DEG and EG testing of glycerin and propylene glycol: how to accept the raw material

A certificate of analysis does not close the DEG and EG risk in glycerin and propylene glycol. Supply chain checks, per-container sampling and a chromatographic limit test do more.

Glycerin and propylene glycol should be accepted on supply chain checks and a diethylene glycol (DEG) and ethylene glycol (EG) limit test run on your own samples, not on the supplier's certificate of analysis (CoA) alone. For drugs, FDA recommends sampling every container and applies a limit of no more than 0.10% for DEG and EG. Cosmetics rules do not go that far: Türkiye and the EU cap DEG traces at 0.1%, and how often you measure is a matter for your own risk assessment.

Why are these two raw materials singled out?

DEG and EG are industrial solvents that have caused fatal poisonings. FDA's May 2023 guidance traces cases from a 1937 US incident with 107 deaths to the present. In January 2023 WHO reported cases from at least seven countries and more than 300 deaths in three of them, mostly children under five.

The cases share a pattern. According to FDA, manufacturers did not run the full identity test that quantifies DEG in incoming material, and relied on the supplier's CoA. That CoA was often not the original manufacturer's document but a copy on a distributor's letterhead, and nobody knew how many times the material had changed hands.

Most of these cases involved oral liquid medicines, so it would be wrong to paint the same picture for cosmetics. But the raw material can come through the same supply chains, and FDA's identity testing requirement also covers topical drugs.

What do the rules require, and from whom?

SourceApplies toRequirement
FDA guidance (May 2023)Drugs for the US, including OTCIdentity test on each shipment and lot; DEG and EG no more than 0.10%; sampling every container recommended
EMA GMP questions and answersHuman medicines in the EUGlycerol, propylene glycol and low molecular weight macrogols treated as high risk; identity and limit test on each container unless justified
Turkish Regulation on Cosmetic Products, Annex II line 1370, Annex III line 186Cosmetics in TürkiyeDEG prohibited; traces from ingredients capped at 0.1%
Same entries added by Directive 2009/6/ECCosmetics in the EUSame prohibition and trace limit
TİTCK GMP Guideline (KDD-KLVZ-01, Rev.02)Cosmetics in TürkiyeSample per batch rather than per container, representative samples, analysis against acceptance criteria

Two details are easy to miss. The 2023 text of the Turkish regulation has no general line for EG as it does for DEG; EG appears as an impurity limit of 0.1% in the ethoxydiglycol entry. So your EG acceptance criterion has to be written into your own procedure. And EMA says the per-container test may be dropped for external-use medicines where a low risk is justified. Testing every container is not automatic for topical products, but skipping it needs a reason on file.

There is a recent example close to home. In a warning letter dated 21 January 2025, FDA wrote that an OTC manufacturer in Türkiye had failed to adequately test each shipment and lot of glycerin and propylene glycol with an identity test that includes the DEG and EG limit test. For any company shipping OTC products to the US, this is a live inspection topic.

How should raw material intake run, step by step?

  1. Document the supply chain. Identify the original manufacturer and any repackers and distributors in between. Check whether the CoA actually comes from the original manufacturer. FDA also recommends a CoA naming the original manufacturer for each lot shipment.
  2. Quarantine. Clause 11.2.3 of the TİTCK guideline holds raw material in quarantine until quality control approval. Record container integrity, closure and agreement between label and delivery note on receipt (11.2.2).
  3. Write the sampling plan by risk. Sample every container where there is a broker, where the material has been repacked, or where the source is new. FDA's reasoning is the wide variation it has seen from container to container in the same lot. If containers will be composited, assess in your procedure whether that could dilute the signal.
  4. Pick the right test. An infrared identity test recognises glycerin, but FDA says it is not suitable for detecting or quantifying DEG and EG. The compendial limit test is gas chromatography, and EMA keeps it as the official method. This determination should be done in an accredited laboratory; TİTCK's analysis guideline requires the principles of TS EN ISO/IEC 17025.
  5. Set acceptance criteria in advance. Your procedure should state the limit, the method and who makes the rejection decision for your product group.
  6. Release or reject, and retain samples. Starting material samples are kept for at least two years from product release (12.4.2). Move non-conforming lots to a separate area and notify the supplier in writing.

The same controls can extend to other at-risk components such as sorbitol, polyethylene glycol and polysorbates, which FDA covers under the same heading.

Frequently asked questions

Is a CoA stating "DEG absent" enough? No. It is the supplier's statement about a sample of that batch, and it may not show the method or the source. For more on its limits, see is a CoA enough.

Is accredited laboratory analysis required for every batch? Cosmetics rules do not spell that out. Testing first shipments, new suppliers and chains with brokers is easy to defend; frequency can drop with a documented rationale once a reliable history builds up. For OTC products, FDA expects each shipment and lot to be tested.

Does a "pharmaceutical grade" label remove the risk? Not on its own. WHO recommends both buying pharmaceutical-grade material from qualified suppliers and testing on receipt, and FDA reports a propylene glycol case where the label and CoA may not have matched the container's contents.

Where does Kozvex fit in this flow?

Kozvex's supplier verification records, before production, whether the delivered material is what was ordered and whether it matches the supplier's certificate. The on-site scan is a preliminary assessment. DEG and EG are determined in a contracted accredited laboratory, not in the field, and the report is kept in the batch record. The acceptance decision stays with your quality control unit. Detail is on the solutions and how we work pages.

Sources

Meeting request

Let us define which batches and raw materials you will measure.

The first meeting reviews your product, sales channels and the risks you face, and recommends a suitable method.

Request a meeting