Decaf Green Tea Extract: Solvent-Free Sourcing Guide
scCO₂ vs. solvent decaffeination — and how to verify caffeine and residual solvent compliance.
A buyer specifies "decaffeinated green tea extract, caffeine <0.1%, solvent-free." The supplier delivers a product with caffeine at 0.08% — technically within specification. The COA is silent on decaffeination method. A residual solvent screen reveals dichloromethane (methylene chloride) at 380 ppm.
Under USP <467> / ICH Q3C, dichloromethane is a Class 1 solvent — a known carcinogen with a limit of 600 ppm in pharmaceutical products. The supplier's product technically passes that limit. But the product will not pass many retailer specifications requiring ND (not detected) for Class 1 solvents, will not qualify for clean-label or organic positioning, and conflicts with the buyer's "solvent-free" claim entirely.
This scenario illustrates the central compliance problem in decaffeinated green tea extract sourcing: the "decaffeinated" claim alone specifies nothing about how caffeine was removed, what solvent class was used, or what residuals remain. This guide covers the three primary decaffeination methods, their process parameters and selectivity limitations, the exact USP <467> solvent residue limits applicable to each, the HPLC caffeine assay method for verification, and the supplier COA audit framework for clean-label compliance.
Two genuinely solvent-free decaffeination methods are applicable to green tea extract: (1) Supercritical CO₂ (scCO₂) extraction — CO₂ above its critical point (T >31.1°C, P >73.8 bar) acts as a tunable non-toxic solvent that selectively extracts caffeine. Adding water as a co-solvent (4–7 wt%) at 40°C and 400 bar achieves caffeine/EGCG selectivity of 2.57 — removing more caffeine per unit of EGCG lost than either water or ethanol extraction. CO₂ leaves zero solvent residue as it fully reverts to gas at ambient conditions. (2) Hot water extraction decaffeination — caffeine is selectively leached from fresh tea leaf using hot water at 100°C for 3 minutes (ratio 1:20 leaf:water); 83% of caffeine is removed while 95% of total catechins are retained. The resulting decaffeinated leaf is then processed to extract catechins. Both methods require GC-Headspace or HPLC verification to confirm residual caffeine ≤0.1% and confirmed absence of Class 1/2 solvents. Methylene chloride and ethyl acetate decaffeination require strict USP <467> residual solvent testing and cannot support clean-label, organic, or "solvent-free" positioning.
- Why Decaffeination Method Matters for Clean-Label Sourcing
- Full Decaffeination Method Comparison: Process Parameters and Trade-Offs
- Supercritical CO₂: Process Physics, Selectivity, and EGCG Retention
- Hot Water Decaffeination: Parameters and Catechin Retention Data
- Methylene Chloride and Ethyl Acetate: USP Class and Residue Limits
- Caffeine HPLC Assay: Verifying <0.1% Residual Caffeine
- USP <467> Solvent Residue Verification for Decaffeinated Extracts
- Supplier COA Audit Checklist for Solvent-Free Decaf Claims
- FAQ
1. Why Decaffeination Method Matters for Clean-Label Sourcing
The supplement industry's growing clean-label, organic, and "natural" positioning creates a specific procurement challenge for decaffeinated green tea extract. Unlike EGCG standardization (which depends on extraction method but not necessarily on whether a solvent was used post-extraction), decaffeination almost always involves a deliberate process step to remove an endogenous compound — and the choice of removal agent carries direct regulatory, commercial, and consumer transparency implications.
A product labeled "decaffeinated green tea extract" carries no regulatory requirement in the US to disclose the decaffeination method or the solvents used. A product decaffeinated with methylene chloride and one decaffeinated with supercritical CO₂ carry identical labels. The only way to determine which process was used is to: (1) require the supplier to declare the decaffeination method in writing as part of the purchase agreement; (2) request and verify the residual solvent data by GC-Headspace (USP <467> Procedure B); (3) confirm that the stated method is operationally plausible for the supplier's facility (scCO₂ requires specialized pressure vessels — a facility without this capital investment cannot credibly claim scCO₂ decaffeination). "Solvent-free decaffeination" without these three verifications is an unsubstantiated label claim.
