Simultaneous Determination of C4-C9 Alkylphenols and Bisphenol a in Environmental Water Using Gas Chromatography-Mass Spectrometry Technique

A liquid-liquid extraction combined with derivatization and gas chromatography-mass spectrometry (GC-MS) method was developed for the determination of C4-C9 alkylphenols (APs, included 4-tert-butylphenol, 4-n-butylphenol, 4-n-pentylphenol, 4-n-hexylphenol, 4-n-heptylphenol, 4-n-octylphenol, 4-tert-octylphenol, 4-n-nonylphenol and 4-nonylphenol) and bisphenol A (BPA) in environmental water. The extraction solvent, extraction number, extraction time, extraction solvent volume, pH value, salt content, concentration degree and derivatization time were optimized. The target compounds were quantied by internal standard method. Results showed that the optimal extraction number, time, solvent volume, pH value and salt content were 2, 5 min, 30 ml, about 2 and 5 g, respectively, and derivatization time was 1 h at room temperature. Dichloromethane was selected as the extraction solvent and was concentrated to 0.5 mL in the process of concentration. The developed method was validated and showed good linearity (R 2 > 0.995), low limit of detection (LOD, 0.002 µg/L–0.006 µg/L), and excellent recovery (68.0 %–122 %) and low relative standard deviation (0.67 %–13.7 %). The developed method was nally applied to groundwater, surface water, seawater and sewage samples in Dongying City located at Yellow River Estuary and some of these target compounds were detected. The measured values of 4-nonylphenol and BPA in these water samples were below their restrictive values of relevant directives at home and abroad.

, however, these methods can be considered de cient and compounds that can be investigated are limited to one or several compounds such as 4-tert-octylphenol (4tOP), 4-nonylphenol (4NP) and BPA. This is due to the lack of understanding of proper use of APs. The plastic industry commonly employs 4NP and 4tOP, and as a result these pollutants was often found in environmental water. Additionally, in the case of other APs, their concentration is very low, hence, researchers cannot determine their presence. Numerous methods have been reported for the detection of APs and BPA in water, among which the liquid-liquid extraction combined gas chromatography-mass spectrometry (GC-MS) method has been extensively studied. However, these studies mostly focus on compounds such as 4tOP, 4NP and BPA (Selvaraj et Martinez and Peñuela 2013). Additionally, with the development of new plastics and liquid crystal products, many new C4-C9 APs have been widely used. 4-tert-Butylphenol (4tBP) is an important raw material for the production of 4-tert-butylphenol formaldehyde resin. 4-n-Pentylphenol (4nPP) is an effective component of disinfectant, food preservative and deodorant. 4-n-butylphenol (4nBP), 4-n-hexylphenol (4nHexP) and 4-n-heptylphenol (4nHepP) are often used as liquid crystal raw materials and intermediates. The wide application of C4-C9 APs leads to the continuous entry of such compounds into environmental water. In 1998, the Ministry of environment of Japan listed 4-pentyl phenol and other alkyl phenols as the endocrine disruptors that affect the performance of human endocrine system (Ministry of the Environment of Japan 1998). However, there is no standard method for the monitoring of C4-C9 APs and BPA in water. In recent years, besides BPA, 4NP and 4tOP, the monitoring and analysis of other short chain alkyl phenols have gradually attracted researcher's attention. For example, Bolivar evaluated 4tBP, 4nHepP, 4tOP, 4NP and BPA using high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS) (Bolivar-Subirats et al. 2021). Yuan evaluated several BPs using solid phase microextraction coupled with GC-MS (Yuan et al. 2016). However, there are no reports on the systematic study of C4-C9 short chain AP and BPA by liquid-liquid extraction, derivatization and GC-MS.
APs are a combination of phenol and alkyl chain. Due to their varied alkyl chain length, the physical and chemical properties differ, such as nonpolarity, adsorption capacity on the material surface and partition coe cient between aqueous phase and organic phase. The shorter the alkyl chain, the stronger the nonpolarity of these target compounds. The quantitative determination of C4-C9 APs and BPA in groundwater, surface water, seawater, sewage and other real water samples involves an analytical technique for trace or ultra-trace mixed components in complex matrix, which is di cult to extract, enrich, separate and detect. Liquid-liquid extraction has the advantages of good separation effect, simple operation and high repeatability. Hence, it is a widely used organic matter extraction and enrichment technology Deng et al. 2018). GC-MS is also a commonly used, advanced and reliable detection technology (Subuhi et al. 2020;Azzouz et al. 2020). After derivatization, the nonpolarity and boiling point of APs and BPA increase, volatility and stability increase, and antiinterference ability increase, hence, the samples can be stored for a long time in case of retesting (Liu et al. 2019;Tan et al. 2019).

