Residual Solvents in Botanical Extracts: Ethanol vs. Ethyl Acetate
ICH Q3C / USP 467 limits, GC-FID headspace validation parameter
Two botanical extract batches land on your QC desk. Both declare 95%+ purity. Both use "solvent extraction" in their manufacturing description. One was extracted with food-grade ethanol; the other with ethyl acetate. Their COAs show no residual solvent data — only "moisture: ≤5% by LOD."
This is the most common residual solvent compliance gap in nutraceutical ingredient sourcing. The extraction solvent is disclosed — if disclosed at all — without the analytical data to confirm it was adequately removed. And the regulatory implications differ significantly between ethanol and ethyl acetate, even though both are ICH Q3C Class 3 solvents sharing the same 5,000 ppm ceiling.
This article provides the exact ICH Q3C / USP <467> limit values, the validated GC-FID headspace parameters required for compliance testing, and the regulatory and clean-label trade-offs that B2B formulation chemists need to evaluate when specifying extraction solvents for botanical ingredient suppliers.
Per ICH Q3C(R8) / USP <467> Option 1 (based on 10 g/day product dose): Ethanol — 5,000 ppm (Class 3; PDE ≥50 mg/day); Ethyl Acetate — 5,000 ppm (Class 3; PDE ≥50 mg/day); Methanol — 3,000 ppm (Class 2; PDE 30.0 mg/day); n-Hexane — 290 ppm (Class 2; PDE 2.9 mg/day); Acetonitrile — 410 ppm (Class 2; PDE 4.1 mg/day). Both ethanol and ethyl acetate share the same 5,000 ppm regulatory ceiling — but differ critically in FDA GRAS status, clean-label acceptability, hydrolysis risk, and screening requirements. Testing methodology: Gas Chromatography with Headspace sampling and Flame Ionization Detection (HS-GC-FID), USP Phase G43 column (DB-624: 30 m × 0.32 mm i.d. × 1.8 µm film), carrier gas helium at ~35 cm/s, equilibration at 80°C for 45–60 min, per USP <467> Procedures A/B/C.
- The ICH Q3C / USP <467> Classification Framework — Exact Limits
- Class 2 Solvent Limits Reference Table — Complete B2B Panel
- Class 3 Solvents: Ethanol vs. Ethyl Acetate — Full Comparison
- GC-FID Headspace Methodology: Validated Instrument Parameters
- USP <467> Procedures A, B, C — When and How to Apply Each
- The Loss on Drying Compliance Trap
- Solvent Screening Panel by Botanical Ingredient Type
- Industrial-Grade and Recycled Solvents: Contamination Risk Forensics
- Clean-Label and "Solvent-Free" Claim Defensibility
- Supplier COA Documentation Checklist
- FAQ
1. The ICH Q3C / USP <467> Classification Framework — Exact Limits
ICH Q3C (Impurities: Guideline for Residual Solvents) classifies solvents into three risk tiers. Limits are calculated using the Option 1 formula: C (ppm) = (1,000 µg/mg × PDE mg/day) / dose g/day, where the standard assumed dose is 10 g/day for Option 1. Class 3 solvents have PDEs of ≥50 mg/day, yielding a universal ceiling of 5,000 ppm without further justification. Class 2 solvents have individual PDE-derived limits. Class 1 solvents are effectively prohibited (benzene: 2 ppm; 1,2-dichloroethane: 5 ppm). ICH Q3C is harmonized into USP <467> (US), EP 5.4 (EU), and JP (Japan). The Option 1 limits apply when daily product dose ≤10 g; Option 2 (summation of daily exposures from all components) applies when daily dose exceeds 10 g or when any ingredient exceeds the Option 1 limit.
| ICH Class | Toxicity Profile | Regulatory Treatment | Limit Basis |
|---|---|---|---|
| Class 1 — Prohibited | Known or suspected human carcinogens; environmental hazards | Should be avoided; if unavoidable, must justify and test extensively — limits set near zero | No safe threshold — effective prohibition |
| Class 2 — Limit Solvents | Non-genotoxic animal carcinogens; possible neurotoxins or teratogens | Controlled to individual PDE values; GC identification and quantification mandatory regardless of level | Individual PDEs → Option 1 ppm per solvent (formula: ppm = 1000 × PDE / 10g) |
| Class 3 — Low Toxicity | Low toxic potential; PDE ≥ 50 mg/day | Acceptable at ≤5,000 ppm without justification; non-specific LOD test acceptable ONLY if level <0.1% (1,000 ppm) | Universal ceiling: 5,000 ppm (50 mg/day PDE basis) |
Table 1 — ICH Q3C Classification Framework. Source: ICH Q3C(R8) Step 4 (April 2021); USP <467> official monograph.
