Yam Extract Specifications and Test Methods: Is 40% Aglycone or Total Saponins?

This article is part of the Yam Extract overviewHerb Wiki Editorial TeamCAS 512-04-9Last updated 2026-10-28

Definition in One Sentence

A yam extract specification is the diosgenin (CAS 512-04-9) content bracket of a standardised extract made from the rhizome of Dioscorea opposita Thunb., measured by the method agreed in the contract, together with the associated identity, ash, moisture, heavy metal and microbiological controls. Because the pharmacopoeial yam monograph sets no assay item, the specification for this category has no official method behind it: the same lot can be quoted as three figures differing by a factor of two depending on whether the basis is ratio, total saponins or aglycone, and only enterprise standards and contract terms can pin it down.

Three Layers of Specification: Ratio, Total Saponins, Aglycone

The confusion in yam extract specifications is not caused by fraud; three layers of definition naturally sit on top of one number. Each layer says something different, and only the third can be compared directly.

LayerHow it appears on a reportWhat is measuredTypical reported valueDirectly comparable?
Ratio grade10:1, 20:1Nothing measured; only the feed ratio is promised—No. A 10:1 product may contain only 1%–5% aglycone
Total saponinsTotal saponins ≥40% (as dioscin)The glycoside with its sugar chain20%–50%Comparable only after conversion
AglyconeDiosgenin ≥40% (after hydrolysis, as aglycone)The aglycone after removal of sugars10%–45%Comparable, but still ask about method and conversion basis

A ratio grade is a number without a denominator. "10:1" says only that 10 kg of raw material yields 1 kg of extract; it promises no constituent level. Fresh yam or dried slices, aqueous or ethanolic extraction, hydrolysed or not — the same 10:1 can describe completely different products. Comparing a 10:1 price with a 40% price is the most common mismatch in this kind of purchasing.

Between total saponins and aglycone lies the weight of a sugar chain, and that is the arithmetic of the next section.

The Pharmacopoeia Cannot Help: No Official Method for This Category

There is no official method for the assay of yam extract: the pharmacopoeial yam monograph contains only description, identification, moisture, total ash and sulfur dioxide residue, with no assay item. The reason is straightforward — the medicinal value of Chinese yam is not built on a single marker compound, so the official standard does not set one.

This contrasts sharply with related species:

SpeciesPharmacopoeial markerMeaning for extract specifications
Chinese yam D. oppositaNo assay itemNo official method to cite; enterprise standards only
Japanese yam D. nipponicaDioscin ≥1.3%A starting point for a glycoside method only; not directly applicable to aglycone
Scutellaria baicalensisBaicalin ≥9.0%An official minimum exists, so extracts can be benchmarked

There are three consequences:

  1. You cannot say "tested to the pharmacopoeia". When a supplier writes "tested against: enterprise standard" on the COA, the buyer has no statutory clause to invoke and must fall back on the contract.
  2. Method differences cannot be dissolved by "we all follow the law". The gap between colorimetry and HPLC, and between hydrolysed and non-hydrolysed testing, all lands in the contract terms.
  3. The specification sheet matters more here than in other categories. For an item with a pharmacopoeial minimum, such as baicalin, the sheet is supplementary; for yam extract, the sheet is everything.

So the first question in this category is not price but method. If the method is unclear, every later price comparison is a comparison of different things.

Conversion: One Factor from Dioscin to Diosgenin

Conversion chain from glycoside to aglycone
Figure 1. Conversion chain from dioscin to diosgenin: after removal of three sugar units the molecular weight falls from about 869 to 414.62, a theoretical mass yield of 0.477. Placeholder image, to be replaced with a photograph or illustration.

Saponins in the yam rhizome occur as dioscin (C₄₅H₇₂O₁₆, molecular weight about 869.05), in which diosgenin carries two rhamnose units and one glucose at C3. Hydrolysis removes these three sugars and leaves the aglycone:

aglycone % = total saponins % × (414.62 ÷ 869.05) = total saponins % × 0.477

The reverse uses the same single factor: total saponins % = aglycone % ÷ 0.477 = aglycone % × 2.096.

Aglycone basisEquivalent total saponins (as dioscin)What this grade is usually called
19%39.8%The industry's most common "40% total saponins"
20%41.9%Same, reported as a round number
30%62.9%Less common; requires stronger enrichment
40%83.8%Our grade
42.02%88.1%Measured on batch 26052501

The most important row in this table is 40% corresponding to 83.8%.

