Thesis Archive
Effect of Particle Size, Solvent, and Extraction Time on Tannin Extract From Spondias Purpurea Bark Through Soxhlet Extraction (2013-2014)
Hilary Y. Baldosano
Ma. Beatriz Micaela G. Castillo
Chantal Danica H. Elloran
Abstract:
-Tannins are phytochemical polyphenols that have a wide range of uses in industries such as the leather tanning, textile dyeing, ink, wood adhesive, and wine industries. Their many uses make them significant and in demand in various industries. Throughout time, many different trees have been studied to determine their tannin content. In this study, the tannin content of Spondias purpurea or siniguelas bark was explored due to the lack of information and studies on the species. This study aimed to study the effects of the solvent type and extraction time on the extraction of tannin from its bark through Soxhlet extraction. The ratios of solvent used were ethanol-water solutions at 0:1, 1:1, and 0.5:9.5. The extraction time was also varied at 4, 6, and 8 hours. The extracts were then analyzed using a Perkin Elmer High-Performance Liquid Chromatography (HPLC) with UV/VIS detector. The HPLC was performed using a reversed phase C-18 column as the stationary phase, and methanol (Solvent A) and 1:25 aqueous acetic acid solution (Solvent B) as the mobile phase. It was confirmed that extraction parameters varied in this study affected the yield of tannin in the extract. Finer particle sizes can produce a higher yield of tannin as long as agglomeration in the extractor does not occur. Moreover, a longer extraction time produces a higher yield. Solvents with higher amount of ethanol gave better yields of tannin. Pure water, as a solvent, gave the lowest amounts of tannin due to the solubility of proteins; the 95% ethanol was the most effective. It was observed that 95% ethanol with the longest extraction time of 8 hours gave the highest percentage yield of 19.19% and 17.13% from 10 grams of bark. The model generated that gave an insignificant lack of fit was quadratic. Using ANOVA, the P-value obtained from the model was less than 0.0001, which implies that it is significant.
Adviser:
Dr. Florinda T. Bacani