Rapid determination of sucrose in chocolate mass using near infrared spectroscopy

Geographical Discrimination of Ground Amazon Cocoa by Near-Infrared Spectroscopy: Influence of Sample Preparation

This work presents the application of the NIR technique associated with exploratory analysis of spectral data by main principal components for the discrimination of Amazon cocoa ground seeds. Cocoa samples from different geographic regions of the state of Pará, Brazil (Medicilândia, Tucumã, and Tomé-Açu), were evaluated. The samples collected from each region were divided into four groups distinguished by the treatment applied to the samples, which were fermented (1-with fat and 2-fat-free) and unfermented (3-with moisture and 4-dried). Each set of samples was analyzed separately to identify the influence of moisture, fermentation, and fat on the geographical differentiation of the three regions. From the results obtained, it can be observed that it was not possible to differentiate the samples of seeds not fermented by geographic origin. However, fermentation was crucial for efficient discrimination, providing more defined clusters for each geographic region. The presence of fat in the seeds was a determinant to obtain the best model of geographic discrimination.

M3GPSpectra: A novel approach integrating variable selection/construction and MLR modeling for quantitative spectral analysis

Quantitative analysis of the physical or chemical properties of various materials by using spectral analysis technology combined with chemometrics has become an important method in the field of analytical chemistry. This method aims to build a model relationship (called prediction model) between feature variables acquired by spectral sensors and components to be measured. Feature selection or transformation should be conducted to reduce the interference of irrelevant information on the prediction model because original spectral feature variables contain redundant information and massive noise. Most existing feature selection and transformation methods are single linear or nonlinear operations, which easily lead to the loss of feature information and affect the accuracy of subsequent prediction models. This research proposes a novel spectroscopic technology-oriented, quantitative analysis model construction strategy named M3GPSpectra. This tool uses genetic programming algorithm to select and reconstruct the original feature variables, evaluates the performance of selected and reconstructed variables by using multivariate regression model (MLR), and obtains the best feature combination and the final parameters of MLR through iterative learning. M3GPSpectra integrates feature selection, linear/nonlinear feature transformation, and subsequent model construction into a unified framework and thus easily realizes end-to-end parameter learning to significantly improve the accuracy of the prediction model. When applied to six types of datasets, M3GPSpectra obtains 19 prediction models, which are compared with those obtained by seven linear or non-linear popular methods. Experimental results show that M3GPSpectra obtains the best performance among the eight methods tested. Further investigation verifies that the proposed method is not sensitive to the size of the training samples. Hence, M3GPSpectra is a promising spectral quantitative analytical tool.

Assessing the flavor of cocoa liquor and chocolate through instrumental and sensory analysis: a critical review

The performance of appropriate instrumental and/or sensory analyses is essential to gain insights into the flavor profile of cocoa products. This three-part review is compiled of an overview of the most commonly used instrumental techniques to study cocoa liquor and chocolate flavor, their perception by a trained panel and the potential relationship between them. Each part is the result of a thorough literature study, principally focusing on the assumptions, features and limitations of these techniques. Reviewing of the literature revealed that cocoa matrix effects and methodology restraints were not always considered when instrumentally analyzing cocoa flavor. With respect to sensory analyses, various studies lacked reporting of accomplished trainings and performance of panelists. Moreover, a discrepancy was noticed in the descriptive flavor lexicon employed. Finally, when linking instrumental and sensory data, linear modeling is regularly applied, which might not always be appropriate. This review paper addresses the challenges associated with flavor assessment, intending to incite researchers to critically study cocoa flavor and apply standardized protocols and procedures.

