Rimayi Cornelius, Odusanya David, Mtunzi Fanyana, Tsoka Shepherd
Department of Water Affairs, Resource Quality Services (RQS) Roodeplaat, P. Bag X313, 0001 Pretoria, South Africa.
Department of Water Affairs, Resource Quality Services (RQS) Roodeplaat, P. Bag X313, 0001 Pretoria, South Africa.
Chemosphere. 2015 Jan;118:35-43. doi: 10.1016/j.chemosphere.2014.05.075. Epub 2014 Jun 24.
This paper investigates the efficiency of application of four different multivariate calibration techniques, namely matrix-matched internal standard (MMIS), matrix-matched external standard (MMES), solvent-only internal standard (SOIS) and solvent-only external standard (SOES) on the detection and quantification of 20 organochlorine compounds from high, low and blank matrix water sample matrices by Gas Chromatography-Mass Spectrometry (GC-MS) coupled to solid phase extraction (SPE). Further statistical testing, using Statistical Package for the Social Science (SPSS) by applying MANOVA, T-tests and Levene's F tests indicates that matrix composition has a more significant effect on the efficiency of the analytical method than the calibration method of choice. Matrix effects are widely described as one of the major sources of errors in GC-MS multiresidue analysis. Descriptive and inferential statistics proved that the matrix-matched internal standard calibration was the best approach to use for samples of varying matrix composition as it produced the most precise average mean recovery of 87% across all matrices tested. The use of an internal standard calibration overall produced more precise total recoveries than external standard calibration, with mean values of 77% and 64% respectively. The internal standard calibration technique produced a particularly high overall standard deviation of 38% at 95% confidence level indicating that it is less robust than the external standard calibration method which had an overall standard error of 32% at 95% confidence level. Overall, the matrix-matched external standard calibration proved to be the best calibration approach for analysis of low matrix samples which consisted of the real sample matrix as it had the most precise recovery of 98% compared to other calibration approaches for the low-matrix samples.
本文研究了四种不同的多元校准技术，即基质匹配内标法（MMIS）、基质匹配外标法（MMES）、仅溶剂内标法（SOIS）和仅溶剂外标法（SOES），在采用气相色谱 - 质谱联用仪（GC - MS）结合固相萃取（SPE）对高、低和空白基质水样中的20种有机氯化合物进行检测和定量分析时的应用效率。进一步使用社会科学统计软件包（SPSS）进行多变量方差分析（MANOVA）、t检验和Levene's F检验的统计测试表明，与所选校准方法相比，基质组成对分析方法的效率影响更为显著。基质效应被广泛描述为GC - MS多残留分析中主要的误差来源之一。描述性和推断性统计证明，基质匹配内标校准是用于不同基质组成样品的最佳方法，因为在所有测试基质中它产生的最精确平均回收率为87%。总体而言，使用内标校准产生的总回收率比外标校准更精确，平均值分别为77%和64%。内标校准技术在95%置信水平下产生的总体标准偏差特别高，为38%，这表明它不如外标校准方法稳健，外标校准方法在95%置信水平下的总体标准误差为32%。总体而言，基质匹配外标校准被证明是分析低基质样品（由实际样品基质组成）的最佳校准方法，因为与低基质样品的其他校准方法相比，它的回收率最精确，为98%。
