Chemosphere 129 (2015) 110–117 Contents lists available at ScienceDirect Chemosphere journal homepage: www.elsevier.com/locate/chemosphere A 17-fold increase of trifluoroacetic acid in landscape waters of Beijing, China during the last decade Zihan Zhai a, Jing Wu a,c, Xia Hu a, Li Li a, Junyu Guo a, Boya Zhang b, Jianxin Hu a, Jianbo Zhang a,⇑ a Collaborative Innovation Center for Regional Environmental Quality, State Key Joint Laboratory of Environmental Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, China b School of Public Health, Peking University, Beijing 100191, China c China Waterborne Transport Research Institute, Beijing 100088, China h i g h l i g h t s ! Concentrations of TFA in landscape waters after 10 years are reported. ! Atmospheric deposition is the main contribution route for increased TFA in waters. ! Distribution of TFA in waters is simulated by QWASI model. a r t i c l e i n f o Article history: Received 25 March 2014 Received in revised form 3 September 2014 Accepted 3 September 2014 Available online 26 September 2014 Handling Editor: I. Cousins Keywords: Trifluoroacetic acid Concentration Deposition Water QWASI model a b s t r a c t The concentrations of trifluoroacetic acid (TFA) were measured in urban landscape waters, tap water and snows in Beijing, China in 2012. Compared with the 2002 measurements, a 17-fold increase from 23–98 ng L"1 to 345–828 ng L"1 was observed for TFA concentrations in urban landscape waters, and an obvious increase from not detected (n.d.) to 155 ng L"1 occurred to TFA in tap water. By flux estimation between air and water interface, the remarkable increase of TFA was attributable to dry and wet deposition. The quantitative water–air–sediment interaction (QWASI) model simulated TFAs in various environmental media and showed that, over 99% of TFA distributed in water bodies. Our results recommend that measures are needed to control the increase of TFA in China. ! 2014 Elsevier Ltd. All rights reserved. 1. Introduction As a result of the phase-out of ozone-depleting substances such as chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs), hydrofluorocarbons (HFCs) have been substituted as the main class of chemicals with no potential for ozone depletion. Among them, the atmospheric dry mole fractions of 1,1,1,2-tetrafluoroethane (HFC-134a) shows significant growth (NOAA, 2014). The mixing ratio of HFC-134a increased from #0 parts per trillion (ppt) to #70 ppt in the past 20 years (NOAA, 2013). HFC-134a is mainly used as a refrigerant for mobile air conditioning systems (McCulloch et al., 2003). In the natural atmosphere, reacting with hydroxyl radical (Vasil’ev et al., 2011) and finally transforming into trifluoroacetic acid (CF3COOH, TFA) was revealed as the ⇑ Corresponding author. Tel.: +86 10 62753438; fax: +86 10 62760755. E-mail address: jbzhang@pku.edu.cn (J. Zhang). http://dx.doi.org/10.1016/j.chemosphere.2014.09.033 0045-6535/! 2014 Elsevier Ltd. All rights reserved. major eliminating route of HFC-134a, with the molecular yield of #7–20% (Wallington et al., 1992, 1996). As a kind of perfluorinated acid (PFA), TFA has attracted attention in recent years (Kazil et al., 2011; Russell et al., 2012), attributed to both its existence in a wide range of water bodies at concentrations ranging from 0.5 ng L"1 to 40 900 ng L"1 (Zehavi and Seiber, 1996; Scott et al., 2002), and aquatic toxicity (Davison and Pearson, 1997; Wiegand et al., 2000). Water bodies were regarded as the principal sink of TFA because TFA had low Henry constant [1.1 $ 10"2 Pa m"3 mol"1] (Bowden et al., 1996), and was miscible in all proportions (Feenstra-Bieders and Olthof, 1992). TFA was resistant to physical and chemical degradation (Visscher et al., 1994; Ellis et al., 2001). Microbial degradation may contributed to TFA removal, but the initial results were proven difficult to repeat by experiments (Visscher et al., 1994; Matheson et al., 1996). TFA tended to accumulate in aquatic ecosystem, especially in some seasonal wetlands (Tromp et al., Z. Zhai et al. / Chemosphere 129 (2015) 110–117 1995). The no observed effect concentration (NOEC) of TFA for aquatic ecosystem was 0.1 mg L"1 (Berends et al., 1999). Although the monitored concentrations of TFA in most water bodies (Zehavi and Seiber, 1996; Berg et al., 2000; Scott et al., 2002, 2006; Zeng et al., 2004; Zhang et al., 2005) were below 0.1 mg L"1, some studies suggested that the extensive use of 2,3,3,3-tetrafluoropropene (HFO-1234yf) could dramatically increase the amount of TFA in wetlands (Luecken et al., 2010). There might still be some unknown natural sources in deep waters (Kim and Kannan, 2007; Scott et al., 2010). Simulation studies showed that TFA concentrations would reach 0.1 mg L"1 in seasonal wetlands in 50 years (Tromp et al., 1995). Accumulation of TFA is potential to occur in unique waters, where there is only inflow, deposition and evaporation but with little outflow or groundwater recharge. The phase-out of HFC-134a can potentially lead to the increase of TFA concentration if the alternative substances have higher molecular yields. In nowadays, the use of HFC-134a has been rigorously restricted under Kyoto Protocol due to its high global warming potential value as high as 1430 (Metz et al., 2005; Montzka et al., 2011). 