建立了高效液相色谱-串联质谱法（HPLC-MS/MS）测定电子烟烟液及气溶胶中4种烟草特有N-亚硝胺类化合物（TSNAs）[包括N-亚硝基去甲基烟碱（NNN）、N-亚硝基假木贼碱（NAB）、N-亚硝基新烟草碱（NAT）和4-（N-甲基-N-亚硝胺）-1-（3-吡啶基）-丁酮（NNK）]含量的方法。取1 g电子烟烟液，加入0.1 mol·L^-1乙酸铵溶液20 mL及4 mg·L^-1混合内标溶液50 μL，超声萃取30 min后，经0.2 μm聚四氟乙烯（PTFE）滤膜过滤后进样。将在最大功率下抽吸20口电子烟释放的气溶胶捕集在一张直径为44 mm剑桥滤片上，用上述提取烟液的方法提取气溶胶中的TSNAs。所得滤液中的目标物用Agilent Eclipse Plus C₁₈ RRHD色谱柱在60 ℃柱温下固定，以不同体积比的0.01 mol·L^-1甲酸铵溶液和含0.025%（体积分数）甲酸的乙腈溶液的混合溶液为流动相在0.3 mL·min^-1流量下进行梯度洗脱；用配有电喷雾正离子源（ESI⁺）的串联质谱仪在多反应监测（MRM）模式下检测，以4种TSNAs的氘代试剂作内标物进行内标法定量。结果显示：4种TSNAs的质量浓度均在0.010~5.0 μg·L^-1内与其对应的峰面积与内标物峰面积的比值呈线性关系，检出限（3S/N）分别为0.002~0.015 ng·g^-1（烟液）和0.001~0.008 ng/20口（气溶胶）；日内测定值的相对标准偏差（RSD，n=5）分别为0.78%~1.5%（烟液）和0.96%~2.5%（气溶胶），日间测定值的RSD（n=5）分别为1.3%~2.1%（烟液）和4.9%~6.3%（气溶胶）；对实际样品进行3个浓度水平的加标回收试验，回收率分别为90.7%~103%（烟液）和92.0%~108%（气溶胶）；方法用于14个实际样品的分析，NNN、NAT、NNK、NAB在烟液中的最大检出量分别为40.57，31.94，11.22，8.33 ng·g^-1，在气溶胶中的最大检出量分别为6.40，5.03，1.89，1.28 ng/20口。

A method was established for determination of 4 tobacco-specific N-nitrosamines (TSNAs) in e-cigarette liquid and aerosol by HPLC-MS/MS, including NNN, NAB, NAT and NNK. Analytes in 1 g of e-cigarette liquid was extracted by 20 mL of 0.1 mol·L^-1 ammonium acetate solution containing 50 μL of mixed internal standard solution ultrasonically for 30 min and then filtered through a 0.2 μm PTFE filter. 20 puffs of aerosol realeased at maximum power were collected by a 44 mm diameter Cambridge filter, and the TSNAs in aerosol were extracted by the same method as e-cigarette liquid. The analytes were fixed on an Agilent Eclipse Plus C₁₈ RRHD chromatographic column at 60℃, and mixed solutions of 0.01 mol·L^-1 ammonium formate solution and acetonitrile solution containing 0.025% (φ) formic acid with the different volume ratios were used as mobile phase for gradient elution at flow rate of 0.3 mL·min^-1. A tandem mass spectrometer equipped with ESI⁺ was used for detection under multi-reaction monitoring (MRM) mode. Four deuterated reagents of TSNAs were used as internal standards for quantitative analysis. The results showed that linear relationships between mass concentrations of 4 TSNAs and the ratios of their corresponding peak areas to those of internal standards were kept in the same range of 0.010-5.0 μg·L^-1. Detection limits (3S/N) were 0.002-0.015 ng·g^-1 (e-cigarette liquid) and 0.001-0.008 ng/20 puffs (aerosol) respectively; RSDs (n=5) of the determined values for the intra-day were 0.78%-1.5% (e-cigarette liquid) and 0.96%-2.5% (aerosol) respectively; RSDs (n=5) of the determined values for inter-day were 1.3%-2.1% (e-cigarette liquid) and 4.9%-6.3% (aerosol) respectively. The spiked recovery test was made on actual samples at the 3 concentration levels, giving results of recovery in the ranges of 90.7%-103% (e-cigarette liquid) and 92.0%-108% (aerosol) respectively. This method was used to analyze 14 actual samples, with maximum detectable amounts of NNN, NAT, NNK and NAB in e-cigarette liquid of 40.57, 31.94, 11.22, 8.33 ng·g^-1 and in aerosol of 6.40, 5.03, 1.89, 1.28 ng/20 puffs.