Plus de 65 000 laboratoires dans le monde, et ça continue
Test Methods

Botanical Identity Testing — HPTLC, Microscopy, DNA Barcoding

8 min read Updated June 11, 2026

Identity testing answers one question: is this the botanical ingredient you paid for? The supplement industry loses millions annually to adulteration, substitution, and misidentification. HPTLC, microscopy, and DNA barcoding are the three primary identity verification methods. Each has strengths and blind spots. Using the right combination prevents expensive reformulations, Amazon listing takedowns, and FDA warning letters.

Quick answer

HPTLC is the gold standard for botanical identity — it produces a chemical fingerprint showing the full spectrum of compounds in the plant. Microscopy identifies plant parts by cellular structure (e.g., confirming you received root powder, not stem powder). DNA barcoding identifies species but cannot distinguish plant parts or detect non-DNA adulterants. Most quality supplement brands use HPTLC plus microscopy. Cost per test: HPTLC $100-250, microscopy $75-200, DNA barcoding $150-400.

HPTLC: the chemical fingerprint

High-Performance Thin-Layer Chromatography separates the chemical constituents of a botanical extract on a silica-coated glass plate. The sample is spotted at the bottom, the plate is placed in a developing chamber with solvent, and compounds migrate up the plate at different rates. After development, the plate is viewed under UV light (254 nm and 366 nm) and/or derivatized with a staining reagent.

The result is a visual fingerprint — a pattern of bands at specific positions and colors that matches the authenticated botanical reference material. HPTLC answers several questions simultaneously:

  • Identity: Does the sample match the reference? Are the expected marker compounds present in the correct relative proportions?
  • Adulteration: Are there extra bands that should not be there? Goldenseal root (Hydrastis canadensis) has a specific HPTLC fingerprint. Oregon grape root (Mahonia aquifolium), a common adulterant, produces different bands.
  • Quality: Are the marker bands the expected intensity? Weak bands may indicate diluted, over-processed, or improperly harvested material.

HPTLC is recognized by USP, AOAC, and the American Herbal Pharmacopoeia (AHP). It is included in USP monographs for hundreds of botanicals and is explicitly recommended in FDA's Guidance for Industry on Botanical Drug Products.

Microscopy: seeing the plant

Microscopic identification examines the cellular structure of the botanical material. A trained botanist or pharmacognosist mounts a small sample on a slide, adds a clearing or staining agent, and examines it at 100-400x magnification.

Key features they look for:

  • Cell types: Parenchyma, sclerenchyma, collenchyma
  • Tissue fragments: Vascular bundles, trichomes (plant hairs), stomata
  • Crystalline inclusions: Calcium oxalate crystals (druses, raphides, prismatic crystals)
  • Starch grains: Shape, size, and hilum pattern
  • Pollen grains: Species-specific size and surface ornamentation

Microscopy is the only method that can definitively identify which plant part is present. Many adulteration schemes substitute cheaper plant parts for the labeled part. For example, Echinacea angustifolia root commands a premium; Echinacea purpurea aerial parts (stem, leaf, flower) are cheaper. HPTLC may not distinguish root from aerial parts of the same species. Microscopy does.

Microscopy also detects non-plant adulterants: sawdust, powdered grain hulls, starch from non-botanical sources, and inorganic fillers.

DNA barcoding: species-level identification

DNA barcoding amplifies and sequences a short standardized region of the plant genome — typically the ITS2 (internal transcribed spacer 2) or matK/rbcL chloroplast regions. The sequence is compared against reference databases (GenBank, BOLD) to identify the species.

Strengths of DNA barcoding:

  • Species-level resolution, often distinguishing closely related species that HPTLC cannot separate
  • Works on processed material where morphological features are destroyed
  • Useful for identifying powdered blends where microscopic features are unrecognizable
  • Emerging validation: USP has adopted DNA-based methods for several botanicals

Weaknesses of DNA barcoding:

  • Cannot distinguish plant parts (root vs. leaf vs. fruit have the same DNA)
  • Cannot detect non-DNA adulterants (starch, silica, synthetic compounds, solvents)
  • DNA may be degraded in extracts, tinctures, or heat-processed powders
  • Cannot quantify the amount of DNA from each species in a blend (semi-quantitative at best)
  • Does not confirm that the expected phytochemicals are present at therapeutic levels

⚠️ Note

DNA barcoding alone is insufficient for botanical identity testing under FDA GMP (21 CFR 111). 21 CFR 111.75 requires at least one "scientifically valid" identity test. HPTLC plus microscopy fulfills this. DNA barcoding is best used as a complementary method, not a replacement.