2,3,3,3-Tetrafluoropropene (HFO-1234yf) is regarded as the most promising alternatives of HFC-134a (Leck, 2009; Kajihara et al., 2010; Luecken et al., 2010). However, HFO-1234yf degrades to TFA in the atmosphere with a 100% molecular yield (Hurley et al., 2008), much higher than the 7–20% yield of HFC134a (Wallington et al., 1996). Regional simulations for the degradation of HFO-1234yf were conducted. For instance, there would be 6 to 8-fold increase of TFA concentration in precipitation in Europe by 2020 if HFO-1234yf was substituted for HFC-134a (Henne et al., 2012). Other results also showed that the alternative substances would lead to increased concentrations of TFA in local atmosphere and precipitation (Kajihara et al., 2010; Luecken et al., 2010), and subsequently increase TFA concentrations in surface water (Russell et al., 2012). Therefore, with the increasing consumption of HFO-1234yf, the accumulation of TFA in unique waters will be likely to occur. According to the ‘‘Zhejiang Chemical Engineering Report, 2007’’, HFC-134a is mainly used in automotive air conditioning industry, room air conditioner industry (HFC-407A) and the production of medical aerosols in China. Because room air conditioner by HFC134a is mainly used for export, its emissions occur in the corresponding consumers. Medical aerosols possess a small portion of HFC-134a usage. Therefore, automotive air conditioning industry is the most main emission source of HFC-134a in China. Beijing, the capital of China, is regarded as the HFC-134a emission hotspot because the number of automobiles is up to 5.2 million in 2012 (Beijing Statistical Yearbook, 2013). Based on the monitoring results, the emission and atmospheric concentration of HFC-134a in Beijing is significantly higher than any other cities in China (Fang et al., 2012; Wu et al., 2013). Moreover, evapotranspiration capacity in Beijing is greater than rainfall in most years (Dan et al., 2011). Therefore, the accumulation of TFA is likely to happen in Beijing, especially in urban landscape waters, where there are semi-closed tranquil flow and little evapotranspiration. To carry out continual monitoring of TFA in urban landscape waters is necessary. As far as we know, Zeng et al. (2004) first determined the concentration of TFA in Chinese waters, and Zhang et al. (2005) reported TFA concentrations in some waters of Beijing in 2002. However, the TFA concentrations in these waters have not been updated in the past decade. Studies about source analysis and environmental distribution of TFA are limited. The concentrations of TFA in current is still unknown. In this study, the concentrations of TFA in four urban landscape waters in Beijing were reported and compared with results (Zhang et al., 2005) obtained in the same locations in 2002. Furthermore, the increases of TFA concentration in tap water during the past 111 decade were recorded. Increasing scenario was assumed to explain the increase of TFA concentrations. A water quality model was used to analyze the sources and the contribution proportions of each source. Finally, quantitative water–air–sediment interaction (QWASI) model was used to simulate the transmission rate of TFA in various media, and to determine the distribution characteristics and residence time. To our knowledge, this study is the first time to describe the change of TFA concentrations in the same waters after a decade. Also, by fluxes estimation and model simulation, more information about TFA’s behavior in China is given for further studies. 2. Materials and methods 2.1. Materials and reagents Standard TFA and perfluoropropionic acid (PFPA) were purchased from Acros Organics Co. (Geel, Belgium). The derivatization agent 2,4-difluoroaniline (2,4-DFAn) was obtained from J&K Chemical Co. (Greensboro, GA, USA). N,N0 -dicyclohexylcarbodiimide (DCC) was supplied by the Fluka Chemical Co. (Milwaukee, WI, USA). Organic residue analysis grades of ethyl acetate, hexane, acetone, and methanol were obtained from J.T. Baker Co. (Phillipsburg, NJ, USA). Optima-grade anhydrous sodium sulfate was supplied by Tianjin Jinke Co. (Tianjin, China). Reagent-grade sodium bicarbonate and sodium chloride were obtained from Xilong Chemical Co. (Beijing, China). Silica gel (60 mesh) was purchased from Merck & Co. (Rahway, NJ, USA). The purities of all standards and reagents were >99%. Anhydrous sodium sulfate was baked overnight at 600 "C. Sodium bicarbonate was immersed in methanol for 10 min, and then in ethyl acetate for 10 min. It was tiled in evaporating dish at room temperature overnight, and then baked at 80 "C for 24 h. The same treating processes applied to sodium chloride except that it was baked at 450 "C for 24 h. Silica gel was immersed in dichloromethane for 1 h, evaporated to dryness at room temperature, and then baked overnight at 550 "C. All glassware was washed with acetone and n-hexane in sequence. All water was deionized and further purified twice by distillation. 