06 Identity & adulteration

How we prove a soy protein lot is the isolate

Technical document 06 · Reviewed 2026-11-19 · Soy Protein Isolate

On this page
  1. The four checks that build identity
  2. Nitrogen, ash and the arithmetic of identity
  3. Where XRD and FTIR stop
  4. Adulteration routes worth testing for
  5. Specification lines that make identity testable
  6. Five questions that expose the file

Four independent measurements have to line up before a powder can be called soy protein isolate: the protein result and the arithmetic behind it, the non-protein fraction, the amino acid pattern, and a test that specifically detects soy. Each one can be satisfied by a different material, which is why the file needs all four.

In one line

Soy isolate identity is a set, not a number: the protein figure, the ash–fat–fiber profile, the amino acid pattern and a soy-specific test have to agree.

The four checks that build identity

Start with the protein result. Kjeldahl gives nitrogen; the N × 6.25 factor turns it into protein; the dry basis removes the effect of moisture. The 90.0% floor is a statement about protein, not about soy, so this layer answers only the first question.

Next comes everything that is not protein. Moisture at ≤7.0%, ash at ≤5.0% and fat at ≤1.0% are the limits that place the powder in the isolate band, and fiber plus carbohydrate by difference completes the mass balance. A concentrate carries more fiber and more ash because the process stops before those fractions are washed out.

The amino acid pattern is the third check. Soy protein is rich in aspartate, glutamate, arginine and lysine and comparatively low in methionine and cysteine. Adding whey, pea or gluten protein shifts that pattern; adding a non-protein nitrogen source leaves it unchanged.

The fourth check is a soy-specific method. PCR is used to look for soy DNA in the retained sample, while ELISA detects soy protein and can be applied to the powder or the finished matrix. A positive result establishes that soy material is present; it does not establish the grade. The protein and ash–fiber layers are what place the material in the isolate band.

Nitrogen, ash and the arithmetic of identity

The 6.25 factor rests on the assumption that nitrogen is about 16% of soy protein; a 90.0% result therefore corresponds to roughly 14.4% nitrogen on the same dry basis. At the catalogue ceilings, ash is 5.0% and fat 1.0% against 90.0% protein, so the mineral and lipid load is small by design. A sheet showing 90% protein beside 8% ash and 6% fiber is describing something other than an isolate.

A second arithmetic check is the dry-basis mass balance. Protein at 90.0% plus ash at 5.0% plus fat at 1.0% leaves about 4% for fiber and carbohydrate. At 65.0% protein, a concentrate leaves about 35% for the other components, many of them fiber and carbohydrate. This gap separates the two grades more reliably than the protein number by itself.

The nitrogen gap is the third check. The amino acid sum, expressed as protein, usually falls a little below the Kjeldahl result because acid hydrolysis loses part of the tryptophan and cysteine. Two or three points of difference can be explained analytically; five or more suggests nitrogen that is not protein nitrogen.

The factor can also disguise a genuine lot. A certificate using 5.71 instead of 6.25 will show about 82.2% for a material that is really 90.0%. Nothing is wrong with the powder; the convention is different. Settle the factor before using any protein result as identity evidence.

Where XRD and FTIR stop

Spray-dried soy isolate is amorphous, so an X-ray diffraction scan shows a broad halo rather than sharp peaks. That halo is consistent with a spray-dried protein, though it does not identify soy. Sharp reflections point to a crystalline addition: an amino acid used to lift the nitrogen figure, an inorganic salt, a sugar that has crystallised in storage, or a mineral carrier.

FTIR gives a functional-group fingerprint. Absorbance near 1650 cm⁻¹ (amide I) and 1545 cm⁻¹ (amide II) comes from the protein backbone; the carbohydrate region carries fiber and residual sugar signals. FTIR confirms the material class quickly, but it cannot tell soy from whey or pea, and it does not quantify protein. Use it as a screen.

What each technique can settle is described in the XRD and FTIR method notes. Neither replaces the nitrogen result, the amino acid profile or the soy-specific test.

Adulteration routes worth testing for

Adulteration routes · what changes and which test catches it
RouteWhat it changesTest that catches it
Concentrate sold as isolateProtein below 90%, fiber and ash upProtein, ash, crude fiber, amino acid profile
Other protein addedAmino acid pattern moves away from soyAmino acid profile, PCR or ELISA
Maltodextrin or starchProtein down, carbohydrate by difference upProtein, carbohydrate and fiber, iodine test
Non-protein nitrogenKjeldahl up, amino acid sum flatAmino acid profile, melamine by LC-MS/MS
Added amino acidsProtein up, amino acid pattern distortedAmino acid profile, mass balance
Soy hull or fiberFiber up, protein down, ash upCrude fiber, ash, protein

The first route is commercial: a lower grade sold as the higher one. The rest matter for safety and the label, which is why a protein result on its own cannot carry an identity file.

Specification lines that make identity testable

  • “Soy protein isolate (Glycine max), spray-dried, 80 mesh; agglomerated or instantized if required.”
  • “Protein at ≥90.0% dry basis, Kjeldahl with N × 6.25, excluding added amino acids, vitamins and minerals.”
  • “Moisture ≤7.0%, ash ≤5.0%, fat ≤1.0%, pH 6.5–7.5, each with method and basis.”
  • “Crude fiber and carbohydrate by difference reported; no other protein or nitrogen source added.”
  • “Soy allergen statement; non-GMO or identity-preserved status where the label relies on it.”
  • “PCR or ELISA on request; retained sample traceable to the lot on the drum.”

If the specification covers only the protein, moisture and ash lines, it describes a white powder rather than soy protein isolate.

Five questions that expose the file

  1. Which nitrogen factor and basis are on the certificate, and can you show the nitrogen result itself?
  2. Will you provide a lot-specific amino acid profile, and does its total reconcile with the Kjeldahl result?
  3. What are the ash, fat and crude fiber results, and do they sit in the isolate band?
  4. Has the lot been screened for melamine or another non-protein nitrogen source, and by what method?
  5. Do you have a soy-specific PCR or ELISA result, and does the retained sample carry the same lot code as the drum?

These are ordinary questions for a protein supplier. A supplier who cannot answer them is offering a material, not a documented one. The COA page shows how the numbers fit together; the isolate-versus-concentrate page sets out the grade arithmetic; the compliance page covers the allergen and label file; and the product page carries the catalogue limits.

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