Sheehan and Bowman, 2001).

The secondary particle concentration increases with ozone concentration under the same concentrations of VOCs (Vu et al., 2011a, 2011b). Concentrations and the chemical composition of VOCs also affect secondary aerosol formation (Odum et al., 1996; Cocker, 2001). Evaporation of VOCs from household products mostly depends on temperature and the surface area to which they are applied. Nearly 100% of methyl benzoate, an ingredient used in perfumes, was reported to evaporate after 55 min (Aggarwal et al., 1997), and the evaporation of fragrances in gel-type air fresheners occurred readily at temperatures <100°C (Aggarwal et al., 1998). Normally, common organic pollutants are two to five times higher inside homes than outside (US. EPA, 2012). Indoor VOC concentrations are <50 Hg/m³, with average levels <10 Hg/m³ in Europe and North America (Wolkoff and Nielsen, 2001). In addition, outdoor background- level ozone concentrations are affected by the time of day and season, with peak concentrations on hot and sunny days. A large proportion (30 70%) of – indoor ozone is derived from the outdoor ozone source (Weschler, 2000). The United States National Ambient Air Quality Standard set an 8-h average outdoor ozone concentration limit of 75 ppb. The WHO guideline value is 50 ppb, which implies that the indoor ozone concentration should be <35 ppb. Secondary aerosol formation, however, can occur at low VOC and ozone concentrations (Liu et al., 2004; Langer et al., 2008).

In this work, we investigated the potential for secondary aerosol formation during the reactions of ozone and VOCs emitted from a commercial household insecticide in a 1-m³ reaction chamber. The effects of three levels of injected ozone concentration (50, 100, and 200 ppb) were observed. The particle number size distribution was continuously monitored. The consumed ozone was compared with the amount of secondary aerosol formed during the test period.

2. Experimental methods

2.1 Identification of VOCs in a test specimen

A commercial liquid-type insecticide, marketed for home use, was chosen as a test product. This product contained prallethrin, a chemical widely used to kill mosquitoes, with a content of 1.33 g/100 mL product. The product’s labeled constituents were naturally extracted ingredients, such as orange and pine oils. According to the label, the product contained sufficient active ingredient for about 45 days of use (10 h/day use time). A headspace test using a solid-phase microextraction (SPME) method was performed to identify VOCs emitted from the insecticide. SPME is a very simple, efficient, and solventless sample preparation method with sensitive detection limits (Vas and Vekey, 2004). VOCs emitted from 20 μL insecticide contained in a glass vial were adsorbed on an SPME fiber (carboxen/polydimethylsiloxane, 75-μm thickness; Supelco, USA) for 1 h at 25°C and analyzed by gas chromatography/mass spectrometry (GC/MS; Model 6890N; Agilent Technologies, USA). A 30-m fused silica capillary column (HP-5MS, 0.25-mm internal diameter, 0.25-μm film thickness; Agilent Technologies) was used to separate the target analytes. The initial GC oven temperature was set at 40°C for 2 min, and then programmed to increase to 100°C at 10°C/min intervals and from 100°C to 250°C at 5°C/min intervals. The MS was operated in full-scan mode from 35 to 350 m/z at a 0.5-s scan interval.

2.2 Ozone reaction experiment

The experimental system consisted of a reaction chamber, ozone and pure air generators, a heater to control temperature, and monitoring instruments for ozone, particle, temperature and relative humidity. The cube-shaped reaction chamber was made of 2-mil (~51-μm) Teflon fluorinated ethylene propylene film and had a volume of 1 m³. The surface-to-volume ratio was 6.0 m^-1. The chamber was installed in an indoor wooden box equipped with a door to avoid any light disturbance from the outside (Fig. 1). A pure air generation system (Model 737-15; Aadco Instruments, USA) was used to supply pure air into the reaction chamber (Bae et al., 2003). The confirmed concentrations of O₃, NO, NO₂, NO_x, and SO₂ impurities in the pure air were <1 ppb. GC/MS analysis of pure air detected no limonene, α-pinene, β-pinene, or terpinene.

Fig. 1. Schematic of the experimental setup.

Prior to each experiment, the chamber was flushed twice with pure air, and then with ozone (200 ppb) under ultraviolet (UV) irradiation for 4 h to minimize the effects of bag contamination. Thereafter, background particle and ozone concentrations of the air-filled chamber were measured for 4 h, resulting in a final ozone concentration <1.5 ppb and particle mass concentration <0.02 μg/m³.

