Direct and two-stage data analysis procedures based on PCA, PLS-DA and ANN for ISE-based electronic tongue-Effect of supervised feature extraction

A novel strategy of data analysis for artificial taste and odour systems is presented in this work. It is demonstrated that using a supervised method also in feature extraction phase enhances fruit juice classification capability of sensor array developed at Warsaw University of Technology. Comparison of direct processing (raw data processed by Artificial Neural Network (ANN), raw data processed by Partial Least Squares-Discriminant Analysis (PLS-DA)) and two-stage processing (Principal Components Analysis (PCA) outputs processed by ANN, PLS-DA outputs processed by ANN) is presented. It is shown that considerable increase of classification capability occurred in the case of the new method proposed by the authors.     

HPLC determination of organic acids, sugars, phenolic compositions and antioxidant capacity of orange juice and orange wine made from a Turkish cv. Kozan

Organic acids, sugars, phenolic compositions and antioxidant capacities of orange juice and orange wine obtained from the cv. Kozan of Turkey were determined. High-performance liquid chromatographic methods were used to identify and quantify of these compounds. Three organic acids (citric, malic and ascorbic acids) and three sugars (sucrose, glucose and fructose) were determined. The major organic acid was found as citric acid. With regard to sugars, sucrose was present in the largest amounts for orange juice and wine. A total of 13 phenolic compounds were identified and quantified in orange juice and wine, including hydroxybenzoic acids (2), hydroxycinnamic acids (5), and flavanones (6). Hesperidin, narirutin and ferulic acid were the most abundant phenolic compounds in orange juice and wine. Antioxidant activities of orange juice and wine were measured using the DPPH^• (2,2-diphenyl-1-picrylhydrazyl) assay, and the antioxidant capacity of orange juice was found to be higher than that of orange wine.     

Application of multivariate calibration to the simultaneous routine determination of ethanol, glycerol, fructose, glucose and total residual sugars in botrytized-grape sweet wines by means of near-infrared reflectance spectroscopy

The simultaneous determination of ethanol, glycerol, fructose, glucose and residual sugars in botrytized-grape sweet white wines was performed by means of near-infrared reflectance using 19 interference filters and a partial least squares (PLS) model in latent variables as a multivariate calibration technique. The results were compared with those obtained using other multivariate calibration techniques, such as MLR, SWR and PCR, by means of a validation set of samples with known compositions. Ethanol, fructose and residual sugars were well predicted by all the multivariate techniques. Glycerol and glucose showed the highest prediction residuals. The technique may be of practical interest in the routine analysis of these types of wines with low cost in terms of samples, time and personnel. Received: 12 June 1995 / Revised: 5 February 1996 / Accepted: 7 February 1996     

Determination of saccharides in fruit juices by capillary electrophoresis and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry

A fast method for the detection of cheap sweeteners is presented. Detecting the adulteration of foods rich in carbohydrates is complicated by the presence of variety of commercial sweeteners that are designed to match exactly the major carbohydrate profiles of these foods. Electrophoretic and mass spectrometric assays for the determination of fruit juice authenticity were developed. Capillary zone electrophoresis with indirect detection was employed to detect adulteration of juices demonstrated by the ratio of the concentrations of major low molecular mass saccharides (glucose, fructose and sucrose). Traces of oligosaccharides, which are not present in the sugar profiles of citrus fruits but are present in inexpensive sweeteners, were evaluated as the other group of target compounds. The fast determination of oligomeric starch hydrolysates in a complex matrix was tested by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS) and applied to orange juice. MALDI-TOFMS was shown to be a suitable method for the identification of adulteration of fruit juices by starch hydrolysates. The effects of the presence of salts and low molecular mass saccharides on the detection of oligosaccharides by MALDI-TOFMS were studied. Low molecular mass saccharides and organic acids decrease the detectability of oligosaccharides by MALDI-TOFMS, but the concentration of maltooligosaccharides present in juices sweetened with starch hydrolysates is high enough to be detected with good sensitivity.     

