Multivariate Calibration Strategy Research Articles

Electronic waste analysis using laser-induced breakdown spectroscopy (LIBS) and X-ray fluorescence (XRF): Critical evaluation of data fusion for the determination of Al, Cu and Fe

BackgroundElectronic waste (e-waste) proliferation and its implications underscore the imperative for advanced analytical methods to mitigate its environmental impact. It is estimated that e-waste production stands at a staggering 20–50 million tons yearly, of which merely 20–25% undergo formal recycling. The e-waste samples evaluated contain computers, laptops, smartphones, and tablets. ResultsForty-one samples were processed, involving the disassembly and separation of components. Subsequently, two analytical techniques, laser-induced breakdown spectroscopy (LIBS) and energy dispersive X-ray fluorescence (ED-XRF), were applied to quantify aluminum (Al), copper (Cu), and iron (Fe) in the e-waste samples. The samples were then analyzed after acid mineralization with 50% v v−1 aqua regia in a digester block and finally by ICP OES. A solid residue composed of Si and Ti was observed after the digestion of the samples. Multivariate calibration strategies such as partial least-squares regression (PLS), principal component regression (PCR), maximum likelihood principal component regression (MLPCR), and error covariance penalized regression (ECPR) were used for calibration. Finally, the figures of merit were calculated to verify the most suitable models. The results revealed robust models with notable sensitivity, varying from 8.98 to 35.04 Signal (a.u.)(% w w−1) −1, low Limits of Detection (LoD) within the range of 0.001–0.2 % w w−1, and remarkable relative errors ranging from 2% to 33%, particularly for Cu and Fe. SignificanceNotably, the models for Al faced inherent challenges, thus highlighting the complexities associated with its quantification in e-waste samples. In conclusion, this research represents an important step toward a more sustainable and efficient future for electronic waste recycling, signifying its relevance to global environmental welfare and resource conservation.

Multivariate calibration strategies for the simultaneous quantification of aluminium and vanadium in Ti6Al4V alloys

The detection of wear tribocorrosion residues released from orthopaedic implants, such as aluminium and vanadium liberated from Ti6Al4V alloys, is essential to assess their clinical performance. In this work, for the first time to our knowledge, a simple, effective, pre-treatment-free and low-cost methodology for the simultaneous detection of aluminium and vanadium in biological matrices is developed. Both metals, present in a simulated body fluid matrix, form a complex with quercetin, which present overlapping UV–Vis spectra. To overcome this problem, we analyse and compare different chemometric tools, including Partial Least-Squares (PLS), interval Successive Projections Algorithm coupled with Partial Least Squares (iSPA-PLS) and Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS). iSPA-PLS provided the best results for the determination of aluminium and vanadium, using pre-processed data with baseline correction offset. For vanadium, the best performance resulted in ratio performance to deviation for prediction (RPDPred) values of 7.9847, detection limit (LoD) of 0.4725 ppm, quantification limit (LoQ) of 1.4318 ppm, and REP of 4.9708 employing two sets of 20 total intervals, while for aluminium, the best performance resulted in RPDPred values of 3.8331, LoD of 0.2111 ppm, LoQ of 0.6398 ppm, and relative error of prediction (REP) of 10.3546 employing three sets of 23 total intervals. Additionally, the methodology was employed in the detection of aluminium and vanadium residues released from 1 mm thick, 10 mm diameter Ti6Al4V alloy discs in Hank's Solution (simulated biological fluid), detecting residues after twelve weeks of immersion.

Response of runoff and its components to climate change in the Manas River of the Tian Shan Mountains