2. Full Decaffeination Method Comparison: Process Parameters and Trade-Offs
| Parameter | Supercritical CO₂ (scCO₂) | Hot Water Decaffeination | Methylene Chloride (DCM) | Ethyl Acetate (EA) |
|---|---|---|---|---|
| Process Principle | CO₂ above critical point (T >31.1°C, P >73.8 bar) acts as tunable non-polar solvent; water co-solvent (4–7 wt%) added to increase caffeine selectivity | Caffeine selectively leached from fresh tea leaf by hot water at near-boiling temperature; catechins are less water-soluble at high temperatures and short contact times | Chlorinated organic solvent selectively binds caffeine; counter-current extraction for 12h; steam stripping removes residual solvent; air/vacuum drying | Organic ester solvent (natural or synthetic) selectively extracts caffeine; similar process to DCM but lower toxicity profile |
| Operating Conditions | T: 40–80°C; P: 200–400 bar; water co-solvent: 4–7 wt%; extraction time: 60–180 min | T: 100°C; leaf:water ratio: 1:20 (w/v); contact time: 3 min; multiple cycles may be used for deeper removal | T: 40–60°C (ambient to warm); P: atmospheric; contact time: 12h counter-current; steam stripping at 80–100°C post-extraction | T: 40–60°C; P: atmospheric; contact time: variable; steam stripping required for solvent removal |
| Caffeine Removal Efficiency | 54–97% (highly parameter-dependent); water co-solvent at 40°C/400 bar: 54%; ethanol co-solvent at 80°C: up to 92.8% | 83% removal in single 3-min hot water treatment at 100°C (1:20 ratio); multiple cycles can achieve >90% | 96–97% caffeine removal | 96–99% caffeine removal |
| EGCG Retention | Water co-solvent at 40°C/400 bar: 79% EGCG retained (only 21% co-extracted); highest selectivity for caffeine vs. EGCG of all solvent methods | 95% total catechin retention in single hot water treatment; EGCG is less water-extractable than caffeine under short high-temp conditions | EGCG retained in tea matrix if process is properly controlled; risk of catechin loss during extended steaming steps | Moderate EGCG retention; ethyl acetate has intermediate polarity — some catechin co-extraction possible at higher contact time or temperature |
| Caffeine/EGCG Selectivity | Water co-solvent at 40°C/400 bar: 2.57 (highest documented selectivity); pure CO₂ at 70°C, ≥25 MPa: highest selectivity per Hansen parameter analysis (2025) | Caffeine/EGCG selectivity ≈ 0.88 with water at same conditions (SC-CO₂ reference); requires careful temperature/time optimization | Caffeine-selective; chlorinated solvents have high selectivity for methylxanthines; EGCG less soluble in DCM | Caffeine/EGCG selectivity ≈ 0.24 with ethanol-type solvents; EA has similar polarity to ethanol — lower selectivity than scCO₂ |
| Solvent Residue | Zero — CO₂ fully reverts to gas at ambient conditions; no residue possible | Zero — water is not a regulated residual solvent under USP <467> | Dichloromethane (DCM) — ICH Q3C Class 1 solvent; USP <467> limit 600 ppm; FDA food use limit 10 ppm for finished product | Ethyl acetate — ICH Q3C Class 3 solvent; USP <467> limit 5,000 ppm; off-notes above ~1,000 ppm; no GRAS as direct food ingredient |
| Clean-Label Compatibility | Fully compatible — "CO₂ decaffeinated," "solvent-free," organic, non-GMO positioning all supported | Fully compatible — "water decaffeinated," "solvent-free" supported; requires process documentation | Incompatible with "solvent-free," "natural," or "clean label" positioning; Class 1 carcinogen concern | Conditional — technically compliant at ≤5,000 ppm; "solvent-free" claim not defensible; many retailers specifically prohibit ethyl acetate |
