Distribution and pollution characteristics of organophosphate esters: reflected by tree rings of arbor species

Organophosphate esters (OPEs) are emerging pollutants. Currently, research on OPEs in tree rings is still limited. In this study, tree rings of five arbor species from Sichuan Province, China, were sampled to study the occurrence and distribution of six OPEs, which were quantitatively analyzed by gas chromatography–mass spectrometry (GC–MS). The total concentrations of OPEs in all samples ranged from 189.79 (Fir species) to 341.23 ng/g (Toona sinensis), with average concentration of 284.77 ± 46.66 ng/g. So, arbor could be used as good passive samplers for OPEs. The levels of OPEs among five arbor species showed no significant difference (p = 0.668 > 0.05), suggesting that the pollution status of OPEs in a region or country could be roughly assessed by any arbor tree species. In this study area, tris(2-butoxyethyl) phosphate (TBEP) was the dominant OPEs followed by tri(2-chloroethyl) phosphate (TCEP). Tris(2-ethylhexyl) phosphate (TEHP) and tri-n-butyl phosphate (TnBP) showed relatively stable concentrations in each arbor species, while the other four OPEs including TBEP, triphenyl phosphate (TPhP), tri(chloropropyl) phosphate (TCPP) and TCEP had significantly different concentrations. Interestingly, the absorption and accumulation of OPEs by tree rings of arbor species were quite different from that of inorganic elements reported by other studies.

Dendrochemistry is based on the fact that elements in the environment can be transported through plant tissues under the physiological activity of trees and accumulate in the xylem. Their existence is relatively stable, and no more obvious migration occurs. Therefore, the use of chemical element content in the annual ring can trace the history of environmental pollution and the migration characteristics of elements in the environment. Many studies have shown that tree rings can serve as a useful passive air sampler to monitor the long-term trends of pollutants, such as inorganic pollutants, heavy metals and PAHs, and reveal the historical changes in the bioavailability level of polluting elements in the environment (Kuang et al., 2011;Rauert et al., 2020;Wang et al., 2021). Contaminants can enter tree rings in many ways (Wen et al., 2004), and tree rings of different age can truly reflect the environmental information of that year (Yang et al., 2011). Therefore, the study of tree rings can reproduce the high-resolution history of environmental pollution and determine the source and spatial path of pollutants entering the environment as well (Scanlon et al., 2020). However, large number of studies have focused on heavy metals in tree rings (He et al., 2021;Xu, 2005;Yin et al., 2011), while the research on trace organic pollutants, such as OPEs, is almost blank. Whether a new method for evaluating the pollution history of emerging pollutants in various regions could be proposed? Based on this hypothesis, this study was devoted to use tree rings to reflect the pollution characteristics of emerging organic pollutant-OPEs in the study area, and explored whether there are significant differences in OPEs pollution reflected by different tree species.
As a commonly used passive sampler, arbor has the characteristics of wide distribution, easy availability, high resolution, large time span and strong continuity. Compared with shrub growth rings, arbor growth rings are generally wider with clear boundaries and are easy to distinguish. When compared with other plants commonly used in biological monitoring (such as lichens and mosses), arbor growth rings have a clear temporal trend. Therefore, arbor was selected as the research object. In addition, OPEs only have anthropogenic sources, and tree rings over 5 years can fully provide the information on the use and pollution history of OPEs in the region (Yin et al., 2011). Therefore, monitoring the pollution level and distribution characteristics of OPEs in tree rings are conducive to inferring the pollution degree and profile of OPEs as well as providing basic data for the study of organic pollution in this area. It can take a good large area sample and evaluate the pollution level of certain POPs in a large area.
In this study, five arbor species were selected in a township far from industrial point sources were selected and the objectives are to: (a) observe if the