2. Class 2 Solvent Limits Reference Table — Complete B2B Panel
Every botanical extract supplier COA must be screened against these values. For any Class 2 solvent detected above LOQ — even if the extraction process did not declare it — USP <467> requires identification and quantification. Undisclosed Class 2 solvents are a supplier qualification failure.
| Solvent | ICH Class | PDE (mg/day) | Option 1 Limit (ppm) | GRAS? | Primary Risk | Botanical Relevance |
|---|---|---|---|---|---|---|
| Benzene | Class 1 | — | 2 ppm | No | Human carcinogen | Effectively prohibited — must be absent (<2 ppm) |
| 1,2-Dichloroethane | Class 1 | — | 5 ppm | No | Human carcinogen | Effectively prohibited |
| n-Hexane | Class 2 | 2.9 mg/day | 290 ppm | No | Peripheral neuropathy; neurotoxin | High risk — commonly used for defatting in curcumin and plant protein extraction; USDA Organic prohibits hexane |
| Acetonitrile | Class 2 | 4.1 mg/day | 410 ppm | No | Non-genotoxic; possible cardiovascular effects | Used in HPLC purification; may appear as cross-contamination in multi-product facilities |
| Dichloromethane (DCM) | Class 2 | 6.0 mg/day | 600 ppm | No | Possible carcinogen; CNS effects | Used in some alkaloid and lipid-soluble compound extraction; chlorinated solvent — major clean-label disqualifier |
| Methanol | Class 2 | 30.0 mg/day | 3,000 ppm | No | Optic nerve damage; metabolic acidosis | Critical risk — commonly used in berberine, ashwagandha withanolide extraction; also appears as industrial ethanol denaturing agent |
| Toluene | Class 2 | 8.9 mg/day | 890 ppm | No | Reproductive toxin; CNS | May appear as cross-contamination from recycled solvents in mixed-product facilities |
| Cyclohexane | Class 2 | 38.8 mg/day | 3,880 ppm | No | CNS; irritant | Less common in botanical extraction; may appear in facilities handling petroleum-derived products |
| Tetrahydrofuran (THF) | Class 2 | 7.2 mg/day | 720 ppm | No | Irritant; possible carcinogen at high exposure | Used in some resin and polymer extraction; cross-contamination risk |
| N-Methylpyrrolidone (NMP) | Class 2 | 5.3 mg/day | 530 ppm | No | Reproductive toxin; teratogen | Used as high-boiling analytical diluent in HS-GC; should never appear in final botanical extract |
Table 2 — Class 2 Solvent Limits per ICH Q3C(R8) / USP <467> Option 1. All limits verified against ICH Q3C(R8) Step 4 (2021) and pharmabeginers.com ICH annexure table.
n-Hexane (ICH Q3C Class 2; limit: 290 ppm; PDE: 2.9 mg/day) is routinely used in the defatting step of curcumin 95% production. It is a documented peripheral neurotoxin. Hexane is prohibited under USDA National Organic Program (NOP). Many brands specify "hexane-free" or "ND hexane" as a non-negotiable requirement. A standard curcumin COA showing only ethanol in the extraction step may conceal a hexane defatting step that was not disclosed. The only way to confirm hexane-free status is GC-HS with a validated hexane LOQ ≤29 ppm (10% of the 290 ppm limit). "ND hexane" on a COA is only meaningful if the stated LOQ is confirmed at or below 29 ppm.