Determination of ethanol in fuel ethanol and beverages by Fourier transform (FT)-near infrared and FT-Raman spectrometries

Fourier transform-near infrared (FT-NIR) and FT-Raman spectrometries have been used to design partial least squares (PLS) calibration models for the determination of the ethanol content of ethanol fuel and alcoholic beverages. In the FT-NIR measurements the spectra were obtained using air as reference, and the spectral region for PLS modeling were selected based on the spectral distribution of the relative standard deviation in concentration. In the FT-Raman measurements hexachloro-1,3-butadiene (HCBD) has been used as an external standard. In the PLS/FT-NIR modeling for ethanol fuel analysis 50 ethanol fuel standards (84.9-100% (w/w)) were used (25 in the calibration, 25 in the validation). In the PLS/FT-Raman modeling 25 standards were used (13 in the calibration, 12 in the validation). The PLS/FT-NIR and FT-Raman models for beverage analysis made use of 24 standards (0-100% (v/v)). Twelve of them contained sugars (1-5% (w/w)), one-half was used in the calibration and the other half in the validation. Different spectral pre-processing were used in the PLS modeling, depending on the type of sample investigated. In the ethanol fuel analysis the FT-NIR pre-processing was a 17 points smoothed first derivative and for beverages no spectral pre-processing was used. The FT-Raman spectra were pre-processed by vector normalization in the ethanol fuel analysis and by a second derivative (17 points smoothing) in the beverage analysis. The PLS models were used in the analysis of real ethanol fuel and beverage samples. A t-test has shown that the FT-NIR model has an accuracy equivalent to that of the reference method (ASTM D4052) in the analysis of ethanol fuel, while in the analysis of beverages, the FT-Raman model presents an accuracy equivalent to the reference method. The limits of detection for NIR and Raman calibration models were 0.05 and 0.2% (w/w), respectively. It has also been shown that both techniques, present better results than gas chromatography (GC) in evaluating the ethanol content of beverages.     

Simultaneously calibrating solids, sugars and acidity of tomato products using PLS2 and NIR spectroscopy

In this work, the development of a robust spectroscopic procedure for determining, simultaneously and non-destructively, relevant quality parameters of processed tomato products (total and soluble solids, total acidity, total sugars, glucose and fructose), is described. Samples of tomato concentrate products with total solids content ranging from 6.9 to 35.9% were collected from Latin America, the US and Europe and NIR spectra were acquired in the 4000-10,000 cm(-1) region. The original spectra were pre-processed by mean-smoothing or by Fourier filter, followed by multiplicative signal correction (MSC) or derivatives. Partial least squares (PLS2 and PLS1) models were built and their predictive abilities were compared through the RMSEP of external validation. The PLS2 regression had better predictive abilities for four out of the six properties under study, namely total solids, total sugars, glucose and fructose. Besides, the model was less complex than the PLS1 models in the sense that only four factors were demanded whilst from 4 to 11 factors were necessary for building the PLS1 models. The standard error of prediction (SEP%) of the PLS2 model for each property was: total solids, 2.67; soluble solids, 1.14; total acidity, 9.60; total sugar, 18.69; glucose, 11.60; and fructose, 13.45.     

高效阴离子交换色谱-脉冲安培检测法测定烤烟中的水溶性葡萄糖、果糖和蔗糖

用高效阴离子交换色谱-脉冲安培检测法（HPAEC-PAD）测定了烤烟中的水溶性葡萄糖、果糖和蔗糖。采用水浸取及膜过滤法处理烤烟样品,以Dionex CarboPac PA-1阴离子交换柱为色谱柱,0.2 mol/L NaOH水溶液为淋洗液进行分离测定。葡萄糖、果糖和蔗糖的含量与其峰面积的线性关系良好,回收率均在97%以上。方法简便易行,灵敏度高,重现性良好,可以实现对烟草中单糖的快速分离和测定。     

The prediction of total anthocyanin concentration in red-grape homogenates using visible-near-infrared spectroscopy and artificial neural networks

This study compares the performance of partial least squares (PLS) regression analysis and artificial neural networks (ANN) for the prediction of total anthocyanin concentration in red-grape homogenates from their visible-near-infrared (Vis-NIR) spectra. The PLS prediction of anthocyanin concentrations for new-season samples from Vis-NIR spectra was characterised by regression non-linearity and prediction bias. In practice, this usually requires the inclusion of some samples from the new vintage to improve the prediction. The use of WinISI LOCAL partly alleviated these problems but still resulted in increased error at high and low extremes of the anthocyanin concentration range. Artificial neural networks regression was investigated as an alternative method to PLS, due to the inherent advantages of ANN for modelling non-linear systems. The method proposed here combines the advantages of the data reduction capabilities of PLS regression with the non-linear modelling capabilities of ANN. With the use of PLS scores as inputs for ANN regression, the model was shown to be quicker and easier to train than using raw full-spectrum data. The ANN calibration for prediction of new vintage grape data, using PLS scores as inputs, was more linear and accurate than global and LOCAL PLS models and appears to reduce the need for refreshing the calibration with new-season samples. ANN with PLS scores required fewer inputs and was less prone to overfitting than using PCA scores. A variation of the ANN method, using carefully selected spectral frequencies as inputs, resulted in prediction accuracy comparable to those using PLS scores but, as for PCA inputs, was also prone to overfitting with redundant wavelengths.     

