How to Read a Food-Grade GDL COA: Assay Basis, Moisture and Lot Checks

A GDL certificate of analysis is useful only when its numbers can be tied to the right product, the right lot, and the right specification. “Assay 99%” on its own cannot tell you whether the result is reported as received or on a dried basis, whether it is a result for your shipment or a catalogue limit, or whether the method matches the market you serve. That makes a review auditable.

For a first review, use this order: match the identity and lot; identify the applicable legal or customer specification; align the assay basis; compare each result with the limit and method; then resolve missing or ambiguous data before release. JECFA’s food-additive monograph sets a useful reference point, but it does not replace the requirements of the destination market, your customer, or your own approved specification. A supplier’s sheet is not automatically a batch COA.

Related reading on GDL

For the process side of GDL — sourness, whey release and coagulant route choice — see diagnosing sour or weeping GDL silken tofu and GDL versus calcium sulfate for tofu.

Start with identity and traceability, not the headline purity

The label may say glucono delta lactone, gluconolactone, GDL, or E575. Check that the material is the food-grade ingredient you intended to buy, with CAS 90-80-2 and, where used, INS 575. JECFA describes it as “the cyclic 1,5-intramolecular ester of D-gluconic acid.” That identity matters because similar commercial names can describe a blend, a different gluconate, or a formulated product designed for a particular process.

A batch COA should connect unambiguously to the shipment. Look for the manufacturer’s product code and grade, lot or batch number, manufacturing or retest date where provided, test date, issue date, and the supplier’s document identifier. The same lot identifier should appear on the bag label or pallet paperwork. If the certificate says only “GDL food grade” and has no lot number, it is not yet evidence about the material delivered to your dock.

Establish who made the material. A trader may issue its own certificate or forward the manufacturer’s COA; either can work if the document identifies who performed or authorized the tests. If reissued by a distributor, request the original lot record or a traceable explanation. Confirm that producer, facility, product code, and batch agree across the COA, label, and shipping documents. “Factory direct” is not traceability.

Treat specification limits and batch results as different documents

A product specification describes the agreed acceptance limits or targets for a grade. A COA reports results for a particular batch, or sometimes states that a parameter was not tested. Suppliers often place a grid of limits on their product page and call it a COA; buyers should request the dated, lot-specific report instead.

“Assay: not less than 99.0%” is a limit. “Assay: 99.7%” is a batch result only if linked to the lot and supported by an identified method. “Pass” can work for a controlled incoming check when the limit and method are clear, but numeric results are more useful for qualification, trend analysis, and supplier comparison.

JECFA’s monograph for glucono-delta-lactone specifies an assay of at least 99.0% on the dried basis, loss on drying not more than 1.0% (105°C for 2 hours), sulfated ash not more than 0.1%, reducing substances not more than 0.5% expressed as D-glucose, and lead not more than 2 mg/kg. These are source-specific JECFA limits—not universal purchase criteria. The monograph records evaluation in 1998 and publication in the 2006 Compendium; cite the edition adopted in your quality agreement.

The European Union’s E575 purity criteria use a different presentation, including assay on an anhydrous basis, water not more than 0.2%, reducing substances, and a lead limit of 1 mg/kg. The U.S. regulation at 21 CFR 184.1318 recognizes glucono delta-lactone under specified conditions and references the Food Chemicals Codex (FCC) third edition. These are not interchangeable checklists. Confirm the rule and edition applicable to your product category, destination, and customer before comparing a supplier’s result with a limit.

The assay-basis trap: dried, anhydrous, or as received

Assay values cannot be compared safely until their basis is understood. A dried-basis result removes the measured moisture contribution mathematically. An anhydrous-basis result is likewise corrected for water under the referenced method or specification. An as-received result describes the sample in its current condition. Those figures can differ even when they refer to the same material.

Suppose a lot has 0.5% water and an assay of 99.2% as received. A simplified moisture correction gives about 99.7% on a dry basis: 99.2 ÷ (1 − 0.005). This illustration assumes the water measurement and assay basis are compatible and that the specified correction is appropriate. It is not a substitute for the method in the governing standard. If the COA prints “99.2%” with no basis, do not silently treat it as a JECFA dried-basis result.

Ask the supplier to state the basis beside the value, the analytical method, and whether moisture is measured separately or accounted for in the assay calculation. Keep the original reported figure in your record. If your internal specification is on a dry basis and the certificate is as received, document the calculation, inputs, rounding rule, and approval. Do not “correct” a result when the standard defines a specific method or calculation that differs from your shorthand formula.

Water and loss on drying are related but not identical labels. A Karl Fischer water result measures water by a specified chemical method. Loss on drying measures mass lost under defined heating conditions; that loss can include volatile material as well as water. A supplier sheet that lists both values is not necessarily contradictory, but you need the methods and conditions before treating one as a substitute for the other. For GDL, moisture can also change after opening or during poor storage, so a certificate at manufacture does not guarantee the condition of a bag exposed to humid air in the warehouse.