Samples preparation and Standard calibration curves
According to the relevant environmental standards (HJ 91.1 2019; HJ/T 164 2020; HJ/T 493 2009), groundwater, river water, Bohai Seawater and sewage were collected in Dongying City. 500 mL water was measured into a separating funnel, followed by addition of 100 µL substitute working solution. 0-30 g of sodium chloride was added and the solution shaken until complete dissolution was achieved. Then 20-50 mL of dichloromethane was added and the solution shaken vigorously for 5-15 min. This was allowed to stand for 10 min in order to separate the layers. The organic phase was collected and aqueous phase was extracted again 1-2 times with the same volume of dichloromethane. The combined organic phases were dried over anhydrous sodium sulfate, ltered with glass wool and concentrated using a laboratory type rotary evaporator (Heidolph Rotary Evaporator vv-2000, Germany) under vacuum at 40 °C. The extraction solvent was selected from dichloromethane, nhexane, ethyl acetate and toluene for high extraction e ciency. The extraction conditions were optimized and results are presented and discussed from section 2.1 to section 2.8. The concentrated sample was transferred to 1 mL volumetric ask, washed the concentrated ask with a small amount of dichloromethane to ensure all sample was removed, and combined the washing solution into the volumetric ask. 100 µL internal standard working solution and 100 µL derivatization reagent were added, and diluted to 1 mL with dichloromethane. The samples were obtained by derivatization at room temperature for 1 h. All measured values were the average of three measurements.
The effect of derivatization time was investigated at room temperature. 100 µL APs and BPA working solution were measured into a volumetric ask, followed by 100 µL internal standard working solution and 100 µL derivatization reagent, which was diluted to 1 mL with dichloromethane, and derivatized at room temperature for 5-120 min.
The calibration solutions were prepared by adding 100 μL internal standard working solution, 100 μL derivatization reagent, appropriate amount of APs and BPA working solution and same volume of substitute working solution to 1 mL volumetric ask. The mixed liquid was diluted to 1 mL with dichloromethane. The concentration of calibration solutions ranges from 5.0 μg/L to 100 μg/L. After derivatization at room temperature for 1 h, the target compounds were determined from low concentration to high concentration according to section 1.3. The retention time and quantitative ion response values of target compounds and corresponding internal standards were recorded.

GC-MS Analysis
A 7890B-5977B Agilent GC-MS equipped with a HP-5MS capillary column (30 m×0.25 mm i.d., 0.25 μm lm thickness) was used to analyze the target compounds. The GC column temperature was raised from 50 °C (initial equilibrium time 2 min) to 100 °C at 20 °C/min, 100 °C to 200 °C at 10 °C/min, and 200 °C to 300 °C (equilibrium time 5 min) at 20 °C/min. The temperature of the injection port was 300 °C and a 1 µL volume was injected in splitless mode. The carrier gas (Helium) pressure was raised from 40 kPa (initial equilibrium time 5 min) to 70 kPa (equilibrium time 5 min) at 2 kPa/min. The mass spectrometer was operated in electron ionization mode with an ionizing energy of 70 eV, ion source temperature 230 °C, MS quadruple temperature 150 °C, and solvent delay 6.0 min. The transmission line temperature was set at 280 °C. Analysis was performed in selected ion monitoring mode (SIM) with the mass/charge ratio ranging from m/z 35 to 400. The quantitative and qualitative ions are shown in Table 1. In uence of extraction solvent 500 mL puri ed water was added into a separating funnel, and then APs and BPA working solution were added to produce the desired concentration of target compounds in the water sample (0.20 µg/L). The water sample was extracted three times using 30 mL different organic solvent with different polarities (n-hexane, toluene, dichloromethane or ethyl acetate). The effects of different extraction solvents on the recovery of target compounds are shown in Table 2. Dichloromethane displays the best extraction effect recovery rate of -target compounds, whereas the extraction e ciency of the other solvents is very low, especially for BPA (13.5 %-76.2 %). These results are consistent with standard methods (ASTM D7065 2017; JIS K 0450-10-10 2006), in which the extraction solvents are all dichloromethane. Dichloromethane was used as the extraction solvent in this method. The effects of different extraction number using of dichloromethane on the recovery of target compounds were examined (Table 3). According to the results, C4-C9 APs can be extracted completely after one extraction, and BPA can be extracted mainly after two extractions. Hence, when dichloromethane is selected as the extraction solvent, the second extraction is su cient to adhere to the necessary requirements.

