Watermelon Seed Oil

Two studies of Citrullus lanatus seed oil that agree closely set every range here. Raziq et al. 2012 (Grasas y Aceites 63(4):365-372, open access), four Pakistani varieties - Sugar Baby, Q-F-12, D-W-H-21 and Red Circle-1885, hexane-extracted: saponification index 190.20-205.57 mg KOH/g, iodine value 97.10-116.32 g I2/100 g, unsaponifiable matter 0.54-0.82 %, free fatty acids 1.17-2.10 % as oleic, peroxide 2.90-5.06 meq O2/kg, refractive index 1.4665-1.4670 at 40 C, oil 28.25-35.65 % of the seed. Cheikhyoussef et al. 2017 (3 Biotech 7:297), Namibian Kalahari melon seed - the same species, and what is sold as kalahari melon seed oil - by cold press, traditional method and Soxhlet: saponification 181.12-189.26, iodine 110.28-118.61, acid value 0.95-1.63 mg KOH/g, specific gravity 0.910-0.922, refractive index 1.4665-1.4726 at 25 C, oil 40.16 % of the seed. The stored fatty-acid ranges are the union of Raziq's Table 5 and Cheikhyoussef's composition table: palmitic 14.3-16.58, stearic 11.2-14.58, oleic 10.51-23.0, linoleic 45.1-56.94, arachidic 1.03-1.38 %. TWO CAVEATS ON THE PRIMARY TABLE. First, Raziq's Table 5 quantifies only palmitic, stearic, oleic and linoleic and books a further 1.5-2.0 % as unidentified, which is why the midpoints stored here total about 97.3 % rather than 100 - no acid has been dropped from a fuller table. Second, that paper's English abstract gives palmitic 15.1-16.9 and stearic 11.5-14.4 %, which does not match its own Table 5 (14.3-16.2 and 11.2-13.8); the table is used, and it was read from the published PDF, not from the abstract. Table 5 is also laid out with its columns displaced in the PDF, and the assignment used here is the one that reproduces the paper's TSFA, TUFA and TEFA subtotals exactly for all four varieties. A THIRD STUDY DISAGREES and is not merged in: Han et al. 2024 (Food Chem. X 24:101816) report a commercial Chinese watermelon seed oil at palmitic 8.46, stearic 5.83, oleic 9.60, linoleic 73.68 % - far less saturated than any other work on the species - together with an iodine value of 63.65 g I2/100 g. That iodine value cannot belong to a 73.7 % linoleic oil, whose composition implies roughly 136, so the paper's physicochemical table is internally inconsistent and neither its profile nor its constants are used here. It is flagged because it ranks high in literature searches for this oil. FOR SOAP: 26-31 % of this oil is palmitic plus stearic, which is high for a seed oil and makes watermelon behave more like a hardening oil than its linoleic content suggests; the balance is mostly linoleic, so it still carries the usual rancidity risk. The recorded composition implies about 193 mg KOH/g by stoichiometry, within a fraction of a percent of the centre of the published saponification band. Melting point not reported. Egusi oil is sometimes given as a synonym, but egusi is also pressed from Cucumeropsis mannii, a different species, so it is not listed as an alias. INCI not verified against CosIng in this session, so left null.

This record is part of the open ingredient database (ODbL). Found an error or a better primary source? The dataset is on GitHub; corrections with a document link are welcome.