A warming–wetting climate trend has led to increased runoff in most watersheds in the Tian Shan Mountains over the past few decades. However, it remains unclear how runoff components, that is, rainfall runoff (Rrain), snowmelt runoff (Rsnow), and glacier meltwater (Rglacier), responded to historical climate change and how they will evolve under future climate change scenarios. Here, we used a modified Hydrologiska Byråns Vattenbalansavdelning (HBV) model and a detrending method to quantify the impact of precipitation and temperature changes on runoff components in the largest river (Manas River) on the northern slope of the Tian Shan Mountains from 1982 to 2015. A multivariate calibration strategy, including snow cover, glacier area, and runoff was implemented to constrain model parameters associated with runoff components. The downscaled outputs of 12 general circulation models (GCMs) from the Sixth Coupled Model Intercomparison Project (CMIP6) were also used to force the modified HBV model to project the response of runoff and its components to future (2016–2100) climate change under three common socio-economic pathways (SSP126, SSP245, and SSP585). The results indicate that Rrain dominates mean annual runoff with a proportion of 42%, followed by Rsnow (37%) and Rglacier (21%). In terms of inter-annual variation, Rrain and Rsnow show increasing trends (0.93 (p < 0.05) and 0.31 (p > 0.05) mm per year), while Rglacier exhibits an insignificant (p > 0.05) decreasing trend (−0.12 mm per year), leading to an increasing trend in total runoff (1.12 mm per year, p > 0.05). The attribution analysis indicates that changes in precipitation and temperature contribute 8.16 and 10.37 mm, respectively, to the increase in runoff at the mean annual scale. Climate wetting (increased precipitation) increases Rrain (5.03 mm) and Rsnow (3.19 mm) but has a limited effect on Rglacier (−0.06 mm), while warming increases Rrain (10.69 mm) and Rglacier (5.79 mm) but decreases Rsnow (−6.12 mm). The negative effect of glacier shrinkage on Rglacier has outweighed the positive effect of warming on Rglacier, resulting in the tipping point (peak water) for Rglacier having passed. Runoff projections indicate that future decreases in Rglacier and Rsnow could be offset by increases in Rrain due to increased precipitation projections, reducing the risk of shortages of available water resources. However, management authorities still need to develop adequate adaptation strategies to cope with the continuing decline in Rglacier in the future, considering the large inter-annual fluctuations and high uncertainty in precipitation projection.

Laser-induced breakdown spectroscopy as an analytical tool for total carbon quantification in tropical and subtropical soils: evaluation of calibration algorithms

The demand for efficient, accurate, and cost-effective methods of measuring soil carbon (C) in agriculture is growing. Traditional approaches are time consuming and expensive, highlighting the need for alternatives. This study tackles the challenge of utilizing laser-induced breakdown spectroscopy (LIBS) as a more economical method while managing its potential accuracy issues due to physical–chemical matrix effects. A set of 1,019 soil samples from 11 Brazilian farms was analyzed using various univariate and multivariate calibration strategies. The artificial neural network (ANN) demonstrated the best performance with the lowest root mean square error of prediction (RMSEP) of 0.48 wt% C, a 28% reduction compared to the following best calibration method (matrix-matching calibration – MMC inverse regression and multiple linear regression – MLR at 0.67 wt% C). Furthermore, the study revealed a strong correlation between total C determined by LIBS and the elemental CHNS analyzer for soils samples in nine farms (R² ≥ 0.73). The proposed method offers a reliable, rapid, and cost-efficient means of measuring total soil C content, showing that LIBS and ANN modeling can significantly reduce errors compared to other calibration methods. This research fills the knowledge gap in utilizing LIBS for soil C measurement in agriculture, potentially benefiting producers and the soil C credit market. Specific recommendations include further exploration of ANN modeling for broader applications, ensuring that agricultural soil management becomes more accessible and efficient.

Evaluation of univariate and multivariate calibration strategies for the direct determination of total carbon in soils by laser-induced breakdown spectroscopy: tutorial

The correct choice of the calibration strategy method is a step that can ensure the accuracy of carbon determination in soils by laser-induced breakdown spectroscopy. In this paper, we evaluate and discuss the intrinsic characteristics of univariate and multivariate calibration strategies in overcoming soil matrix effects and their influence on the magnitude of calibration and prediction error values. The matrix-matching calibration with previous correction of Al spectral interference in the C I line at 193.03 nm was the best strategy of the 13 univariate models evaluated [root-mean-square error of calibration—(RMSEP)=0.2wt.%C]. In the multivariate evaluation of the 22 models obtained, the artificial neural network allowed obtaining a lower value of prediction error for C (RMSEP=0.1wt.%C) with very good accuracy (98%) for the set samples of validation. The observation of these results was justified considering the advantages and limitations of each calibration strategy explored.