A 1.0-mL liquid test specimen of the household insecticide was contained in a 50-mL glass Petri dish with a diameter of 65 mm, and was then placed in the bottom of the 1-m³ chamber filled with pure air. The chamber was maintained at a temperature of 20 ± 1°C and a relative humidity <20%. Ozone injection into the center of the chamber through the top surface using a photometric O₃ calibrator (API 401; A Teledyne Technologies Company, USA) commenced at 4 min after the test specimen was set. The ozone calibrator was operated at three different concentrations (50, 100, and 200 ppb) at a flow rate of 4.0 L/min. Each experiment lasted for approximately 4 h.

A Teflon sampling line was connected to an O₃ analyzer with a flow rate of 1.5 L/min. A stainless-steel sampling line was connected to a wide-range particle spectrometer (WPS; 1000XP; MSP Corporation, USA) with a flow rate of 1.1 L/min. The total sampling flow rate for monitoring concentrations of particles and ozone in the chamber was controlled to be similar to the flow rate of ozone injected into the chamber using mass flow controllers. The control of inflow and outflow produced an approximate air exchange rate of 0.24 h^-1.

During the experiments, the ozone concentration was monitored once per minute using a UV photometric O₃ analyzer (TEI 49i; Thermo Electron Corporation, USA) with an O₃ lower detection limit of 1.0 ppb. The analyzer was calibrated with 700-ppb span gas. The particle number size distribution was determined using the WPS over a scanning time of 153.6 s. The aerosol and sheath flows in the electrostatic classifier were set to 0.3 and 3.0 L/min, respectively, to detect particles ranging from 10 to 10,000 nm in diameter. The particle mass concentration was calculated from the particle number size distribution obtained from the WPS data, with an assumption of a 1-g/cm³ unit density. The air temperature and relative humidity in the Teflon chamber were monitored every 5 min using a small sensor with a data logger (SK-L200Th; Sato Keiryoki Mfg. Co., Ltd., Japan). This data logger was placed in the bottom of the chamber and had a lower detection limit for relative humidity of 20%.

3. Results and Discussion

3.1 VOCs emitted from a home insecticide

The major chemical ingredient of the test product was prallethrin. This product is applied in the home using an electrically heated evaporator. According to the product guidelines, it has an effective mosquito prevention or elimination area of 7.29–12.96 m². The product contains a large number of VOCs (Fig. 2). Toluene is well known for its human toxicity. Other VOCs present include α-pinene, cymene, d-limonene, α-terpinene, and α-thujone. Some of these VOCs, such as α-pinene and d-limonene, react readily with ozone to form secondary aerosols (Grosjean et al., 1993; Hoffmann et al., 1997; Jang and Kamens, 1999; Fan et al., 2003; Liu et al., 2004; Sarwar and Corsi, 2007). The reaction of limonene and ozone can occur at relatively low concentrations (~10 ppb each) (Langer et al., 2008). The reaction of α-pinene and ozone is significantly effective in forming secondary organic aerosols (Chen and Hopke, 2009). This reaction can contribute to secondary organic aerosol yields of 13.7–81.3% (Hoffmann et al., 1997; Cocker, 2001), depending on environmental conditions such as parent reactant concentrations, temperature, relative humidity, irradiation, ventilation, and the presence of OH and NO₃ radicals. In addition, these reactions produce OH radicals, which contribute 19–29% of the particle mass formed (Fan et al., 2003). The final products of these reactions are acids and aldehydes (Jang and Kamens, 1999; Leungsakul et al., 2005), which are formed by the contribution of OH radicals to the chemical transformations of intermediates.

Fig. 2. Chromatogram of volatile organic compounds emitted from the home insecticide tested in this study.

3.2 Effect of ozone concentration on secondary aerosol formation

Prior to conducting reactions of VOCs and ozone, we needed to determine the ozone concentration supplied to the chamber and the background potential for secondary aerosol particle formation. A mass balance equation was considered to calculate the variation in ozone concentrations in the chamber for these experiments, as described in a previous study (Vu et al., 2011b). Based on the variations in ozone concentration in the chamber with and without the test home insecticide, the reacted ozone concentration in each experiment was estimated and compared with the secondary particle mass. Reactions of VOCs and ozone are well known to form secondary aerosols, depending on various factors such as the type of VOCs present in the air (Odum et al., 1996; Hoffmann et al., 1997; Cocker, 2001; Fan et al., 2005), the reacted ozone concentration (Hoffmann et al., 1997; Vu et al., 2011a, 2011b), light (Hoffmann et al., 1997; Sheehan and Bowman, 2001), temperature (Jang and Kamens, 1999; Sheehan and Bowman, 2001; Takekawa et al., 2003; Lamorena and Lee, 2008), and air exchange rate (Sarwar and Corsi, 2007). Here, we investigated the effect of ozone and VOCs emitted from a home insecticide on secondary aerosol formation.