3. Class 3 Solvents: Ethanol vs. Ethyl Acetate — Full Regulatory and Practical Comparison
| Comparison Parameter | Ethanol (Food-Grade) | Ethyl Acetate |
|---|---|---|
| ICH Q3C Class | Class 3 | Class 3 |
| PDE | ≥50 mg/day | ≥50 mg/day |
| Option 1 Limit (ppm) | 5,000 ppm (0.5% w/w) | 5,000 ppm (0.5% w/w) |
| CAS Number | 64-17-5 | 141-78-6 |
| Boiling Point | 78.4°C | 77.1°C |
| FDA GRAS Status | Yes — 21 CFR §182.1180 (direct food ingredient) | No GRAS as direct food ingredient; 21 CFR §173.228 (indirect additive only) |
| EU Regulatory Status | E1510 — permitted food additive; widely accepted for botanical extraction in EU | Permitted flavoring solvent (EC 1334/2008); acceptable for extraction with residual testing |
| Hydrolysis Risk | Stable under normal extraction conditions; does not produce toxic hydrolysis products | Hydrolyzes to acetic acid + ethanol in presence of water/acid, especially at elevated temperature. Acetic acid in finished extract causes off-taste and pH depression — test pH of 1% aqueous solution as screening tool |
| Sensory Threshold | Low impact at ≤5,000 ppm; ethanol evaporates readily; clean sensory profile | Distinct fruity/acetone odor above ~1,000 ppm; organoleptic impact on finished product — most brands set in-house limit ≤1,000 ppm |
| Extraction Selectivity | Broad — tunable with water (e.g., 70% EtOH for polyphenols; 95% for lipophilic compounds); excellent for catechins, curcuminoids, anthocyanins, flavonoids, alkaloids | Narrower and more selective — excellent for moderately non-polar polyphenols (EGCG purification, curcumin enrichment); used in liquid–liquid partitioning to remove chlorophyll and sugars from crude extracts |
| Clean-Label / NSF / Organic Compatibility | Universally accepted; fully compatible with NSF, Informed Sport, USDA Organic, and all major clean-label programs | Varies by program; some certification bodies flag ethyl acetate or require explicit disclosure; check with your specific certification body before specifying |
| Typical Role in Botanical Extraction | Primary extraction solvent for GTE, curcumin, most standardized botanical extracts — the default global food supplement extraction solvent | Secondary purification/partitioning solvent: EGCG ≥90% production, curcumin enrichment, chlorophyll removal from crude polyphenol extracts |
| Recommended In-House Limit (Premium Brands) | ≤1,000–3,000 ppm (sensory and clean-label driven; tighter than USP 5,000 ppm) | ≤500–1,000 ppm (organoleptic + certification body driven) |
Table 3 — Ethanol vs. Ethyl Acetate: Complete Head-to-Head Comparison. Same ICH Q3C Class and ppm limit; significantly different regulatory and commercial implications.
For standard food supplement applications requiring clean-label positioning, natural claims, or certification body compliance: specify food-grade ethanol as the extraction solvent. For products where ethyl acetate was used in a purification step (e.g., EGCG ≥90%, curcumin enrichment): require (1) explicit process disclosure; (2) GC-HS data showing ethyl acetate ≤1,000 ppm in-house limit; (3) acetic acid pH screening; and (4) certification body confirmation that EA use is compatible with your certification program.
4. GC-FID Headspace Methodology: Validated Instrument Parameters
USP <467> specifies static headspace gas chromatography with flame ionization detection (HS-GC-FID) as the standard analytical technique for residual solvent determination. The primary column is USP Phase G43 — a 6% cyanopropylphenyl / 94% dimethylpolysiloxane fused silica capillary column (trade name: Agilent DB-624, Phenomenex ZB-624, Restek Rtx-624). Procedure A uses the G43 column for initial identification/limit testing. Procedure B uses a G16 phase (polyethylene glycol wax-type, e.g., DB-WAX) as the orthogonal confirmation column. Procedure C is the quantitative procedure using G43 or G16 depending on which produced fewer coelutions. Key system suitability: Resolution Rs ≥1.0 between acetonitrile and dichloromethane (Class 2 Mixture A); RSD ≤15% for 6 replicate injections; S/N ≥10 for reporting limit solution.
Column (Procedure A / C — G43 Phase): DB-624 (Agilent) / ZB-624 (Phenomenex) / Rtx-624 (Restek) — 6% cyanopropylphenyl–94% dimethylpolysiloxane fused silica capillary; 30 m × 0.32 mm i.d. × 1.8 µm film thickness (standard); alternatives validated: 30 m × 0.25 mm × 1.4 µm; 30 m × 0.53 mm × 3.0 µm
Column (Procedure B — G16 Phase, confirmatory): DB-WAX UI (Agilent) — polyethylene glycol; 30 m × 0.32 mm × 0.25 µm
Carrier Gas: Helium (primary); linear velocity ~35 cm/s at column temperature; constant flow mode; ~1.5–2.0 mL/min for 0.32 mm i.d. column. Nitrogen acceptable with adjustment and verification; hydrogen acceptable with retention time shift confirmation.