| Capital Cost | High — requires specialized high-pressure vessels, compressors, and CO₂ recycling systems; not available at commodity extraction facilities | Low-moderate — hot water leaching equipment is simple; additional processing required to extract catechins from decaffeinated leaf | Low — atmospheric pressure solvent tanks; widely available at commodity processing facilities | Low-moderate — similar to DCM facilities; additional safety controls needed for flammability |
| Cost to Buyer (relative) | Premium — $15–30/kg price premium over solvent-decaffeinated grades | Moderate — process is scalable but adds cost vs. non-decaffeinated extraction | Lowest — most cost-competitive decaffeination method | Low to moderate |
Table 1 — Decaffeination Method Comparison: Full Process Parameter and Clean-Label Compatibility Assessment
3. Supercritical CO₂: Process Physics, Selectivity, and EGCG Retention
Carbon dioxide becomes supercritical above its critical point: Tc = 31.1°C, Pc = 73.8 bar (7.38 MPa). Above these conditions, CO₂ exists as a single phase with the density of a liquid but the diffusivity of a gas — combining the solvating power of a liquid solvent with excellent mass transfer. Its polarity is tunable by adjusting temperature and pressure: at lower temperatures and higher pressures, scCO₂ is more polar and extracts more caffeine; at higher temperatures and lower pressures, it becomes less polar and is more selective for lipophilic compounds. For green tea decaffeination, the addition of water as a co-solvent (4–7 wt%) is critical for achieving caffeine/EGCG selectivity — water increases the polarity of the scCO₂ phase, improving caffeine dissolution while EGCG (more polar, larger molecular weight) remains less soluble. Published research (Kim et al., 2008) established that at 40°C, 400 bar, and 7 wt% water co-solvent, caffeine/EGCG extraction selectivity reaches 2.57 — meaning for every unit of EGCG co-extracted, 2.57 units of caffeine are removed. A 2025 Hansen Solubility Parameter analysis confirmed that highest selectivity favoring caffeine removal and EGCG retention occurs at 70°C and ≥25 MPa using pure CO₂.
Park et al. (2007, Food Chemistry) used scCO₂ with 95% ethanol co-solvent at 300 bar and 70°C: caffeine extraction yield 92.8%, EGCG extraction yield 69.5% — not ideal for EGCG retention. Kim et al. (2008, Journal of Food Engineering) demonstrated that water co-solvent dramatically improves selectivity: at 40°C, 400 bar, 7 wt% water — caffeine yield 54%, EGCG yield only 21% — selectivity of 2.57. Huang et al. reported 95.6% caffeine removal with water co-solvent at 200–300 bar, with caffeine/catechin ratio of 4.8 (low initial caffeine content matrix). Current consensus (2025 Hansen parameter review): highest caffeine/EGCG selectivity at 70°C and ≥25 MPa (250 bar) with pure CO₂; water facilitates substrate swelling for improved internal mass transfer. The challenge: achieving caffeine <0.1% with maximum EGCG retention requires multi-stage extraction with careful optimization — single-pass scCO₂ at moderate conditions may achieve only 54–75% caffeine removal, insufficient for <0.1% residual caffeine if starting material contains 2.5–4% caffeine (typical green tea: 2.5–4.1% caffeine dry weight).