In Situ Monitoring of Sugar Content in Breakfast Cereals Using a Novel FT-NIR Spectrometer

This research demonstrates simultaneous predictions of individual and total sugars in breakfast cereals using a novel, handheld near-infrared (NIR) spectroscopic sensor. This miniaturized, battery-operated unit based on Fourier Transform (FT)-NIR was used to collect spectra from both ground and intact breakfast cereal samples, followed by real-time wireless data transfer to a commercial tablet for chemometric processing. A total of 164 breakfast cereal samples (60 store-bought and 104 provided by a snack food company) were tested. Reference analysis for the individual (sucrose, glucose, and fructose) and total sugar contents used high-performance liquid chromatography (HPLC). Chemometric prediction models were generated using partial least square regression (PLSR) by combining the HPLC reference analysis data and FT-NIR spectra, and associated calibration models were externally validated through an independent data set. These multivariate models showed excellent correlation (Rpre ≥ 0.93) and low standard error of prediction (SEP ≤ 2.4 g/100 g) between the predicted and the measured sugar values. Analysis results from the FT-NIR data, confirmed by the reference techniques, showed that eight store-bought cereal samples out of 60 (13%) were not compliant with the total sugar content declaration. The results suggest that the FT-NIR prototype can provide reliable analysis for the snack food manufacturers for on-site analysis.

Total lipid prediction in single intact cocoa beans by hyperspectral chemical imaging

•

Quantitative calibrations were built from shelled and in-shell single cocoa beans by HSI.

•

The fat content of commercial batches of cocoa beans varies by up to 15% within batches.

•

HSI prediction of the total lipid content was successful for shelled and unshelled beans.

•

Segregation using HSI fat calibration enhanced cocoa bean fat content by 6%.

This work aimed to explore the possibility of predicting total fat content in whole dried cocoa beans at a single bean level using hyperspectral imaging (HSI).

170 beans randomly selected from 17 batches were individually analysed by HSI and by reference methodology for fat quantification. Both whole (i.e. in-shell) beans and shelled seeds (cotyledons) were analysed. Partial Least Square (PLS) regression models showed good performance for single shelled beans (R² = 0.84, external prediction error of 2.4%). For both in-shell beans a slightly lower prediction error of 4.0% and R² = 0.52 was achieved, but fat content estimation is still of interest given its wide range. Beans were manually segregated, demonstrating an increase by up to 6% in the fat content of sub-fractions.

HSI was shown to be a valuable technique for rapid, non-contact prediction of fat content in cocoa beans even from scans of unshelled beans, enabling significant practical benefits to the food industry for quality control purposes and for obtaining a more consistent raw material.

FT-NIR Analysis of Intact Table Grape Berries to Understand Consumer Preference Driving Factors

Fourier-transform near infrared spectroscopy (FT-NIR) is a technique used in the compositional and sensory analysis of foodstuffs. In this work, we have measured the main maturity parameters for grape (sugars and acids) using hundreds of intact berry samples to build models for the prediction of these parameters from berries of two very different varieties: “Victoria” and “Autumn Royal”. Together with the chemical composition in terms of sugar and acidic content, we have carried out a sensory analysis on single berries. Employing the models built for sugars and acids it was possible to learn the sweetness and acidity of each berry before the destructive sensory analysis. The direct correlation of sensory data with FT-NIR spectra is difficult; therefore, spectral data were exported from the spectrometer built-in software and analyzed with R software using a statistical analysis technique (Spearman correlation) which allowed the correlation of berry appreciation data with specific wavelengths that were then related to sugar and acidic content. In this article, we show how it is possible to carry out the analysis of single berries to obtain data on chemical composition parameters and consumer appreciation with a fast, simple, and non-destructive technique with a clear advantage for producers and consumers.

Chocolate Quality Assessment Based on Chemical Fingerprinting Using Near Infra-red and Machine Learning Modeling

Chocolates are the most common confectionery and most popular dessert and snack across the globe. The quality of chocolate plays a major role in sensory evaluation. In this study, a rapid and non-destructive method was developed to predict the quality of chocolate based on physicochemical data, and sensory properties, using the five basic tastes. Data for physicochemical analysis (pH, Brix, viscosity, and color), and sensory properties (basic taste intensities) of chocolate were recorded. These data and results obtained from near-infrared spectroscopy were used to develop two machine learning models to predict the physicochemical parameters (Model 1) and sensory descriptors (Model 2) of chocolate. The results show that the models developed had high accuracy, with R = 0.99 for Model 1 and R = 0.93 for Model 2. The thus-developed models can be used as an alternative to consumer panels to determine the sensory properties of chocolate more accurately with lower cost using the chemical parameters.