Split Ratio: 5:1 (standard, USP-specified baseline for Class 2/3 screening); 2:1 to 10:1 for trace Class 1 solvents; 20:1 to 50:1 for abundant Class 3 solvents at 5,000 ppm to prevent column overload
Injector Temperature: 140–200°C (instrument-specific; Shimadzu validated: 140°C SPL; literature range: 170–200°C for DB-624); set to prevent condensation while minimizing thermal degradation of labile analytes
GC Oven Temperature Program (Procedure A — Shimadzu validated baseline): 40°C hold 20 min → ramp at 10°C/min → 240°C hold. Alternative validated programs: 35°C/2 min → 4°C/min → 80°C → 40°C/min → 230°C/12 min (Springer 2019 peer-reviewed); 50°C/3 min → 5°C/min → 80°C → 30°C/min → 230°C/2 min (LEAN method, LCGC 2026)
FID Detector: Temperature 250–300°C; H₂ flow 40 mL/min; air flow 400 mL/min; N₂ makeup gas 25–30 mL/min
| Headspace Parameter | USP <467> Specified Value | Validated Literature Range | Critical Notes |
|---|---|---|---|
| Vial Equilibration Temperature | 80°C | 80–140°C (application-dependent) | Higher temperature increases volatile partitioning into headspace; 80°C is USP baseline for water-soluble samples; 100–120°C for botanical matrices with higher-boiling solvents |
| Vial Equilibration Time | 45–60 min (USP) | 10–60 min (experimentally validated as equivalent at 10–15 min for most solvents) | 60 min conservatively ensures equilibrium for all matrix types; peer-reviewed data shows 10–15 min sufficient for most volatile solvents — validate per specific matrix |
| Transfer Line Temperature | 110°C (USP baseline) | 110–180°C | Must be ≥10°C above the boiling point of the highest-boiling analyte to prevent condensation; higher settings (130–180°C) required for DMSO/DMA diluents and high-boiling residuals |
| Headspace Vial Size | 20 mL | 10–20 mL | 20 mL standard with 1 mL sample volume (phase ratio β ≈ 19); smaller vials increase sensitivity but reduce equilibrium time |
| Sample Volume in Vial | ~1 mL in 20 mL vial | 0.5–5 mL | 1 mL is optimal for most analyses; >1 mL does not always increase signal (vial capacity effect); USP recommends 5 mL for some applications |
| Sample Diluent | DI water (water-soluble); DMSO or DMA (water-insoluble) | DMA or DMSO — DMA preferred in literature (cleanest blanks, thermally stable to ~120°C) | DMA preferred over DMSO: DMSO reacts with chloride salts to form dimethylsulfide (coelutes with acetone); DMSO is nucleophilic and may react with some analytes |
| Sample Concentration | ~250 mg / 25 mL (~10 mg/mL) | 50–250 mg/mL depending on solubility | Higher concentration increases sensitivity; must ensure complete dissolution; finely powdered samples + sonication + longer equilibration for botanical matrices |
| Loop Fill Time | Not specified exactly | 0.25–0.5 min | Ensure at least 3× loop volume flushed before injection; too short = incomplete loop fill; too long = sample dilution/loss |
| Injection Time | Not specified exactly | 0.5–1.0 min | Sufficient for complete injection of 1 mL loop into injector; adjust for specific headspace instrument model |
| System Suitability — %RSD | ≤15.0% (6 replicate injections of working standard) | Literature typically achieves <5% RSD | Must confirm before each analytical run; poor precision indicates equilibration issues, septum leaks, or inlet contamination |
| System Suitability — Resolution | Rs ≥ 1.0 (acetonitrile / dichloromethane in Class 2 Mix A) | DB-Select 624 UI achieves Rs = 3.1 for this pair | Also verify: Rs ≥ 0.9 between MEK and ethyl acetate (LEAN method requirement); resolution confirms column suitability and method integrity |
| LOQ (typical, 100 mg sample) | <100 ppm for most solvents | 1–100 ppm depending on K value and split ratio | Hexane: LOQ easily <10 ppm (favorable headspace partition); DMF, NMP: LOQ ~870 ppm (high K, poor headspace partition — may require liquid injection for these specific analytes) |
Table 4 — GC-Headspace FID Parameters for USP <467> Residual Solvent Analysis. Sources: USP <467> official monograph (uspnf.com 20190927); Shimadzu application note GC-005; Agilent AN 5994-6020EN; LCGC LEAN method (2026); Springer Link validation study (2019); Teledyne Tekmar HT3 USP <467> application (Chromatography Online).
