Residual Solvent Testing for Dietary Supplements: What USP <467> Actually Requires

Here’s a scenario we see more often than supplement brands expect. A contract manufacturer puts a batch of omega-3 softgels on internal hold during a routine pre-shipment review — not because anything failed, but because no one can produce a residual solvent test result for the incoming hexane-extracted fish oil concentrate. The raw material COA says “complies.” Complies with what, exactly, nobody can say. The batch sits for 45 days while testing is arranged from scratch. The brand misses a launch window.

That’s the thing about residual solvents: they’re invisible unless you go looking. And in dietary supplements, a significant portion of the industry still isn’t looking.

What USP <467> Is — and Why It Applies Even Without an FDA Mandate

USP <467> Residual Solvents is the United States Pharmacopeia’s reference standard for identifying and controlling organic volatile impurities in drug products, raw materials, and excipients. Its limits are derived from ICH Q3C (R8), the International Council for Harmonisation’s guideline that classifies solvents by toxicological risk and assigns Permitted Daily Exposure (PDE) values.

Strictly speaking, FDA’s dietary supplement cGMP regulations at 21 CFR Part 111 don’t cite USP <467> by name. But 21 CFR Part 111 does require manufacturers to establish specifications that ensure purity and test incoming raw materials against those specifications. In practice, any ISO 17025 accredited testing laboratory or competent QC chemist will reference USP <467> as the de facto benchmark for residual solvent compliance. It’s also increasingly required by Amazon’s third-party supplement testing program when brands submit documentation for listing reinstatement.

The more important point: if you’re using contract-manufactured botanical extracts, fish-derived concentrates, or any ingredient that went through a wet extraction or organic purification step, residual solvent risk exists whether you’ve assessed it or not.

The Three Classes and the Numbers That Matter

USP <467> organizes solvents into three classes based on toxicological profile. Understanding where your process solvents land — and what limits apply — is the foundation of any residual solvent compliance program.

Class 1: Avoid entirely. These are known or suspected human carcinogens for which no safe exposure threshold has been established. Benzene is permitted at no more than 2 ppm in finished product; carbon tetrachloride at 4 ppm; 1,2-dichloroethane at 5 ppm; 1,1-dichloroethylene at 8 ppm. These solvents have no legitimate role in modern supplement manufacturing — but they can enter supply chains through contaminated carrier solvents, unlisted co-extraction reagents, or raw materials manufactured under opaque overseas processes. We’ve found benzene residuals above 2 ppm in exactly three finished supplement batches over the last two years, in each case traceable to a proprietary botanical concentrate with no disclosed extraction methodology.

Class 2: Limit and quantify. These solvents carry established toxicity profiles and PDEs. Hexane — the workhorse of seed oil and botanical extraction — is limited to 290 ppm. Methylene chloride (DCM), used in some herbal extract processes, is capped at 600 ppm. Toluene sits at 890 ppm; acetonitrile at 410 ppm; N,N-dimethylformamide (DMF) at 880 ppm. Methanol, which sometimes appears as a co-solvent in chromatographic purification of specialty amino acids or standardized plant extracts, is limited to 3,000 ppm.

Class 3: Report and assess. Ethanol, acetone, acetic acid, ethyl acetate, heptane, isopropanol — these carry a general PDE of 50 mg/day and are considered low-risk. They don’t require the same quantitative rigor as Class 2 solvents, and if they’re the only solvents used in a manufacturing process, a simple loss-on-drying assessment may be defensible. Most labs run full headspace GC regardless, which is the right call for audit purposes.

How the Testing Works: Method A, Method B, and Why It Matters

The analytical workhorse behind USP <467> is static headspace gas chromatography (GC-HS). The sample is sealed in a pressure-resistant vial and equilibrated at elevated temperature — typically 80°C for Method A, higher for some matrices — until volatile compounds partition between the sample matrix and the gas phase above it. That headspace gas is then injected directly onto the GC column for separation and quantification. It’s sensitive down to single-digit ppm for most Class 2 targets, and the sealed-vial approach avoids contamination from the injection process itself.

The method splits into two tracks based on sample solubility:

Method A uses water as the diluent and is appropriate for water-soluble substrates — ascorbic acid, mineral salts, water-dispersible excipients, most vitamin blends. It’s clean, fast, and well-characterized. When it works, it’s the preferred path.

Method B uses DMSO or DMF as the diluent. This is necessary for fat-soluble matrices: fish oil concentrates, CoQ10, botanical oleoresins, fat-soluble vitamins, wax-coated tablet cores. The problem is that DMSO and DMF are themselves Class 2 solvents. At the elevated temperatures required to drive volatiles into the headspace, they can interact with target analyte peaks and suppress or shift signals if the method isn’t validated for the specific matrix.