Evaluation of a Peak-Free Chemometric Laser-Induced Breakdown Spectroscopy Method for Direct Rapid Cancer Detection via Trace Metal Biomarkers in Tissue

The ability to perform direct rapid analysis in air and at atmospheric pressure is a remarkable attraction of laser-induced breakdown spectroscopy (LIBS) for the diagnostic quantification of disease biomarker metals in body tissue. However, accurate trace analysis is limited by matrix effects and a pronounced background that masks the subtle (peak-free) analyte signals because tissue plasma is dense and most lines are optically thick. In this work, a peak-free chemometric LIBS method based on a single-shot (for rapidity and nondestructiveness) and an artificial neural network multivariate calibration strategy with spectral feature selection was evaluated for its utility for direct trace quantitative analysis of copper (Cu), iron (Fe), manganese (Mg), magnesium (Mg), and zinc (Zn) in model soft body tissue. The spectral signatures corresponding to the biometals (so-called because the metals are intrinsic to tissue biochemistry) were generated by spiking their known human-body-representative concentrations in molten paraffin wax. The developed multivariate analytical model achieved ≥95% accuracy as determined from the analysis of oyster tissue-certified reference material. The analytical models were tested on the liver, breast, and abdominal tissue biopsies. The results of applying the model to the clinical tissues indicated the absence or presence (including severity) of cancer as either malignant or benign, in agreement with the pathological examination report.

Multivariate calibration strategy in simultaneous determination of temperature properties of petroleum diesel by near infrared spectrometry

It is essential to control the quality of diesel products so that they comply with relevant fuel specifications, however, the quality assessments rely upon conventional wet chemical analyses that are costly and time consuming. Rapid, simultaneous quality measurement enabling immediate online optimisation for process control and blending offers tremendous cost savings by minimising product quality give-away, shipment demurrage, tank inventory, and laboratory analysis. In this study, the use of near infrared spectroscopy and chemometrics demonstrates a straightforward workflow for simultaneous determination of the petroleum diesel’s boiling point at 95% recovery (T95), flash point (FP), cloud point (CP), and cetane index (CI) calibration development. It involved appropriate spectral region selection, calibration/validation set partition, data pre-processing, regression modelling and validation. Based on the calibration and validation results, the supervised learning models that are obtained from a combination region of 4000–4800 cm−1 on a randomly selected calibration set managed to deliver promising predictive performance in terms of coefficient of determination for prediction (r2P/T95 ≥ 0.94, r2P/FP ≥ 0.89, r2P/CP ≥ 0.89, r2P/CI ≥ 0.993), root mean square error of prediction (RMSEP (T95) ≤ 5.2°C, RMSEP (FP) ≤ 2.0°C, RMSEP (CP) ≤ 2.4°C, RMSEP (CI) ≤ 0.3), and ratio of performance deviation (RPD (T95) ≥ 3.7, RPD (FP) ≥ 3.0, RPD (CP) ≥ 2.9, RPD (CI) ≥ 11). Regardless of principal component regression or partial least square regression on either the multiplicative scattering corrected spectra or Savitzky-Golay second derivative spectra, the developed models met respective ASTM reproducibility requirements, and were considered adequate for immediate quality assessment of diesel.

Chemometrics aided energy dispersive X-ray fluorescence and scattering for diagnostic modeling of trace biometals as cancer biomarkers

Direct analysis of biometals in biomedical samples by energy dispersive X-ray fluorescence (EDXRF) for disease diagnostics has hardly been fully explored due to dark matrix analytical challenges. In this study, we exploited multivariate chemometrics modeling of cancer diagnostics in model human tissue simulates and cultures using selected biometals’ (Mn, Fe, Cu, Zn and Se) fluorescence and Compton scatter profiles. PCA successfully reduced the correlated data dimension to uncorrelated datasets for the characterization of the cell cultures. Artificial neural network (ANN) enhanced the classification of cancer staging and the development of a multivariate calibration strategy for the quantification of trace elements. ANN characterized cancer into early, intermediate, and advanced stages of development. Low concentrations of Fe (101 ± 28 ppm), Zn (59 ± 4 ppm) and Cu (21 ± 1 ppm) were evident in SV10 due to the lag phase stage of cancer development. Further, strong correlation (0.976) was evident in early-stage cancer between Zn and Se but with strong negative correlations between Mn and Se (−0.973) and between Mn and Zn (−0.900) probably due to their antioxidant effects. The results show predictable and systematic associations between the concentrations of Fe, Cu, Zn, Se and Mn as cancer biomarkers with the potential to be used for cancer diagnosis at the early stage of development.