5. USP <467> Procedures A, B, C — When and How to Apply Each
| Procedure | Column | Purpose | When to Use | Outcome |
|---|---|---|---|---|
| Procedure A | G43 (DB-624): 30 m × 0.32 mm × 1.8 µm | Initial identification and limit test. Semi-quantitative screening of Class 1 and Class 2 residual solvents against limit concentrations. | First-line screening of all incoming botanical extract batches. Use when the full solvent history is unknown or when performing routine QC against USP <467> compliance. | If all peaks are below limit concentrations → material meets requirements. If any peak exceeds a limit → proceed to Procedure B for confirmation. If positive: proceed to Procedure C for quantification. |
| Procedure B | G16 (DB-WAX): 30 m × 0.32 mm × 0.25 µm (orthogonal confirmation column) | Confirmation using orthogonal stationary phase. Confirms whether a peak identified in Procedure A is actually the suspected solvent (different polarity may resolve coelutions). | Required when Procedure A shows a peak above the limit concentration for a specific solvent. Provides identity confirmation through second-column retention time agreement. | If confirmed above limit on Procedure B → proceed to Procedure C for quantification. If not confirmed (peak at different retention time on G16) → investigate further; peak may be a different compound. |
| Procedure C | G43 or G16 — whichever provided better resolution (fewer coelutions) in Procedures A/B | True quantitative determination. Single-point standard addition or external standardization with validated calibration. This is the definitive compliance test. | Required when Procedures A and B confirm a solvent above the applicable limit. Also used directly when the solvent panel is known and defined (bypassing A/B screening is acceptable per ICH Q3C if the compound list is definitive). | Reports quantitative ppm value for each detected solvent; compared against Option 1 limits or Option 2 calculated PDE limits. Generates the data that appears on compliant COAs. |
Table 5 — USP <467> Procedures A, B, C: Application Hierarchy. Note: If the compound list is known and definitive (e.g., ethanol + ethyl acetate only), the analyst may proceed directly to Procedure C quantification without the A/B screening steps.
6. The Loss on Drying Compliance Trap
ICH Q3C states: "If only Class 3 solvents are present, a non-specific method such as loss on drying may be used — provided the level of Class 3 solvent is less than 0.1% (1,000 ppm)." This narrow exception is frequently misapplied by suppliers. Loss on drying measures total volatile content including water, free moisture, and any volatile solvents — it cannot distinguish between water content and ethanol residuals, cannot detect Class 2 solvents at the relevant sensitivity levels, and cannot quantify individual solvents against their ICH limits. A COA stating "LOD: ≤1.0%" or "moisture and volatile matter: ≤5% by gravimetric method" provides NO information about residual solvent identity or individual solvent compliance. It does NOT demonstrate ICH Q3C compliance. GC-FID headspace analysis is required for all botanical extracts where any solvent was used in manufacturing.
| COA Statement | ICH Q3C Compliant? | What It Actually Tells You | Action Required |
|---|---|---|---|
| "Residual solvents: NMT 0.5% by LOD" | No | Total volatiles (water + solvents) below 0.5% — no individual solvent identity or quantification | Reject — request GC-HS with individual solvent ppm values |
| "Residual solvents: Not detected by LOD method" | No | Meaningless for residual solvent compliance — LOD method cannot detect trace solvents at the relevant levels | Reject — request GC-HS Procedure C quantification |
| "Ethanol: NMT 5,000 ppm by GC-HS; Ethyl Acetate: NMT 1,000 ppm by GC-HS; Methanol: ND (LOQ ≤100 ppm) by GC-HS; Hexane: ND (LOQ ≤29 ppm) by GC-HS" | Yes — compliant | Individual solvent quantification with stated LOQ; USP <467> Procedure C compliant data; LOQ confirmed at ≤10% of limit for each solvent | Accept — retain as QC record; verify batch number matches physical shipment |
| "Residual solvents: Complies with USP <467>" (no values stated) | Conditional | Supplier declares compliance but no numerical data provided — cannot verify | Request the actual GC-HS data with ppm values and LOQ confirmation — compliance declaration without data is not acceptable for regulated product documentation |
Table 6 — COA Residual Solvent Statement Acceptability for ICH Q3C / USP <467> Compliance Documentation
7. Solvent Screening Panel by Botanical Ingredient Type
| Botanical Ingredient | Likely Primary Solvent | Undisclosed Solvent Risk | Mandatory GC-HS Panel | Special Requirements |
|---|---|---|---|---|
| Green Tea Extract 40–70% EGCG | Ethanol (water-ethanol) | Methanol (from industrial ethanol); ethyl acetate (undisclosed purification step) | Ethanol, methanol, ethyl acetate; acetone, isopropanol screen | Methanol mandatory — industrial ethanol denaturing risk |
| EGCG 90–98% Ultra-Pure | Water → Ethanol → Ethyl Acetate (partitioning) → Chromatographic purification | Acetonitrile (if HPLC purification used); ethyl acetate always expected | Ethanol, ethyl acetate, acetonitrile, methanol; acetic acid (EA hydrolysis product) by pH and GC-MS | Ethyl acetate presence expected and must be quantified; in-house limit ≤1,000 ppm recommended |
