Evaluation of Portable Near-Infrared Spectrometer for Rapid and Non-Destructive Determination of Quality and Authenticity of Baobab (Adansonia Digitata l.) Fruit Pulp in Kenya

Abstract

The demand for baobab fruit pulp (BFP) is growing significantly due to increasing popularity of natural, organic, and nutritious ingredients. This demand has created the need for quality control to ensure quality and safety of the pulp. Conventional methods of assessing the quality of BFP are subject to human error, destructive and costly. Therefore, this study evaluated the ability of portable NIR spectrometer for rapid and non-destructive determination of key quality attributes of BFP. The study also evaluated the potential of the technique to monitor quality changes of stored BFP and to detect adulteration. A portable NIR spectrometer (Model: NIR-S-G1, Tellspec, Toronto, Canada) was used to acquire BFP spectra. Reference measurements on total titratable acidity (TTA), total soluble solids (TSS), vitamin C, and moisture content were immediately collected through specific wet-chemistry procedures. Chemometrics of partial least square regression (PLSR) was used to correlate between NIR spectra and reference measurements. Prediction model specific to each parameter was constructed and validated. This study proved that portable NIR spectrometer could be used for rapid, accurate, and non-destructive determination of BFP quality parameters with R2 of above 0.63 and RPD of above 2.00. A 23 factorial design storage experiment establishing the effect of storage duration (six months), storage conditions (25°C/75%RH and 35°C/83%RH), and packaging materials (unbleached kraft paper, UbKP and low-density polyethylene, LDPE) on quality of BFP. Constructed models were used to monitor changes in stored BFP. Additionally, microbial safety of stored pulp (total aerobic counts, TAC and total yeast and mold counts, TYMC) was determined through specific wet-chemistry procedures. Results indicated that the nutrient composition of stored BFP deteriorated regardless of the effect of packaging material and the storage conditions used. The TTA declined insignificantly (P>0.05) while TSS and vitamin C significantly reduced (P<0.05) at the end of the storage period. The moisture content of stored samples also increased significantly regardless of the protection offered by the packaging material. There was a significant growth of TAC over time (P>0.05) in all samples regardless of packaging and storage conditions. Yeast and molds were not detected in samples kept in LDPE bags until the end of storage. However, the increase in moisture content and microbial load of stored pulp did not surpass the upper limits stipulated by KEBS. Finally, the potential of portable NIR spectrometer to detect and quantify the adulterants in BFP was evaluated. Partial Least Square Discriminant Analysis (PLS-DA) was adopted for classification purpose. The device was sensitive and precise in the discrimination of pure and adulterated BFP. It detected rice, wheat, and maize fours adulterants with sensitivity and specificity of above 0.982 and error of below 0.009 for all two-class PLS-DA models. Finally, the PLSR was also used to establish predictive modes for quantifying the amount of adulterants present in BFP. The models proved to be efficient with prediction R2 and RMSE of above 0.88 and below 6.20% respectively. The models also resulted in reasonably low limits of detection (LODs) of 8.79%, 11.01%, 13.79% for rice, wheat and maize flours, respectively. Therefore, portable NIR spectrometer paired with chemometrics could be used for rapid, non-destructive, and cost-effective quality assessment of BFP. The adoption of portable NIR spectrometers by baobab value chain actors could help reduce post-harvest losses by enabling rapid, non-destructive quality screening of fruits to identify and reject immature or poor-quality batches. Furthermore, this cost-effective technique could be utilized to monitor changes in stored BFP, and rapidly screen for adulteration, thereby maintaining product authenticity and quality.