Search

The optimization, analytical characterization and validation of a novel ion-selective electrode for the highly sensitive and selective determination of creatinine in urine is presented. A newly synthesized calix[4]pyrrole-based molecule is used as an ionophore for the enhanced recognition of creatininium cations. The calculation of the complex formation constants in the polymeric membrane with creatininium, potassium and sodium confirms the strong selective interactions between the ionophore and the target. The optimization of the potentiometric sensor presented here yields an outstanding analytical performance, with a linear range that spans from 1µM to 10mM and limit of detection of 10-6.2M. The calculation of the selectivity coefficients against most commonly found interferences also show significant improvements when compared to other sensors already reported. The performance of this novel sensor is tested by measuring creatinine in real urine samples (N=50) and comparing the values against the standard colorimetric approach (Jaffe's reaction). The results show that this sensor allows the fast and accurate determination of creatinine in real samples with minimal sample manipulation.

Hydroponics is a water, energy, space, and cost efficient system for growing plants in constrained spaces or land exhausted areas. Precise control of hydroponic nutrients is essential for growing healthy plants and producing high yields. In this article we report for the first time on a new computer-operated analytical platform which can be readily used for the determination of essential nutrients in hydroponic growing systems. The liquid-handling system uses inexpensive components (i.e., peristaltic pump and solenoid valves), which are discretely computer-operated to automatically condition, calibrate and clean a multi-probe of solid-contact ion-selective electrodes (ISEs). These ISEs, which are based on carbon nanotubes, offer high portability, robustness and easy maintenance and storage. With this new computer-operated analytical platform we performed automatic measurements of K(+), Ca(2+), NO3(-) and Cl(-) during tomato plants growth in order to assure optimal nutritional uptake and tomato production.

Adenosine monophosphate (AMP) has been covalently linked to single-walled carbon nanotubes (SWCNTs) for the first time. SWCNT carboxylic acid moieties were activated through carbodiimide chemistry and acylation procedure. On the one hand, the phosphoramidate salts 12 and 13 were used to directly anchor AMP, and on the other hand a two-step procedure with grafting first the diaminopropane to SWCNTs followed by coupling of AMP. Thermogravimetric as well as elemental analysis, Fourier-transformed infrared and Raman spectroscopies give hints on the characterization of SWCNT–AMP.

Potentiometric sensing represents the preferred technique in many routine measurements of pH and ions. Unfortunately, the simplicity of the technique has not been exploited so far in high throughput biomolecular sensing. In this work, we demonstrate the capabilities of the hybrid functional material carbon nanotubes/aptamer for the creation of a new generation of nuclease-resistant aptasensors using the potentiometric transduction capabilities of single-walled carbon nanotubes in combination with the recognition capabilities of a protein-specific RNA aptamer. The aptasensor was used to detect and identify disease-related proteins at attomolar concentration values in a rapid and non-expensive way. The variable surface glycoprotein from African Trypanosomes was chosen as an ideal model system for a pathogenic exoantigen protein in a clinical sample. Variations in the electromotive force are achieved in real-time upon the direct addition of diluted real blood samples containing the target protein thus eliminating the need of preliminary matrix removal. This work would open the door to real-time diagnostic assays for a wide range of diseases, but also to the rapid molecular detection of several proteins in truly customizable protein biosensing platforms.

The behaviour of double stranded DNA molecules (dsDNA) attached to single walled carbon nanotubes (SWCNTs) was investigated by molecular dynamics simulations. We observe that disruption of short dsDNA helices is accelerated by the presence of SWCNT due to the establishment of π-stacking interactions between the nucleobases and the sidewall of the nanotubes. These results are in good agreement with experimental evidences and confirm the mechanism of assembly between a single stranded DNA molecule and a SWCNT. Longer DNA molecules maintain the structure of a double helix although with the hydrogen-bonding network slightly altered on the relatively short time scale.