2.2. Sampling and chemical analysis Fig. 1 shows the sampling sites. Tap water was collected from the Geological Building at Peking University, Beijing (39"590 2600 N, 116"180 3100 E) on July 8, 2012. The landscape waters were sampled from the same locations (Chaoyang Park, Qingnian Lake, Beihai Park, The Summer Palace) as reported previously by Zeng et al. (2004). The samples of landscape waters were collected between July 8, 2012 and July 10, 2012. For details of the water bodies, please see Table S1 in supporting information. Samples of landscape waters were collected directly with 1000-mL glass collectors at a depth of 1.0 m below the surface and a distance of 3.0 m off the lakeshore. The water collectors were prewashed by TFA-free water and sonicated for 1 h, then rinsed three times by ultrapure water in the laboratory. At the sampling sites, the collectors were rinsed three times by lake water before collection. After collection, all samples were immediately stored in 1000-mL polyethylene bottles at 4 "C and transported to the laboratory before analysis. Snow and tap water samples from Peking University were collected for comparative purposes. Snow samples were collected on December 14, 2012 and December 21, 2012, respectively. They were taken above ground immediately after the snowfall ended. After removing topsoil and sundries on the surface the snow samples were stored in 1000-mL glass flasks before analysis. Snow samples were melted at room temperature. 112 Z. Zhai et al. / Chemosphere 129 (2015) 110–117 Fig. 1. Location of sampling sites. All the samples were filtered by qualitative filter paper and pH values were adjusted to 9 before analysis. Vacuum distillation by rotary evaporator was applied to concentrate samples from 1000 mL to 25 mL. Optimum conditions were determined to be a temperature of 35–55 "C, a vacuum of 60–100 k Pa, and a rotation speed of 60–199 r min"1. The temperature was set at 50 "C to ensure an adequate time period with no bumping. The vacuum was set at 60 kPa, and was gradually adjusted to 100 kPa. At the rotation speed of 150 r min"1, a stable state could be maintained over an adequate time period. A gas chromatography-mass spectrometry (GC–MS) (Shimadzu, QP-2010) equipped with an autosampler was applied for analysis. More detailed information has been reported by Wu et al. (2014). 2.3. Quality assurance and quality control Procedural blanks were set along with site samples during analysis. For calibration and recovery tests, ultrapure water spiked with TFA was applied. Briefly, two groups of spiked blank solutions (containing 15.3 and 30.5 ng TFA, respectively) were prepared with three parallel samples in each group. Similarly, 10, 40, 80 lL of TFA standard solution was added to 100-mL water samples from Qingnian Lake. A blank value was subtracted from the sample measurement for calibration. The calibration curve was linear in the range of 0–4910 ng L"1 with a coefficient of R2 P 0.999. Table S2 lists the results of tests using spiked samples. The relative standard deviation (RSD) calculated from all single samples (n = 15) ranged from 4.3% to 7.5%, and the recovery was between 88.2% and 95.1%, indicating good collection efficiency. The paired-sample t test was performed to calculate the recovery of the samples spiked with ultrapure water and the environmental samples, showing that there was no difference in recovery between the samples, indicating a good agreement. 2.4. Water quality model description In some semiarid tourism regions of China, additional water was artificially added into landscape water body annually, for the Table 1 Water volumes of evapotranspiration, leakage, rainfall and supplementary water (2002–2012). Beihai The Summer Palace VE ($105 m3) VL ($105 m3) VR ($105 m3) VS ($105 m3) 1.73 13.40 2.32 17.90 2.99 23.19 1.06 8.19 purpose of maintaining a fixed water level and better scenery. Tap water was used as supplementary water in Beijing. The TFA in this supplementary water would contribute to the increasing levels of TFA. The calculation method for this portion of TFA was shown as follows: From the mass balance Eq. (1) and the principle of the water balance Eq. (2), the amount of TFA supplied by this supplementary water can be calculated. V dC ¼ J $ A þ Q in $ C in " k $ C $ V dt ð1Þ Q in ¼ ½ET þ L " P* $ A 3 ð2Þ where V (m ) is the volume of water in lakes; C (g m ) is the concentration of TFA in water; t (yr) is the time; J (g m"2 yr"1) is the annual average deposition flux; A (m2) is the superficial area of water; Qin (m3 yr"1) is the annual average supplementary water volume; ET, L, P represents the variation in evapotranspiration, bottom leakage, and precipitation with the units of m yr"1, respectively; Cout is the annual leakage volume. It equals ‘‘L $ A’’ with units of m3 yr"1. The concentration in leakage is assumed to be the same as the concentration in the water body; Cin (g m"3) is the TFA concentration in the supplementary water; k (yr"1) is the first-order loss rate and is assumed to be 0 in this study because TFA has a low log Kow and no effective degradation pathways (Boutonnet et al., 1999; Ellis et al., 2001); The annual evapotranspiration capacity of Beijing was 447 mm y"1. For details about the computational formula of evapotranspiration, please see text S1 in "3 Z. Zhai et al. / Chemosphere 129 (2015) 110–117 supporting information. Groundwater recharge was estimated to be 600 mm yr"1 based on the head of the water body and