Use of Deconvolution Methods for the Analysis of Sugars in Kiwi Juice by HPLC

The automated interpretation of quantitative results in chromatography is always a challenge. We have tested three deconvolution algorithms so that the concentration of sugars in samples of kiwi juice can be calculated accurately. The concentrations of glucose and fructose increase almost 150% at the end of six months of storage and increase nearly 131% for sucrose. Corsican kiwi fruit can be stored for six months in a cold room. This study falls within the sphere of the evaluation of Corsican kiwi fruit.     

An integrated bienzyme glucose oxidase-fructose dehydrogenase-tetrathiafulvalene-3-mercaptopropionic acid-gold electrode for the simultaneous determination of glucose and fructose

A bienzyme biosensor for the simultaneous determination of glucose and fructose was developed by coimmobilising glucose oxidase (GOD), fructose dehydrogenase (FDH), and the mediator, tetrathiafulvalene (TTF), by cross-linking with glutaraldehyde atop a 3-mercaptopropionic acid (MPA) self-assembled monolayer (SAM) on a gold disk electrode (AuE). The performance of this bienzyme electrode under batch and flow injection (FI) conditions, as well as an amperometric detection in high-performance liquid chromatography (HPLC), are reported. The order of enzyme immobilisation atop the MPA-SAM affected the biosensor amperometric response in terms of sensitivity, with the immobilisation order GOD, FDH, TTF being selected. Similar analytical characteristics to those obtained with single GOD or FDH SAM-based biosensors for glucose and fructose were achieved with the bienzyme electrode, indicating that no noticeable changes in the biosensor responses to the analytes occurred as a consequence of the coimmobilisation of both enzymes on the same MPA-AuE. The suitability of the bienzyme biosensor for the analysis of real samples under flow injection conditions was tested by determining glucose in two certified serum samples. The simultaneous determination of glucose and fructose in the same sample cannot be performed without a separation step because at the detection potential used (+0.10 V), both sugars show amperometric response. Consequently, HPLC with amperometric detection at the TTF-FDH-GOD-MPA-AuE was accomplished. Glucose and fructose were simultaneously determined in honey, cola softdrink, and commercial apple juice, and the results were compared with those obtained by using other reference methods.     

Measurement of total fructan in foods by enzymatic/spectrophotometric method: collaborative study

An AOAC collaborative study was conducted to evaluate the accuracy and reliability of an enzyme assay kit procedure for measuring oligofructans and fructan polysaccharide (inulins) in mixed materials and food products. The sample is extracted with hot water, and an aliquot is treated with a mixture of sucrase (a specific sucrose-degrading enzyme), alpha-amylase, pullulanase, and maltase to hydrolyze sucrose to glucose and fructose, and starch to glucose. These reducing sugars are then reduced to sugar alcohols by treatment with alkaline borohydride solution. The solution is neutralized, and excess borohydride is removed with dilute acetic acid. The fructan is hydrolyzed to fructose and glucose using a mixture of purified exo- and endo-inulinanases (fructanase mixture). The reducing sugars produced (fructose and glucose) are measured with a spectrophotometer after reaction with para-hydroxybenzoic acid hydrazide. The samples analyzed included pure fructan, chocolate, low-fat spread, milk powder, vitamin tablets, onion powder, Jerusalem artichoke flour, wheat stalks, and a sucrose/cellulose control flour. Repeatability relative standard deviations ranged from 2.3 to 7.3%; reproducibility relative standard deviations ranged from 5.0 to 10.8%.     

Amperometric Determination of Sugars at Activated Barrel Plating Nickel Electrodes

We report amperometric determination of sugars by using a disposable barrel plating nickel electrode (Ni‐BPE) in this study. The activated Ni‐BPE possesses good reproducibility in flow injection analysis of sugars with a relative standard deviation of 3.16% for 10 consecutive injections of glucose. The electrocatalytic mechanism for the detection of sugars as well as the use as a detector in high‐performance liquid chromatography (HPLC) is investigated. We achieve a good separation of four sugars (glucose, fructose, sucrose, and maltose) in HPLC with favorable sensitivity at a detection potential of +0.55 V vs. Ag/AgCl. The results of wide linear calibration ranges and detection limits in the μM range meet the need of real sample analysis. This detection method is successfully used for quantitative determination of sugars in honey to identify its authentication.     