The singular chemical structure and the size of carbon nanotubes (CNTs) provide their extraordinary properties. Many researchers have used the electronic characteristics of CNTs to develop electrochemical analytical methods and sensing devices with improved characteristics. In this overview, we discuss the most significant innovations in electrochemical sensing based on CNTs, as reported in the past five years. We highlight the advantages of this type of nanostructured material and the main performance parameters achieved. We also discuss several problems that hinder the widespread use of these devices in practical applications and describe some possible future trends.

In this study we have investigated the morphology and electrical characteristics of protein layers non-covalently adsorbed onto an irregular network of carbon nanotubes (CNT). The layer system presents a prototype for an ion-sensitive field-effect transistor based on CNT-networks. The complementary characterization techniques AFM and ellipsometry give the overall morphology of the functionalized layer system and in combination with concentration dependent measurements a detailed image of the adsorption dynamics. The advantage of CNT-based FETs is their huge surface area, which makes them extremely sensitive even to weak adsorption processes. The here-presented comparative investigations clearly show that significant changes in the transport properties of the CNTs occur much below one monolayer. This sensitivity is an important condition for the future development of efficient biodevices with optimal performance parameters for the detection of pathogenic microorganisms.

We report a field effect transistor (FET) based on a network of single-walled carbon nanotubes (SWCNTs) that for the first time can selectively detect a single gaseous molecule in air by chemically functionalizing the SWCNTs with a selective molecular receptor. As a target model we used SO2. The molecular synthetic receptor is a square-planar NCN-pincer platinum (II) complex (NCN is the N,C,N′-terdentate-coordinating monoanionic [C6H3(CH2NMe2)3-2,6]− ligand) to which SO2 selectively binds. Because of the strong electronwithdrawing character of SO2, it withdraws negative charge from the synthetic receptor thus affecting the electronic properties of the functionalized SWCNTs. The minimum concentration detected is 0.05% SO2 in air at room temperature. Interferences like CH4 and CO2 need to be present in a higher concentration than SO2 to give a significant response.

In this paper we report a biosensor for the fast, ultrasensitive and selective determination of bisphenol A in water. It is based on a field effect transistor (FET) in which a network of single-walled carbon nanotubes (SWCNTs) acts as the conductor channel. SWCNTs are functionalized for the first time with a nuclear receptor, the estrogen receptor alpha (ER-alpha), which is adsorbed onto the SWCNTs and acts as the sensing part of the biosensor. SWCTNs are subsequently protected to prevent the non-specific binding of interferences. With this biosensor we can detect picomolar concentrations of BPA in only 2 min of analysis. Selectivity has been tested against possible interferences such as fluoranthene, pentacloronitrobenzene and malathion, and this is the first device that experimentally shows that small molecules can also be selectively detected at ultralow concentrations using a CNTFET biosensor.

A biosensor for detecting Candida albicans has been developed to improve the performance characteristics of the currently available methods. It is based on a field-effect transistor (FET) in which a network of single-walled carbon nanotubes (SWCNTs) acts as the conductor channel. Monoclonal anti-Candida antibodies were adsorbed onto the SWCNT to provide specific binding sites for fungal antigens. Tween 20 at 0.5% was used as a blocking agent to prevent the non-specific binding of other yeasts or proteins. Our FET devices were exposed to increasing concentrations of C. albicans and were able to detect at least 50 cfu/mL in only 1 h. The sensor response remained stable for more than 10 days. To evaluate the selectivity of our FET devices, Cryptococcus albidus and Saccharomyces cerevisiae were tested as potential competing yeasts for C. albicans. The results showed that biosensors based on a carbon nanotube field-effect transistor are useful tools to detect pathogenic yeasts like C. albicans at low concentrations displaying high selectivity.

