基于响应面法的吹扫捕集-气相色谱-质谱法测定生活饮用水中54种挥发性有机物的含量

Determination of 54Volatile Organic Compounds in Drinking Water by Gas Chromatography-Mass Spectrometry with Purge and Trap Based on Response Surface Method

提出了题示方法,并采用单因素法初步确定吹扫流量、吹扫时间、吹扫温度、解吸时间和解吸温度5个吹扫捕集条件的范围,采用响应面法建立回归模型并给出上述5个条件的最优结果。在40 mL棕色玻璃瓶中装满水样,加入2.0 mg·L^(-1)氟代苯内标溶液20μL。水样中54种挥发性有机物(VOCs)经吹扫捕集后,在HP-VOCs色谱柱上进行程序升温分离,用附电子轰击离子源的质谱仪检测,内标法定量。结果显示:基于单因素法试验结果和响应面法中心组试验设计原理,采用Design-Expert 8.0.6软件进行五因素三水平试验建立的回归模型为五元二次方程,回归模型具有统计学意义(P<0.0001),拟合程度较好(失拟项P=0.2926),可信度高(变异系数小于10%);单因素显著性影响程度大小顺序为解吸时间、吹扫温度、吹扫流量、解吸温度和吹扫时间,交互因素显著性影响程度大小顺序为吹扫流量与吹扫时间、吹扫温度与吹扫流量、吹扫温度与吹扫时间和解吸时间与解吸温度;采用Design-Expert 8.0.6软件对回归模型求解,所得吹扫流量为40 mL·min^(-1)、吹扫时间为9.0 min、吹扫温度为20℃、解吸时间为2.2 min、解吸温度为240℃,此条件下6种典型VOCs(二氯甲烷、三氯乙烯、1,3-二氯丙烷、异丙基苯、1,3-二氯苯、萘)的总质量浓度(30.00±0.41)μg·L^(-1)和预测值(30.36μg·L^(-1))基本一致。52种VOCs的质量浓度均在0.05~10.00μg·L^(-1)内与对应的峰面积与内标峰面积的比值呈线性关系(间/对二甲苯的线性范围为0.10~20.00μg·L^(-1)),检出限(3S/N)为0.01~0.03μg·L^(-1)。以自来水为空白基体进行加标回收试验,54种VOCs的回收率为87.1%~107%,测定值的相对标准偏差(n=7)为1.2%~9.2%。方法用于兰州市城区管网水水样的分析,检出了23种VOCs,检出量在45.7μg·L^(-1)以内,低于GB 5749-2022规定的限值。

The method mentioned by the title was proposed,and single-factor test was used to determine the ranges of five conditions of purge and trap,such as purging flow,purging time,purging temperature,desorption temperature and desorption time,and the regression model was established by response surface method for obtaining the optimum results of the above 5conditions.The 20μL of 2.0mg·L^(-1) fluorobenene internal standard solution was added into a 40mL-brown glass bottle which was filled with water sample.After purge and trap,the 54volatile organic compounds(VOCs)were separated on the HP-VOC column by temperature-programming,detected by mass spectrometer with electron impact ion source,and quantified by internal standard method.As shown by the results,the regression model established by the five-factor three-level test using Design-Expert 8.0.6 software was a five-element quadratic equation on the basis of the test results of single-factor method and the center group design principle of response surface method.The regression model had statistical significance(P<0.0001),the fitting degree was good(Pof the misfit term was 0.2926)and the confidence was high(coefficient of variation was less than 10%).The order of influence degree of single-factor significance was desorption time,purging temperature,purging flow,desorption temperature and purging time,and the order of influence degree of interactive-factor significance was purging flow and purging time,purging temperature and purging flow,purging temperature and purging time and desorption time and desorption temperature.Using Design-Expert 8.0.6software to solve the regression model,and the purging flow was 40mL·min^(-1),the purging time was 9.0min,the purging temperature was 20℃,the desorption time was 2.2min,and the desorption temperature was 240℃.Under these conditions,the total mass concentrations[(30.00±0.41)μg·L^(-1)]of the 6typical VOCs(dichloromethane,trichloroethylene,1,3-dichloropropane,isopropylbenzene,1,3-dichlorobenzene and naphthalene)was basically consistent with the predicted value(30.36μg·L^(-1)).Linear relationships between values of the mass concentration of 52VOCs and the peak area ratio of VOCs to internal were kept in the range of 0.05-10.00μg·L^(-1)(the linear range of m-xylene/p-xylene was 0.10-20.00μg·L^(-1)),with detection limits(3S/N)in the range of 0.01-0.03μg·L^(-1).Using tap water as the blank matrix,test for recovery was made by standard addition method,and values of recovery were found between 87.1%and 107%,with RSDs(n=7)of the determined values in the range of 1.2%-9.2%.The proposed method was applied to the analysis of water samples from urban pipe network in Lanzhou city,and 23VOCs were detected,and the detection amounts were not more than 45.7μg·L^(-1),lower than the limits specified in the GB 5749-2022.