Target outcome: Residual caffeine ≤0.1% (w/w) in finished decaffeinated GTE; EGCG retention ≥70% of original content
Recommended conditions (water co-solvent approach for selectivity): T = 40–50°C; P = 350–400 bar; water co-solvent: 5–7 wt%; multiple extraction stages (3–5 passes); CO₂ flow rate: optimized to balance contact time and extraction efficiency
Alternative high-efficiency approach (ethanol co-solvent, higher caffeine removal): T = 70–80°C; P = 300 bar; ethanol co-solvent: 7 g per 100 g CO₂; caffeine removal: 92.8%; EGCG loss: ~37.8% — trade-off must be evaluated against EGCG retention requirement; ethanol residual must be tested per USP <467> Class 3
Verification requirement: Finished extract tested by HPLC for caffeine ≤0.1% AND GC-Headspace USP <467> to confirm zero organic solvent residuals; CO₂ leaves no residue but co-solvent (ethanol or water) must be documented and verified
Facility verification: Request equipment specification for CO₂ extraction system — vessel pressure rating (>500 bar capacity); CO₂ recycling system (commercial scCO₂ requires CO₂ recovery to be economically viable); if supplier cannot provide equipment specifications, the scCO₂ claim is unverifiable
4. Hot Water Decaffeination: Parameters and Catechin Retention Data
Hot water decaffeination exploits the differential solubility of caffeine vs. catechins in water at elevated temperatures and short contact times. At 100°C for 3 minutes with a 1:20 leaf:water ratio: caffeine concentration decreases from 23.7 to 4.0 mg/g (83% removal); total tea catechins decrease from 134.5 to 127.6 mg/g — only 5% catechin loss. This selectivity arises from the kinetics of extraction: caffeine is a small, highly water-soluble molecule that diffuses rapidly from the leaf matrix; catechins (particularly the galloylated forms EGCG and ECG) are larger, less water-soluble, and extract more slowly. The process is entirely solvent-free and leaves no chemical residues. Limitations: single-pass treatment at 83% caffeine removal is typically insufficient to achieve <0.1% caffeine if starting material contains 2.5–4% caffeine; multiple cycles are required; each additional cycle may increase catechin losses. The decaffeinated leaf must then undergo a second extraction step to produce the actual green tea extract.
The hot water method decaffeinates the tea leaf — it does not produce a ready-to-use green tea extract. After hot water decaffeination of the leaf, a separate extraction step (typically ethanol/water or water extraction) is required to produce the catechin-rich extract. This two-step process adds cost and complexity. Critically, the second extraction step uses a solvent — so the "water decaffeinated" claim for the leaf does not necessarily mean the finished extract is solvent-free. If ethanol is used for the catechin extraction step following water decaffeination, the finished extract must be tested for ethanol residuals per USP <467> Class 3. Request the full manufacturing process flowchart from the supplier — not just the decaffeination step description.
5. Methylene Chloride and Ethyl Acetate: USP Class and Residue Limits
| Solvent | ICH Q3C Class | USP <467> Limit | FDA Food Use Limit | Toxicity Basis | Clean-Label Impact |
|---|---|---|---|---|---|
| Dichloromethane (Methylene Chloride; DCM) | Class 1 — Should Be Avoided | 600 ppm (pharmaceutical products; USP <467> Option 1) | 10 ppm FDA limit for decaffeinated food products (21 CFR §173.255) | Probable human carcinogen (EPA, WHO, IARC Group 2A); CNS depressant; banned by US EPA from consumer paint strippers (2019); classified as hazardous air pollutant | Completely incompatible with "solvent-free," "natural," "clean label," or "organic" positioning |
| Ethyl Acetate (EA) | Class 3 — Low Toxicity | 5,000 ppm (USP <467>; ICH Q3C Class 3 default) | No specific FDA food limit for EA; GRAS as indirect food additive (21 CFR §173.228) | Low systemic toxicity; no carcinogenicity concerns; characteristic fruity/acetone odor above ~1,000 ppm; hydrolysis to acetic acid in presence of water | Incompatible with "solvent-free" claims; many premium retailers prohibit EA; clean label — conditional on disclosure |
| Ethanol (if used as scCO₂ co-solvent or catechin extraction after water decaff) | Class 3 — Low Toxicity | 5,000 ppm | GRAS (21 CFR §182.1180) | Food-grade solvent; widely accepted; no carcinogenicity concerns | Compatible with most clean-label programs when disclosed; "ethanol extraction" is generally acceptable |
| CO₂ (supercritical) | Not classified — not a residual solvent concern | No limit — CO₂ leaves no residue | GRAS | No toxicity; naturally present in atmosphere; reverts completely to gas at ambient conditions | Fully compatible with all clean-label claims; "CO₂ extracted" is a premium positioning claim |
| Water | Not classified — no residual concern | No limit | GRAS | No toxicity | Fully compatible |
Table 2 — Decaffeination Solvents: ICH Q3C Classification, USP <467> Limits, and Clean-Label Compatibility
Methylene chloride (dichloromethane) faces two separate regulatory limits that are often confused: (1) USP <467> ICH Q3C Class 1 limit: 600 ppm — applies to pharmaceutical products and dietary supplement ingredients per USP standards. A supplier testing to USP <467> may report "DCM: 380 ppm — Passes ICH Q3C Class 1 limit." (2) FDA food use limit: 10 ppm (21 CFR §173.255) — the maximum residue level permitted in decaffeinated coffee and tea for direct human consumption. A product at 380 ppm DCM would be 38× above the FDA food limit even while technically passing USP Class 1 criteria. For dietary supplement ingredients, the governing standard depends on how the product is positioned — pharmaceutical ingredient vs. food ingredient. For all green tea extract positioned as a food supplement in the US, the FDA 10 ppm limit is the stricter applicable standard, not the USP 467 600 ppm pharmaceutical limit.