Method comparison table

FeatureHPTLCMicroscopyDNA Barcoding
Confirms species identityYes (chemical)Partial (cellular)Yes (genetic)
Distinguishes plant partLimitedYesNo
Detects non-plant adulterantsYesYesNo
Works on extractsYesLimitedDepends (DNA may degrade)
Works on powdersYesYes (with skill)Yes (if DNA intact)
USP-recognizedYesYesEmerging
Cost per sample$100-250$75-200$150-400
Turnaround2-5 days2-5 days5-10 days
Requires reference materialYes (authenticated voucher)No (reference slides/keys)Yes (database)

Real adulteration examples detected by each method

HPTLC caught this: Goldenseal root adulterated with Oregon grape root. The HPTLC fingerprint showed berberine (present in both) plus hydrastine and canadine (unique to goldenseal). The absence of hydrastine proved adulteration despite the presence of berberine.

Microscopy caught this: Echinacea purpurea "root" that was mostly stem and leaf tissue. The microscopist identified abundant trichomes and vascular tissue characteristic of aerial parts, not the starch-rich parenchyma of root tissue.

DNA barcoding caught this: Black cohosh (Actaea racemosa) substituted with Asian Actaea species. Three species of Asian Actaea look similar and have overlapping HPTLC profiles but are genetically distinct. DNA barcoding confirmed the substitution.

When to use which method

ScenarioRecommended approach
New supplier qualificationHPTLC + microscopy
Routine incoming identity testHPTLC (most cost-effective single method)
Suspected plant part substitutionMicroscopy
Suspected species substitution (closely related species)DNA barcoding
Extract or tincture that has been heat-processedHPTLC (DNA degraded)
Powdered proprietary blendAll three if possible; at minimum HPTLC + microscopy
Amazon or FDA documentationHPTLC with authenticated reference material comparison
Research-grade botanical identificationDNA barcoding for definitive species ID

💡 Note

Build your own authenticated botanical reference library. Purchase voucher specimens from AHP, USP, or a university herbarium. Send a retained sample of every accepted lot for HPTLC. When your next lot arrives, compare against your own historical fingerprint, not just the lab's generic reference.

FAQ

Q: Does FDA require all three identity methods?

A: No. 21 CFR 111.75 requires at least one scientifically valid identity test per incoming ingredient lot. HPTLC alone, conducted with authenticated reference material, satisfies this requirement. Most reputable brands use HPTLC plus microscopy for additional confidence. FDA does not mandate DNA barcoding.

Q: Can I skip identity testing if my supplier provides a COA?

A: No. 21 CFR 111.75 explicitly requires you to conduct at least one identity test on each incoming lot before use in manufacturing. A supplier COA alone does not fulfill this requirement. You may reduce testing frequency under 111.75(d) with documented justification, but you cannot eliminate identity testing entirely. See our guide on skip lot testing.

Q: Which method detects fillers like maltodextrin or silica?

A: Microscopy and HPTLC both detect fillers. Under microscopy, maltodextrin appears as amorphous non-plant material. Silica and other inorganic fillers are visible as crystalline fragments that do not match any plant tissue. HPTLC shows fillers as the absence or attenuation of expected botanical bands.

Q: Is DNA barcoding accepted by Amazon?

A: Amazon's Supplement Compliance Requirements do not explicitly name DNA barcoding. They require identity testing by a scientifically valid method. HPTLC and microscopy are more established for Amazon documentation. DNA barcoding data may support your submission but should be presented alongside conventional identity test results.

Q: How much does botanical identity testing cost per lot?

A: HPTLC costs $100-250 per sample. Microscopy costs $75-200 per sample. Combined HPTLC plus microscopy costs $175-400 per sample. DNA barcoding costs $150-400 per sample. For a product with 5 botanicals, expect $500-2,000 per lot for identity testing depending on method selection and lab rates.

Quick Reference

Lab Category: Botanical Identity Testing / Pharmacognosy

Methods:

MethodWhat it detectsCost per sample
HPTLCChemical fingerprint of plant compounds$100-250
MicroscopyCellular structure and plant part identity$75-200
DNA BarcodingSpecies-level genetic identification$150-400

Sample requirements: 5-25 g of raw botanical powder or ground finished product.

Turnaround: HPTLC and microscopy 2-5 business days. DNA barcoding 5-10 business days.

Accreditation: Look for ISO 17025 with botanical identity methods on the scope. Labs specializing in botanicals include Alkemist Labs, Eurofins, and NSF International.

Pricing summary: $175-400 for HPTLC plus microscopy per botanical ingredient. Budget $500-2,000 for identity testing on a typical multi-botanical product.

Country/Region: United States (FDA 21 CFR 111.75), EU (Traditional Herbal Medicinal Products Directive), Canada (NHPD), Australia (TGA).

Get lab testing quotes →

Ready to get your products tested?

Build a basket of the tests you need and compare quotes from ISO 17025–accredited labs in one place. Free to start.

Get lab quotes