Matrix-Matching Calibration Using Solid Standards: A Comparison between Univariate and Multivariate Strategies for the Determination of Calcium and Magnesium in Bean Seed Samples Employing Laser-Induced Breakdown Spectroscopy (LIBS)

The aim of this study was to evaluate solid calibration standards for matrix-matching comparing with multivariate calibration strategies for calcium and magnesium determination in bean seed samples employing laser-induced breakdown spectroscopy combined with direct solid analysis. Four calibration strategies, namely, (i) partial least square (PLS), (ii) principal component regression (PCR), (iii) multi-energy calibration (MEC) and (iv) one-point gravimetric standard addition (OP GSA), were assessed for fourteen bean seed samples. From all calibration strategies used in this study, univariate calibrations (MEC and OP GSA) led to better results, where for MEC, the trueness values ranged from 89 to 188% for Ca, and from 75 to 116% for Mg, and for OP GSA ranged from 76 to 112% for Ca, and 87 to 127% for Mg. Therefore, the use of solid standard combined with univariate calibration strategies has improved the accuracy by minimizing the matrix effects with appropriate blank and standard dilution.

Potential of satellite and reanalysis evaporation datasets for hydrological modelling under various model calibration strategies

Twelve actual evaporation datasets are evaluated for their ability to improve the performance of the fully distributed mesoscale Hydrologic Model (mHM). The datasets consist of satellite-based diagnostic models (MOD16A2, SSEBop, ALEXI, CMRSET, SEBS), satellite-based prognostic models (GLEAM v3.2a, GLEAM v3.3a, GLEAM v3.2b, GLEAM v3.3b), and reanalysis (ERA5, MERRA-2, JRA-55). Four distinct multivariate calibration strategies (basin-average, pixel-wise, spatial bias-accounting and spatial bias-insensitive) using actual evaporation and streamflow are implemented, resulting in 48 scenarios whose results are compared with a benchmark model calibrated solely with streamflow data. A process-diagnostic approach is adopted to evaluate the model responses with in-situ data of streamflow and independent remotely sensed data of soil moisture from ESA-CCI and terrestrial water storage from GRACE. The method is implemented in the Volta River basin, which is a data scarce region in West Africa, for the period from 2003 to 2012.Results show that the evaporation datasets have a good potential for improving model calibration, but this is dependent on the calibration strategy. All the multivariate calibration strategies outperform the streamflow-only calibration. The highest improvement in the overall model performance is obtained with the spatial bias-accounting strategy (+29%), followed by the spatial bias-insensitive strategy (+26%) and the pixel-wise strategy (+24%), while the basin-average strategy (+20%) gives the lowest improvement. On average, using evaporation data in addition to streamflow for model calibration decreases the model performance for streamflow (-7%), which is counterbalance by the increase in the performance of the terrestrial water storage (+11%), temporal dynamics of soil moisture (+6%) and spatial patterns of soil moisture (+89%). In general, the top three best performing evaporation datasets are MERRA-2, GLEAM v3.3a and SSEBop, while the bottom three datasets are MOD16A2, SEBS and ERA5. However, performances of the evaporation products diverge according to model responses and across climatic zones. These findings open up avenues for improving process representation of hydrological models and advancing the spatiotemporal prediction of floods and droughts under climate and land use changes.

Modelling of bioprocess non-linear fluorescence data for at-line prediction of etanercept based on artificial neural networks optimized by response surface methodology

In the last years, regulatory agencies in biopharmaceutical industry have promoted the design and implementation of Process Analytical Technology (PAT), which aims to develop rapid and high-throughput strategies for real-time monitoring of bioprocesses key variables, in order to improve their quality control lines. In this context, spectroscopic techniques for data generation in combination with chemometrics represent alternative analytical methods for on-line critical process variables prediction. In this work, a novel multivariate calibration strategy for the at-line prediction of etanercept, a recombinant protein produced in a mammalian cells-based perfusion process, is presented. For data generation, samples from etanercept processes were daily obtained, from which fluorescence excitation-emission matrices were generated in the spectral ranges of 225.0 and 495.0 nm and 250.0 and 599.5 nm for excitation and emission modes, respectively. These data were correlated with etanercept concentration in supernatant (measured by an off-line HPLC-based reference univariate technique) by implementing different chemometric strategies, in order to build predictive models. Partial least squares (PLS) regression evidenced a non-linear relation between signal and concentration when observing actual vs. predicted concentrations. Hence, a non-parametric approach was implemented, based on a multilayer perceptron artificial neural network (MLP). The MLP topology was optimized by means of the response surface methodology. The prediction performance of MLP model was superior to PLS, since the first is able to cope with non-linearity in calibration models, reaching percentage mean relative error in predictions of about 7.0% (against 12.6% for PLS). This strategy represents a fast and inexpensive approach for etanercept monitoring, which conforms the principles of PAT.