the hydraulic conductivity of the sediment layer; the amount of precipitation in the year 2012 was 733 mm (Beijing Statistical Yearbook, 2013). Therefore Qin = 0.313A. Water volumes of evapotranspiration, leakage, rainfall and supplementary water from 2002 to 2012 were calculated and showed in Table 1. Eq. (3) gives the concentration of TFA in a water body. Here, the TFA concentration in supplementary water is assumed to be consistent with that in the tap water, i.e., Cin = Ctap and V/A = Z. Z (m) is the average depth of the water. ! " A Q C ðtÞ ¼ C 0 þ J $ þ C in $ in t V V ð3Þ However, the increase of water volumes in Qingnianhu Park and Chaoyang Park rely completely on natural precipitation, with no artificial supplementary water. Eq. (4) was used to calculate the concentration of TFA in these two bodies. C ðtÞ ¼ C 0 þ 1 $t Z ð4Þ 2.5. QWASI model description The quantitative water–air–sediment interaction (QWASI) model (Mackay et al., 1983) was used to simulate TFA distribution in water bodies of the Summer Palace. Factors of advection, chemical reaction, diffusion, mass transformation, and dilution were considered. The input parameters of the model include basic information regarding water bodies, chemical properties, and TFA concentrations in different media. Details about migration rate and other input parameters can be found in Table S3 in supporting information (SI). 3. Results and discussion 3.1. TFA concentration in water bodies Fig. 2 shows the TFA concentrations in landscape waters, tap water, and snow in 2012. The TFA concentrations were 643 ± 169 ng L"1, 155 ± 7 ng L"1 and 282 ± 68 ng L"1 in landscape 113 waters, tap water, and snow, respectively. The highest TFA level was observed in landscape water. The TFA concentration in tap water here was lower than that in tap water from 15 other Chinese cities (Wang and Wang, 2011). A more than 10-folds increase was found between our measurement and that in previous studies (Zeng et al., 2004; Zhang et al., 2005) conducted 10 years ago. In 2001–2002, the TFA concentrations were measured to be 23– 107 ng L"1 in landscape water bodies and 148–169 ng L"1 in snow, and TFA in tap water were below the detection limit (Zeng et al., 2004; Zhang et al., 2005). The 2012 TFA concentrations in Beijing were comparable to those reported in other countries (14– 360 ng L"1) (Berg et al., 2000; Scott et al., 2006) and no extreme values were observed. Compared with the concentrations reported by Zhang et al. (2005) and Zeng et al. (2004), the 2012 TFA concentrations in Beijing showed significant increase. There was an average of 12-fold increase in surface water samples, with the maximum of 17fold increase (from 43 ng L"1 to 726 ng L"1) in The Summer Palace. An average of 0.776 fold increase was observed in snow samples. The sharp TFA increase during the past decade may be attributed to the large use of its precursor, HFC-134a. During the last decade, there is a rapid increase in the production and consumption of automobiles in China (Gan, 2003); the total number of registered air-conditioned automobiles using HFC-134a increased from 6 $ 104 in 1995 to 3.67 $ 106 in 2002, and the emissions of HFC-134a were estimated to increase from 7321 tonnes (t) in 2005 to 40 194 t in 2015 (Hu et al., 2010). The regional atmospheric concentration of HFC-134a in China has increased from 23 (parts per trillion by volume) pptv in 2001 to 87 pptv in 2010 (Barletta et al., 2006; Hu et al., 2010). The global annual average background concentration of HFC-134a had increased from 1.7 ppt in 1995 to 57.3 ppt in 2010 (NOAA/ESRL, 2014). More HFC-134a will bring more TFA, therefore, accumulation of TFA is easy to occur in unique aquatic waters when deposition plus inflow effect is greater than outflow effect, especially in stagnant waters. Xiang et al. (2009) set up a multiphase mass transfer model to study the environmental accumulation. Based on their simulations, the accumulation of TFA in the study area would rise up to 100 ng L"1, or even higher than 300 ng L"1 in some small water bodies in 10 years (Xiang et al., 2009). In our study, the concentrations of TFA in 2012 has reached to this level. The increasing rate is worthy of attention. 3.2. Sources of the increase in TFA The cumulative mass of increased TFA (M2012 " M2002) in each landscaped water body (Table S1) was calculated by multiplying the total water volume in individual landscape water body with TFA concentrations, assuming the water area remained constant during the last decade while the depth fluctuated seasonally. The Kunming Lake of the Summer Palace had the highest TFA increment of 4358 ± 1089 g, whereas Qingnianhu Park has the lowest increment of only 70 ± 23 g. Given that there was no obvious point source of emission at the sampling sites, the TFA increment was attributed to the two sources: (a) dry and wet deposition from the atmosphere, and (b) TFA in the supplementary water (supplementary water means water supplied by external source of water to maintain a certain amount of water in lakes or ponds. In Beijing, tap water was used as supplementary water). To separately assess the contribution of these two sources, there are three assumptions: (a) no physical, chemical, or biological degradation of TFA occurred in water; (b) the dry and wet deposition flux were identical across Beijing; (c) the water was mixed uniformly and the concentration recorded was therefore representative of the entire water body. Fig. 2. Comparison of TFA concentration in landscape water, tap water and snow collected in Beijing (2002–2012). 