Species and characteristics of volatile organic compounds emitted from an auto-repair painting workshop

Volatile organic compounds (VOCs) are secondary pollutant precursors having adverse impacts on the environment and human health. Although VOC emissions, their sources, and impacts have been investigated, the focus has been on large-scale industrial sources or indoor environments; studies on relatively small-scale enterprises (e.g., auto-repair workshops) are lacking. Here, we performed field VOC measurements for an auto-repair painting facility in Korea and analyzed the characteristics of VOCs emitted from the main painting workshop (top coat). The total VOC concentration was 5069–8058 ppb, and 24–35 species were detected. The VOCs were mainly identified as butyl acetate, toluene, ethylbenzene, and xylene compounds. VOC characteristics differed depending on the paint type. Butyl acetate had the highest concentration in both water- and oil-based paints; however, its concentration and proportion were higher in the former (3256 ppb, 65.5%) than in the latter (2449 ppb, 31.1%). Comparing VOC concentration before and after passing through adsorption systems, concentrations of most VOCs were lower at the outlets than the inlets of the adsorption systems, but were found to be high at the outlets in some workshops. These results provide a theoretical basis for developing effective VOC control systems and managing VOC emissions from auto-repair painting workshops.

of pollution. In addition, the VOC data obtained are used for profiling, grouping, and analysis of individual compounds, which are then further studied to determine tracer indicators to identify their concentration patterns under different environmental conditions.
Kamal et al. 20 conducted a study assessing the potential health risks to workers, but detailed information on VOCs emitted from auto-repair painting facilities, as a small-scale VOC emission source, has not been provided. VOCs are either directly emitted in large quantities from the outlets of painting facilities or via diffusion from the solvent content in paints 21 . Various paints and complex processes (pretreatment, primer, base coat, top coat, drying, and polishing) exist in auto-repair painting, and some of the VOC emissions are likely to be unaccounted for because of the lack of specific measurements of VOC chemical profiles from the painting facilities. This makes identifying the main VOC emission sources and devising efficient VOC emission control strategies difficult. Hence, a control strategy for reducing VOC emissions from the source facilities, including auto-repair painting facilities, is essential for reducing the ambient O 3 levels as well as SOA formation 22,23 .
To effectively control the indoor and outdoor pollutants emitted from painting facilities, we performed a chemical profile analysis of the VOCs emitted from the main painting workshop in an auto-repair painting facility at Korea. In particular, we compared the VOC species emitted from the use of oil-and water-based paints and analyzed them by grouping into BTEX (benzene, toluene, ethylbenzene, and xylene), ketone, and O 3 precursor groups. In addition, we measured the concentrations of the individual VOCs at the inlet and outlet of the adsorption systems installed in the auto-repair painting booths. This study will help understand the characteristics of VOCs emitted from small-scale auto-repair painting facilities and support development of efficient VOC emission control strategies.

Results
Total VOC (TVOC) concentration: comparison between workshops and interpretation based on the experimental setup. TVOC (sum of TO-14 VOCs) concentrations ranged from 5069 to 8058 ppb in the main workshop and 24-35 VOC species were detected, including precursors of O 3 formation (Supplementary Table S1). Table 1 shows the 16 VOC species that were detected to be emitted in the highest concentrations from the top coat workshop. The observed concentrations of the individual VOCs ranged from 0 to 3378 ppb (butyl acetate had the highest concentration in workplace E); the differences between the workshops were mainly influenced by conditions such as the amount of paint used in each facility, operator characteristics, and type of paint 24 . For all workshops, butyl acetate, toluene, ethylbenzene, m,p-xylene, o-xylene, and 1,2,3-trimethylbenzene were the most commonly detected species, and workshops A and C had lower concentrations of these compounds than the other workshops. Workshops A, C, and E used water-based paint, whereas workshops B and D used oil-based paint. According to Can et al. 25 , higher VOC concentrations are found in painting workshops using oil-based paints than in those using water-based paints. The species composition of VOC emission is detailed in the following sections.

