Optimization of headspace experimental factors to determine chlorophenols in water by means of headspace solid-phase microextraction and gas chromatography coupled with mass spectrometry and parallel factor analysis.

In this work an analytical procedure based on headspace solid-phase microextraction and gas chromatography coupled with mass spectrometry (HS-SPME-GC/MS) is proposed to determine chlorophenols with prior derivatization step to improve analyte volatility and therefore the decision limit (CCα). After optimization, the analytical procedure was applied to analyze river water samples. The following analytes are studied: 2,4-dichlorophenol (2,4-DCP), 2,4,6-trichlorophenol (2,4,6-TrCP), 2,3,4,6-tetrachlorophenol (2,4,6-TeCP) and pentachlorophenol (PCP). A D-optimal design is used to study the parameters affecting the HS-SPME process and the derivatization step. Four experimental factors at two levels and one factor at three levels were considered: (i) equilibrium/extraction temperature, (ii) extraction time, (iii) sample volume, (iv) agitation time and (v) equilibrium time. In addition two interactions between four of them were considered. The D-optimal design enables the reduction of the number of experiments from 48 to 18 while maintaining enough precision in the estimation of the effects. As every analysis took 1h, the design is blocked in 2 days. The second-order property of the PARAFAC (parallel factor analysis) decomposition avoids the need of fitting a new calibration model each time that the experimental conditions change. In consequence, the standardized loadings in the sample mode estimated by a PARAFAC decomposition are the response used in the design because they are proportional to the amount of analyte extracted. It has been found that block effect is significant and that 60°C equilibrium temperature together with 25min extraction time are necessary to achieve the best extraction for the chlorophenols analyzed. The other factors and interactions were not significant. After that, a calibration based in a PARAFAC2 decomposition provided the following values of CCα: 120, 208, 86, 39ngL(-1) for 2,4-DCP, 2,4,6-TrCP, 2,3,4,5-TeCP and PCP respectively for a probability of false positive set at 5%. Also, the accuracy (trueness and precision) of the procedure is assessed. Finally, river water samples have been analyzed with the proposed method showing the absence of the chlorophenols studied.

Usefulness of a PARAFAC decomposition in the fiber selection procedure to determine chlorophenols by means SPME-GC-MS.

In this work, a procedure based on solid-phase microextraction and gas chromatography coupled with mass spectrometry is proposed to determine chlorophenols in water without derivatization. The following chlorophenols are studied: 2,4-dichlorophenol; 2,4,6-trichlorophenol; 2,3,4,6-tetrachlorophenol and pentachlorophenol. Three kinds of SPME fibers, polyacrylate, polydimethylsiloxane, and polydimethylsiloxane/divinylbenzene are compared to identify the most suitable one for the extraction process on the basis of two criteria: (a) to select the equilibrium time studying the kinetics of the extraction, and (b) to obtain the best values of the figures of merit. In both cases, a three-way PARAllel FACtor analysis decomposition is used. For the first step, the three-way experimental data are arranged as follows: if I extraction times are considered, the tensor of data, X, of dimensions I × J × K is generated by concatenating the I matrices formed by the abundances of the J m/z ions recorded in K elution times around the retention time for each chlorophenol. The second-order property of PARAFAC (or PARAFAC2) assesses the unequivocal identification of each chlorophenol, as consequence, the loadings in the first mode estimated by the PARAFAC decomposition are the kinetic profile. For the second step, a calibration based on a PARAFAC decomposition is used for each fiber. The best figures of merit were obtained with PDMS/DVB fiber. The values of decision limit, CCα, achieved are between 0.29 and 0.67 µg L(-1) for the four chlorophenols. The accuracy (trueness and precision) of the procedure was assessed. This procedure has been applied to river water samples.

Usefulness of parallel factor analysis to handle the matrix effect in the fluorescence determination of tetracycline in whey milk.