Total saponins in dried yam slices are only a few percent; most of the mass is starch, polysaccharide and protein. To reach 83.8% total saponins means more than nine tenths of those non-saponin components have been removed — a product that, in chemical composition, is closer to a crude saponin than to what is normally meant by an extract. So the difference between "40% aglycone" and "40% total saponins" is not a level of content but whether the hydrolysis and deep-refinement steps are performed, with a completely different cost structure.

That yields a shortcut for judging the basis: look at the order of magnitude of the price. If a "40% yam extract" is quoted at the same level as "40% total saponins", it is almost certainly not aglycone. True 40% aglycone approaches the saponin monomer in raw-material consumption and processing steps, and its price cannot match.

Test Method: Colorimetry for Extracts, 203 nm Only for the Monomer

The same diosgenin molecule needs two completely different methods in a monomer and in an extract, and the reason is the matrix.

Diosgenin has no conjugated system and can only be detected in the terminal-absorption region at 203 nm. That wavelength is barely usable for a pure crystalline sample and unusable for an extract:

  • starch hydrolysates, uronic acids, proteins and pigments all absorb strongly between 200 and 210 nm, lifting the whole baseline;
  • saponins make up only 40% of the extract (or just 19% on a total-saponin basis), and the remaining 60%-plus all produces peaks at 203 nm, drowning the main peak;
  • molar absorptivity in the terminal-absorption region is intrinsically low, and once matrix interference is added the signal-to-noise ratio cannot support quantitation.

The industry therefore splits the work: monomers by HPLC-UV at 203 nm, extracts by vanillin–perchloric acid colorimetric spectrophotometry (about 540 nm).

ParameterExtract (colorimetry)Monomer (HPLC-UV)
PrincipleSteroidal sapogenins react with vanillin–perchloric acid and are measured at 540 nmTerminal absorption at 203 nm, external standard
Reference standardDiosgenin reference standardDiosgenin reference standard
HydrolysisUsually hydrolysed before colour developmentAlready hydrolysed; injected directly
Separation powerNoneYes
PrecisionLower; duplicate RSD usually controlled at ≤2%Higher; usually controlled at ≤1.5%
Equipment thresholdA spectrophotometer is enoughHPLC-UV

Why this division makes sense: a plant running quick controls on dozens of samples a day gets a spectrophotometric result in ten minutes at almost no cost, and an extract is a mixture product by nature — the buyer wants to know whether the total saponin is right, not whether the peak at 7.3 minutes is an isomer. HPLC comes out only where structural information is genuinely needed: high-purity orders, disputed re-tests and regulatory filings.

But division is not equivalence. Signing a "diosgenin ≥40%" contract on colorimetric data means taking on structural responsibility with a number that has no separation power. The method must be written into the contract, or a dispute has no resolution.

What the Colorimetric Method Actually Measures

The vanillin–perchloric acid reaction does not target diosgenin as a single compound; it targets the chromophore of steroidal sapogenins. Any molecule with a similar skeleton will develop colour, which defines four limitations:

Source of interferenceConsequence
25R and 25S epimers (yamogenin)Cannot be separated at all; either counts toward the total
Related steroidal sapogenins (diosgenin-type, pennogenin and others)Counted together, biasing the result above the true diosgenin value
Incompletely hydrolysed dioscinThe sugar chain does not give this colour, so the reading is low; but incomplete hydrolysis also leaves related substances, so the direction is uncertain
Sugars, proteins and pigments at high concentrationPositive interference; the "dirtier" the sample, the higher the inflated reading

The strict wording is therefore "total steroidal sapogenins, calculated as diosgenin", not "diosgenin content". Where a COA says only "Diosgenin ≥40% Assay" without method or reference standard, at least half of what that 40% means is unknowable.

Three questions that must be asked; without any one of them the figure cannot be interpreted:

  1. the colour-development conditions (vanillin concentration, acid type and concentration, water-bath temperature and time);
  2. which reference standard the calibration curve used, and whether its range covers the measured value;
  3. whether the assay follows hydrolysis, and under what hydrolysis conditions.

Without answer 2, even the reference basis of the 40% is unstated; without answer 3, the most basic fork — glycoside or aglycone — is left hanging.

Batch 26052501: Item-by-Item Reading and Consistency Check

Batch 26052501 was tested on 2026-05-25 and reported on 2026-06-03 against the enterprise standard. The overview page lists every item; here we look only at two things it does not address: completeness of the report and consistency between the values.