In this paper we report a fast, sensitive and label-free biosensor for the selective determination of Salmonella Infantis. It is based on a field effect transistor (FET) in which a network of single-walled carbon nantotubes (SWCNTs) acts as the conductor channel. Anti-Salmonella antibodies were adsorbed onto the SWCNTs and subsequently the SWCNTs were protected with Tween 20 to prevent the non-specific binding of other bacteria or proteins. Our FET devices were exposed to increasing concentrations of S. Infantis and were able to detect at least 100 cfu/mL in 1 h. To evaluate the selectivity of our FET devices, Streptococcus pyogenes and Shigella sonnei were tested as potential competing bacteria for Salmonella. At a concentration of 500 cfu/mL, neither Streptococcus nor Shigella interfered with the detection of Salmonella. Therefore, these devices could be used as useful label-free platforms to detect S. Infantis and, by using the suitable antibody, other bacteria or viruses.

A weight parameter (α) was introduced into the generalized rank annihilation method (GRAM) to calculate and reduce the bias in the predicted concentration. Here we show that α can be used as an indicator to determine whether the trilinearity assumptions are met and to select the right number of factors to calculate the model. The procedure is to calculate several GRAM models by varying α and the number of factors. When the experimental data are trilinear and the right number of factors is used, α does not affect the predicted concentration. If the condition of trilinearity is not met or the correct number of factors is not used, the predicted concentration changes when α changes. A graph shows this behavior. Both simulated and real data were checked for trilinearity. Deviations from the ideal mathematical model, such as the time shift or different shape of the chromatographic profiles, were simulated. These parameters have an effect but the greatest effect was produced by the selection of the number of factors.

Discrimination between virgin olive oils and pure olive oils is of primary importance for controlling adulterations. Here, we show the potential usefulness of two multiway methods, unfold principal component analysis (U-PCA) and parallel factor analysis (PARAFAC), for the exploratory analysis of the two types of oils. We applied both methods to the excitation–emission fluorescence matrices (EEM) of olive oils and then compared the results with the ones obtained by multivariate principal component analysis (PCA) based on a fluorescence spectrum recorded at only one excitation wavelength. For U-PCA and PARAFAC, the ranges studied were λ ex =300–400 nm, λ em =400–695 nm and λ ex =300–400 nm, λ em =400–600 nm. The first range contained chlorophylls, whose peak was much more intense than those of the rest of species. The second range did not contain the chlorophylls peak but only the fluorescence spectra of the remaining compounds (oxidation products and Vitamin E). The three-component PARAFAC model on the second range was found to be the most interpretable. With this model, we could distinguish well between the two groups of oils and we could find the underlying fluorescent spectra of three families of compounds.