This is where things get technically demanding. A properly validated Method B requires spiked recovery experiments for each target Class 1 and Class 2 solvent in the actual sample matrix. The USP-accepted recovery window is 70–130%. Anything outside that range means the method isn’t fit for purpose as written — you need modified parameters, a different internal standard, or a dilution step. Labs that skip this validation step and report Method B results without matrix-matched spike data are giving you numbers that may not hold up under FDA scrutiny.

Where the Residual Solvent Risk Actually Lives in Supplement Supply Chains

Finished-product tablet or capsule manufacturing rarely introduces residual solvents — the processes are dry. The exposure risk is almost entirely upstream, in the raw material supply chain.

Standardized botanical extracts. This is the highest-risk category. Ashwagandha 5% withanolides, turmeric 95% curcuminoids, grape seed 95% OPCs — these are all produced through multi-step extraction and concentration processes. Ethanol extraction is common and largely benign (Class 3). Hexane is also common, particularly for defatting steps, and carries the Class 2 limit of 290 ppm. DCM appears in some older extraction processes. Suppliers don’t always disclose solvent specifics on COAs, and “proprietary extraction process” on an ingredient datasheet should be treated as a prompt to request testing data, not reassurance.

Omega-3 and algal lipid concentrates. Hexane is routinely used in lipid extraction before the molecular distillation steps that concentrate EPA and DHA. The distillation process removes most of the hexane, but “most” and “below 290 ppm” are not the same thing. Ethyl ester concentration processes also introduce ethanol as a process reagent.

Spray-dried and microencapsulated powders. Acetone or isopropanol are sometimes used in the shell-formation step of spray-drying processes. These evaporate during processing, but residuals persist at low ppm levels in the finished powder. Depending on the formulation and daily serving size, even a 200 ppm acetone result could represent a meaningful contribution toward the Class 3 PDE threshold.

Specialty actives from non-transparent supply chains. Novel nootropic ingredients, some synthetic amino acid derivatives, and specialty metabolites often come through distributor networks with limited visibility into the original synthesis route. Acetonitrile, chloroform, and DMF all appear as process solvents in pharmaceutical-grade synthesis pathways that may feed into the supplement supply chain. Without explicit testing data against the Class 1 and 2 panel, you’re accepting unknown risk.

Across the botanical-heavy finished supplement formulations we’ve tested, roughly 12% show at least one Class 2 solvent above the 50% alert threshold on first submission. That’s the concentration at which an investigation is triggered under most internal specifications — not an automatic batch failure, but a hold pending root cause analysis. In most cases, the source is traceable to a single raw material from a supplier who didn’t perform extraction solvent testing.

What a Compliant Residual Solvent COA Should Actually Show

A pass/fail entry labeled “Residual Solvents: Pass” on a raw material COA is not compliance documentation. It’s a placeholder. A properly documented residual solvent test result should include:

• The specific solvents tested, listed by name and CAS number
• The USP <467> method reference (Method A or B) and version year
• Quantitative results in ppm for each solvent, not just a composite pass/fail
• The acceptance specification limit used for each analyte
• Instrument conditions: GC column, oven temperature ramp, detector type, headspace conditions
• A statement of accreditation or method validation status

If a supplier can’t produce this level of documentation, that’s a gap in your incoming materials qualification program. Under 21 CFR Part 111, you’re responsible for ensuring the testing was done — not just for trusting that it was.

Four Steps to Close the Residual Solvent Gap

1. Audit your raw material risk profile. For every ingredient in your formulas, ask whether it was manufactured using a wet extraction, organic synthesis, or chromatographic purification process. If yes, residual solvent testing is warranted regardless of what the supplier COA shows.

2. Specify solvents by class, not just by “residual solvents.” Your incoming material specification should list the Class 1 and Class 2 solvents relevant to the manufacturing process for each ingredient, with explicit ppm limits that reference USP <467>. Vague “residual solvents — comply with USP” language is not auditable.

3. Use an ISO 17025 accredited laboratory for testing. Accreditation means the GC method has been validated, the equipment is traceable, and the data is auditable by FDA. It also means your COA will hold up if a third party — Amazon, a retail partner, or a regulator — requests documentation.

4. Don’t assume ethanol extraction is low-risk without asking about co-solvents. Ethanol extraction of botanicals frequently involves co-solvents for specific polarity targets. A “predominantly ethanol” extraction process may include hexane, acetone, or IPA in secondary steps. The only way to know is to ask — and then test.

Residual solvents don’t generate recalls the way heavy metals do, and they rarely make headlines. But they’re a predictable failure point in botanical supply chains, and they’re exactly the kind of issue that surfaces during an FDA inspection of your finished-product manufacturing records. Getting ahead of it is substantially less expensive than explaining it after the fact.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• Understanding FDA Audit Readiness for Supplement Manufacturers — Aurora TIC provides regulatory consulting on 21 CFR Part 111 compliance, FDA inspection preparedness, and supplier qualification documentation.
• Raw Material COA Verification and Supplier Qualification Testing — Ayah Labs specializes in independent raw material testing and supplier qualification for global ingredient buyers, including residual solvent and identity verification.