6. Caffeine HPLC Assay: Verifying <0.1% Residual Caffeine
Residual caffeine in decaffeinated green tea extract is quantified by the same reversed-phase HPLC-UV method used for catechin profiling — caffeine co-elutes with the catechin separation on a C18 column with acidified mobile phase and is detected at 272–280 nm. Caffeine has a characteristic UV absorption maximum at approximately 272 nm; detection at 272 nm provides better sensitivity for caffeine quantification while 280 nm is preferred for catechin quantification. In a simultaneous catechin + caffeine HPLC method, caffeine elutes between EC and EGCG in most reversed-phase gradient systems (approximate retention time 12–15 min on standard C18 gradient). Caffeine content is calculated from a certified caffeine reference standard calibration curve and expressed as % w/w. For a specification of "caffeine <0.1%," the HPLC LOQ must be ≤0.01% (100 ppm) to confirm compliance reliably.
Column: Reversed-phase C18 (ODS); 150–250 mm × 4.6 mm; 5 µm particle size; same column as catechin method
Mobile Phase A: Water with 0.1% orthophosphoric acid (H₃PO₄) or 0.1% acetic acid
Mobile Phase B: Acetonitrile (HPLC grade)
Gradient: Same gradient as catechin method; caffeine elutes between EC and EGCG in standard gradient (approximately 12–16 min); PDA detection confirms identity by UV spectrum match at 272 nm
Detection Wavelength: 272 nm (primary for caffeine); 280 nm (catechins simultaneously)
Column Temperature: 35°C (consistent with catechin epimerization prevention specification)
Reference Standard: Certified caffeine reference standard (anhydrous); purity ≥99%; lot-specific COA
Calibration: External standard calibration; at minimum 5-point curve from 0.01% to 0.5% caffeine (equivalent in solution); R² ≥ 0.9990
LOQ requirement: ≤0.01% (100 ppm w/w) for specification of "caffeine <0.1%" — LOQ must be ≤10% of specification limit
Sample Preparation: Dissolve 100 mg GTE in 100 mL 50% methanol/water; sonicate 10 min; filter 0.45 µm PTFE; inject 10 µL
Result Reporting: Caffeine reported as % w/w; result compared to specification of <0.1%; "ND (LOQ ≤0.01%)" acceptable if caffeine is truly absent
7. USP <467> Solvent Residue Verification for Decaffeinated Extracts
| Decaffeination Method | Solvents to Test | USP <467> Class | Required LOQ (ppm) | Compliance Claim Supported |
|---|---|---|---|---|
| Supercritical CO₂ only (no co-solvent) | No organic solvent to test; CO₂ leaves no residue. Confirm by GC-HS blank run showing ND for all Class 1/2/3 solvents | N/A — no solvent residue expected | LOQ ≤500 ppm (Class 3 screen confirmatory) | Fully supports "solvent-free" and "CO₂ decaffeinated" claims |
| scCO₂ with ethanol co-solvent | Ethanol (primary); also screen acetone, isopropanol (Class 3 cross-contamination check) | Ethanol: Class 3 (5,000 ppm limit) | LOQ ≤500 ppm for ethanol | Cannot claim "solvent-free" — must disclose ethanol co-solvent use; "CO₂ + ethanol decaffeinated" claim is defensible |
| scCO₂ with water co-solvent | No organic solvent residue expected; confirm by Class 3 GC-HS screen showing ND | N/A | LOQ ≤500 ppm screen | Supports "solvent-free" and "CO₂ + water decaffeinated" claims |
| Hot water decaffeination → ethanol extraction | Ethanol (from extraction step); methanol (industrial ethanol contamination check) | Ethanol: Class 3; Methanol: Class 2 (3,000 ppm limit) | Ethanol LOQ ≤500 ppm; methanol LOQ ≤300 ppm | Cannot claim "solvent-free" — ethanol used in extraction step; "water decaffeinated, ethanol extracted" is accurate disclosure |