3.2.1. Input of supplementary water The TFA concentration increment in tap water need to be estimated first before assessing the influence of supplementary water. 114 Z. Zhai et al. / Chemosphere 129 (2015) 110–117 Zeng et al. (2004) reported that TFA was below the detection limit in tap water in 2002, whereas this study measured TFA in tap water at a concentration of 155 ng L"1 in 2012. Therefore, a linear growth scenario was assumed to explain the increasing amount of TFA in the supplementary water. In this scenario, it was assumed that TFA concentration in tap water of 2002 equaled to zero. There was a linear growth of the TFA concentration in tap water from 2002 to 2012. Fig. 3 shows the changes in precipitation, deposition flux, and TFA concentration in supplementary water over time under linear growth scenario. The results showed that the contribution of supplementary water accounted for 22.9% and 16% of the total increase in the TFA concentrations over the 10-year period in Beihai Park and Kunming Lake, respectively. The analysis below carried out under the scenario. Dan et al. (2011) reported that water evapotranspiration in Beijing is greater than rainfall for most years. This infers that the water level will decrease gradually for landscaped waters whose influents come merely from the natural precipitation (e.g., Qingnian Lake and Kunming Lake) but with high rates of evapotranspiration. Due to the low volatility (KH = 1.11 $ 10"7 atm m"3 mol"1), TFA is readily accumulated in terminal water bodies (Russell et al., 2012). Therefore, it can be concluded that water evapotranspiration is the main reason for the TFA increase when no artificial supplementary water is provided. Surface runoff following rainfall would contribute some TFA to water bodies, but it can be ignored here when compared to the effect of precipitation. 3.2.2. Dry and wet deposition In general, organic acids are removed from the atmosphere in two ways: reaction with OH radicals and wet and dry deposition (Poisson et al., 2001). The annual dry and wet deposition flux of TFA could be calculated by deposition model (Eq. (5)) (Atkinson, 2000). The average monthly Henry coefficient (KH), the erosion ratio of gas phase (Wg), and the average monthly wet and dry deposition flux (Fwet, Fdry) were calculated using Eq. (6) throughout Eq. (8): lnðK H Þ ¼ 9:328 $ 103 $ ! " 1 1 " " 4:105 Tr T ð5Þ W g ¼ RT=K H ð6Þ F wet ¼ ðC g W g C p W p Þ $ P ð7Þ F dry ¼ V dg C g þ V dp C p ð8Þ where KH (Pa m3 mol"1) is the Henry coefficient, Tr (298.15 K) is the reference temperature; T (K) is the actual temperature; R (8.314 J mol"1 K"1) is the gas constant; Cg and Cp (pg m"3) are monthly-averaged concentrations of gaseous and particulate TFA, respectively; Wg and Wp (m s"1) are the erosion ratios of TFA in the gas and particle phase, respectively; P (mm) is the monthly rainfall; and Fwet and Fdry (lg m"2 yr"1) are the average monthly wet and dry deposition fluxes, respectively. Cg and Cp (pg m"3) are cited from Hu et al. (2013), where TFA in the gaseous phase was recorded at a concentration of 368–5026 pg m"3 and in the particulate phase was present in the range of 106–2421 pg m"3. The total concentration was within the range of 501–7447 pg m"3 and the average concentration was 1.816 ng m"3. The detailed information about parameters and calculation process can be found in Wu et al. (2014). According to the deposition model, annual wet and dry deposition fluxes of TFA were 372 ± 171 lg m"2 yr"1 and 247 ± 93 lg m"2 yr"1, respectively in 2012. The total deposition flux was 619 ± 264 lg m"2 yr"1. Conservatively, a low flux value was adopted here to calculate the contribution of atmospheric deposition to the increased concentration of TFA. The predicted growth of Chinese ownership of cars using HFC-134a from 2005 to 2015 has been given (Hu et al., 2010). Wet and dry deposition scenario was assumed to explain the significant differences in TFA concentrations in landscaped waters. The deposition flux and atmospheric concentration of TFA are correlative. The atmospheric concentration of TFA in Beijing, 2012 was 1.816 ng m"3 (Hu et al., 2013), but the atmospheric concentration of TFA in 2002 was not available. Therefore, assumptions were made. Given that the atmospheric concentration of TFA increased by n folds from 2002 to 2012, the atmospheric concentration of TFA in 2002 is 1/n + 1 fold of that in 2012. According to the deposition model, annual wet and dry deposition fluxes of TFA in 2012 was calculated. In order to estimate the annual wet and dry deposition fluxes during the 10 years, an exponential growth scenario (i.e. y = Aebx, x and y represent year and flux respectively) was assumed. Because of individual differences between water bodies, the values of A and b are not the same. Fig. 3 presents the results of deposition fluxes. For example, for the Kunming Lake in the Summer Palace, the results of the