GIIAVA Group Technical Team · 2026-06-18
Four numbers tell you almost everything about a lecithin batch: acetone insolubles (AI), hexane insolubles (HI), phosphatidylcholine (PC), and moisture. Each has a correct test method; each has a cheaper shortcut that gives misleading results. This page explains both, so a buyer can read a COA without having to take the supplier's word for it.
The active ingredient in lecithin is the family of phospholipids: phosphatidylcholine (PC), phosphatidylethanolamine (PE, historically called cephalin), phosphatidylinositol (PI), phosphatidic acid (PA), and phosphatidylserine (PS). All of them share two properties: they are soluble in hexane and insoluble in acetone.
AI exploits the acetone half of that pair. The standard method (AOCS Ja 4-46) takes a sample, washes it with acetone, dries the residue, and weighs it. The residue is the “acetone-insoluble” fraction — the part of the sample that didn’t dissolve. For a clean lecithin, that residue is almost entirely phospholipid.
For a less clean lecithin, the residue is also silica carriers (the industry workhorse is Sipernat, a precipitated silica brand widely used as a flow agent), mineral fillers, undigested solids, or other carriers that the supplier added intentionally or carried through from process. The AI test cannot tell phospholipid from filler. They both fail to dissolve in acetone, so they both count.
This is why a supplier can quote a powder at “95% AI” that is actually 65% phospholipid plus 30% silica. The number is technically correct. The implication — “95% active phospholipid” — is not.
The second method (AOCS Ja 3-87) is the mirror image. Wash the sample with hexane, dry the residue, weigh it. Phospholipids dissolve in hexane, so they wash away. What stays behind is the silica, the mineral, the carrier — everything that wasn’t phospholipid in the first place.
HI is essentially a direct measurement of the non-phospholipid filler load. For a clean liquid lecithin or a clean de-oiled powder, HI should be under 0.3%. For lyso-lecithin carried on silica (the powder form of a normally liquid product), HI can be 25–30% — but in that case the carrier is declared on the spec sheet and is the entire point of the product. Hidden silica is the problem. Declared silica is just chemistry.
True phospholipid content (%) = AI − HI
That subtraction is the punchline. The arithmetic is what strips out the carrier and leaves only the active ingredient. A serious COA reports both numbers explicitly so the buyer can do the subtraction themselves. A COA that reports only AI is either lazy or hiding something.
Quick reference targets:
| Grade | AI | HI | True phospholipid (AI−HI) |
|---|---|---|---|
| Liquid lecithin | ≥60% | <0.3% | ≥~59.7% |
| De-oiled powder lecithin | ≥95% | <0.3% | ≥~94.7% |
| Lyso-lecithin on silica (powder) | ~95% by design | ~25–30% (declared carrier) | ~65–70% |
Within the phospholipid family, PC is the workhorse. It drives emulsification, instantization, wettability, and the choline nutritional claim. Standard de-oiled powder lecithin runs PC ≥22% (i.e. of the bag, not of the phospholipid fraction). High-PC specialty grades can exceed 25% for applications where the functional or nutritional density matters — for example, choline-claim supplements or sensitive baked-good systems.
PC is measured by HPLC, not by any solubility method. If a supplier quotes “PC content” without an HPLC reference on the COA, ask for the chromatogram.
For most food ingredients, oven-drying (heat the sample to 105°C, weigh the mass loss, call it moisture) is the cheap default. For lecithin, it is wrong.
Lecithin degrades at temperatures above approximately 105°C — the phospholipids oxidize, the sample darkens, and the mass loss becomes a mix of evaporated water and thermal decomposition products. Whatever number comes out of the oven is not moisture — it is moisture plus an unknown amount of damage.
The correct method is Karl Fischer titration: a chemical reaction that quantifies water directly, with no heating, in a few minutes. Every credible lecithin COA reports moisture by Karl Fischer. If a supplier’s COA says “moisture: 0.8% (oven)”, the number is meaningless for this product class.
Acid value (AOCS Ja 6-55) indicates the level of free fatty acids — a measure of hydrolytic degradation. Peroxide value (AOCS Ja 8-87) indicates oxidative degradation. Both should be reported per batch and held within the published TDS bounds. Rising trends over consecutive batches suggest storage or handling problems upstream of you.
A serious lecithin supplier provides, on every batch:
A COA missing any of these — particularly HI or Karl Fischer moisture — is a COA that is hiding what would be revealed by the missing test.
Tell us the grade and the application. We’ll send a recent batch COA with every parameter above, the methods named, and the historical batch spread — not a generic spec sheet.
Talk to GIIAVA → See the 10 supplier questions →