| Methylene chloride (DCM) decaffeination | DCM (primary); chloroform and other Class 1 solvents (cross-contamination screen) | Class 1 (600 ppm USP <467> limit; 10 ppm FDA food limit) | LOQ ≤1 ppm (to confirm ≤10 ppm FDA food limit) | Cannot support "solvent-free," "natural," or "clean label" claims |
| Ethyl acetate (EA) decaffeination | Ethyl acetate (primary); acetic acid (EA hydrolysis product — check by pH and GC-MS); ethanol (co-extraction) | EA: Class 3 (5,000 ppm limit) | LOQ ≤500 ppm for EA; pH screen for acetic acid | Cannot claim "solvent-free"; many premium programs prohibit EA; "EA decaffeinated" requires explicit disclosure |
Table 3 — USP <467> Residual Solvent Testing Requirements by Decaffeination Method
8. Supplier COA Audit Checklist for Solvent-Free Decaf Claims
Required Documentation for Genuine Solvent-Free Decaffeinated GTE
- Decaffeination method explicitly stated in writing: "Supercritical CO₂ decaffeination at [pressure] bar, [temperature]°C, [co-solvent type and concentration]" — or "Hot water decaffeination at 100°C, 3 min, 1:20 ratio; catechin extraction by [stated solvent]." Generic "solvent-free" without process description is not verifiable.
- Caffeine content by HPLC, LOQ stated: Result ≤0.1% w/w; LOQ ≤0.01%; reference standard identified (caffeine anhydrous, certified purity); individual chromatogram available for batch.
- GC-Headspace residual solvent data per USP <467>: Class 1 solvent screen (DCM, chloroform, benzene — all ND with LOQ ≤1 ppm for DCM); Class 3 screen (ethanol, EA, acetone — ND or within stated limits); method LOQ stated for each solvent.
- scCO₂ facility verification (if claimed): Equipment specifications for CO₂ extraction vessel (rated pressure >500 bar); presence of CO₂ recycling/recovery system; manufacturing process flowchart showing CO₂ step. A facility without high-pressure CO₂ equipment cannot perform scCO₂ decaffeination.
- Full catechin profile post-decaffeination: HPLC showing EGCG, ECG, EGC, EC, GCG individually quantified — confirms catechin retention after decaffeination process; EGCG should be ≥70% of pre-decaffeination content for a compliant scCO₂ process.
- Batch-specific documentation: All results linked to specific production lot; test dates within 6 months of shipment.
Supplier Disqualification Triggers
- Decaffeination method not disclosed — "decaffeinated" claim only, no process description
- GC-Headspace residual solvent data absent — only LOD/moisture stated
- DCM detected at any level in a product claiming "solvent-free" or "natural"
- EA detected above 1,000 ppm in a product with clean-label positioning
- Caffeine result stated as "ND" without LOQ — cannot confirm <0.1% specification
- scCO₂ claimed but facility cannot provide equipment documentation confirming high-pressure CO₂ capability
- EGCG content in decaffeinated product is <50% of the declared grade — excessive catechin co-extraction during decaffeination indicates poor process control
Request a Free 20g Sample for Lab Testing
Verify our decaffeination process with HPLC caffeine assay, full GC-Headspace USP 467 solvent data, and complete catechin profiles. Batch-specific COA with method documentation available upon request.