VALIDACIO DE METODES DANALISIS QUALITATIVA La informacio quimica sobre la composicio duna mostra pot ser molt diversa: des de saber de quins analits es composa un cert material a saber exactament en quina quantitat shi troben o de quina forma hi son presents, si estan relacionats estructuralment entre ells, etc. Dacord amb tota aquesta varietat, els metodes analitics es classifiquen en dos grans grups: els metodes danalisi qualitativa i els metodes danalisi quantitativa. Segons les caracteristiques del problema analitic es triara un o altre tipus de metode danalisi. Quan lobjectiu es saber que hi ha en una mostra desconeguda, un metode qualitatiu sera el mes adequat. En els darrers temps, aquests metodes han estat objecte destudi, i sutilitzen avui dia, en molts camps daplicacio. Per exemple, en lanalisi daliments es habitual lus dun metode qualitatiu per determinar si un o mes analits es troben presents en la mostra per sobre o per sota duna determinada concentracio. Pero si linteres es saber la quantitat dun determinat component en una mostra, lopcio dun metode quantitatiu sera la mes adient. Aquesta tesi sha centrat en els metodes danalisi qualitativa pels nombrosos avantatges que presenten. Aquests metodes poden ajudar a destriar mostres en funcio de si aquestes presenten una quantitat dun cert analit al voltant dun valor de concentracio previament establert, abans de ser quantificades. Es a dir, sutilitzen com a pas previ a laplicacio del metode quantitatiu, implicant un estalvi de feina, de temps i de diners important si es tracta de quantificar contaminants, detectar adulteracions o qualsevol altra situacio en la que no es pugui sobrepassar una certa concentracio. En aquests casos nomes sha de quantificar la mostra que en el metode qualitatiu revela un resultat en el que es sobrepassa aquesta certa concentracio. En daltres ambits daplicacio, els metodes qualitatius estan perfectament integrats en el procediment estandard doperacions, pel que, llevat en situacions molt especifiques, un resultat positiu no necessita ser confirmat mitjancant una analisi posterior amb un metode quantitatiu. A mes de la importancia de triar un metode analitic adequat a cada problematica, cal destacar que es igual dimportant tenir fiabilitat sobre el resultat trobat i, per tant, sobre el metode emprat. Aixo vol dir que qualsevol metode analitic ha de tenir definits els seus requeriments i qualitats analitiques i que sha de comprovar que aquests parametres previament definits, realment tenen el valor que sels ha assignat. Daquesta confirmacio sen diu Validacio, i es una condicio indispensable per a poder emprar un metode analitic. Daquesta manera es poden garantir els resultats demanats pels clients/usuaris. A mes, des de laprovacio de la norma ISO 17025 aquesta comprovacio del metode analitic i dels seus resultats encara sha fet mes recomanable. Fins fa poc temps, la validacio de metodes analitics sha centrat en els metodes quantitatius. El resultat ha estat una serie de guies/pautes perfectament establertes dus molt comu. Pero no hi ha cap protocol general per a validar un metode qualitatiu. Amb aquesta tesi es vol contribuir a millorar aquesta situacio. Es comenca amb una revisio de les classificacions i de les definicions lligades a aquests metodes, a mes dun repas sobre quines institucions han fet esment daquest tema. Es segueix amb una proposta de classificacio daquests metodes i, finalment, es defineixen aquells parametres de qualitat que es consideren mes importants en la validacio. En les tres aplicacions practiques presentades es descriuen les caracteristiques intrinseques del metode danalisi qualitativa. Despres, es defineixen els parametres que sadequen millor als requeriments del metode i, finalment, es proposa un protocol de validacio que permet el seu establiment. El cas de la revisio de les classificacions i definicions emprades en aquest ambit, com en el cas de la presentacio de les contribucions corresponents a diferents institucions, shan traduit en dues publicacions que sadjunten en la tesi. Pel que fa a les aplicacions practiques, una delles tambe sinclou com a article publicat i les altres dues, sinclouen com a articles acceptats VALIDATION OF QUALITATIVE ANALYTICAL METHODS The chemical information about the composition of a sample can be of different nature: which species are in the sample, their concentration or if they are structurally related, etc. In order to fit any of these requirements, either a qualitative or a quantitative analytical method may be used. If the aim is to identify species, a qualitative method will suit the problem at hand. These types of methods have been recently studied and nowadays are being increasingly used in several fields of analysis. For example, it is common to use qualitative methods as far as food analysis is concerned. On the contrary, if the aim is to quantify one or more analytes of a sample, a quantitative method will be very useful. This thesis has focused on qualitative analytical methods because they provide several advantages and they are being increasingly used. These types of methods can screen samples according to the presence or absence of certain analytes with regard to a pre-set level of concentration. That is to say, they are used as a step before the quantitative method and results in lower analysis time and costs because analyte quantification is not required in all situations. There are some particular analysis fields where qualitative methods are used as routine methods. Therefore, analyte quantification is not always necessary. Moreover, it is also important to provide reliable results, that is to say, to assure that the method performs with reliability. Any analytical method must have its requirements and its analytical properties previously defined, and their values must be proven. To confirm that the requirements and the analytical properties are the right ones and to confirm that they have the right values is to validate the analytical method. This is a necessary condition to use an analytical method. In this sense, the reliability of the results given to the clients or to the users is assured. Moreover, the ISO Standard 17025 strongly encourages method validation. Method validation has focused on quantitative methods. Therefore and as a result, there are more standards or guidelines addressed to quantitative methods validation. These guidelines are commonly used by several communities of practitioners. However, there is no generally accepted standard or validation procedure addressed to qualitative methods. In this sense, this thesis aims to contribute with the development of several validation procedures. The starting point is to provide an overview as a result of a bibliographic search concerning qualitative methods validation. This overview includes the criteria existing for qualitative methods classification as well as the institutions committed the validation of these methods. After that, a classification of these methods is suggested and the most relevant performance parameters in the validation process are defined. The subsequent practical applications describe the intrinsic characteristics of the corresponding qualitative analytical method. After that, the performance parameters that best fit the requirements and the characteristics of the method are defined and, finally, a validation strategy is proposed. Bear in mind, that the strategy considers the intrinsic characteristics of the analytical method. The overview including relevant aspects such as qualitative methods classification, performance parameters definitions and the institutions committed to qualitative method validation, among others, are presented as two publications included in the thesis. Regarding the three practical applications, they are presented as three accepted papers.