Sample collection
Liang'an Town is located in Ziyang City, Sichuan Province, China which is in a temperate region, 46 km away from the county seat and 110 km away from Chengdu. The trees here grow faster in spring and summer to form loose xylem (earlywood) and slower in autumn and winter to form dense xylem (latewood). The alternate distribution of loose and dense ring makes it easy to distinguish tree rings. At the same time, Liang'an town is a typical traditional agricultural town dominated by agriculture, animal husbandry and aquaculture activities. There are no industrial facilities, so the impact of OPEs point source pollution could be neglected. Therefore, the forest farm in Liang'an town (30° 30′ 28.35″ N 104° 56′ 34.43″ E) was a good proxy for typical pollution in the area and was selected as the sampling site ( Fig. 1), and five common perennial arbor tree species: Toona sinensis, Cypress, Fir, Pterocarya stenoptera and Dongmu were selected as the research objects. They were in good growth condition, with similar tree ages. All of them were 8-10 years old, and the number of tree rings could meet the analysis needs of the latest 8 years.
After the collected wooden piles being air dried, the bark was removed and the rough surface of the tree disk was polished smoothly to distinguish the growth rings. By taking a complete circle of earlywood and latewood near the outermost layer as the first ring (the samples were collected in 2020, so the latest ring was formed in 2019). Every two growth rings from sapwood to heartwood were segmented for one study period, and the samples were numbered from 1 to 4 by year. More details of sampling are illustrated in Table 2.
Ultrapure water was obtained from Milli-Q equipment.

Sampling and pretreatment
The tree samples selected in the wood processing plant should be of different types, with straight trunks and clear growth rings. According to the workers, all the trees were cut in the local artificial timber forest. Three samples for each tree species were selected which were all about 5 cm thick disks. The samples were wrapped in aluminum foil, transported back to the laboratory and stored at − 20 °C. Two parallel samples for each disk were analyzed. In the process of sample preparation, the circular wooden pile was divided into fan-shaped areas with an electric saw. Considering the difference in growth speed between the north and south sides of the tree, a fan-shaped area of the same size was taken from the sparse surface (sunny surface) and dense surface (shady surface), respectively, so as to ensure the preciseness. The divided piles were mixed and crushed with an electric grinder and screened through 100 mesh; then, the samples were wrapped in tin foil paper and labeled.
1.0000 g sample was accurately weighed and added into the test tube, soaked in 20 mL ethyl acetate: acetone (3:2 V/V) for 12 h, and then, ultrasonic extracted for 30 min twice. The extraction solution was combined and centrifuged at 3000 r/min. The supernatant was transferred to a concentration bottle and concentrated to about 1 mL with a vacuum concentrator. 5-10 mL n-hexane was added for solvent conversion, and the supernatant continued concentrating to about 1 mL again. Then, it was separated and purified by silica gel/alumina/anhydrous sodium sulfate chromatography column activated with n-hexane. After 10 mL of n-hexane was added for removing the co-extracts, the target compounds were eluted with 20 mL of ethyl acetate: acetone (3:2 V/V), and the eluent was concentrated to 200 μL.