The determination of tetracycline by fluorescence spectrophotometry in complex matrices has some difficulties, because the presence of other compounds in the matrix affects the analytical signal. In this work, the effect of some inorganic species that are present in whey milk on the fluorescence signal of tetracycline is studied using a D-optimal experimental design. Next, an experimental strategy is proposed in conjunction with Parallel Factor Analysis, PARAFAC, modeling that leads to suitably modeling the severe matrix effect in the determination of tetracycline in whey milk. A specific design is performed in such a way that the lack of trilinearity due to the effect of the presence of interferents on the signal is obviated. Then, ten test samples from three brands of milk, spiked with different quantities of tetracycline and measured in 2 days were analysed using this methodology (mean of the absolute value of the relative errors: 5.1%). The developed analytical method fulfils the property of trueness, the relative errors being, both in calibration and prediction, inside the interval set by Commission Decision 2002/657/EC at these concentration levels. Decision limits (CCalpha) at x(0)=0 microg L(-1) and at x(0)=100 microg L(-1) were 13.2 and 112.4 microg L(-1) respectively, for alpha=0.05; whereas detection capabilities (CCbeta) were 25.9 microg L(-1) and 124.4 microg L(-1) respectively for alpha=beta=0.05.

Usefulness of D-optimal designs and multicriteria optimization in laborious analytical procedures. Application to the extraction of quinolones from eggs.

An analytical method has been developed to extract ciprofloxacin and enrofloxacin from eggs. The aim of this work is to determine the experimental conditions of extraction providing high recoveries with small standard deviations. An experimental design based on the D-optimality criterion and replicated three times was built to evaluate the effect of five factors related to the extraction which is the most inaccurate stage of the procedure. This non-classical design is needed because there are several practical constraints: (i) the extraction procedure is time-consuming, quinolones are not stable and the design must be performed in a single working session. (ii) The tube capacity of the centrifuge is 6, so the number of experiments will be 6 or a multiple of 6. In the optimal experimental conditions, the extraction is performed once with 5 ml of methanol. Then, fatty acids are removed with a mixture of hexane/ether. Analytes are finally separated and detected by HPLC-fluorescence without the additional step of purification by solid-phase extraction (SPE). Under these conditions, the mean recovery is 64% and 70% and the standard deviation 5% and 4% for ciprofloxacin and enrofloxacin, respectively. The capability of decision, CCalpha, is 3.1 and 2.8 microg kg(-1) of ciprofloxacin and enrofloxacin, respectively. The capability of detection, CCbeta, is 7.8 and 7.0 microg kg(-1) of ciprofloxacin and enrofloxacin, respectively. In both cases the probabilities of false positive, alpha, and of false negative, beta, were fixed at 0.05.

Robust and non-parametric statistics in the evaluation of figures of merit of analytical methods. Practices for students.

A set of laboratory practices is proposed in which evaluation of the quality of the analytical measurements is incorporated explicitly by applying systematically suitable methodology for extracting the useful information contained in chemical data. Non-parametric and robust techniques useful for detecting outliers have been used to evaluate different figures of merit in the validation and optimization of analytical methods. In particular, they are used for determination of the capability of detection according to ISO 11843 and IUPAC and for determination of linear range, for assessment of the response surface fitted using an experimental design to optimize an instrumental technique, and for analysis of a proficiency test carried out by different groups of students. The tools used are robust regression, least median of squares (LMS) regression, and some robust estimators as median absolute deviation (m.a.d.) or Huber estimator, which are very useful as an alternatives to the usual centralization and dispersion estimators.

An optimization procedure for determination of indomethacin and acemethacin by differential pulse adsorptive stripping voltammetry. Application on urine samples.

Procedures for the determination of indomethacin and acemethacin by differential pulse adsorptive stripping voltammetry with a mercury electrode have been described and optimised. The selection and optimization of the experimental parameters was done using factorial and central composite designs. Indomethacin and acemethacin in urine were determined by this method with good results and without the need for tedious prior separation. For routine calibration and calculation of the 'capacity to detect', the robust regression method least median squares (LMS) has been proposed.