The influence of geographical origin on the physical and bioactive characteristics of dark chocolate.
Dark chocolate presents unique characteristics, that make it a food product with worldwide consumption and rich in bioactive compounds. Therefore, the purpose of a recent study by a team of Portuguese researchers (Cartas et al., 2024), was the quantification of the bioactive compounds in samples of different types of cocoa from different geographical origins in order to recognize the importance of single origin dark chocolate from the nutritional point of view.
Chocolate varieties from Africa (Amelonado) and Peru (Piura Blanco and Chunco) were used for the tests. The results showed that Amelonado chocolate has higher hardness, plastic viscosity and yield values than other samples. On the other hand, both chocolates from Peru presented higher results in total phenolic compounds, antioxidant capacity, caffeine, and vitamin E compared to Africa’s chocolate.
Additionally, Piura Blanco samples were characterized by a higher content of theobromine, lactic acid, succinic acid and oxalic acid, while Amelonado samples were characterized by higher levels of sucrose and saturated fatty acids than other products. In conclusion, the results obtained are in line with previous studies referring to the impact of variety, geographical origin and processing operations on the nutritional balance of dark chocolate. These results are of great importance for the sector’s industry to support the selection process of the most suitable type of cocoa for specific markets with specific needs.
Use of spectroscopy and machine learning techniques for the sensory quality assessment of cocoa-based products.
Monitoring the sensory quality of cocoa-based products is time-consuming and requires expert panellists. Therefore, the aim of a recent study by a group of international researchers (Collazos-Escobar et al., 2024), was to apply mid-infrared spectroscopy (MIR) in combination with machine learning (ML) techniques to predict the quality of cocoa. The samples were evaluated using ATR-FTIR (Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy) in the 4000 to 650 cm-1 range, and four trained panellists.
Spectral data were pre-processed using a Multiplicative Scatter Correction and combined with the sensory data. Secondly, a Principal Component Analysis (PCA) was performed to obtain uncorrelated regressors as input to the supervised ML techniques. According to the testing panel results, cocoa beans from different growing areas have similar sensory characteristics.
However, using PCA, a distinction was identified in the northern beans. The statistical results revealed the potential of MIR spectroscopy, coupled with ML techniques, for the rapid (calculation time, CT, <0.02 s) and accurate prediction (coefficient of determination, R2, >99.9%) of the overall sensory quality of cocoa-based products. According to the Authors, the implementation of artificial intelligence tools can support decisions in the evaluation of cocoa beans quality, allowing segmentation according to origin and characteristics.
References: Cartas et al., European Food Research and Technology, 250, 2024, 2569-2580; G.A. Collazos-Escobar et al., Infrared Physics and Technology, 141, 2024, 105482.