Hybridized robust chemometrics approach for direct rapid determination of trace biometals in tissue utilizing energy dispersive X-ray fluorescence and scattering (EDXRFS) spectrometry

Direct rapid energy dispersive X-ray fluorescence and scattering (EDXRFS) analysis of trace biometals in soft body tissues is important because it has an immense potential for biomedical applications. Unfortunately this is challenging because soft body tissues are characterized by dark matrix problems, weak analyte fluorescence, scattering, poor signal-to-noise ratio (SNR) of the analyte and spectral overlaps due to the properties of the detector and detection process. We report on hybridized utility of robust chemometrics approach for spectral preprocessing towards improving the quality of spectra towards quantitative analysis of trace biometals in soft body tissue. The study was based on (5–20 µm thick) paraffin wax model ‘standards’ spiked with biometals Fe, Cu, Mn, Zn, Co, Na and Mg. Wavelet transform (WT) and principal component analysis (PCA) were used conjointly for de-noising and mathematical enhancement of resolution. There was improved SNR of spectra by a factor of 3 compared to use of WT alone. The preprocessed spectra were used as input to artificial neural network (ANN) and partial least squares (PLS) models for developing multivariate calibration strategies for quantitative analysis. Both models predicted the concentrations of the biometals better than when raw spectra were utilized (R2 ~ 0.892–0.954 before, and ~ 0.990–0.998 after preprocessing for ANNs; and R2 ~ 0.876–0.931 before, and ~ 0.977–0.992 after preprocessing for PLS). There was also improvement in prediction of Na and Mg in model tissue when both fluorescence and scatter were utilized conjointly (EDXRFS) i.e. R2 = 0.970 for fluorescence alone and R2 = 0.998 for both fluorescence and scatter for Na; and R2 = 0.934 for fluorescence alone and R2 = 0.993 for both fluorescence and scatter for Mg for ANN model. The accuracy of the calibration model was tested using Oyster tissue (NIST 1566b). The results of all analyzed elements were in agreement with certified values to ≤ 6%. This shows proof-of-concept for use of hybridized robust chemometrics approaches for direct rapid determination of trace biometals in soft tissue utilizing EDXRFS spectrometry; an approach that has potential for biomedical applications of EDXRF.

Univariate and multivariate calibration strategies in combination with laser-induced breakdown spectroscopy (LIBS) to determine Ti on sunscreen: A different sample preparation procedure

Sunscreen is a cream which the intended use is to protect the human skin from ultraviolet radiation (UV) from the sun. Physical (inorganic) or chemical (organic) filters could be used to scatter the UV radiation, being TiO2, one physical filter widely used in its formulation. To determine Ti concentrations on sunscreen, laborious and unsafe procedures must be performed involving strong acids and high temperatures in furnaces. With laser-induced breakdown spectroscopy (LIBS) the possibility of determining this element with minimal sample preparation is feasible. In this study, Ti concentration on sunscreen was determined through several calibration approaches (univariate and multivariate) with LIBS and an unusual sample preparation for sunscreen using a solution of poly(vinyl)alcohol (PVA) to immobilize the sample, converting its matrix to solid. Five samples were prepared and analyzed by inductively coupled plasma optical emission spectrometry (ICP OES) and the Ti concentration ranged from 0.45 to 1.58%. Standard errors for cross validation (SECV) and recoveries were 0.07% and 97–104% using multiple linear regression (MLR) with information from two Ti emission lines: I 498.17 nm and I 499.11 nm.