wet deposition flux interannual variability of the wet deposition flux was y = 5E"245e0.2827x, R2 = 0.9972, and for the dry deposition flux it was y = 4E"245e0.2827x, R2 = 0.9985. The total mass of TFA deposited into Beihai and Fig. 3. Changes in precipitation, deposition flux, and TFA concentration in supplementary water over time under linear growth scenario. Z. Zhai et al. / Chemosphere 129 (2015) 110–117 115 Fig. 4. Sketched routes of the simulation results produced by the QWASI model. Kunming Lake in the past decade was 312.44 g and 3800.62 g, respectively, accounting for 79.6% and 87.2% of the average increase in total TFA, respectively. Therefore, inputs from atmospheric deposition was the main reason for the increase of TFA levels in water bodies. Our result is in agreement with the previous studies (Wujcik et al., 1999; Peña et al., 2002; Martin et al., 2003). The increase in TFA concentrations attributable to deposition and supplementary water inputs accounted for 102.5–103.2% of the actual increase, i.e., Dwet and dry + Msupplementray water > M2012 " M2002. This minor difference is acceptable because: (1) the atmospheric TFA concentration data in 2002 is unavailable, here we just simply assumed an exponential increase for wet and dry deposition and linear growth of the TFA concentration in supplementary water. The situation in reality may result in a lower increase than the scenarios assumed; (2) an undulating but increasing trend of precipitation occurred (Fig. 3), with additional 363 nm higher in 2012 than 2002. TFA in the atmosphere is partly removed by wet deposition, and frequent continuous rain will lead to decreased concentrations of TFA. In this study, average values of precipitation and concentration were adopted for calculation, which might lead to the bias of wet deposition to some extent; (3) the surface areas of the waters in each lake is small, with the largest one being only 300 hectares. The atmospheric deposition was mainly designed for simulations on a larger scale, therefore some deviations would be inevitable when applied to smaller regions is inevitable; (4) the leakage occurred in the base of the sandy soil will cause a decrease in the TFA concentrations. In this section only input sources was considered while the loss by leakage was not included. 3.3. Simulation results by the QWASI model We hypothesized that the TFA concentration in the supplementary water was consistent with that in tap water (155 ng L"1) before the simulation by QWASI model. Supplementary water flowing into the Summer Palace was considered to be a direct emission source of TFA. Thus, in this study, the direct emission source input was estimated to be 0.1791 kg y"1, and the atmospheric concentration was set to be 1.816 ng m"3 (Hu et al., 2013). The simulation results (Fig. 4) showed that 99.5% of TFA distributed in water and only 0.5% was in sediment. Our findings are consistent with those of a previous study, where 90.28% of TFA was observed to accumulate in the water while very little was present in sediment (Russell et al., 2012). The predominant distribution of TFA in water is attributed to its high solubility and a low Henry coefficient (Feenstra-Bieders and Olthof, 1992; Kutsuna and Hori, 2008). The transmission rate from the atmosphere to water was 0.7917 kg year"1, while that from water to the atmosphere was 0.0844 kg year"1, the net transmission rate from air to water was, therefore, calculated to be 0.7073 kg year"1. A positive value indicates that the water bodies are a major sink of atmospheric TFA. In a state of mobile equilibrium, the transmission rate between water and sediment was 6.95 kg year"1. The transmission rate for the three sources, TFA in the atmosphere, TFA in supplementary water, and TFA in surface runoff, was 0.7917 kg year"1, 0.1790 kg year"1, and 0.2716 kg year"1, respectively. By the simulation, the residence time of TFA in the ecosystem was 11 710 h. Its persistence and high water solubility indicate that TFA has a significant pollution potential for aquatic ecosystems. Besides, by inputting the concentration data of 2012, the simulation model produced a concentration of 661 ng L"1 in water, close to the measured 726 ng L"1 in Kunming Lake of the Summer Palace. 4. Conclusions This study is the first report of the changes in TFA concentrations in landscaped waters collected in Beijing. Results of the deposition model and the QWASI model indicated that dry and wet deposition was the major contributor for aquatic TFA. The research provided an understanding of how TFA concentrations vary in water bodies over a long period and a theoretical basis for the implementation of a control policy for HFCs. 116 Z. Zhai et al. / Chemosphere 129 (2015) 110–117 Acknowledgements This work was funded by the National Natural Science Foundation of China through Project 41275156. The authors would like to thank Zhi HUANG and Tianyu GAO from Peking University for offering help in the sampling, Yanna LIU from University of Alberta for her constructive comments in English. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chemosphere. 