We have established a procedure for calculating limits of detection for second-order data. One of the steps involves curve resolution by iterative target transformation factor analysis (ITTFA) and we have checked some experimental factors that affect the efficiency of resolution by ITTFA. Therefore, they directly affect the estimation of the limits of detection (LOD). In this paper, we describe the quality of the LOD estimator as a function of the performance characteristics of a determination with high performance liquid chromatography (HPLC)-diode array detection (DAD) (sensitivity and selectivity of spectra and chromatograms) and advise the end user about how he can improve it by modifying these experimental variables.

When the generalized rank annihilation method (GRAM) is applied to liquid chromatographic data with diode-array detection, an important problem is the time shift of the peak of the analyte in the test sample. This problem leads to erroneous predictions. This time shift can be corrected if a time window is selected so that the chromatographic profile of the analyte in the test sample is trilinear with the peak of the analyte in the calibration sample. In this paper we present a new method to determine when this condition is met. This method is based on the curve resolution with iterative target transformation factor analysis (ITTFA). The calibration and test matrices are independently decomposed into profiles and spectra, and aligned before GRAM is applied. Here we study two situations: first, when the calibration matrix has one analyte and second, when it has two analytes. When the calibration matrix has two analytes, we selectively determine the time window for the analyte to be quantified. There were considerably fewer prediction errors after correction.

Over the past few years linear regression taking into account the errors in both axes has become increasingly important in chemical analysis. It can be applied for instance in method comparison studies at several levels of concentration (where each of the two methods normally present errors of the same order of magnitude) or at calibration straight lines using reference materials as calibration standards, such as in X-ray fluorescence for analysing geological samples. However, the results obtained by using a regression line may be biased due to one or more outlying points in the experimental data set. These situations can be overcome by robust regression methods or techniques for detecting outliers. This paper presents a graphical criterion for detecting outliers using the bivariate least squares (BLS) regression method, which takes into account the heteroscedastic individual errors in both axes. This graphical criterion is based on a modification of Cook’s well-known test for detecting outliers. This new technique has been checked using two simulated data sets where an outlier is added, and one real data set corresponding to a method comparison analysis.

This paper discusses a new stepwise approach for comparing the results from several analytical methods which analyse a set of analytes at different concentration levels, taking into account all the individual uncertainties produced by measurement errors. This stepwise comparison approach detects the methods that provide outlying concentration results. The concentration results from each one of the remaining analytical methods are then compared to the ones from the others taken together, by using linear regression. To do this, the concentration results from the methods considered together and their individual uncertainties, are decomposed at each step to obtain a vector of concentrations. This is achieved by a maximum likelihood principal component analysis (MLPCA), which takes into account the measurement errors in the concentration results. The bivariate least squares (BLS) regression method is then used to regress the concentration results from the method being tested at a given step on the scores generated from the MLPCA decomposition (which have the information of the other remaining methods), considering the uncertainties in both axes. To detect significant differences between the results from the method being tested at a given step and the results from the other methods (MLPCA scores), the joint confidence interval test is applied on the BLS regression line coefficients for a given level of significance α. We have used four real data sets to provide application examples that show the suitability of the approach.

