Abstract:
In the low-temperature regions of Sri Lanka, coconut oil vendors frequently resort to repetitive
heating practices to maintain the liquid state of the oil for commercial purposes. Recurrent
heating of coconut oil leads to the formation of toxic compounds like peroxides and hydrocarbons,
despite health risks. This research investigated the impact of repeated heating on the quality of
various coconut oil varieties in Sri Lanka, addressing a gap in existing methodologies. These
include virgin, refined, milk-extracted, copra-extracted, and commercially heated coconut oils
that have been heated repeatedly. With the aid of Soft Independent Modelling of Class Analogy
(SIMCA) and Partial Least Squares (PLS) models and Near-Infrared (NIR) spectroscopy, the study
investigated the compositional changes of different coconut oil categories, revealing varying
predictive capacities for different types. Virgin coconut oil exhibits robust resistance to
degradation, evidenced by a Standard Error of Prediction (SEP) of 0.018 and a correlation
coefficient of 0.904. Conversely, refined coconut oil displays increased susceptibility to quality
alterations upon repetitive heating, as indicated by a SEP of 0.027 and a correlation coefficient of
0.742. Milk-extracted coconut oil shows a SEP of 0.015 with a correlation coefficient of 0.767,
while copra-extracted coconut oil yields a SEP of 0.017 and a correlation coefficient of 0.734.
Commercially scaled repetitively heated coconut oil registers a higher SEP of 0.037 and a
correlation coefficient of 0.781, signaling substantial quality degradation under commercial
heating conditions. Critical wavelengths for predicting peroxide value and heating durations are
identified at 1012.751nm and 663.229nm, respectively. Notably, SIMCA models for each coconut
oil type demonstrate 100% validation accuracy without inconsistencies and highlights the use of
NIR spectroscopy for rapid quality assessment, detecting significant quality alterations caused by
repetitive heating, especially elevated peroxide values. It emphasizes the potential of NIR
spectroscopy as a rapid assessment tool for evaluating repetitively heated coconut oil quality and
understanding its chemical transformations during repetitive heating. The research highlights
the importance of adopting prudent culinary practices to mitigate potential health hazards
associated with deteriorated oil quality.