| Curcumin 95% | Ethanol or ethyl acetate; hexane for defatting | Hexane (defatting step often undisclosed); dichloromethane (some producers) | Ethanol, ethyl acetate, hexane (mandatory), methanol, dichloromethane | Hexane ≤290 ppm (ICH Class 2); for organic: require ND hexane with LOQ ≤29 ppm |
| Ashwagandha (Withanolides) | Water-ethanol; some producers use methanol | Methanol (Class 2!); ethyl acetate for withanolide purification | Ethanol, methanol, ethyl acetate; acetone | Methanol mandatory — Class 2, 3,000 ppm limit; significant enforcement risk category |
| Berberine | Ethanol, water-ethanol | Methanol (common in cost-driven production) | Ethanol, methanol, ethyl acetate | Methanol mandatory screen — commonly used by budget producers |
| Resveratrol | Ethanol | Methanol; ethyl acetate (purification) | Ethanol, methanol, ethyl acetate, acetone | Standard panel; methanol mandatory |
| Ginseng (Ginsenosides) | Water-ethanol (60–80% EtOH) | Methanol; butanol (some saponin extraction processes) | Ethanol, methanol, n-butanol, ethyl acetate | n-Butanol is Class 3 (5,000 ppm) but should be declared if used |
| Marine Collagen Peptides | Water extraction only (enzymatic hydrolysis) | No extraction solvent expected — but check for sanitizing agent residuals (isopropanol, peracetic acid) | Ethanol, isopropanol, acetone — CIP (cleaning-in-place) solvent screen | Low risk but relevant for high-hygiene facilities using solvent sanitizers |
Table 7 — GC-HS Screening Panel by Botanical Ingredient Type. "Mandatory" indicates the panel minimum for first-order supplier qualification.
8. Industrial-Grade and Recycled Solvents: Contamination Risk Forensics
(1) Industrial-grade (denatured) ethanol: Denatured alcohol contains methanol (typically 2–5%), isopropanol, or benzene derivatives as denaturing agents added to avoid beverage tax. Methanol from denatured ethanol will appear in the botanical extract at concentrations well above the Class 2 limit of 3,000 ppm. GC-HS showing both ethanol and methanol simultaneously is the forensic indicator. (2) Recycled/reclaimed solvents without adequate purification: Cross-contamination from previous extraction batches or other products introduces unexpected solvents — toluene from aromatic extraction, acetonitrile from HPLC purification steps, THF from polymer processing. Multiple unexpected solvents at significant concentrations is the forensic indicator. (3) Undisclosed multi-solvent processes: A producer may use hexane for defatting, then ethanol for extraction, then ethyl acetate for purification — but disclose only "ethanol extraction" on the COA. The GC-HS result revealing undisclosed solvents not mentioned in the manufacturing process description is the forensic indicator — and a supplier qualification failure.
Forensic Interpretation of GC-HS Results
9. Clean-Label and "Solvent-Free" Claim Defensibility
| Label Claim | Defensible? | Documentation Required | Key Risk |
|---|---|---|---|
| "Ethanol-Extracted" / "Alcohol-Extracted" | Yes — most favorable | Food-grade ethanol SDS; GC-HS confirming ethanol ≤5,000 ppm (ICH) or in-house tighter limit; no Class 2 solvents detected; no undisclosed ethyl acetate | Must disclose if ethyl acetate was used in any purification step — claiming "ethanol-extracted" while concealing EA use is a material misrepresentation |
| "Water-Extracted" / "Aqueous Extract" | Yes — if accurate | Process documentation confirming water-only extraction; GC-HS confirming ND for all organic solvents (with stated LOQ); no blending of water-extracted with solvent-extracted fractions | Some suppliers blend water-extracted and ethanol-extracted material to hit standardization targets — blended product cannot be called "water-extracted" |
| "CO₂ Extracted" / "Supercritical Extract" | Yes — premium claim | Supercritical CO₂ leaves no residual solvent; GC-HS confirms ND for all organic solvents; CO₂ extraction process documentation; higher cost premium should be verifiable in cost structure | Some suppliers blend CO₂-extracted with conventional material and claim "CO₂ extracted" for the full batch — verify that the entire batch was CO₂ extracted |
| "Solvent-Free" Extract | Conditional | GC-HS showing all solvents ≤10 ppm (effectively ND with stated LOQ); process documentation confirming no organic solvent use at any stage; or documented CO₂ extraction | "Solvent-free" cannot mean "not tested" — requires analytical evidence. False "solvent-free" claims are actionable under FTC truth-in-advertising rules and FDA label accuracy requirements |
| "USDA Organic Certified" Extract | Yes — with USDA NOP compliance | USDA NOP certification from accredited certifier; NOP prohibits hexane; ethanol permitted; CO₂ permitted; ethyl acetate generally not used in NOP-certified organic extraction | Any hexane use — including in a defatting step — disqualifies USDA Organic certification; verify explicitly for curcumin and protein-containing botanical extracts |
Table 8 — Extraction Claim Defensibility and Substantiation Requirements
10. Supplier COA Documentation Checklist
Mandatory GC-HS Documentation for Residual Solvent Compliance
- Extraction solvent(s) fully disclosed: Full IUPAC or common name + ICH Q3C class for every solvent at every stage (primary extraction, liquid–liquid partitioning, purification, precipitation). "Solvent extraction" without specifying the solvent is not acceptable.