Rapid Nuclear Forensics Analysis via Laser Based Microphotonic Techniques Coupled with Chemometrics

Nuclear forensics (NF) is an important tool for analysis and attribution of nuclear and radiological materials (NRM) in support of nuclear security. The critical challenge in NF currently is the lack of suitable microanalytical methodologies for direct, rapid and minimally-invasive detection and quantification of NF signatures. Microphotonic techniques can achieve this task particularly when the materials are of limited size and under concealed condition. The purpose of this paper is to demonstrate the combined potential of chemometrics enabled LIBS and laser Raman spectromicroscopy (LRS) for rapid NF analysis and attribution. Using LIBS, uranium lines at 385.464 nm, 385.957 nm and 386.592 nm were identified as NF signatures in uranium ore surrogates. A multivariate calibration strategy using artificial neural network was developed for quantification of trace uranium. Principal component analysis (PCA) of LIBS spectra achieved source attribution of the ores. LRS studies on UCl3, UO3(NO3)2.6H2O, UO2SO4.3H2O and UO3 in pellet state identified the bands associated with different uranium molecules as varying in the range of (840 to 867) ± 15 cm-1. Using this signature, we have demonstrated spectral imaging of uranium under concealed conditions.

Coping with matrix effects in headspace solid phase microextraction gas chromatography using multivariate calibration strategies

SPME is extremely sensitive to experimental parameters affecting liquid–gas and gas–solid distribution coefficients. Our aims were to measure the weights of these factors and to design a multivariate strategy based on the addition of a pool of internal standards, to minimize matrix effects. Synthetic but real-like wines containing selected analytes and variable amounts of ethanol, non-volatile constituents and major volatile compounds were prepared following a factorial design. The ANOVA study revealed that even using a strong matrix dilution, matrix effects are important and additive with non-significant interaction effects and that it is the presence of major volatile constituents the most dominant factor.A single internal standard provided a robust calibration for 15 out of 47 analytes. Then, two different multivariate calibration strategies based on Partial Least Square Regression were run in order to build calibration functions based on 13 different internal standards able to cope with matrix effects. The first one is based in the calculation of Multivariate Internal Standards (MIS), linear combinations of the normalized signals of the 13 internal standards, which provide the expected area of a given unit of analyte present in each sample. The second strategy is a direct calibration relating concentration to the 13 relative areas measured in each sample for each analyte. Overall, 47 different compounds can be reliably quantified in a single fully automated method with overall uncertainties better than 15%.

Multivariate optimization of the synthesis of titania biomorphic ceramics and development of a FT-IR method for quantification synthesis yield

Abstract Titania biomorphic ceramics were manufactured from wood preforms by a two steps process, i.e. infiltration and then pyrolysis in an inert atmosphere to produce porous TiC ceramics, followed by infiltration and pyrolysis in an air atmosphere where TiC was oxidized to TiO2. This technology provides a cost effective and eco-friendly route to advanced ceramic materials. The properties of biomorphic ceramics depend decisively on the synthesis parameters, that were optimized using a multivariate methodology for the design of experiments. Four variables (infiltration time, titanium isopropoxide and acetic acid proportion, and number of infiltrations before and after pyrolysis) were considered as factors in the synthesis optimization process. Interactions between these factors and their optimal levels were investigated using a two level factorial design. For evaluating the yields, a new method by Fourier transform infrared spectrometry (FTIR) has been developed for the direct determination of TiO2 by absorbance measurements in KBr pellets. The procedure is based on the use of the ratio between the absorbance of the characteristic band of titania and those of a nitrate internal standard added to samples. A multivariate calibration strategy based on inverse least squares approach was employed for quantification.

Correlation of quantitative sensorial descriptors and chromatographic signals of beer using multivariate calibration strategies.

In this study, two important sensorial parameters of beer quality - bitterness and grain taste - were correlated with data obtained after headspace solid phase microextraction - gas chromatography with mass spectrometric detection (HS-SPME-GC-MS) analysis. Sensorial descriptors of 32 samples of Pilsner beers from different brands were previously estimated by conventional quantitative descriptive analyses (QDA). Areas of 54 compounds systematically found in the HS-SPME-GC-MS chromatograms were used as input data. Multivariate calibration models were established between the chromatographic areas and the sensorial parameters. The peaks (compounds) relevant to build each multivariate calibration model were determined by genetic algorithm (GA) and ordered predictors selection (OPS), tools for variable selection. GA selected 11 and 15 chromatographic peak areas, for bitterness and grain taste, respectively; while OPS selected 17 and 16 compounds for the same parameters. It could be noticed that seven variables were commonly pointed out by both variable selection methods to bitterness parameter and 10 variables were commonly selected to grain taste attribute. The peak areas most significant to the evaluation of the parameters found by both variable selection methods fed to the PLS algorithm to find the proper models. The obtained models estimated the sensorial descriptors with good accuracy and precision, showing that the utilised approaches were efficient in finding the evaluated correlations. Certainly, the combination of proper chemometric methodologies and instrumental data can be used as a potential tool for sensorial evaluation of foods and beverages, allowing for fast and secure replication of parameters usually measured by trained panellists.