We propose a new procedure for estimating the uncertainty in quantitative routine analysis. This procedure uses the information generated when the trueness of the analytical method is assessed from recovery assays. In this paper, we assess trueness by estimating proportional bias (in terms of recovery) and constant bias separately. The advantage of the procedure is that little extra work needs to be done to estimate the measurement uncertainty associated to routine samples. This uncertainty is considered to be correct whenever the samples used in the recovery assays are representative of the future routine samples (in terms of matrix and analyte concentration). Moreover, these samples should be analysed by varying all the factors that can affect the analytical method. If they are analysed in this fashion, the precision estimates generated in the recovery assays take into account the variability of the routine samples and also all the sources of variability of the analytical method. Other terms related to the sample heterogeneity, sample pretreatments or factors not representatively varied in the recovery assays should only be subsequently included when necessary. The ideas presented are applied to calculate the uncertainty of results obtained when analysing sulphides in wine by HS-SPME-GC.

Testing for lack of fit of the experimental points to the regression line is an important step in linear regression. When lack of fit exists, standard deviations for both regression line coefficients are overestimated, and this gives rise, for instance, to confidence intervals that are too large. If these confidence intervals are then used in hypothesis tests, bias may not be detected so there is a greater probability of committing a β error. In this paper, we present a statistical test, which analyses the variance of the residuals from the regression line whenever the data to be handled have errors in both axes. The theoretical expressions developed were validated by applying the Monte Carlo simulation method to two real and nine simulated data sets. Two other real data sets were used to provide examples of application.

This work presents a new approach for calculating multivariate detection limits for the commonly used classical or direct calibration models. The derived estimator, which is in accordance with latest IUPAC recommendations, accounts for the different sources of error related to the calibration and prediction steps. Since the multivariate detection limit for a given analyte is influenced by the presence of other components in the sample, a different detection limit is calculated for each analyte and analysed sample, at the chosen significance levels α and β . The proposed methodology has been experimentally validated by determining four pesticides in water using a FIA method with diode-array detection. The results compare favourably with the ones obtained using previously proposed estimators.

This paper reports a new approach for validating bias in analytical methods that provide simultaneous results on multiple analytes. The validation process is based on a linear regression technique taking into account errors in both axes. The validation approach is used to individually compare two different chromatographic methods with a reference one. Each of the two methods to be tested is applied on a different set of data composed of two real data sets each. In addition, three different kinds of simulated data sets were used. All three methods are based on RP-high-performance liquid chromatography (HPLC) and are used to quantify eight biogenic amines in wine. The two methods to be tested use different derivatizing procedures: precolumn 6-aminoquinolyl-n-hydroxysuccinimidyl carbamate (AQC) and oncolumn o-phtalaldehyde (OPA), respectively. On the other hand, the reference method uses derivatization with OPA precolumn. Various analytes are determined in a set of samples using each of the methods to be tested and their results are regressed independently against the results of the reference method. Bias is detected in the methods to be tested by applying the joint confidence interval test to the slope and the intercept of the regression line which takes into account uncertainties in the two methods being compared. The conclusions about the trueness of the two methods being tested varied according to whether the joint confidence interval test was applied to data obtained from various biogenic amines considered simultaneously or individually.

Constant or proportional bias in method comparison studies using linear regression can be detected by an individual test on the intercept or the slope of the line regressed from the results of the two methods to be compared. Since there are errors in both methods, a regression technique that takes into account the individual errors in both axes (bivariate least-squares, BLS) should be used. In this paper, we demonstrate that the errors made in estimating the regression coefficients by the BLS method are fewer than with the ordinary least-squares (OLS) or weighted least-squares (WLS) regression techniques and that the coefficient can be considered normally distributed. We also present expressions for calculating the probability of committing a β error in individual tests under BLS conditions and theoretical procedures for estimating the sample size in order to obtain the desired probabilities of α and β errors made when testing each of the BLS regression coefficients individually. Simulated data were used for the validation process. Examples for the application of the theoretical expressions developed are given using real data sets.