2014.09.033. References Atkinson, R., 2000. Handbook of Property Estimation Methods for Chemicals. Lewis publishers, Boca Raton, FL. Barletta, B., Meinardi, S., Simpson, I.J., Sherwood Rowland, F., Chan, C.Y., Wang, X., Blake, D.R., 2006. Ambient halocarbon mixing ratios in 45 Chinese cities. Atmos. Environ. 40 (40), 7706–7719. Beijing Statistical Yearbook, 2013. (in Chinese). Berends, A.G., Boutonnet, J.C., de Rooij, C.G., Thompson, R.S., 1999. Toxicity of trifluoroacetate to aquatic organisms. Environ. Toxicol. Chem. 18, 1053–1059. Berg, M., Muller, S.R., Muhlemann, J., Wiedmer, A., Schwarzenbach, R.P., 2000. Concentrations and mass fluxes of chloroacetic acids and trifluoroacetic acid in rain and natural waters in Switzerland. Environ. Sci. Technol. 34, 2675–2683. Boutonnet, J.C., Bingham, P., Calamari, D., de Rooij, C., Franklin, J., Kawano, T., Libre, J.M., McCulloch, A., Malinverno, G., Odom, J.M., Rusch, G.M., Smythe, K., Sobolev, I., Thompson, R., Tiedje, J.M., 1999. Environmental risk assessment of trifluoroacetic acid. Hum. Ecol. Risk. Assess. 5, 59–124. Bowden, D.J., Clegg, S.L., Brimblecombe, P., 1996. The Henry’s law constant of trifluoroacetic acid and its partitioning into liquid water in the atmosphere. Chemosphere 32, 405–420. Dan, L., Yang, F., Wu, J., 2011. Urbanization effects on the variation of pan evaporation in Beijing during 1960–2009. J. Meteor. Sci. 31 (4), 405–413 (In Chinese). Davison, A.W., Pearson, S., 1997. Toxicity of TFA to plants. Final Report to AFEAS (Project SP91–1823/BP96–31). University of Newcastle. Ellis, D.A., Hanson, M.L., Sibley, P.K., Shahid, T., Fineberg, N.A., Solomon, K.R., Muir, D.C.G., Mabury, S.A., 2001. The fate and persistence of trifluoroacetic and chloroacetic acids in pond waters. Chemosphere 42 (3), 309–318. Fang, X., Wu, J., Su, S., Han, J., Wu, Y., Shi, Y., Wan, D., Sun, X., Zhang, J., Hu, J., 2012. Estimates of major anthropogenic halocarbon emissions from China based on interspecies correlations. Atmos. Environ. 62, 26–33. Feenstra-Bieders, G.C., Olthof, J.A., 1992. Determination of the solubility of trifluoroacetic acid and sodium trifluoroacetate in various aqueous media. Solvay Duphar. Report 56630/182/91. Gan, L., 2003. Globalization of the automobile industry in China: dynamics and barriers in greening of the road transportation. Energy policy 31 (6), 537–551. Henne, S., Shallcross, D.E., Reimann, S., Xiao, P., Brunner, D., O’Doherty, S., Buchmann, B., 2012. Future emissions and atmospheric fate of HFC-1234yf from mobile air conditioners in Europe. Environ. Sci. Technol. 46 (3), 1650– 1658. Hu, J., Wan, D., Li, C., Zhang, J., Yi, X., 2010. Forecast of consumption and emission of HFC-134a used in the mobile air-conditioner sector in China. Adv. Climate Change. Res. 1 (1), 20–26. Hu, X., Wu, J., Zhai, Z., Zhang, B., Zhang, J., 2013. Determination of gaseous and particulate trifluoroacetic acid in atmosphere environmental samples by gas chromatography–mass spectrometry. Chin. J. Anal. Chem. 41 (8), 1140–1145. Hurley, M.D., Wallington, T.J., Javadi, M.S., Nielsen, O.J., 2008. Atmospheric chemistry of CF3CF@CH2: products and mechanisms of Cl atom and OH radical initiated oxidation. Chem. Phys. Lett. 450, 63–267. Hideo Kajihara, Kazuya Inoue, Kikuo Yoshida, Ryuichi Nagaosa, 2010. Estimation of environmental concentrations and deposition fluxes of R-1234-YF and its decomposition products emitted from air conditioning equipment to atmosphere. In: 2010 International Symposium on Next-generation Air Conditioning and Refrigeration Technology. Tokyo, Japan. Kazil, J., McKeen, S.A., Kim, S., Ahmadov, R., Grell, G.A., Talukdar, R.K., Ravishankara, A.R., 2011. WRF/chem study of dry and wet deposition of trifluoroacetic acid produced from the atmospheric degradation of a few short-lived HFCs. AGU Fall Meet. Abstr. 1, 08. Kim, S.K., Kannan, K., 2007. Perfluorinated acids in air, rain, snow, surface runoff, and lakes: relative importance of pathways to contamination of urban lakes. Environ. Sci. Technol. 41 (24), 8328–8334. Kutsuna, S., Hori, H., 2008. Experimental determination of Henry’s law constants of trifluoroacetic acid at 278–298 K. Atmos. Environ. 42 (7), 1399–1412. Leck, T.J., 2009. Evaluation of HFO-1234yf as a potential replacement for R-134a in refrigeration applications. In: Proceedings 3rd Conference on Thermophysical Properties and Transfer Processes of Refrigerants, IIR and NIST, Boulder, Colorado June 26. Luecken, D.J., Waterland, R.L., Papasavva, S., Taddonio, K.N., Hutzell, W.T., Rugh, J.P., Andersen, S.O., 2010. Ozone and TFA impacts in north America from degradation of 2,3,3,3-tetrafluoropropene (HFO-1234yf), a potential greenhouse gas replacement. Environ. Sci. Technol. 44 (1), 343–348. Mackay, D., Joy, M., Paterson, S., 1983. A quantitative water, air, sediment interaction (QWASI) fugacity model for describing the fate of chemicals in lakes. Chemosphere 12, 981–997. Martin, J.W., Mabury, S.A., Wong, C.S., Noventa, F., Solomon, K.R., Alaee, M., Muir, D.C.G., 2003. Airborne haloacetic acids. Environ. Sci. Technol. 37, 2889–2897. Matheson, L.J., Guidetti, J.R., Visscher, P.T., Schaefer, J.K., Oremland, R.S., 1996. Summary of research results on bacterial degradation of trifluoroacetate (TFA). USGS Open File Rep., 96–219. McCulloch, A., Midgley, P.M., Ashford, P., 2003. Releases of refrigerant gases (CFC12, HCFC-22 and HFC-134a) to the atmosphere. Atmos. Environ. 