Instrumental analysis
The GC was equipped with a capillary column Rti-5MS (30 m × 0.25 mm × 0.25 µm), with a 280 °C inlet temperature using splitless injection. The MS source was electron impact (EI) and operated in selected ion monitoring (SIM) mode. Helium was used as a carrier gas with a flow rate of 1.00 mL/ min. The GC oven temperature was held at 50 °C for 1 min, increased to 200 °C at 15 °C/min and held for 1 min, increased to 250 °C at 4.00 °C/min and then, increased to 260 °C at 20 °C/min and held for 4 min. The interface temperature was 280 °C, and the ion source temperature was 200 °C.
Quality assurance and quality control (QA/QC) Based on the related concentration in actual samples, seven-point calibration curves (0.50, 1.00, 5.00, 10.00, 50.0,0 100.00 and 500.00 ng/mL) were created using a mixed standard solution of six OPEs to quantify OPEs, and the R 2 of each OPEs monomer was greater than 0.99. According to the target concentration in the samples, two spiked concentrations (50.00 ng/mL, 500.00 ng/mL), three times in parallel for each level and the recoveries of each monomer were in the range of 75.90%-115.81%. In addition, surrogate standard (TPhP-d 15 : 100 ng) was added to each sample to determine the recoveries, and the recoveries ranged from 63.27 to 83.84%. The method detection limits (MDLs) were derived from three times the signal-to-noise ratio (S/N = 3), and method Quantification Limits (MQLs) were derived from ten times the signal-to-noise ratio (S/N = 10). The MDLs and MQLs of OPEs were in the range of 0.02-0.40 ng/g and 0.09-1.62 ng/g, respectively. The instrument precision ranged from 1.9 to 8.3%. A parallel sample was added for every 10 samples, and a solvent blank (100% n-hexane) was run every 20 samples to check the carryover and background contamination. Only TBEP was detected in the blank experiment, and its maximum concentration was lower than 5% of the sample concentration. More information (about quantification curves, R 2 , recoveries, MDLs and MQLs) has been attached to supplementary material (Table S1).
The pictures in this paper were drawn by Origin 2017. Statistical analysis was performed with SPSS Statistics version 25 for Windows. Correlations between OPEs monomers concentrations in five tree species were analyzed by Pearson correlation coefficients. Nonparametric test was used to analyze the difference between Σ 6 OPEs concentrations of five tree species. In order to explore the distribution differences of each monomer in five tree species, multivariate analysis was used for multiple comparison.

Concentration of OPEs in tree rings
As shown in Fig. 2 (see Table 2 for more sample number details), the detection rate of six OPEs in all samples was 100%, indicating that OPEs pollution was widespread in trees in the sampling area. High potential of long-range atmospheric transport and persistence of OPEs would be responsible for their presence in areas where there are few pollution sources . In addition, even though their reported half-life were short at atmosphere, OPEs could bind with air particles, which greatly enhance their persistence in the environment (Wei et al., 2015). This can explain the high-detected frequencies of each individual OPEs in present study. Σ 6 OPEs concentration in the tree rings ranged from 189.79 to 341.23 ng/g, with an average of 284.77 ± 46.66 ng/g. The order of Σ 6 OPEs concentration in various arbor tree species was: Toona s ine nsi s( 320.64 ± 13.63 ng/g) > Cy pre ss (301.46 ± 33.28 ng/g) > Pterocarya stenoptera (275.26 ± 48.53 ng/g) > Dongmu (274.12 ± 40.44 ng/g) > Fir (264.70 ± 50.82 ng/g). The concentration of OPEs in different arbor tree species in the sampling area was all in the hundreds of ng/g level, so the arbor trees have a good absorption of atmospheric OPEs, which can be used to evaluate the OPEs pollution in a region. Compared with OPEs concentrations in leaves reported by other studies, the pollution degree of Σ 6 OPEs in tree rings in this study could not be ignored. For example, Santos et al., (2020) reported OPEs concentration in leaves of Bitter orange in Seville (TPhP: 47.5-265 ng/g, TCPP: 321-1058 ng/g) and Chen et al., (2019) reported that the concentration of Σ 7 OPEs in leaves of Magnolia and Loquat was 191.3 ng/g and 83.1 ng/g, respectively. Moreover, OPEs in different arbor tree species were at the same level, indicating that different arbor tree species had little effect on OPEs concentration. The significant differences in the tree ring levels of OPEs among five arbor species were examined using multivariate analysis (MANOVA). The results showed that there was no significant difference in the tree ring levels of OPEs among five arbor species (p = 0.668 > 0.05). Therefore, different types of arbors can reflect the pollution degree of local OPEs. In terms of the average concentration of Σ 6 OPEs over the last 8 years, OPEs in the Toona sinensis had the highest Σ 6 OPEs concentration which was 1.2 times that of the Fir with the lowest concentration. This may be because stoma and lenticel are the main pathways for organic pollutants to enter the xylem. The larger the leaf size and the more the number of stomata and lenticels, the easier it is for organic matter to enter in the growth rings (Kuang et al., 2011).
Compared with other literatures, Xu, (2005) studied the accumulation of 12 inorganic elements in five trees in Xiangshan, Beijing and found that there were significant differences in the absorption and accumulation of inorganic elements among different tree species . Wu, (2017) found that a single tree species had different absorption capacities for different heavy metal elements by comparing the concentration of eight heavy metal elements in five tree rings, and the ability of five trees absorbing the same heavy metal element was also different. Xue et al., (2016) measured the accumulation of Zn, Cu, Pb, Cd, Hg and As in the trunk of seven plantations in Harbin and found that there were differences in the comprehensive accumulation ability of different plantations to heavy metals. Therefore, the absorption and accumulation of arbors to inorganic elements was inconsistent with that of organic pollutants in this paper. Compared with the data reported by Odabasi et al., (2015) that the concentration of other POPs in tree rings in industrial areas (Σ 16 PAHs: 2526 ng/g, Σ 41 PCBs: 15.6 ng/g, Σ 32 PCNs: 2.35 ng/g, Σ 7 PBDEs: 16.65 ng/g) was higher than that in background areas (Σ 16 PAHs: 1767 ng/g, Σ 41 PCBs: 6.7 ng/g, Σ 32 PCNs: 1.31 ng/g, Σ 7 PBDEs: 6.51 ng/g) in Aliaga, Turkey, the concentration of OPEs in this study was higher than many POPs (PCBs, PBDEs, PCNs, etc.), suggesting that OPEs pollution should be paid attention to and could be used as an excellent indicator of air pollution. Therefore, arbor are suitable species to be used to roughly compare the use and pollution level of OPEs between different regions or countries.

