USP <467> Residual Solvent Testing for Botanical Extracts: What Your Supplier's COA Isn't Telling You

A colleague forwarded a COA last month for a hexane-extracted saw palmetto concentrate from an Indian supplier — 45% fatty acids, heavy metals within USP <232>/<233> limits, total aerobic microbial count compliant. The residual solvent section? A single dash. Not tested.

That dash is a compliance gap, and it’s one we see routinely on botanical extract certificates of analysis that come through our Chicago-area receiving facility. Hexane carries a Class 2 residual solvent limit of 290 ppm under USP <467>. A 45% fatty acid concentrate requires substantial hexane exposure during extraction. What actually remains in that powder is unknown without testing — and “unknown” isn’t a specification. Under 21 CFR Part 111, your purity spec needs a number, not a hope.

Here’s what supplement brands sourcing botanical extracts need to understand: which solvents to watch for, which extracts carry the highest risk, and what an analytical testing laboratory actually does when it runs USP <467>.

---

What USP <467> Classifies — and Why the Solvent Class Determines Your Risk

USP <467> divides residual solvents into three classes based on toxicological risk, and the class determines both the acceptable limit and how seriously a deviation should be treated.

Class 1 solvents are known or suspected human carcinogens — benzene (2 ppm), carbon tetrachloride (4 ppm), 1,2-dichloroethane (5 ppm). No legitimate botanical extract process should use them, but trace Class 1 contamination from solvent impurities or cross-contamination in poorly controlled facilities is possible. Any detection warrants immediate rejection.

Class 2 solvents are where botanical extracts live. These are non-genotoxic animal carcinogens or compounds with demonstrated human toxicity at sufficient exposure. Limits are derived from permitted daily exposure (PDE) calculations — the amount a 50 kg adult can absorb daily over a lifetime without adverse effect. The Class 2 limits you’ll encounter most often in botanical extract manufacturing:

Solvent	USP Class 2 Limit	PDE
Hexane	290 ppm	2.9 mg/day
Chloroform	60 ppm	0.6 mg/day
Dichloromethane	600 ppm	6.0 mg/day
Methanol	3,000 ppm	30 mg/day
Acetonitrile	410 ppm	4.1 mg/day
Toluene	890 ppm	8.9 mg/day

Class 3 solvents — ethanol, ethyl acetate, isopropyl alcohol, acetone, acetic acid — have low inherent toxicity and a general limit of 5,000 ppm (0.5%). For botanical extracts processed with food-grade ethanol, Class 3 testing still needs to be documented, but failures are uncommon with standard drying protocols.

USP <467> is harmonized with the ICH Q3C guideline, now in its eighth revision (R8). Pharmaceutical-grade botanical extracts destined for both supplement and drug product applications are evaluated against both documents. If your supplier claims pharmaceutical-grade quality, ask for data against both standards.

---

Which Botanical Extracts Carry the Highest Residual Solvent Risk

The risk follows the extraction chemistry. Certain botanicals almost universally require non-aqueous, Class 2 solvents to achieve their standardized potency markers at commercial scale.

Saw palmetto extract (45–90% fatty acids): The lipid-rich berries demand a lipophilic extraction solvent. Hexane is the commodity standard. Supercritical CO2 extracts exist and are generally cleaner, but they carry a significant price premium and remain a smaller share of the market. Most hexane-extracted saw palmetto we test comes in under 290 ppm — but that result requires a test to mean anything.

Berberine HCl from Berberis or Phellodendron bark: Methanol is more efficient than ethanol for alkaloid yield, and many Chinese contract manufacturers use methanol-ethanol blends. The 3,000 ppm methanol limit is more forgiving than hexane’s 290 ppm, but inadequate drying or spray-drying at insufficient temperatures can still produce out-of-limit material.

St. John’s Wort extract (0.3% hypericin): Standard ethanol extraction is common, but methanol or methanol-ethanol blends are used by some suppliers to optimize hypericin concentration. The hyperforin-standardized variants introduce additional complexity — hyperforin degrades rapidly, and some extraction processes run more aggressive solvent conditions to maximize initial potency.

Ginkgo biloba extract (24% flavonol glycosides / 6% terpene lactones): Acetone and ethanol are both used. Some fractionation stages involve dichloromethane, which at 600 ppm is one of the tighter Class 2 limits. Given that ginkgo is one of the most globally traded botanical extracts, supply chain transparency is inconsistent.

Curcumin 95% from turmeric: Getting from whole turmeric to a 95% curcuminoid concentrate requires aggressive solvent extraction and concentration. Ethanol, ethyl acetate, and methanol are all used at various stages. The concentration process itself can increase residual solvent levels by reducing water activity, which is counterintuitive — more processed doesn’t mean cleaner.

Kava extract (30–70% kavalactones): Traditional water-extracted kava is generally clean. But acetone-based extraction is used by some manufacturers to achieve higher kavalactone standardization, and CO2-extracted kava oleoresins are increasingly common. Acetone is Class 3 (5,000 ppm limit), so it’s lower priority, but still worth confirming in writing.

---

How an Analytical Testing Laboratory Runs USP <467>

The analytical method is headspace gas chromatography — most commonly GC with a flame ionization detector (FID), though GC-MS is deployed when identity confirmation of a trace peak is needed. Headspace injection is essential: it isolates volatile compounds from the matrix without direct injection of organic solvents, which would degrade the column and produce interference patterns that mask real residuals.

