Javier Saurina

Second-order multivariate curve resolution applied to rank-deficient data obtained from acid-base spectrophotometric titrations of mixtures of nucleic bases

Abstract Rank-deficient data matrices, obtained from simulated spectrophotometric acid-base titrations of mixtures of up to four nucleic bases (adenine, cytosine, hypoxanthine and uracil), were analyzed by second-order multivariate curve resolution. The analysis of these individual mixture data matrices gives a rank value of n + 1, where n is the number of nucleic bases present in the system. This number is, however, lower than 2n, the number of spectrometrically active species theoretically present in the systems under study, since each nucleic base is expected to give two species, a protonated and a deprotonated species. This rank deficiency is solved when more than one titration is simultaneously analyzed by second-order multivariate curve resolution. Full rank recovery is achieved when the titration of the mixture of n nucleic bases and other n - 1 titrations, each one corresponding to a different base, are simultaneously analyzed. Results obtained by second-order multivariate curve resolution indicate that for a total resolution of the system full rank is necessary. However, resolution and quantitative determinations of individual nucleic bases in mixtures in the presence of interferences can be achieved (with a prediction error lower than 2% in most cases) even in the case of rank deficiency.

Multivariate resolution of rank‐deficient spectrophotometric data from first‐order kinetic decomposition reactions

The effect of a rank deficiency upon curve resolution in simple kinetic reaction‐based systems is studied. Firstly, simulated rank‐deficient spectrophotometric data of a mixture of two reagents, each one yielding its own reaction product by a first‐order kinetic reaction, are analysed. Four different situations are considered according to the differences in the spectral responses of the reaction constituents and to the differences in the rate constants between the two kinetic processes. A variation of the rate constant between runs for a certain kinetic process is also taken into account. Secondly, the resolution of a real rank‐deficient data system, corresponding to the study of the pH‐dependent decomposition of 1,2‐naphthoquinone‐4‐sulphonate, is investigated. All these studies were carried out using a multivariate curve resolution method based on the alternating least squares optimization of the kinetic and spectral profiles of the species present in the system.

Calibration methods for complex second-order data

In this paper, different three-way methods are tested for their power and shortcomings to solve complex second-order calibration problems. The generic calibration problem is quantifying for an analyte in the presence of an unknown interferent: a second-order calibration problem. Due to rank restrictions of the data, standard second-order calibration methods like Generalized Rank Annihilation cannot be used to solve the type of complex second-order calibration problems shown in this paper. Different real examples are tested in which it is shown that the three-way methods can, to a certain extent, deal with the complex calibrations. This stresses the fact that all second-order calibration methods should be regarded as three-way methods, and when put in this framework, can be compared with respect to their performance.

Determination of biogenic amines in wines by pre-column derivatization and high-performance liquid chromatography coupled to mass spectrometry

A new HPLC method for determining biogenic amines in wines is developed. This method is based on pre-column amine derivatization, further separation of derivatives and on-line hyphenation of HPLC to atmospheric pressure chemical ionization mass spectrometry (APCI-MS). Biogenic amines have been derivatized with 1,2-naphthoquinone-4-sulfonate at 65 degrees C and pH 9.2 for 5 min. The separation of derivatives has been accomplished in a C(18) analytical column using an elution gradient based on increasing the percentage of methanol. Derivatives have been ionized in positive mode and detected by selected ion monitoring. The operating conditions of the APCI-MS system (voltages, temperatures and gases) have been thoroughly optimized to obtain the maximum sensitivity for all analytes. In the selected conditions, APCI-MS spectra display little fragmentation and good signal-to-noise ratio. Depending on the amine characteristics, the main spectral peaks are due to mono- and di-derivative products. Figures of merit of the method have been established under the selected conditions using red wine samples. Recoveries ranging from 94% to 106% have been obtained which prove excellent accuracy of the method in the determination of histamine, putrescine, cadaverine, tryptamine, phenylethylamine, tyramine and serotonin in red wines. The proposed method has been applied to the analysis of commercial wines from different Spanish regions.

Determination of calcium and total hardness in natural waters using a potentiometric sensor array

Abstract The determination of calcium and total hardness in natural waters is carried out with a potentiometric sensor array which consists of a series of ion-selective electrodes (ISEs) for Ca 2+ , Mg 2+ , NH 4 + , K + , Na + , Li + , and H + . The selectivity of the calcium and magnesium ISEs is not fully achieved as other cation species may interfere with the analysis. The proposed sensor array device can overcome this drawback since it can take advantage of the cross-selectivities of cation species towards each ISE. In this approach, the multivariate data generated by the sensor array results in a richer source of analytical information which allows the quantification of calcium and total hardness in the water samples by means of chemometric methods. Results obtained are in reasonable concordance with those given by the standard method based on complexometry.