Detection of orange juice dilution by canonical correlation analysis

Canonical correlation analysis is used to detect orange juice dilutions masked by addition of citric acid and sugars. The values of 28 analytical characteristics (experimental variables) were determined in 105 pure orange juices, and canonical correlation analysis was applied to two groups of experimental variables (citric acid, sucrose, glucose, and fructose versus the other variables). The first pair of canonical correlation variables (one from each group) showed a correlation coefficient of 0.966. By dividing the whole set of 105 juices into reference sets (to compute predictive equations) and test sets (to check the efficiency of the equations), dilutions of 10, 20, and 30% were detected in, respectively, 28, 62, and 91% of the juices from the test sets.     

Determination of carbohydrates and sweeteners in foods and biologically active additives by high-performance liquid chromatography

A method for the determination of large amounts of carbohydrates (glucose, lactose, maltose, mannose, sucrose, and fructose) and sweeteners (xylitol and sorbitol) by reversed-phase liquid chromatography with refractive index detection without derivatization has been developed. The limits of determination for glucose, fructose, and sucrose in liquid samples were 0.1 g/L, and for xylite, lactose, maltose, mannose, and sorbite, 1 g/L. In solid samples the limits of determination for glucose, fructose, and sucrose were 0.1%, and for xylite, lactose, maltose, mannose, and sorbite, 0.6%. The method is applicable to the analysis of samples of wine, juice, honey, cookies, dairy products, and biologically active additives. The developed method for the determination of carbohydrates and sweeteners in foods and biologically active additives was certified in the Mendeleev Institute for Metrology (St. Petersburg).     

Changes in sugar composition and cellulose content during the secondary cell wall biogenesis in cotton fibers

Two cotton cultivars TX19 and TX55 (Gossypium hirsutum L. cv.) were planted in the greenhouse and fibers were harvested at different stages of development. The percentage of sugars present on the fibers was determined by High Performance Liquid Chromatography and the cellulose content was determined using the anthrone method. The percentage of sugars (sucrose, glucose, fructose, and galacturonic acid) showed statistically significant changes during fiber development. The decrease in the percentages of these sugars as the secondary cell wall develops was associated with an increase in the cellulose content. It is important to point out that these analyses were done on intact fibers, no cell wall extractions and purifications were performed.     

Classification of Rhizoma et Radix Notoperygii by HPLC Fingerprints with the Aid of Chemometrics

Chromatographic profiles of Rhizoma et Radix Notoperygii (RRN, “Qianghuo” in Chinese), a complex traditional Chinese medicine (TCM), were collected by high-performance liquid chromatography with diode array detection (HPLC-DAD) at 330 nm. These data profiles were used as fingerprints to investigate quality control classification modeling of the RRN samples. In contrast to the classical methods for discrimination of TCMs, that is, just using common HPLC peaks, all chromatographic profile data were pre-processed by the correlation optimized warping method and polynomial functions; then, these data were submitted as fingerprints (variables) for classification on the basis of sample origin. Chemometrics methods used for calibration modeling and subsequent sample classification-least square support vector machine (LS-SVM), artificial neural network (ANN), and partial least square discriminant analysis (PLS-DA); all produced satisfactory calibrations as well as classification results.     

Quantitative Analysis of Active Pharmaceutical Ingredients (APIs) Using a Potentiometric Electronic Tongue in a SIA Flow System

An advanced potentiometric electronic tongue and Sequential Injection Analysis (SIA) measurement system was applied for the quantitative analysis of mixtures containing three active pharmaceutical ingredients (APIs): acetaminophen, ascorbic acid and acetylsalicylic acid, in the presence of various amounts of caffeine as interferent. The flow‐through sensor array was composed of miniaturized classical ion‐selective electrodes based on plasticized PVC membranes containing only ion exchangers. Partial Least Squares (PLS) analysis of the steady‐state sensor array responses, measured in API mixtures prepared by the SIA system permitted a correct quantitative analysis of acetylsalicylic acid and ascorbic acid. Further optimization using multiway PLS fed by dynamic responses without additional feature extraction did not improve significantly the resolution of acetaminophen. Lastly, the chemometric treatment, involving the extraction of dynamic components of the transient response employing the Wavelet transform, the removal of less‐significant coefficients by means of Causal Index pruning and training of an Artificial Neural Network (ANN) with the selected coefficients, allowed the simultaneous determination of all the three studied APIs, while counterbalancing any interference due to caffeine.