37 (7), 889– 902. Metz, B., Kuijpers, L., Solomon, S., Andersen, S.O., Davidson, O., Pons, J., de Jager, D., Kestin, T., Manning, M., Meyer, L., 2005. IPCC/TEAP Special Report on Safeguarding the Ozone Layer and the Global Climate System: Issues Related to Hydrofluorocarbons and Perfluorocarbons. IPCC/TEAP, United Kingdom. Montzka, S.A., Reimann, S., Engel, A., Kruger, K., O’Doherty, S., Sturges, W.T., Blake, D., Dorf, M., Fraser, P., Froidevaux, L., Jucks, K., Kreher, K., Kurylo, M.J., Mellouki, A., Miller, J., Nielsen, O.J., Orkin, V.L., Prinn, R.G., Rhew, R., Santee, M.L., Stohl, A., Verdonik, D., 2011. Scientific Assessment of Ozone Depletion: 2010. Global Ozone Research and Monitoring Project. World Meteorological Organization, Geneva, Switzerland. NOAA/ESRL/GMD Flask Sampling Network, 2013. National Oceanic and Atmospheric Administration-Earth System Research Laboratory. (last access 06.05.13). NOAA, 2014. (last access 21.02.14). NOAA/ESRL Global Monitoring, 2014. (last access 27.08.14). Peña, R.M., Garcı́a, S., Herrero, C., Losada, M., Vázquez, A., Lucas, T., 2002. Organic acids and aldehydes in rainwater in a northwest region of Spain. Atmos. Environ. 36 (34), 5277–5288. Poisson, N., Kanakidou, M., Bonsang, B., Behmann, T., Burrows, J.P., Fischer, H., Gölz, C., Harder, H., Lewis, A., Moortgat, G.K., Nunes, T., Pio, C.A., Platt, U., Sauer, F., Schuster, G., Seakins, P., Senzig, J., Seuwen, R., Trapp, D., Volz-Thomas, A., Zenker, T., Zitzelberger, R., 2001. The impact of natural non-methane hydrocarbon oxidation on the free radical and ozone budgets above a eucalyptus forest. Chemosphere – Global Change Sci. 3 (3), 353–366. Russell, M.H., Hoogeweg, G., Webster, E.M., Ellis, D.A., Waterland, R.L., Hoke, R.A., 2012. TFA from HFO-1234yf: accumulation and aquatic risk in terminal water bodies. Environ. Toxicol. Chem. 31 (9), 1957–1965. Scott, B.F., Spencer, C., Marvin, C.H., MacTavish, D.C., Muir, D.C.G., 2002. Distribution of haloacetic acids in the water columns of the Laurentian Great Lakes and Lake Malawi. Environ. Sci. Technol. 36 (9), 1893–1898. Scott, B.F., Spencer, C., Mabury, S.A., Muir, D., 2006. Poly and perfluorinated carboxylates in north American precipitation. Environ. Sci. Technol. 40, 7167– 7174. Scott, B.F., De Silva, A.O., Spencer, C., Lopez, E., Backus, S.M., Muir, D.C.G., 2010. Perfluoroalkyl acids in Lake Superior water: trends and sources. J. Great Lakes Res. 36, 277–284. Tromp, T.K., Ko, M.K.W., Rodriguez, J.M., Sze, N.D., 1995. Potential accumulation of a CFC-replacement degradation product in seasonal wetlands. Nature 376, 327– 330. Vasil’ev, E.S., Knyazev, V.D., Savelieva, E.S., Morozov, I.I., 2011. Kinetics and mechanism of the reaction of fluorine atoms with trifluoroacetic acid. Chem. Phys. Lett. 512 (4–6), 172–177. Visscher, P.T., Culbertson, C.W., Oremland, R.S., 1994. Degradation of trifluoroacetate in oxic and anoxic sediments. Nature 370 (6488), 391. Wallington, T.J., Hurley, M.D., Ball, J.C., Kaiser, E.W., 1992. Atmospheric chemistry of hydrofluorocarbon 134a: fate of the alkoxy radical 1,2,2,2-tetrafluoroethoxy. Environ. Sci. Technol. 26, 1318–1324. Wallington, T.J., Hurley, M.D., Fracheboud, J.M., Orlando, J.J., Tyndall, G.S., Sehested, J., Mogelberg, T.E., Nielsen, O.J., 1996. Role of excited CF3CFHO radicals in the atmospheric chemistry of HFC-134a. J. Phys. Chem. 100, 18116–18122. Wang, Q., Wang, X., 2011. Preliminary study of TFA in surface water of some cities over China. Environ. Sci. 32 (11), 3223–3228 (in Chinese). Wiegand, C., Pflugmacher, S., Giese, M., Frank, H., Steinberg, C., 2000. Uptake, toxicity, and effects on detoxication enzymes of atrazine and trifluoroacetate in embryos of zebrafish. Ecotoxicol. Environ. Saf. 45 (2), 122–131. Wu, J., Fang, X., Xu, W., Wan, D., Shi, Y., Su, S., Zhang, J., 2013. Chlorofluorocarbons, hydrochlorofluorocarbons, and hydrofluorocarbons in the atmosphere of four Chinese cities. Atmos. Environ. 75, 83–91. Wu, J., Martin, J., Zhai, Z., Lu, K., Li, L., Fang, X., Jin, H., Hu, J., Zhang, J., 2014. Airborne trifluoroacetic acid and its fraction from the degradation of HFC-134a in Beijing, China. Environ. Sci. Technol. http://dx.doi.org/10.1021/es4050264. Wujcik, C.E., Cahill, T.M., Seiber, J.N., 1999. Determination of trifluoroacetic acid in 1996–1997 precipitation and surface waters in California and Nevada. Environ. Sci. Technol. 33 (10), 1747–1751. Xiang, X.W., Zheng, G., Sheng, Z.Q., Wu, P., 2009. Multiphase mass transfer model on the accumulation of TFA in regional environments. Res. Environ. Sci. 22 (9), 1109–1112. Z. Zhai et al. / Chemosphere 129 (2015) 110–117 Zehavi, D., Seiber, J.N., 1996. An analytical method for trifluoroacetic acid in water and air samples using headspace gas chromatographic determination of the methyl ester. Anal. Chem. 68, 3450–3459. Zeng, Z., Zhang, J., Hu, J., Liu, X., Zhu, Y., Sun, Z., 2004. Determination of trifluoroacetic acid in surface water of Beijing. Res. Environ. Sci. 17 (5), 64–67 (in Chinese). 117 Zhang, J., Zhang, Y., Li, J., Hu, J., Ye, P., Zeng, Z., 2005. Monitoring of trifluoroacetic acid concentration in environmental waters in China. Water Res. 39 (7), 1331– 1339. Zhejiang Chemical Engineering Report, 2007 (in Chinese).