Quantitative determination of ZrC in new ceramic materials by Fourier transform infrared spectroscopy

Zirconium carbide-based biomorphic ceramics have been manufactured by vacuum infiltration of zirconium-oxychloride (ZrOCl 2·8H 2O) sol into natural wood. After vacuum-assisted infiltration, the specimens were dried in air and pyrolysed at 800 °C in Ar-atmosphere. The infiltration and drying process were repeated up to four times. Finally, the specimens were hold at 1550 °C in a vacuum furnace to form ZrC. The results of the process of synthesis have been studied using scanning electron microscopy (SEM). For evaluating the yield of the synthesis, a new method by Fourier transform infrared spectrometry (FTIR) has been developed for the direct determination of ZrC by absorbance measurements in KBr pellets. The procedure was based on the use of the ratio between the absorbance of the characteristic band of zirconium carbide and those of an oxalate internal standard added to samples. A multivariate calibration strategy based on inverse least squares and the standard addition approach was employed for quantification. The results obtained for all ZrC-ceramics studied were satisfactorily compared with those obtained by X-ray diffractometry (XRD). The proposed method was also applied to the analysis of synthetic samples prepared by mixing pyrolyzed wood with ZrC, the results indicated good recovery in all instances.

The Importance of Choosing the Right Validation Strategy in Inverse Modelling

Inverse modelling techniques, such as principal component regression, partial least squares regression and support vector machines, are very powerful multivariate calibration strategies which are widely used in near infrared spectroscopy. However, these techniques are so efficient in finding correlations between the spectral variables and the parameter to be predicted that great care should be taken to avoid over-optimistic results by use of a proper validation strategy. In this study, different validation strategies were investigated on a dataset that was acquired during various measurement days. The goal was to predict albumen freshness based on spectral measurements. Validation procedures frequently applied in practice, i.e. 10-fold cross-validation (10-fold CV) and validation based on random subdivision in calibration and validation set (RS) were compared to a cross-validation across measuring day (MD). Whereas 10-fold CV and RS validation suggested that prediction of albumen freshness is possible, MD validation on the same dataset indicated that albumen freshness cannot be predicted from the spectral measurements. It is shown that inverse modelling is very sensitive to unspecific correlations between the spectral measurements and the dependent variable, which might be artifacts of the measurement protocol and will not be persistent in the future. Therefore, selection of the right validation strategy for a given application and critical evaluation of the obtained results are crucial steps in inverse modelling to obtain useful calibration models. More specifically, in the context of process analytical technology where spectra are acquired over time, great care should be taken to break the unspecific correlation between the dependent variable and the variations in the spectral measurements over time.

Quantitative determinations of SiC and SiO 2 in new ceramic materials by Fourier transform infrared spectroscopy

Silicon carbide-based biomorphic ceramics have been fabricated by the pyrolysis and infiltration of natural wood (mukali and pine) with molten silicon. The results of the process of synthesis have been studied in this and other biomorphic ceramics using thermogravimetric analysis (TGA), X-ray photoelectron spectroscopy (XPS) analysis and scanning electron microscopy (SEM). For evaluating the yield of the synthesis, a new method by Fourier transform infrared spectrometry (FTIR) has been developed for the direct determination of SiC and the simultaneous determination of SiC and SiO 2 by absorbance measurements in KBr pellets. The procedure was based on the use of the ratio between the absorbance of the characteristic band of silicon carbide or silica and those of an acetate internal standard added to samples. A multivariate calibration strategy based on inverse least squares and the standard addition approach were employed for quantification. The results obtained for all biomorphic ceramics studied and synthetic samples prepared by mixing pyrolyzed wood with pure SiC were satisfactory. The relative standard deviation for all samples was lower than 2.9%.