In uence of extraction time
The in uence of different extraction time on the recovery rate of target compounds is shown in Table 4. The duration of the extraction slightly effected extraction e ciency. As the time was prolonged, the recovery rate of the target substance increased slightly. As shown in Table 4, the results indicated that min was su cient to fully recover the target substance from the water sample. The effects of acidity in water on the recovery rate of target compounds are shown in Table 5. At pH values higher than 6, the recovery of 4NP and 4nOP decreased, but that of other target compounds was not affected. Considering that acidic conditions are conducive to inhibition of bacteria in water and prevention of bacteria from consuming AP organics, pH value of water sample should be adjusted to about 2 (Martinez and Peñuela 2013). The effects of various volumes of dichloromethane on the recovery of target compounds are shown in Table 6. 20 mL of dichloromethane was determined as optimal volume to meet the extraction requirements. However, in the actual extraction process, 30 mL of extraction solvent was selected to ensure complete extraction. Phenols are water-soluble and need salting out to improve extraction e ciency. Hence, the effects of different NaCl amount on the recovery rate of target compounds are shown in Table 7. The results showed that addition of NaCl slightly improved the recovery rate of the target compounds. This is mainly due to the recovery rate already being signi cantly high without addition of salt. Considering the weak salting out effect, 5 g of NaCl was added in water sample in extraction procedure. It is necessary to concentrate the organic extraction solution to a certain volume or redissolve the target compounds to a speci c volume by solvent after being concentrated to dryness prior to analysis. In the case of C4-C9 APs, certain target compounds are liquid and volatilization occurs when the vacuum degree is appropriate. Additionally, some solid targets are powder or velvet with low density, and have drift loss with increasing vacuum degree. In our study, the extraction solution was concentrated by rotary evaporation under vacuum to either 0.5 mL concentrate or to dryness and maintained in vacuum for a period of time (from 1 min to 10 min). The in uence of these methods is shown in Table 8. The recovery rate of target compounds were close to 100 %, when the organic extraction solution was concentrated to 0.5 mL. But the measured values were signi cantly reduced when extraction solution was concentrated to dryness and maintained in vacuum for 1 min, especially for some APs that are liquid at room temperature. Moreover, with the extension of vacuum time, the recovery rate of target compounds further decreased. Finally, in the process of concentration, the extraction solution was concentrated to 0.5 mL.

In uence of derivatization time
The derivatization e ciency of C4-C9 APs and BPA for different derivatization time is shown in Table 9. The derivatization rate of the target compounds tended to be stabilized after 30 min, and the measured value was consistent with the theoretical value (100 µg/mL). Although target compounds were derivatized completely at this time, 60 min was selected as the derivatization time to ensure robustness of the method. Table 9 The derivatization e ciency of C4-C9 APs and BPA for different derivatization time   Table 1).

Linear correlation coe cient and limit of detection
The developed method using the above optimized conditions was validated with respect to linear correlation coe cient (R 2 ) of standard calibration curve and limit of detection (LOD). The linear correlation coe cient (R 2 ), LOD and limit of quanti cation (LOQ) of target compounds to be measured are shown in Table 10. The linear correlation coe cient (R 2 ) values of target compounds were not less than 0.995 and standard calibration curves showed satisfactory linearity based on internal standard method. values of the remaining target compounds were determined by analyzing seven puri ed water samples added with APs and BPA working solution. The concentration of remaining target compounds in these seven puri ed water samples is 0.010 μg/L, whereas LOQ is 4 times that of LOD (Table 10). LOD ranged from 0.002 μg/L to 0.006 μg/L, while LOQ ranged from 0.008 μg/L to 0.024 μg/L. For comparison, LOD values of related compounds in several typical literatures and standards are listed in Table 11. It can be seen that LOD values of 4tOP, 4NP and BPA are 0.001 μg/L-0.2 μg/L, 0.001 μg/L-0.9 μg/L and 0.002 μg/L-0.3 μg/L, respectively, and LOD values of 4tBP, 4nBP and 4nNP range from 0.034 μg/L to 0.077 μg/L, and LOD of this developed method is at low level. To further validate the developed method, precision experiment and recovery experiment were carried out. Puri ed water was taken as the blank matrix to carry out precision experiment, and C4-C9 APs and BPA standard working solutions was added so that the concentration of target compounds was 0.020 µg/L, 0.500 µg/L and 2.00 µg/L, respectively. These three concentration levels basically cover the concentration range of target compounds in groundwater, surface water, seawater, sewage and wastewater. The pretreatment and test were carried out according to section 1.2 and 1.3. Each concentration level was determined six times in parallel. The Average Value (AV) and RSD results are shown in Table 12. Surface water, seawater and sewage were taken as the matrix to carry out recovery experiment, and C4-C9 APs and BPA standard working solution was added so that the concentration of added target compounds was 0.5-3 times of that in the original matrix. The average recovery results are shown in Table 13. The RSD of the target compounds ranged from 0.67 % to 13.7 %, and the average recovery ranged from 68.0 % to 122 %, which indicated that the developed method showed excellent recovery and low relative standard deviation.

Determination of real water samples
The contents of target compounds in groundwater, surface water, seawater and sewage samples are shown in Table 14. It can be seen that BPA (0.016 μg/L) was detected in groundwater, 4nBP, 4nHepP, 4NP, 4nOP and BPA (0.002 μg/L-0.077 μg/L) was detected in Bohai Seawater, 4NP and BPA (0.026 μg/L-0.032 μg/L) was detected in Yellow River water and the detection quantity of these target compounds was low. 4nPP and 4nHexP were not detected in all water samples. The measured values of 4NP (0.139 μg/L-0.965 μg/L) and BPA (0.024 μg/L-0.226 μg/L) in Guangli River water, Xiaoqing River water and two sewage samples were high. The higher pollutant content in two river water samples was attributed to the two river's function of containing pollutants discharged by chemical enterprises. However, the higher content of pollutants in two sewage samples was attributed to the large amount of detergents used in people's lives entering the sewage.  Total ion ow diagram of C4-C9 APs and BPA (SIM)