Huile d'andiroba

Les notes sourcées ci-dessous, ainsi que les titres des documents qu’elles citent, sont conservées dans leur langue d’origine. Une référence traduite ne peut plus être retrouvée, et un résumé technique traduit n’est plus le résumé qui a été vérifié par rapport au document.

Two independent, directly measured saponification values bracket the range: 189.36 +/- 3.56 mg KOH/g in Lozano-Garzon et al. 2023 (Biomolecules 13(6):985, Table 1, modified USP titration, Colombian Amazon oil) and 194 mg KOH/g in Xavier et al. 2025 (ACS Omega 10(49):60374, Table 1, AOCS Cd 1c-85, Brazilian oil). The iodine range is from the same two documents, 59.2 +/- 2.98 and 71 g I2/100 g (AOCS Cd 3a-94), and that spread is wider than is comfortable: the composition recorded here implies about 67 g I2/100 g by stoichiometry, so the ACS Omega end is the better-supported one and the Biomolecules figure looks low, which is consistent with the fact that their own fatty-acid column for andiroba accounts for only 80.9 % of the fat. The profile is Araujo-Lima et al. 2018 (Oxidative Medicine and Cellular Longevity 2018:3246719), Table 1, three commercial Brazilian crabwood seed oils by GC-FID; typical is their Oil 1 and min/max span all three, which differ by less than a percentage point on every acid. Margaric acid (C17:0, 0.12-0.13 %) is dropped for want of a schema key, so the profile sums to 99.88 %. A warning about that same paper, because it is the one a reader will find first: the iodine and saponification values printed in its Table 1 were, by its own methods section, CALCULATED from the profile rather than titrated, and they are printed against the wrong labels, iodine value 196.6 and saponification value 59.5-59.7. Neither is possible for a C16/C18 triglyceride; the lowest saponification value such a fat can have is around 170, and an oil about 61 % unsaturated, nearly all of it monoene, cannot absorb 196 g I2/100 g. Read the two rows swapped and the 196.6 lands within two points of the 194.8 that the same table's composition implies. Those figures are not used here, and anyone copying them into a calculator would be dosing lye from a number roughly three times too small. The ACS Omega paper's own GC-MS resolved only four acids (oleic 44.68, palmitic 37.30, stearic 7.44, and a 10.57 % peak labelled linolenic whose stated identification, 9,12-octadecadienoic acid, is in fact linoleic), so it corroborates the shape of the profile but is too coarse to use. Andiroba is bought for its unsaponifiable fraction of limonoids, described in Oliveira et al. 2018 (BioMed Research International 2018:5032816), but none of the three quantitative documents printed a percentage for it, so unsaponifiablePercent is null. For soap: this is a palmitic-heavy oil, about 38 % saturated with palmitic alone near 28 %, so it behaves more like a mild palm oil than like a soft seed oil and contributes real hardness at modest percentages. Crude lots carry high free fatty acid, acid value 8.90 mg KOH/g in ACS Omega and 8.31 in Biomolecules, both well above refined-oil levels, which is alkali consumed as free acid rather than as triglyceride and worth allowing for in a small batch. No melting point is published in any of the documents read. INCI not verified against CosIng in this session.

Cette fiche fait partie de la base d’ingrédients ouverte (ODbL). Vous avez trouvé une erreur ou une meilleure source primaire ? Le jeu de données est sur GitHub ; les corrections accompagnées d’un lien vers un document sont bienvenues.