Distribution of OPEs in tree rings
The distribution of OPE monomers in samples from five arbor species is shown in Fig. 3. The concentration of TBEP (73.77-184.57 ng/g) in each sample was the highest, accounting for 31.60-61.33% of Σ 6 OPEs, followed by TCEP, TCPP, TPhP and TnBP, respectively, and TEHP was the lowest covering 2.11-6.12% of Σ 6 OPEs. In order to explore the distribution differences of each OPE monomer in five arbor tree species, Friedman test of SPSS Statistics version 25 was used for multiple comparisons of six OPEs in samples. p = 0.1027 > 0.05 indicated that there was no significant difference in OPEs composition distribution among five tree species. Therefore, the trunk rings of different tree species collected in this study have no obvious selectivity to the accumulation of OPEs, suggesting that different types of arbor species can be used to indicate OPEs contamination in regional environment.
According to different substituents, the six OPEs can be divided into alkyl phosphate (TnBP, TBEP, TEHP), chlorinated phosphate (TCEP and TCPP) and phenyl phosphate (TPhP). The proportion of different kinds of OPEs in this five arbor species was alkyl phosphate(680.03 ± 89.00 ng/g) > chlorinated phosphate(337.49 ± 42.13 ng/g) > phenyl phosphate (129.26 ± 34.98 ng/g) (Fig. 4). The accumulation potential of all kinds of arbor species to alkyl phosphate was the highest which presented that the alkyl phosphate was the dominant OPEs in this area. The profile of OPEs in Fir was alkyl phosphate: chlorinated phosphate: phenyl phosphate≈5:4:1, while the general profile of OPEs in the other four tree rings was alkyl phosphate: chlorinated Phosphate: phenyl phosphate≈6:3:1, and the concentration of chlorinated phosphate in Fir (396.20 ng/g) was also higher than that in the other four arbor species (Toona sinensis: 349.04 ng/g > Cypress: 340.55 ng/g > Pterocarya stenoptera: 321.10 ng/g > Dongmu: 280.56 ng/g). It showed that among the various arbor species, the Fir had higher enrichment ability for chlorinated phosphate. However, in general, the profile of OPEs in different arbor species were similar which further prove that arbors are suitable species to be used to compare the pollution of OPEs between different regions or countries.