- Solvent grade documented — SDS provided: Food-grade, pharmaceutical-grade, or industrial-grade for each solvent. Request the Safety Data Sheet (SDS). Industrial-grade ethanol SDS will disclose methanol or isopropanol denaturing agents — an immediate disqualifier for clean-label applications.
- GC-HS residual solvent data with individual ppm values: Per USP <467> Procedure A/C (G43 phase; DB-624 equivalent). Results stated in ppm for each solvent — not "complies with" or "passes" without a numeric value. Not LOD-based methods.
- LOQ confirmed at or below 10% of applicable ICH limit for each solvent: Ethanol/EA (5,000 ppm limit) → LOQ ≤500 ppm; Methanol (3,000 ppm) → LOQ ≤300 ppm; Hexane (290 ppm) → LOQ ≤29 ppm; Acetonitrile (410 ppm) → LOQ ≤41 ppm
- System suitability reported: %RSD ≤15% (6 replicates of working standard); Rs ≥1.0 between acetonitrile and DCM; S/N ≥10 for reporting limit solution. System suitability confirms method integrity for the specific analytical run.
- ISO 17025 accredited laboratory: Supplier in-house GC data acceptable for ongoing qualified suppliers; ISO 17025 third-party GC-HS required for first-order qualification and annual re-qualification.
- For curcumin 95%: Hexane screen mandatory (LOQ ≤29 ppm) regardless of declared extraction method. For organic curcumin: require ND hexane with stated LOQ.
- For EGCG ≥90%: Ethyl acetate screen mandatory (in-house limit ≤1,000 ppm recommended); acetic acid pH screening of 1% aqueous solution; acetonitrile screen if HPLC purification step used.
- For berberine, ashwagandha, ginseng: Methanol screen mandatory (LOQ ≤300 ppm; Class 2 limit 3,000 ppm); methanol is used by cost-driven producers as primary extraction solvent.
Supplier Disqualification Triggers
- Refusal to disclose extraction solvent type or grade
- COA shows only LOD/moisture data — no GC-HS individual solvent quantification
- Any Class 2 solvent detected above its ICH Q3C Option 1 limit (methanol >3,000 ppm; hexane >290 ppm; acetonitrile >410 ppm)
- Class 2 solvents detected that were not disclosed in the manufacturing process description
- Industrial-grade ethanol SDS showing methanol or benzene denaturing agents
- Multiple undisclosed solvents at significant concentrations — recycled mixed-solvent use indicator
- Hexane above LOD in any curcumin product intended for organic-certified applications
Blue Eyes Biotech — Clean Extraction, Full GC-HS Documentation
Our botanical extracts — green tea extract (EGCG), curcumin 95%, and plant extracts — are produced using food-grade ethanol or supercritical CO₂, validated by batch-specific GC-FID headspace residual solvent analysis per USP <467>. Extraction solvent SDS, individual ppm results, and LOQ confirmation available on request.
11. Frequently Asked Questions
The ICH Q3C limit is the regulatory floor — the minimum compliance threshold — not the ceiling of commercial and clean-label acceptability. The differences that matter beyond the shared 5,000 ppm limit: (1) FDA GRAS status — ethanol has GRAS as a direct food ingredient (21 CFR §182.1180); ethyl acetate does not. (2) Hydrolysis risk — ethyl acetate hydrolyzes to acetic acid + ethanol in the presence of water and acid, particularly at elevated temperatures, producing pH-depressing off-notes in finished extracts. (3) Sensory impact — ethyl acetate above ~1,000 ppm produces distinct fruity/chemical off-notes that ethanol at equivalent concentrations does not. (4) Certification body requirements — NSF, Informed Sport, USDA Organic, and other programs scrutinize ethyl acetate use in ways they do not for food-grade ethanol. (5) Consumer and brand perception — "ethyl acetate" as a label disclosure is commercially unfavorable compared to "ethanol extraction." These factors make the extraction solvent choice significant even when both are technically ICH Q3C compliant.