Emission of individual VOCs and their concentrations and proportions: comparison between
workshops and interpretation based on the experimental setup. The VOC mass concentration at the inlet of the adsorption systems varied considerably, possibly due to emissions from previous operations or leaks in the process equipment. The results are expressed as the fraction of each species relative to the TVOC concentration to account for differences in the VOC composition patterns between the workshops. The proportions of the VOCs in each workshop were calculated (Fig. 1). As shown in Fig. 1, when comparing workshops www.nature.com/scientificreports/ using water-based paint (A, C, and E) and oil-based paint (B and D), similar VOC composition patterns were found depending on the type of paint. In workshops A, C, and E, the concentration of butyl acetate was > 50%, while that of all the other compounds was < 10%. On the contrary, in workshops B and D, butyl acetate, m,pxylene, o-xylene, and toluene were evenly detected at concentrations of 10% or higher. Workshops A, C, and E used water-based paint, whereas B and D used oil-based paint; hence, the components and proportions of VOCs emitted from the painting workshops were different, as noted in previous studies 25,26 . In particular, the VOCs emitted from workshop A comprised butyl acetate (65.5%), m,p-xylene (7.8%), 1,2,3-trimethylbenzene (4.9%), m-ethyltoluene (4.2%), ethylbenzene (3.2%), o-xylene (2.9%), and toluene (2.4%). In contrast, the VOCs emitted from workshop B, which used oil-based paint, comprised butyl acetate (31.1%), m,p-xylene (22.7%), o-xylene (12.5%), ethylbenzene (7.6%), 1,2,3-trimethylbenzene (6.5%), m-ethyltoluene (4.0%), and 1,2,4-trimethylbenzene (3.0%). Therefore, we next compared and analyzed the detailed characteristics of VOCs emitted from the use of water-based and oil-based paints.
Analysis of VOCs emitted from painting workshops based on the use of different painting solvents. Information on the type of paint, type of working process, mixing ratio, amount of paint used, and working time of each workshop is summarized in Table 2. The total working time and working process were the  www.nature.com/scientificreports/ same for workshops A and B, but the mixing ratio and amount of paint used were different. In the base coat and top coat workshops, the amount of paint used in workshop A was 847.8 g, which was higher by 325.0 g than that used in workshop B. As for the mixing ratio, the ratio of paint to diluent in the top coat used in workshop A was higher than that used in workshop B; the total working time was the same, but the base coat time in workshop A was 6 min shorter than that in workshop B. The VOC emission compositions of workshops A and B from the main painting workshop are shown in Fig. 2. A similar pattern of VOC species composition was observed, but the range of concentrations was very different. The top five VOCs, in decreasing order, were butyl acetate (3256 ppb), m,p-xylene (389.1 ppb), 1,2,3-trimethybenzene (244.9 ppb), m-ethyltoluene (209.9 ppb), and ethylbenzene (106.4 ppb) in workshop A, and butyl acetate (2449 ppb), m,p-xylene (177.6 ppb), o-xylene (986.4 ppb), ethylbenzene (597.9 ppb), and 1,2,3-trimethybenzene (514.2 ppb) in workshop B. Workshop B had a higher concentration of all VOC species than workshop A, except for butyl acetate. In particular, the concentrations of m,p-xylene and o-xylene were higher in workshop B than in workshop A, which is considered to be a characteristic of oil-based painting, as xylene is the solvent in paints, thinners, and hardeners 12,27 . These results suggested that the amount of VOC emissions and concentrations of individual VOC species differed depending on the type of paint used.