Estimation of figures of merit using univariate statistics for quantitative second-order multivariate curve resolution

Abstract In this study, a straightforward approach is proposed for determining the figures of merit in multivariate curve resolution. The method is based on the recovery of the pure response profiles of the analytes after a curve resolution procedure. Figures of merit such as limit of detection, sensitivity, precision and accuracy are estimated from these pure analyte responses using univariate statistics directly from a calibration graph, as usual in univariate calibration. Examples of the calculation of the figures of merit in the determination of triphenyltin in synthetic and natural sea water samples by using excitation–emission matrix fluorescence are given.

Quantitative determinations in conventional flow injection analysis based on different chemometric calibration statregies: a review

Abstract This paper shows the possibilities of chemometric techniques for enhancing the analytical capabilities of conventional flow injection analysis (FIA). FIA is an excellent methodology for generating responses of different complexity from which relevant analytical information can be extracted chemometrically. The simplest case involves a single scalar datum per sample, as commonly occurs with FI peak heights or areas. More complex multivariate data consist of data vectors or matrices generated by taking a time array over the FI peak at a certain wavelength, or by recording the whole spectrum at a pre-selected time or at regular steps over the peak. The calibration procedures that can be applied to each type of data (i.e. scalars, vectors or matrices) in order to quantify the analyte in unknown samples are generically called zero-, first- and second-order calibration methods, respectively. The characteristics, advantages and shortcomings of each type of calibration are illustrated from various examples of quantitative determinations of amines and amino acids using 1,2-naphthoquinone-4-sulfonate as a derivatizing reagent. In addition, other FI methods using multivariate calibration procedures are also reviewed.

Derivatization strategies for the determination of biogenic amines in wines by chromatographic and electrophoretic techniques.

This paper revises the derivatization approaches for the determination of biogenic amines in wines. Since most of these amines display poor spectroscopic features to be detected by UV absorption or emission (fluorescence) spectroscopy, derivatization is necessary to attain the desired sensitivity. Reagents such as o-phthaldialdehyde, fluorenylmethylchloroformate, dansyl-Cl and dabsyl-Cl have widely been used for analytical labeling through amino group. A comparison of features of off- and on-line pre- and post chromatographic/electrophoretic labeling is given using 1,2-naphthoquinone-4-sulfonate (NQS) as an example. The evaluation of the influence of the wine sample composition on the derivatization process indicates that pre-column labeling may undergo more severe matrix effects.

Determination of amino acids by ion-pair liquid chromatography with post-column derivatization using 1,2-naphthoquinone-4-sulfonate☆

A new chromatographic method for the determination of amino acids is proposed. The method is based on the separation of amino acids by means of ion-pair liquid chromatography and post-column derivatization using 1,2-naphthoquinone-4-sulfonate. The analytical column was a Spherisorb ODS 2. Amino acids were separated by an elution gradient with four linear steps based on increasing the concentration of 2-propanol. Two eluents were used to create the gradient profile: eluent A was an aqueous solution of 20 mM H3PO4 + 20 mM H2PO4(-) + 15 mM dodecyl sulfate and eluent B was a mixture of aqueous (25 mM H3PO4 + 25 mM H2PO4(-) + 18.5 mM dodecyl sulfate)-2-propanol (1:1, v/v). The injection volume was 100 microl and the total flow-rate for the mobile phase was 0.8 ml/min. The chromatographic outlet was coupled on-line to the two-channel derivatization system which delivered reagent and buffer solutions. The reaction took place at 65 degrees C in a reaction coil of 4 m x 1.1 mm I.D. The spectrophotometric detection was performed at 305 nm. The separation of common amino acids was done in 90 min, although an additional period of 15 min was required to stabilize the column. The repeatability of the method for lysine is 2.1% and the reproducibility is 2.6%. The detection limit for lysine is 0.09 nmol. The linear range for lysine is up to 32 nmol with a correlation coefficient of 0.999. The method was applied to the determination of amino acids in animal feed and powdered milks. The results of the method are in good agreement with those obtained with the standard amino acid autoanalyzer method.

Classification and characterisation of Spanish red wines according to their appellation of origin based on chromatographic profiles and chemometric data analysis

Chromatographic profiles of wines have been used as a fingerprint for the discrimination of Spanish wines based on oenological practices. In order to extract information of different families of phenolic compounds, profiles of different UV-vis absorption wavelengths (280, 310, 370 and 520nm) and fluorescence (ex=260nm; em=360nm) were analysed. A total of thirteen phenolic compounds which allowed the discrimination of wines of three different Spanish appellations (Penedes, Rioja and Ribera del Duero) were selected by means of principal component analysis (PCA). Afterwards, these compounds were used to build partial least squares discriminant analysis (PLS1-DA and PLS2-DA) models which allowed the discrimination of wines according to their appellation with classification rates for independent test sets higher than 96% and 93% for PLS1-DA and PLS2-DA models respectively. Finally, characteristic compounds of each appellation were tentatively identified by means of liquid chromatography-mass spectrometry (LC-MS) analysis. Thus, ten out of thirteen compounds (i.e., gallic acid for Penedes, trans-coumaroyltartaric and trans-caffeoyltartaric acids for Rioja and myricetin for Ribera del Duero wines) have been proposed.