No — a compliance declaration without numeric data is not acceptable documentation for regulated product supply chains. USP <467> Procedure C is a quantitative test that produces specific ppm values for each detected solvent. A COA must state: the name of each solvent tested, its quantified concentration in ppm, the LOQ of the method for each solvent, and the analytical method used (GC-HS, Procedure A/C, column type). "Passes USP <467>" without data cannot be used for regulatory submissions, product audits, or your own cGMP incoming QC documentation. Request the actual GC-HS analytical report — a reputable supplier will be able to provide this without hesitation.
In practical terms: Procedure A is the screening test using a G43 phase (DB-624) column — it identifies whether any solvent exceeds the limit concentration and confirms identity via retention time against standards. If all peaks are below limits, the test is complete. Procedure B is the orthogonal confirmation test using a G16 phase (DB-WAX) column — it confirms the identity of any suspect peak by showing it behaves consistently on a chemically different stationary phase. Procedure C is the definitive quantitative test — it provides the actual ppm value used for compliance reporting. For incoming QC of botanical extracts with a known solvent list, you may proceed directly to Procedure C quantification (bypassing A and B screening) because the compound list is known and defined. This is the most efficient approach for routine QC where you have a documented solvent list from the supplier's manufacturing process disclosure.
The ICH Q3C Class 2 Option 1 limit for n-hexane is 290 ppm (PDE 2.9 mg/day). The required LOQ must be at or below 10% of this limit: LOQ ≤29 ppm. GC-HS with FID detection is highly sensitive for hexane — its favorable headspace partition coefficient (low K value) means it partitions efficiently into the vapor phase, making it one of the easier solvents to detect at low ppm. A validated GC-HS method on a DB-624 column will typically achieve LOQ <10 ppm for hexane from a 100 mg sample, well below the 29 ppm requirement. For products intended for USDA Organic certification, specify ND hexane with LOQ ≤10 ppm confirmed — any quantifiable hexane detection is grounds for organic certification disqualification.
A generic GC-HS method validated against USP <467> parameters (column, carrier gas, temperature program, system suitability) does not need to be re-validated from scratch for each botanical matrix. However, a matrix-specific accuracy assessment (spiking study) is recommended for each new botanical extract type to confirm that the botanical matrix does not suppress or enhance analyte response. Specifically: dissolve your botanical extract in the diluent (DMA or DMSO) and spike it with the target solvents at 100% of the ICH limit concentration; compare the recovery against the standard prepared in diluent alone; acceptable recovery is typically 80–120%. If recovery falls outside this range, adjust the method (diluent concentration, equilibration time, temperature) or apply a matrix-matched calibration curve. This partial validation requirement is much less burdensome than full method development and is standard practice for pharmaceutical API residual solvent methods.
Three-step verification: (1) Request the ethanol SDS from the supplier's ethanol source. Food-grade (or pharmaceutical-grade) ethanol SDS will state purity ≥96% or ≥99.5%, with no methanol, benzene, or other denaturing agents listed as components. Industrial (denatured) ethanol SDS will disclose the denaturing agents. (2) Request GC-HS data from the finished extract showing ethanol quantified and all other solvents at ND (with stated LOQs). Specifically look for methanol — food-grade ethanol should show ND methanol at LOQ ≤300 ppm; industrial ethanol will show methanol at 2,000–5,000 ppm depending on the denaturing formula. (3) Independent third-party verification — for first-order qualification, send a retained sample to an ISO 17025 accredited lab for GC-HS analysis. Request a full Class 2 screen alongside the Class 3 ethanol quantification. Methanol above 500 ppm in an ethanol-extracted botanical extract is a strong indicator of industrial ethanol use, regardless of the supplier's declaration.
Related Knowledge Center Articles
- How to Read & Verify a Nutraceutical Ingredient COA
- EGCG 98% Technical Data Sheet — Extraction Method and Residual Solvent Specifications
- Beyond HPLC: Carbon-14 Testing for Curcumin 95% (Including Solvent Adulteration)
- California Prop 65 Heavy Metal Compliance: Low-Lead Botanical Extracts
- Ultimate Guide to Buying Green Tea Extract / EGCG