Most VOCs in the specimen immediately evaporated into air. The quantity of VOCs in the air was largest 10 min after placing the specimen into the chamber. This finding implies that the reactions of VOCs and ozone occurred most readily at this time. Ozone was injected into the chamber 4 min after introducing the specimen. We considered this point to represent the beginning of the experiment (elapsed time = 0). When ozone was injected into the chamber, the reactions between VOCs and ozone occurred immediately due to the pre-existing VOCs emitted from the specimen. Ozone concentrations in the experiment with the test specimen were clearly lower than those in the control experiment without the test specimen (Fig. 3).

Fig. 3. Comparison of ozone concentrations in pure air experiments with and without the home insecticide.

This result indicates that ozone was consumed during the reaction with VOCs. The reacted ozone rate may also vary with time. Obvious differences were observed between the two ozone concentration curves at 27, 10, and 4 min for injected ozone concentrations of 50, 100, and 200 ppb, respectively. The ozone consumption rates were nearly constant during the reaction periods. These rates were 0.03, 0.09, and 0.10 ppb/min for the 50-, 100-, and 200-ppb treatments, respectively. The ozone concentrations measured at 4 h were 22.0, 34.6, and 89.6 ppb for the 50-, 100-, and 200-ppb treatments, respectively. Reacted ozone concentrations of 8.8, 20.7, and 26.1 ppb for the 50-, 100-, and 200-ppb treatments, respectively, were estimated from the differences between measured ozone concentrations with and without the test specimen.

The typical secondary particle formation and growth phenomenon are plotted in Fig. 4 for the 100-ppb treatment. Fig. 4a shows the typical “banana-shaped” plot of nanoparticle formation and growth during the mixture of VOCs and ozone. Nanoparticle formation occurred very abruptly at about 20 min and proceeded rapidly. These nanoparticles also grew quickly to larger particles through the condensational growth phenomenon (Heaton et al., 2007; Simchi et al., 2007), which increased the particle mass concentration rather than the particle number concentration. Ultrafine particles are very stable and can exist from minutes to hours in the atmosphere (US EPA, 2004). The ultrafine particles generated in this study obviously existed for more than 3 h. The stability of ultrafine particles in indoor environments may contribute to their damaging effects, as they have time to come in to contact with the human lung, exposing tissue surfaces to the surface chemistry of the particles (Donaldson et al., 1998) and ultimately depositing in the deepest regions of the human respiratory system (Fissan et al., 2007). The particle number concentration increased rapidly and peaked at 60 min (Fig. 4b), which is well known as the nucleation burst phenomenon. After reaching a peak concentration (16,608 particles/ cm³), the particle number concentration then decreased due to wall loss (Lee et al., 2004). Although the particle number concentration decreased, the particle mass concentration increased gradually through the condensational growth of pre-existing particles during the experimental period.

Fig. 4. Particle formation and growth phenomena due to the reaction of ozone with volatile organic compounds emitted from the household insecticide at 100-ppb ozone concentration: (a) particle size distribution, (b) measured particle number and mass concentrations, (c) corrected particle number and mass concentrations.

To compare the particle formation potential of this insecticide with previously reported results for other household products, the particle mass concentrations were corrected for deposition loss on the reaction chamber wall and ventilation loss due to air exchange using the following equations:

where N(t, d_p ) and VN(t, d_p) are the measured particle number concentration and the wall loss of the particle number concentration, respectively, during Vt at a given time t for a particle diameter of d_p. In this study, a size-independent wall loss rate of 8.4 × 10^-5 s^-1 was used (Lee et al., 2004). The wall loss of the particle mass concentration was calculated from VN(t, d_p) using Eq. 2 (Hinds, 1999):

where C_m,wall is the particle mass concentration lost on the chamber wall, ρ_p is the assumed particle density (1 g/cm³), and d_p is the particle diameter at time t.

The loss of particle number and mass concentrations due to ventilation can be calculated using Eqs. 3 and 4, respectively: 

where NAER and C_m,AER are the lost particle number and mass concentrations due to air exchange during a time interval of t₂ – t₁, respectively. N(t₁) and C(t₁) are the measured particle number and mass concentrations at time t₁, respectively, and AER is the applied air exchange rate. AER was 0.24 h^-1 in this study.

