What the COA for soy protein isolate actually contains
Soy protein isolate is traded on a 90% protein figure, yet that number is an arithmetic result, not a direct measurement. A certificate that omits the nitrogen factor and the moisture basis has left the reader unable to reconstruct it. The rest of the sheet matters too: moisture, ash, fat, the fiber and carbohydrate left over, the mesh, the contaminants and the microbiology. The catalogue states what the material must meet; the lot certificate records what a specific batch did.
A soy isolate certificate is complete only when the 90% protein figure arrives with its N × 6.25 factor, its dry basis and the ash, fat and moisture that account for the rest of the powder.
The parameter list
CAS 9010-10-0 identifies the soy protein material; it does not point to a discrete compound with one molecular formula. A useful certificate therefore describes the powder through several measurements at once: the protein that the nitrogen calculates to, the water, ash and fat beside it, the flow and dispersion behaviour, and the contaminants and organisms that must be absent. The first table gives the catalogue limits for the 80 mesh grade. These are the levels agreed in the specification and repeated for each lot.
| Parameter | Method | Release limit |
|---|---|---|
| Appearance | Visual | Off-white to light cream powder |
| Protein (dry basis) | Kjeldahl, N × 6.25 | ≥90.0% |
| Moisture | Loss on drying | ≤7.0% |
| Ash | Gravimetric, furnace | ≤5.0% |
| Fat | Solvent extraction | ≤1.0% |
| pH (5% suspension) | Potentiometry | 6.5–7.5 |
| Particle size | Sieve analysis | ≥95% through 80 mesh |
| Parameter | Method | Release limit |
|---|---|---|
| Heavy metals | ICP-MS | Pb ≤3 · As ≤2 · Cd ≤1 · Hg ≤0.1 mg/kg |
| Microbiology | Plate count and enrichment | APC ≤10,000 cfu/g · yeast and mould ≤100 cfu/g · E. coli and Salmonella absent |
The grade belongs in the product name on the certificate. “Soy protein” without “isolate” covers a family with different limits, and a buyer comparing two sheets needs to know which member of the family each one describes. The isolate-versus-concentrate distinction is set out on the grade comparison page.
Kjeldahl, the factor and the dry basis
The Kjeldahl method reports nitrogen; the protein figure is a conversion from it. For soy protein the Codex factor is 6.25, so 90.0% protein corresponds to about 14.4% nitrogen on the same dry basis. Some older certificate templates still use the Jones factor of 5.71, and the same nitrogen result would then read about 82.2%. Without the factor on the page, two protein figures are not comparable.
The basis is the other half of the calculation. The catalogue figure is dry basis. At the 7.0% moisture ceiling, 90.0% dry basis is 83.7% as-is, which is the figure a serving calculation needs. The Codex standard for soy protein products also works on a dry basis and excludes added vitamins, minerals, amino acids and food additives. A powder that reaches 90% by adding amino acids is not the same material as one that reaches it with soy protein alone, even when both sheets print 90%.
The percentage is only portable when the factor and the basis travel with it. Copying the percentage alone into a purchase order leaves the arithmetic unresolved. The loss on drying entry explains why the moisture figure moves the as-is number.
Catalogue range, release limit and lot result
The release limit is the boundary the lot must clear. The lot result is the figure the bench obtained for that batch. A useful sheet shows both, with the method beside each line, because the phrase “conforms to specification” is a conclusion rather than a measurement. Two lots can pass the same ≥90.0% line at different distances above it, and the gap matters when a serving is built on the as-is figure.
The non-protein fraction is not one substance. On a dry basis, 90.0% protein with 5.0% ash and 1.0% fat leaves about 4% for fiber and carbohydrate. If the ash and fat are reported as-is, the arithmetic at 7.0% moisture runs 83.7% protein plus 7.0% moisture, 5.0% ash and 1.0% fat, leaving about 3.3%. Both views are legitimate, but they cannot be mixed in one sentence; every line has to state its basis.
What this page calls a specification is the catalogue level, not a record of any batch. The certificate for the lot that ships is the batch-specific document. We do not publish lot results, lot numbers or customer names.
Where lots drift
The protein and ash figures follow the bean and the separation step. Crop, origin and storage change the protein and mineral load that reaches the plant. Alkaline extraction, isoelectric precipitation and washing decide how much non-protein fraction is removed; neutralisation decides part of the ash; spray drying decides the moisture, the mesh and the colour.
Particle size and dispersibility can move while the protein figure stays put. A lot can meet ≥90.0% and still sit at the coarse edge of the 80 mesh cut-off, wet slowly on a filling line and leave lumps in the bag. Colour can drift from off-white towards cream after a hotter drying run, and a beany or cereal note can strengthen with moisture uptake. Request the retained sample and the storage record; the protein number alone will show neither.
When an outside test earns its place
An outside test is worth ordering in three situations: a new source, a first shipment into a regulated market, or a label statement that the supplier's certificate does not support. A useful panel is targeted rather than exhaustive: it should close the specific gap the certificate leaves. Nitrogen by Kjeldahl or Dumas, the amino acid profile, crude fiber and ash, the ICP-MS contaminant panel and a soy-specific identity test can all run on the same retained sample.
An independent laboratory is also the practical route through a disagreement. Provide the retained sample, the certificate for the lot and the specification both sides agreed on, then ask for the few tests that distinguish the competing explanations instead of repeating the full panel. The quality page explains how the lot file and its retained sample are stored.
The lines that separate two offers
- The protein value, its conversion factor, its basis and the underlying nitrogen result.
- The ash, fat and moisture limits, each with its basis.
- The fiber and carbohydrate fraction, especially for an isolate.
- The mesh and bulk density, with the test method.
- The soy allergen statement and the cross-contact controls.
- The contaminant and microbiological panel, plus the identity evidence.
If two offers match on 90% protein but diverge on the other lines, they describe different materials. Behind the paper sits the record: the plant, the soybean source, the extraction and drying conditions, and the retained sample. The paperwork can be accurate and still belong to another container, so the lot code should match on the certificate, the pack and the retained sample.
For the ranges themselves, see the soy protein isolate product page. The grade choice is set out in the grade comparison, the identity tests in the identity page, and the certificate definition in the COA glossary entry.
Related documents
Send the specification you intend to use — grade, quantity and destination market — and we will reply with the batch COA, the technical data sheet and a quotation.
Request a quotation