USP <467> specifies two sample preparation procedures, and choosing the wrong one for a botanical matrix is a common source of non-USP-compliant results from supplier labs.

Procedure A applies to water-soluble samples. The extract is dissolved in water or a water/DMSO mixture, sealed in a 20 mL headspace vial, and equilibrated at 80°C for 60 minutes before headspace injection. This works for water-soluble powders and extracts, but it’s inadequate for lipophilic matrices — the solvent residuals partition poorly into the headspace phase, producing falsely low results.

Procedure B is required for non-water-soluble samples, which describes the majority of botanical extracts — particularly lipid-rich concentrates and resinous extracts. The diluent shifts to N,N-dimethylformamide (DMF) or dimethyl sulfoxide (DMSO), which solubilize lipophilic materials without contributing volatile interference. The equilibration conditions are the same: 80°C for 60 minutes, headspace injection onto a DB-624 or HP-1 type capillary column.

An ISO 17025-accredited analytical testing laboratory running USP <467> applies lot-specific system suitability criteria before every analytical sequence — verifying resolution between solvent peaks, confirming linearity across the calibration range, and validating sensitivity at levels well below the acceptance limits. Detection capability for hexane typically runs 10–20 ppm under validated conditions, giving roughly a 15-fold safety factor below the 290 ppm limit. That margin matters: it means a borderline result isn’t an artifact of method insensitivity.

Standard turnaround for a full Class 1 and Class 2 residual solvent panel (21 compounds) runs 5–7 business days. Targeted single-solvent screens — hexane-only for a saw palmetto intake, for instance — can move faster.

---

Evaluating a Supplier COA for Residual Solvent Compliance

When a new botanical extract COA comes in, work through these steps before releasing the material to your incoming inventory:

1. Identify the extraction solvent. Ask the supplier directly, in writing, as part of your supplier qualification intake questionnaire: what solvent system is used to produce this extract, at which stage, and is it present in the final dried material? This documentation belongs in your supplier file.

2. Map the solvent to its USP class. Hexane → Class 2, 290 ppm. Ethanol → Class 3, 5,000 ppm. If the supplier doesn’t know what solvent was used — or won’t disclose it — that’s a qualification failure independent of any test result.

3. Locate the residual solvent section on the COA. If it’s absent, blank, or marked “N/T,” you have no data. A missing residual solvent section on a COA doesn’t mean solvents are absent; it means they weren’t measured. These are not equivalent.

4. Verify the test method matches USP <467>. A compliant result requires the COA to specify the procedure (A or B), the diluent used, the equilibration temperature and time, and the column type. “In-house headspace GC method” without these parameters cannot be mapped to USP <467> acceptance criteria.

5. Run independent verification for high-risk extracts. For any new supplier relationship, and for lipophilic concentrates or alkaloid extracts specifically, independent third-party testing is the defensible position. Send a sample from the incoming lot before releasing it to production.

---

The 21 CFR Part 111 Angle Midwest Brands Often Underestimate

FDA’s dietary supplement GMP regulations at 21 CFR Part 111.70 require written specifications for each component, and 111.75 requires verification that those specifications are met before use. “Purity” — one of the four required specification categories alongside identity, strength, and composition — explicitly covers the absence of contaminants at levels that could harm consumers.

FDA hasn’t issued a standalone rule requiring USP <467> compliance for dietary supplement raw materials the way it has for drug product ingredients. But if your incoming material specification for a hexane-extracted botanical extract contains no residual solvent parameter, your purity specification is incomplete. In an FDA GMP inspection, an investigator reviewing raw material specifications for a solvent-extracted botanical and finding no residual solvent limits will write that observation. That observation becomes a Form 483 item. Three or more 483 items from a single inspection can escalate to a warning letter.

Brands in the Chicago area and across the Midwest that are scaling up for national distribution — and approaching their first FDA facility inspection — consistently underestimate how granular the incoming material specification documentation needs to be. Residual solvents are one of the first places where “we buy from a reputable supplier” diverges from “we have verified specifications with supporting data.”

When samples come through our Countryside, IL receiving facility, residual solvent testing runs as part of the standard incoming qualification panel for high-risk botanical extracts — alongside ICP-MS heavy metals (USP <232>/<233>), botanical identity (HPTLC), and microbial enumeration (USP <61>/<62>). The resulting CoA is issued under ISO 17025 accreditation, which means the data is methodologically defensible if an FDA investigator asks to see it.

The fix here isn’t complicated. Audit your current botanical extract specifications and identify which ones involve solvent extraction. For those — especially any using hexane, methanol, or dichloromethane — check whether your incoming material records include USP <467>-compliant test data. If they don’t, prioritize those SKUs for your next qualification round. Start with the highest-volume materials and work backward from there.

---

Written by Nour Abochama, VP Operations, Qalitex | Quality Consultant, Ayah Labs. Learn more about our team

Ship your sample to our Chicago facility — get a Qalitex CoA in 5–7 days. Contact us

Related from our network

• ISO 17025-Accredited Botanical and Supplement Testing at Qalitex Laboratories — Full analytical panel including USP <467> residual solvents, ICP-MS heavy metals, HPTLC identity, and USP <61>/<62> microbiology under California ISO 17025 accreditation.

---