Focus on ketones and BTEX.
Ketones, such as methyl ethyl ketone and methyl isobutyl ketone, are toxic chemicals that are widely used in industrial solvents 28 ; inhalation of their vapors and skin contact with liquid ketones are the main routes of human exposure to these compounds 6 . Several studies have shown BTEX pollutants to have a significant adverse impact on human health and classified them as hazardous because they cause serious pathologic conditions, such as asthma, inflammatory disorders, central nervous system dysfunction, and DNA damage, which can potentially lead to cancer 29,30 . In plants, high O 3 concentrations affect the growth and yield, whereas in human beings it has been known to exert harmful effects on the eyes and respiratory system, resulting in reduced lung capacity, respiratory distress, etc. 31,32 . Therefore, in this study, VOC species emitted from workshops A and B were grouped based on the total concentration of ketones, BTEX, and O 3 precursors. Figure 3a shows a comparison of the ketones in workshops A and B; the concentrations of methyl ethyl ketone and methyl isobutyl ketone in workshop A were 2.554 and 1.082 ppb, respectively, while in workshop B, these were 34.29 and 35.12 ppb, respectively. Comparing the concentration of BTEX (Fig. 3b)  www.nature.com/scientificreports/ www.nature.com/scientificreports/ 3, we inferred that the use of oil-based paint emitted a higher concentration of harmful VOC species than that of water-based paint.  Table S2 shows the concentrations of the 10 main components of VOCs emitted in the booth. The concentrations of the VOCs measured at the outlet were observed to be lower than those of the VOCs emitted from the inlet, but toluene (workshop A inlet: 118.6 ppb, outlet: 294.3 ppb; and workshop B inlet: 192.6 ppb, outlet: 200.2 ppb) was confirmed to exhibit negative efficiencies. This is considered to be due to adsorption systems not being replaced, and detachment of adsorbent or decreased absorption efficiency of activated carbon 21 . Although most VOC species in workshops A and B had a lower outlet concentration than an inlet concentration, it was insufficient to remove the pollutants emitted from the painting workshop.

Discussion
Recently, the Korean government implemented a policy to supply eco-friendly paints to reduce and manage VOCs emitted from paints used in auto-repair painting, which involves switching from oil-based paints to waterbased paints. Therefore, our findings can be used to manage VOC emissions and prepare guidelines to minimize worker exposure to these emissions. In auto-repair painting workshops, workers are exposed to the paint (dermal exposure as well as via inhalation); hence, management is necessary to not only control VOC emissions, but also minimize worker exposure. In particular, the occupational exposure limits for BTEX have been set by the Occupational Safety and Health Administration (OSHA), National Institute for Occupational Safety and Health, and American Conference of Governmental Industrial Hygienists. Based on the OSHA standards, the legal airborne permissible exposure limit-time weighted average for benzene, toluene, ethylbenzene, and xylene is 1, 200, 100, and 100 ppm, respectively, averaged over an 8 h work shift (Supplementary Table S3).
When comparing the concentrations of BTEX emitted from water-based and oil-based paints, the BTEX concentration was higher in workshop B than in workshop A, and in particular, the concentration of m,p-xylene was about 1800 ppb in workshop B. All BTEX compounds, including xylene, did not exceed the exposure limit set by the OSHA, National Institute for Occupational Safety and Health, and American Conference of Governmental Industrial Hygienists (100 ppm, time weighted average). According to the OSHA, the face, eyes, head, hands, and all other exposed parts of the bodies of employees handling such highly volatile paints should be protected using appropriate personal protection equipment for occupational safety and health. However, in the present study, the workers only wore gas masks during painting. Hence, further study should be conducted to lower VOC exposure and improve workers' safety.
Moreover, there is a need for a management plan to improve the efficiency of the adsorption systems installed to control VOC emission into the atmosphere from auto-repair painting facilities. In the future, we intend to investigate detailed characteristics (type of activated carbon, number of prevention facilities, thickness of activated carbon, replacement cycle, maintenance, etc.) of the adsorption systems installed in painting facilities to establish efficient VOC removal in the field.