In this article, we critically describe the different approaches proposed so far for calculating uncertainty in chemical measurements: (1) the ISO approach, adapted for the analytical field by EURACHEM (commonly known as `bottom-up'), and (2) the Analytical Methods Committee approach (commonly known as `top-down'), based on interlaboratory information. We also propose a new procedure, which is totally consistent with the ISO approach in the sense that all the sources of error are identified, quantified and combined, but which is conceptually more similar to the `top-down' approach because of its holistic character. This new procedure estimates uncertainty from the information generated during the process of assessing accuracy.

A new approach for calculating uncertainties of analytical results based on the information from the validation process is proposed. This approach complements the existing approaches proposed to date and can be applied to any validated analytical method. The precision estimates generated during the process of assessment of the accuracy take into account the uncertainties of preprocessing steps and analytical measurement steps as long as the different factors that influence these steps are representatively varied in the whole validation process. Since the accuracy of an analytical method should be always assessed before applying it to future working samples, little extra work needs to be done to estimate the final uncertainty. Other sources of uncertainty not previously considered (e.g. uncertainty associated to sampling, to differences between the test-sample and the working sample, etc.) are subsequently included and mathematically combined with the uncertainty arising from the assessment of the accuracy to provide the overall uncertainty. These ideas are illustrated with a case study of the determination of copper in contaminated land.

In order to establish an adequate analytical system for the quality control of industrially produced titanium dioxide white pigments, two multivariate linear calibration techniques, principal component regression (PCR) and partial least squares (PLS), are used to model the relationship between the important pigment property, change of colour, and its chemical composition. The results, in terms of accuracy, precision, suitability for quality control and analysis time are compared to those obtained with artificial neural networks (ANNs). Two multivariate display techniques, principal component analysis (PCA) and correspondence factor analysis (CFA) together with two hierarchical clustering techniques, divisive and Ward's agglomerative hierarchical clustering, are also applied to the X-ray fluorescence data of the pigments samples so as to extract as much information as possible. Correlation coefficients obtained by PCR and PLS are 0.92 and 0.94, respectively. Both of them are higher than the already achieved correlation coefficient by ANNs [1], but the precision of the model derived by ANNs is better. It should also be pointed out that some important additional information about the relations between independent variables (chemical composition of the pigment samples) and about the influence of different oxide concentrations on the pigment property, which could be used in the controlling of the production process, was found out.

Method comparison is one of the most important activities within the domain of method validation. The results of newly developed analytical methods are regressed against those obtained using reference methods for a series of samples with different concentrations of the analyte of interest. The linear model, based on the bivariate least squares (BLS) calibration method, and developed taking into account the comparable errors in both axes, should fit a straight line where the intercept is not significantly different from zero and the slope not significantly different from one. To check these premises, the joint confidence interval for the intercept and the slope is usually applied. This article compares the results of this method with the joint confidence test for the intercept and the slope based on the ordinary least squares (OLS) and weighted least squares (WLS) methods, using two real data sets.

A graphical criterion approach has been developed to examine the quality of a set of sensors in multicomponent analysis. Criteria such as sensitivity and selectivity, used in wavelength selection problems, can be explained in terms of the confidence interval of the estimated concentrations. These confidence intervals describe (hyper)ellipsoids whose volume, shape and orientation are related to the optimization criteria. The effect of sensor selection on these criteria is discussed and guidelines for wavelength selection are given. The usefulness of the graphical criterion is shown in the simultaneous determination of 2-chlorophenol and 2,4-dichlorophenol in water.

The main characteristics of a Matlab program to select D-optimal subsets of calibration samples for multiple linear regression are described. The performance of Fedorov's exchange algorithm to select samples is compared with the Kennard-Stone algorithm and the random selection of samples into training and test sets.

Mesure des concentrations en mercure, plomb, cadmium, fer, nickel, cuivre et zinc dans les eaux de la zone cotiere des Iles Canaries