You need clean batches.
Fast release.
COAs that survive audits.
Let’s walk the real checklist for fruits & seeds herbs: pesticide residues, aflatoxins, elemental impurities, residual solvents, micro limits, and ID/fingerprint.
No fluff. Just what lets your lot move.
GuoCao runs GMP herbal slicing, ISO 22000, ambient/cool/MA storage, and third-party COA. OEM/ODM ready. Batches ship to 30+ regions. If you’re building a multi-market spec, this saves you time. Browse our catalog here: Wholesale Chinese Medicinal Herbs and Spices.
What it is: The pesticide residue chapter used for herbal drugs. Fruits & seeds fall in scope.
How to write it: “Comply with Ph. Eur. 2.8.13; for residues not listed there, follow EU MRLs.”
Why it matters: Simple, audit-friendly, globally legible.
Where this shows in our range: See Fruit & Seed Chinese Medicines and Dried Spices.
What it is: The MRL backbone in the EU.
Key nuance: Spices and some seed botanicals sit under a grouped approach; regulators have used tMRL logic for spices to reflect real risk and analytical limits.
Why buyers care: It explains why not every SKU should be forced to a blanket 0.01 mg/kg when law allows smarter thresholds.
What it is: Compendial chapter for botanicals.
Aflatoxins (the classic line): AFB1 ≤ 5 ppb; Total ≤ 20 ppb when applicable.
Where to use it: Seeds, nuts, and seed-spice lanes (e.g., bitter apricot seed). Put these limits as a distinct line on the COA.
See a seed-type SKU example: Bitter Apricot Seed (Ku Xing Ren).
Scope reality: Q3D says it doesn’t apply to herbal products.
What good is it then? The PDE-based logic is still great for an internal risk memo: source identification → PDE reasoning → control plan.
How to phrase it: “Q3D not applicable to herbal products; internal limits set via Q3D risk logic and historical data.”
Why this helps: It’s honest and auditor-friendly. Not checkbox quality, real quality.
When this applies: Extracts, concentrates, spray-dried materials.
Your move: Map solvents to Class 1/2/3; validate “minimum feasible” removal; test by GC.
If you only buy dried raw herbs: State N/A clearly to avoid scope creep.
Need extracts later? We can configure OEM/ODM specs—see Herbal factory supply for bulk purchase.
Why we still love it: Practical identity toolkit—macro/micro ID, HPTLC/TLC, foreign matter, moisture.
Use case: Incoming QC and skip-lot justifications once your farm blocks and seasons are stable.
Heads-up: For seeds, watch a.w. (water activity). High a.w. loves trouble.
How EMA frames it: ID (often HPTLC), markers/fingerprint, pesticides, aflatoxins, micro, and finished-form tests (disintegration/dissolution) when you have tablets/capsules.
Good news: With strong GACP and clean history, some tests can be reduced frequency.
Translation: Documented reality → leaner release.
Use this as a starting point. Adjust per SKU and market.
| Item | Spec language (concise) | Method / basis | Frequency |
|---|---|---|---|
| Identity | Pass macro/micro ID + HPTLC fingerprint vs. authentic reference | WHO ID + HPTLC; EMA frame | Each lot |
| Moisture / a.w. | NMT 12% moisture or validated a.w. target per SKU | Loss on drying or a.w. | Each lot |
| Pesticide residues | Comply with Ph. Eur. 2.8.13; for non-listed, follow EU MRLs | GC-MS/MS, LC-MS/MS | Each lot or skip-lot with strong GACP |
| Aflatoxins | AFB1 ≤ 5 ppb; Total ≤ 20 ppb (where applicable) | HPLC/LC-FLD/LC-MS | Each lot for high-risk seeds |
| Elemental impurities | Q3D not applicable; internal limit set by Q3D-style risk memo | ICP-MS; risk assessment | Periodic / on change |
| Residual solvents | Meet ICH Q3C class limits (extracts) | GC-FID/GC-MS | Each extract lot |
| Micro limits | Per pharmacopeial / market | Plate count or validated rapid | Each lot |
| Volatile oils | Internal range for aroma SKUs | Hydrodistillation / GC | Periodic trending |
| Packaging | Barrier performance, clean headspace, no cross-contam | Spec review + visual | Each lot |
You don’t release into a vacuum.
Big picture: Most herbal samples land within legal residue limits. Coordinated programs continue to show high compliance; risk-based programs surface outliers in spice/seed lanes.
Use this to calm stakeholders while you still tighten the hot SKUs. No need to over-test every batch forever. Smart paneling wins.
| Topic | Standard / guidance | Reality check | What you do |
|---|---|---|---|
| Pesticide residues | Ph. Eur. 2.8.13 + EU MRLs | Most pass; outliers sit in spice/seed lanes | Targeted panels + history-based skip-lot |
| Spices tMRL logic | Grouped approach for spices in EU framework | Practical, risk-reflective | Don’t force blanket micro-levels where law set sensible tMRL |
| Aflatoxins | USP <561> (AFB1/Total) | Seeds & nuts stay sensitive | Keep it as a separate COA line |
| Elemental impurities | ICH Q3D scope: not applicable | Still use Q3D logic internally | Write a short PDE memo, test periodically |
| Residual solvents | ICH Q3C | Extracts need proof of “minimum feasible” | Map solvents, validate, show data |
| Identity and fingerprint | WHO + EMA HPTLC | Auditor-friendly and cheap to run | Keep ref materials current |
Start broad for a new origin or a new farmer.
Lock history for two or three harvests.
Then tier your panel:
Use LOQ/LOD language, note your lab’s method (GC-MS/MS, LC-MS/MS).
Call out CAPA when something bumps a spec.
Don’t hide the ball. Buyers can smell that a mile away.
Micro can spike when a.w. and handling drift.
Simple rules that pay off:
We run this daily. Kinda boring, but it works.
HPTLC fingerprints say “this is the plant we promised.”
Pair it with macro/micro ID.
If the market expects a marker (e.g., lignans for schisandra), include a target range or at least a trend chart.
Regulators like it. Your flavor team likes it more.
Do I have to run a huge residue panel every time?
No. Start broad. Build a history file. Move to skip-lot for Tier-2, keep Tier-1 every lot.
Can I cite Q3D for herbs?
Yes, but be clear: not applicable by scope. You can still use its PDE logic internally. That’s the grown-up way.
What if I only sell raw dried herbs, not extracts?
State Q3C not applicable. If you later add extracts, add the Q3C section and GC method. Simple.
How many markers do I need?
Enough to prove identity and consistency. HPTLC + one sensible marker often does the trick. Dont overcook it.