For Direct Inquiries & Sourcing Specifications
Contact our supply chain team directly via email for rapid RFQ processing and customized MOQ arrangements.
Email: [email protected]
9. Frequently Asked Questions
Not necessarily — two scenarios require clarification. (1) If the product is scCO₂ decaffeinated using only water as co-solvent, then yes — no organic solvent residue is possible and the product is genuinely solvent-free. (2) If the scCO₂ process uses ethanol as co-solvent (which improves caffeine extraction yield from 54% to 92.8% but also increases EGCG co-extraction), then ethanol residuals may be present at levels requiring testing per USP <467> Class 3. Additionally, some suppliers label products as "CO₂ extracted" when referring to the subsequent catechin extraction step (using CO₂ as an extraction solvent) — but the decaffeination step may still have used a separate chemical solvent. Always require: (a) explicit statement of co-solvent type in the decaffeination step; (b) GC-Headspace data per USP <467> showing ND for Class 1 solvents and ND or within-limit for Class 3 solvents; (c) equipment documentation confirming the facility has scCO₂ capability.
Both may apply depending on how your product is regulated — but the FDA food limit of 10 ppm (21 CFR §173.255) is almost certainly the more relevant standard for dietary supplement use. The USP <467> 600 ppm Class 1 limit was established for pharmaceutical drug products and excipients. For dietary supplements positioned as food products — which is the legal classification of most supplement ingredients in the US — the FDA's food additive regulations set a maximum of 10 ppm residual DCM in decaffeinated products. A green tea extract with 380 ppm DCM would pass the USP <467> pharmaceutical limit but would be 38× above the FDA food limit. Furthermore, many retailers and clean-label certification programs specifically prohibit any detectable DCM (ND), regardless of limit — because it is classified as a probable human carcinogen by EPA and WHO. For clean-label and EU compliance, treat DCM as requiring ND (not detected at your method LOQ of ≤1 ppm), not merely ≤600 ppm.
The selectivity challenge with scCO₂ is that no single set of extraction conditions maximizes both caffeine removal and EGCG retention simultaneously. At conditions optimized for high caffeine/EGCG selectivity (40°C, 400 bar, 7 wt% water), the caffeine extraction yield is only 54% per pass — meaning if green tea contains 3% caffeine (typical), a single pass leaves approximately 1.38% residual caffeine, far above the 0.1% target. Achieving <0.1% caffeine from a 3% starting concentration requires removing 97%+ of the caffeine — which demands multiple extraction passes, higher CO₂ flow rates, or modified conditions. The conditions needed for near-complete caffeine removal (ethanol co-solvent, higher temperature) unfortunately also extract more EGCG — up to 37.8% EGCG loss at 80°C ethanol conditions. Commercial scCO₂ decaffeination to <0.1% caffeine with maximum EGCG retention requires multi-stage extraction with carefully optimized sequential conditions — not a single-pass process. This is why genuinely compliant scCO₂-decaffeinated EGCG extract commands a significant price premium over chemically decaffeinated grades.
To substantiate a "solvent-free" label claim for decaffeinated green tea extract, you need: (1) Written supplier declaration of decaffeination method — specifically stating that no organic solvents (DCM, ethyl acetate, methanol, or similar) were used at any stage; (2) GC-Headspace USP <467> data showing ND for all Class 1 and Class 2 solvents (with LOQs ≤1 ppm for Class 1), and ND or trace amounts for Class 3 solvents; (3) Manufacturing process flowchart documenting that scCO₂ with water co-solvent (or hot water only) was used for decaffeination AND the catechin extraction step also used only water or scCO₂; (4) Facility capability documentation confirming scCO₂ equipment is present and operational; (5) Independent third-party verification — at least for first-order supplier qualification, have a retained sample tested at an ISO 17025 accredited laboratory for the full Class 1/2/3 solvent panel. "Solvent-free" is a meaningful and auditable claim only when supported by this documentation chain.
Related Knowledge Center Articles