Inter-annual variation characteristics of OPEs in tree rings
In all kinds of trees, the variation trend of Σ 6 OPEs concentration with years was different (Fig. 5). Among them, the concentration of Σ 6 OPEs in Toona sinensis and Dongmu decreased year by year, and there was an obvious linear relationship between Σ 6 OPEs concentration and year (Toona sinensis: p = 0.99 Dongmu: R 2 = 0.92). The difference was that the Σ 6 OPEs concentration decreased faster in Dongmu than in Toona sinensis. However, the Σ 6 OPEs concentration in Cypress basically showed a slowly increasing trend while Σ 6 OPEs concentration in Fir and Pterocarya stenoptera fluctuated greatly which showed no obvious unidirectional trend. For Fir, the concentration of Σ 6 OPEs increased year by year in the first three groups and reached the highest value (341.23 ng/g) in the third group (2015)(2016), then decreased by about 1/3 in the fourth group (2017)(2018). For Pterocarya stenoptera, the concentration of Σ 6 OPEs in the first three groups fluctuated greatly, with a fluctuation range of 189.79-314.82 ng/g (inter-group variation: 9.92-66.14%). For Cypress, the concentration of Σ 6 OPEs fluctuated in the range of 254.00-329.30 ng/g (inter-group variation: 4.47-23.62%), and the maximum value (329.33 ng/g) appeared in the second group (2014)(2015), while the fourth group had a significant decline, which was 23.62% lower than that in the third group. The results showed that the sensitivity of different tree species to OPEs in the same area was different, which may be due to the different leaf size and lipid content of different tree species. Therefore, to better understand the changes of OPEs pollution in different years, the changes of OPEs in different tree ring samples could be further analyzed, and more detailed results could be obtained by combining the physical and chemical properties of OPEs. The variation characteristics of OPEs monomers in different tree species varied greatly with the growth of rings (Fig. 6). Among them, TBEP concentration showed the greatest influence on the variation of Σ 6 OPEs in different tree species in the sampling area. TBEP belongs to alkyl OPEs, which has a short half-life in the atmosphere and is easily biodegraded. However, the content of TBEP was highest in different tree species, indicating that there was continuous non-point source input in this area. The concentrations of TEHP and TnBP were relatively stable, while the concentrations of the other three monomers fluctuated significantly. Compared with other OPE components, TnBP and TEHP have higher logK OA (TnBP: 9.21, TEHP: 14.90) and BCF values (TnBP: 1.03 × 10 3 , TEHP: 1.00 × 10 6 ) ( Table 1), so they were easier to be enriched in trees. Moreover, the half-life of these two compounds (TnBP: 3.26, TEHP: 2.62) were short, and they could not exist in the environment for a long time (Li et al., 2018). Therefore, similar with TBEP, their inter-annual variation characteristics also indicated that there was continuous and stable pollution input in the local area. In addition, TnBP is often added as plasticizer and additive to varnishes, concrete and plastics and is widely used in construction and agricultural production (Wang et al., 2022). The study area in this paper mainly relies on agricultural development, so sewage irrigation and plastic film application in agricultural activities may be important sources of local OPEs pollution (https:// dili. chazi dian. com/ baike-46492/). As mentioned above, the accumulation of OPEs did not differ significantly among the five arbor species. However, the fluctuation of each monomer in Fir was relatively obvious, indicating that Fir was more sensitive to the change of OPEs monomer. According to other studies, Conifer xylem has a high concentration of organic matter (such as Turpentine and Rosin), which showed a strong ability to absorb organic matter from the atmosphere and preserved them over time (Kuang et al., 2011). Therefore, the adaptability of tree species should be considered when selecting trees to reflect the bioavailability and historical changes of pollutants in the environment.

Conclusions
Hundreds ng/g of OPEs were detected in tree rings of the five arbor species, and there was no significant difference of the total concentrations of OPEs, indicating that arbor could be used as a good passive sampler to show the occurrence and distribution of emerging pollutants with similar physicochemical properties to OPEs in the regional environment. Among the five arbor species, Fir species was more sensitive to the change of OPE monomers. Therefore, the use history and pollution status of OPEs in a region or country could be reflected by any arbor tree species. If the same arbor species with sensitive response to OPEs was selected, we could accurately infer the usage history and pollution status of OPEs in a region or country.