Finally, corrected particle mass concentrations can be calculated using Eq. 5: 

where C is the measured particle mass, C_m,wall is the lost particle mass on the wall, and C_m,AER is the lost particle mass due to ventilation. The corrected particle number and mass concentrations are shown in Fig. 4c. The particle mass concentration that formed was 2.1 times higher than the measured particle mass concentration.

The corrected particle mass concentration of the insecticide was lower than those for air fresheners previously tested under the same experimental conditions, although their increasing trends are similar (Vu et al., 2011b). This difference in particle formation potential between the two kinds of household product is likely to result from differences in composition and emission rates of the different VOCs contained in each product.

Reactions of ozone and VOCs, particularly biogenic hydrocarbons such as terpenes, significantly contribute to secondary organic aerosol formation (McMurry et al., 2004). Several studies have reported the effect of ozone concentration on particle mass concentration (Hoffmann et al., 1997; Coleman et al., 2008; Lamorena and Lee, 2008; Vu et al., 2011a, 2011b). In this study, we investigated the effect of ozone concentration on secondary aerosol formation during reactions between ozone and VOCs emitted from a home insecticide in a test chamber. Three levels of ozone concentration (50, 100, and 200 ppb) were set to simulate real- world environments.

Fig. 5a shows that the nucleation burst was delayed as the injected ozone concentration decreased. Nanoparticle formation occurred at 50, 25, and 17 min for ozone concentrations of 50, 100, and 200 ppb, respectively. Clearly, the reaction between ozone and VOCs significantly contributed to secondary aerosol formation. The peak particle number concentration at 100 ppb was twice that at 50 ppb. However, the peak particle number concentrations at 100 and 200 ppb were similar (16,608 and 16,933 particles/ cm³, respectively). These findings indicate that certain amounts of condensable organic vapors, available to nucleate new particles, were produced after a similar reaction time. When the ozone concentration was increased to 200 ppb, more condensational growth, rather than new particle production, occurred. In this case, the increased ozone concentration promoted condensational growth of particles, thereby generating larger particles that produced a higher particle mass concentration (Fig. 5b). The particle mass concentrations corrected for wall loss and ventilation loss gradually increased during the reactions (Fig. 5c). Particle mass concentrations were 7.3, 33.1, and 40.0 Hg/cm³ at 240 min for the 50-, 100-, and 200-ppb treatments, respectively. The highest particle mass concentration at 240 min was 40.0 Hg/m³ for the 200-ppb treatment, which was lower than that (163.5 Hg/m³, corrected for wall loss and ventilation loss) produced from reactions between ozone and VOCs emitted from an air freshener under the same experimental conditions (Vu et al., 2011b). The ratios of reacted ozone to the corrected particle mass concentration were 1.21, 0.63, and 0.65 for the 50-, 100-, and 200-ppb treatments, respectively. These ratios are higher than those recorded in the air freshener experiment (~0.4) under the same conditions (Vu et al., 2011b). These results indicate that the difference in composition of the VOCs emitted from household products is an important factor determining the potential for secondary aerosol formation through ozone reaction.

Fig. 5. Effect of ozone concentration on particle formation: (a) particle number concentration, (b) measured particle mass concentration, (c) corrected particle mass concentration.

4. Conclusions

Home insecticides are commonly used to prevent or eliminate mosquitoes, houseflies, and cockroaches. The health risks associated with the generation of primary air pollutants from the use of home insecticides should be minimized. Secondary air pollution should also be considered. In this study, we conducted a reaction chamber experiment to investigate the potential for secondary air pollution formation following reactions between ozone and VOCs. Among VOCs, toluene was emitted from the tested home insecticide at the highest concentration. However, the product also contained various biogenic VOCs, such as α-pinene, cymene, d-limonene, α-terpinene, and α-thujone. Ozone consumption rates were maintained at 0.03, 0.09, and 0.10 ppb/min for injected ozone concentrations of 50, 100, and 200 ppb, respectively, during the reaction period. The reacted ozone concentrations were estimated to be 8.8, 20.7, and 26.1 ppb and the aerosol mass concentrations were 7.3, 33.1, and 40.0 Hg/m³ after a 4-h reaction time for the 50-, 100-, and 200-ppb treatments, respectively. These results imply that biogenic VOCs emitted from the test home insecticide may contribute to the production of nanoparticles under low ozone concentrations, similar to those in actual indoor environments. This result will be useful to determine the potential for secondary aerosol formation among household products in the management of indoor air quality. Further studies will be conducted to quantify the VOC species emitted from home insecticides and to determine their effects on aerosol formation during subsequent reactions.

ACKNOWLEDGEMENTS

This study was supported by the Korea Ministry of Environment as “The Eco-technopia 21 project” and by the Korea Institute of Science and Technology.