Build a practical parameter review around your risk

Do not demand every conceivable test from every lot if it adds cost without changing a decision. Start with the applicable specification and your use. For a typical food ingredient qualification, buyers commonly review the following fields where they are required by the adopted standard or purchase agreement:

JECFA includes lead among its stated specifications; a buyer may need additional contaminants based on local law, product category, customer standards, manufacturing route, or hazard analysis. A blank field does not prove a contaminant is absent. “ND” is also incomplete without a method and a detection or quantification limit. Ask whether a result is below the reporting limit, what that limit is, and whether the result is batch-tested, periodically monitored, or supported by a supplier declaration. Do not confuse an “ND” value with zero.

Physical properties deserve attention when the application is sensitive to mixing or distribution. A GDL grade with the right chemical assay may still behave differently if particle-size distribution, moisture pickup, or dissolution rate changes. For silken tofu or other applications where local acidification affects texture, the supplier’s granulation and process guidance can inform a plant trial. That is an application qualification question, however; it should not be inferred from assay alone or used to rewrite the legal purity specification.

Sourcing candidate for buyer review

One Alibaba listing for food-grade GDL is linked below as an RFQ starting point. It is a sourcing route, not a qualification — the COA checks above still decide whether a lot is approved.

What a COA cannot tell you by itself

A passing COA does not establish that the supplier is approved, that the ingredient is suitable for every food, or that the final food complies with its labeling and use rules. In the United States, 21 CFR 184.1318 provides conditions for GDL as a direct food substance; in other markets, different food categories, purity criteria, or customer specifications may apply. Regulatory permission is not proof that a specific batch meets your contract. Conversely, an excellent assay does not settle allergen controls, manufacturing aids, food-safety systems, authenticity, or traceability.

The COA also rarely proves the complete manufacturing history. If your qualification requires origin, fermentation organism, GMO status, process aids, allergen statements, or country-specific documentation, request the relevant controlled declaration separately. Ask for the current specification, food-safety certificate, allergen statement, regulatory statement, and traceability documents as needed; check scope and dates rather than accepting logos alone. A generic certificate for a company does not automatically cover the plant or product you are buying.

When quality risk justifies it, compare a retained supplier sample or incoming lot against an agreed method at your own lab or an accredited third-party laboratory. Define sampling plans before ordering. One drum or bag may not represent a heterogeneous or transit-damaged lot. Agree on the dispute procedure, independent lab, sample retention, and time window in the purchase terms. This turns COA review into a control before a problem occurs.

A receiving workflow that closes the gap

Use five steps: freeze the identity (grade, manufacturer, product code, CAS/INS, destination, and food category); attach the acceptance standard and version to the purchase order; obtain the current specification and offered lot’s COA; verify each result’s method, basis, units, and limit; then record approval, conditional acceptance with a named action, or hold.

For repeat orders, preserve a baseline rather than re-reading every certificate from scratch. Track assay basis, moisture or loss on drying, key contaminant results, lot-to-lot variation, and any supplier method changes. Define which tests are required on every lot and which can be reviewed at a justified frequency. Any change to manufacturer, plant, grade, method, or product code should trigger a documented review. A supplier notification alone is not a technical assessment.

If a COA is incomplete, ask a specific question instead of requesting “all documents.” For example: “Please issue the lot-specific COA for lot X, state whether assay is reported on a dried or as-received basis, identify the assay method, and provide the numeric lead result with units and reporting limit.” A precise request is easier to answer and audit. If the seller cannot provide the information, decide whether your quality system permits an alternative test or requires holding the order.

Send a focused request: grade, quantity, destination, food category, packaging, required specification, and documents. Request a sample after commercial and regulatory fit is plausible, and set acceptance criteria before testing. This avoids comparing price quotes for material that cannot meet the plant’s documentation or performance needs.

The decision buyers should make

Approve a GDL lot only when identity, traceability, specification, and batch evidence align. The headline assay is one line in that decision. Its basis matters; water and loss on drying are method-defined; contaminant reporting needs units and limits; and a certificate must refer to the actual lot. Use JECFA, EU, U.S., or customer requirements as applicable, not as interchangeable badges. If evidence is missing, state exactly what is missing and hold or qualify the lot under your documented process.

That discipline makes GDL purchasing repeatable. It also gives procurement, quality, and production a shared reference when a supplier changes a method, a lot moves outside trend, or a finished product behaves differently. A COA is most valuable when it closes those gaps.

References

FAO/WHO Joint FAO/WHO Expert Committee on Food Additives (JECFA). “Glucono-delta-lactone.” Combined Compendium of Food Additive Specifications, Monograph 1 (2006). Official monograph PDF. https://www.fao.org/fileadmin/user_upload/jecfa_additives/docs/Monograph1/Additive-202.pdf. Evaluation recorded in 1998. European Commission. Commission Regulation (EU) No 231/2012, specifications for food additives listed in Annexes II and III to Regulation (EC) No 1333/2008; E575 glucono-delta-lactone. EUR-Lex consolidated regulation. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A02012R0231-20180527.

U.S. Electronic Code of Federal Regulations. 21 CFR §184.1318, Glucono delta-lactone. eCFR. https://www.ecfr.gov/current/title-21/chapter-I/subchapter-B/part-184/subpart-B/section-184.1318. U.S. Food and Drug Administration. Substances Added to Food (formerly EAFUS), glucono delta-lactone listing and regulatory references. FDA database. https://www.cfsanappsexternal.fda.gov/scripts/fdcc/?set=FoodSubstances.