Analytical and sensory data correlation to understand consumers' grape preference

NIR spectroscopy is a rapid, economic and not destructive technique employed in food analysis. Concerning fresh table grape, the analysis is usually limited to juices, homogenates or skin extracts which usually give better NIR prediction models. Scanning of intact berries is challenging since each berry has specific features (berry shape, presence of superficial pigmentation, etc.) and, moreover, there are punctual variations even within the same berry. It would be of great interest to obtain information about maturity parameters and consumer's appreciation directly from intact berries, since it would save both time and money. In this article, near infrared (NIR) spectroscopy and chemometric methods have been employed to search for a correlation between sensory analysis and analytical data. The research findings show how it is possible to use a rapid, economic and not destructive emerging technology such as NIR spectroscopy to understand consumer's preference directly from intact berries.

Merging vibrational spectroscopic data for wine classification according to the geographic origin

The wine making procedure is no longer a secret and it is nowadays well described and repeated around the world. Nevertheless, wines present unique features, strongly associated with their geographic origin. Classification systems were developed to catalogue wines according to the provenance, and are currently established by official authorities in order to ensure wine authenticity. The use of near-infrared (NIR), mid-infrared (MIR) and Raman spectroscopy for tracing the origin of wine samples, has been reported with different levels of success. This work evaluated and compared the performance of these techniques, as well as their joint use, in terms of geographic origin classification. NIR, MIR and Raman spectra of wine samples belonging to four Portuguese wine regions (Vinhos Verdes, Lisboa, Açores and Távora-Varosa) were analyzed by partial least squares discriminant analysis (PLS-DA). Results revealed the better suitability of MIR spectroscopy (87.7% of correct predictions) over NIR (60.4%) and Raman (60.8%). The joint use of spectral sets did not improve the predictive ability of the models. The best models were achieved by combining MIR and NIR spectra resulting in 86.7% of correct predictions. Multiblock partial least squares (MB-PLS) models were developed to further explore the combination of spectral data. Although these models did not improve the percentage of correct predictions, they demonstrated the higher contribution of MIR spectroscopic data, in the development of the models.

Monitoring of Water Spectral Pattern Reveals Differences in Probiotics Growth When Used for Rapid Bacteria Selection

Development of efficient screening method coupled with cell functionality evaluation is highly needed in contemporary microbiology. The presented novel concept and fast non-destructive method brings in to play the water spectral pattern of the solution as a molecular fingerprint of the cell culture system. To elucidate the concept, NIR spectroscopy with Aquaphotomics were applied to monitor the growth of sixteen Lactobacillus bulgaricus one Lactobacillus pentosus and one Lactobacillus gasseri bacteria strains. Their growth rate, maximal optical density, low pH and bile tolerances were measured and further used as a reference data for analysis of the simultaneously acquired spectral data. The acquired spectral data in the region of 1100-1850nm was subjected to various multivariate data analyses - PCA, OPLS-DA, PLSR. The results showed high accuracy of bacteria strains classification according to their probiotic strength. Most informative spectral fingerprints covered the first overtone of water, emphasizing the relation of water molecular system to cell functionality.

Multi-wavelength spectrophotometric determination of acidity constant of some newly synthesized Schiff bases and their QSPR study

The acidity constants of some newly synthesized Schiff base derivatives were determined by hard-model based multivariate data analysis of the spectrophotometric data in the course of pH-metric titration in 50% (v/v) methanol-water binary solvent. The employed data analysis method was also able to extract the pure spectra and pH-dependent concentration profiles of the acid-base species. The molecules that possess different substituents (both electron donating and withdrawing) on the ortho-, meta- and para-positions of one of the phenyl ring showed variable acidity constants ranging from 8.77 to 11.07 whereas the parent molecule had an acidity constant of 10.25. To investigate the quantitative effects of changing of substitution pattern on the acidity constant, a quantitative structure-property relation analysis was conducted using substituent constants and molecular descriptor. Some models with high statistical quality (measured by cross-validation Q(2)) were obtained. It was found that the acidity constant of the studied molecules in the methanol-water mixed solvent not only is affected by electronic features of the solutes but also by the lipophilic interaction between methanol part of solvent and the deprotonated solutes.