Conclusion
The concentration trends of VOC species emitted from auto-repair painting workshops in Korea were identified. In addition, we compared the detailed characteristics of the VOC species emitted from water-and oil-based paints and evaluated the VOC removal efficiency of the adsorption systems. The main components of the VOC mixture were butyl acetate, methyl chloride, acrylonitrile, 1-butene, toluene, octane, ethylbenzene, m,p-xylene, o-xylene, n-propylbenzene, m-ethyltoluene, p-ethyltoluene, 1,3,5-trimethylbenzene, o-ethyltoluene, 1,2,3-trimethylbenzene and 1,2,4-trimethylbenzene; butyl acetate, toluene, m,p-xylene, o-xylene and ethylbenzene were present at relatively high concentrations. The total number of VOC species detected was 24-35. The concentrations of most VOCs emitted from oil-based paints were higher than those of VOCs emitted from water-based paints. In particular, the oil-based paints showed high concentrations of m,p-xylene and o-xylene; individual VOC species were also detected in higher proportions and concentrations from the use of oil-based paints than water-based paints. The concentrations of VOCs emitted from the outlet of the adsorption system were lower than those of VOCs emitted from the inlet; however, some workshops showed higher VOC concentrations at the outlet than the inlet. www.nature.com/scientificreports/ Our findings provide a theoretical basis for managing and developing effective VOC control systems to remove the major VOCs emitted from auto-repair painting, and establishing standards for VOC content in paint application-based products. Our study could lead to further establishment of profiles of the main VOC species, which can be used to establish efficient VOC control strategies in the future.

Methods
Experimental set up. Auto-repair painting facilities (workshops A-E) in Seoul were considered in our study. Workshops B and D used oil-based paint, while workshops A, C, and E used water-based paint for autorepair painting. All auto-repair painting facilities were equipped with an adsorption system for VOC control connected to painting booths (Fig. 4a) 33 . The painting operations occurred within the painting booth, and air samples were collected at the inlet and outlet during the main painting workshop to analyze the concentration of individual VOCs (Fig. 4b) 33 . The measurements were conducted throughout the auto-repair painting operations. In general, auto-repair painting operations start with the primer application, followed by base coat and top coat applications, and finally, drying of the automobile surface (Fig. 5).  www.nature.com/scientificreports/ Measurements and analytical method. For the analysis of VOC components, sampling was conducted as per the solid adsorption method using Tenax-TA (40/60 mesh, Markers, USA) 280 mg sorbent tubes of stainless steel. The tubes were heated for 6 h at 300 °C using a TC-20 device (Markes, USA) prior to the measurements. VOC sampling was conducted for 5 min during the painting and drying processes; the air collection rate was 100 mL/min. The flow rate was corrected using a sampling pump equipped with a sorbent tube, before starting the measurement. Tenax-TA sorbent tubes were refrigerated at 4 °C or below and analyzed using GC/ MSD (Agilent HP-6890, USA). As VOC standards, 10 CHEM and 50 ozone precursors were analyzed, and SUPELCO's TO-14 VOC standard mixture (nominal 1 ppm) containing 42 toxic VOCs was used (Supplementary Table S4). Out of a total of 102 chemicals, 88 substances were classified by excluding overlapping substances. The gas chromatography (GC)/mass spectrometry (MS) analysis conditions (Supplementary Table S5) were as follows: secondary thermal desorption from the thermal desorption device; solvent delay performed for 5 min to minimize analysis of the initial low molecular material; initial temperature of the GC oven maintained at 50 °C for 10 min, gradually increased to 220 °C, and maintained for 10 min; and the post run set to 5 min to reduce the contamination of the equipment due to the inflow of polymers other than the material to be measured. The analysis time was approximately 54 min.
To evaluate the linearity of VOC concentrations, a liquid standard (100 μg/mL; Supelco, USA) was added in small amounts at 100, 300, and 500 ng to three Tenax-TA-using adsorbent tube injector systems (Supelco, USA), respectively, and then, after thermal desorption using Autosampler Thermal Desorption (Ultra-xr, Markes), a calibration curve for VOCs was plotted.
Evaluation of the linearity of the calibration curve revealed the coefficient of determination (R 2 ) as 0.99 or higher and the relative standard deviation, which represents reproducibility of the analysis, was evaluated to range from 0.52 to 4.32% as a result of seven repeated analyses.