Consistency first. The report gives the assay together with two physical constants:

ItemMeasuredValue as a criterion
Assay42.02% (diosgenin ≥40%)Depends on method and reference standard; not verifiable on its own
Melting point206 ℃ (single point, no range)Of limited value for an extract, see below
Specific rotation−123.5°Can rule out a marked excess of 25S, provided the sample is pure enough
Moisture0.57%Low, favouring a dried-basis calculation

Melting point is an awkward item for an extract. A melting point is a property of the mixture: if the sample really contained 57% starch, polysaccharide and protein, the initial melting point should be markedly depressed and the range broadened to ten degrees or more. The report gives a clean single value of 206 ℃ with no range recorded, and there are only two explanations — either the sample has been refined to a crystallisable degree (in which case its actual purity is higher than the reported figure), or the melting point was carried over from a monomer report template: measured, but without the range recorded as required.

Both explanations lead to the same conclusion: on an extract COA, melting point and specific rotation have far less diagnostic value than on a monomer. Their presence suggests the sample is not impure; their presence in such "textbook" form suggests the report may have been carried over from a monomer template. The buyer need not reject on this basis, but should ask for the melting range and the heating rate — a supplier who can answer did the work; one who cannot probably copied it.

Completeness next. The report is missing three items of different kinds:

  • Ash / residue on ignition missing: this directly decides whether a mass balance can be done, as the next section shows. It is the item least worth omitting.
  • Salmonella missing: microbiological testing covered only total aerobic count, yeast and mould, and E. coli. With yam's high starch content, this is a hard gap for export and infant-formula-related applications.
  • Sulfur dioxide residue missing: Chinese yam is a traditionally sulfur-fumigated species, one of those in the 400 mg/kg tier of the pharmacopoeia; the extract is not directly subject to that limit, but fumigation of the raw material carries into the finished product.

Mass Balance: Three-Term Addition to Detect Carrier Dilution

Mass balance in three terms
Figure 2. The relationship between assay, moisture, ash and the residual: different residual bands point to carrier dilution, a wrong basis or normal polysaccharide residue. Placeholder image, to be replaced with a photograph or illustration.

Given assay, moisture and ash, yam extract lets you back out the mass that is not accounted for in the report, and thus judge whether the lot has been diluted with a carrier.

100% = assay + moisture + ash + other organic matter (starch, polysaccharide, protein, pigment)

Of these four terms, assay, moisture and ash are all available from the COA (provided nothing is missing), and the residual is what remains. Take batch 26052501, assuming an ash of 2.0%:

ItemValue
Assay (aglycone)42.02%
Moisture0.57%
Ash (assumed)2.00%
Residual: other organic matter55.41%

Where the residual falls points to completely different problems:

Residual bandIndicatesNext step
< 5%The assay basis is wrong, or ash (inorganic filler) is abnormally highRequest the full method; re-test ash
10%–25%Signs of carrier dilutionAsk about carrier type and addition rate
30%–70%Normal; mostly residual polysaccharide and starchAccept, but ask whether a carrier was added
> 75%The assay figure does not fit the sampleRequire re-testing; consider changing supplier

At 55.41% the value falls in the normal band, so the composition of this lot is plausible.

One qualification is essential, or the method will be misused: a mass balance cannot detect a switch between glycoside and aglycone bases.

If the same lot were actually "40% total saponins" (equivalent to 19.08% aglycone), the residual would become 78.35% — a figure equally plausible for a yam extract, because starch and polysaccharide are the bulk anyway. That is:

  • Carrier dilution (maltodextrin blending a low assay up to a high one): the mass balance can see it, because the dextrin alters the composition of the residual while ash and moisture are often abnormal;
  • Switching the basis (reporting the glycoside as aglycone): the mass balance cannot see it; both residuals fall in plausible ranges.

The basis problem can only be checked three other ways: ask for the full method and the reference-standard lot number, ask for a written statement of the conversion factor, and re-test yourself by HPLC with an external standard. This also explains why the conversion table in section 3 is worth memorising — it is the only tool that lets you judge whether a quote is reasonable without seeing the report.

As for whether the sugars in the residual are native to the yam or added dextrin, a routine COA cannot tell. Only a polysaccharide composition analysis (the glucose-to-mannose ratio) or starch gelatinisation characteristics can distinguish them — extra testing beyond routine acceptance, done only on large-value orders or in quality disputes.

Sampling: Blended Powders Segregate More Than Crystals

Yam extract is a blended powder of "saponin crystals plus a starch-polysaccharide carrier". The two components differ greatly in density and particle size, so transport vibration separates them: fine, light particles rise and heavy crystals sink.

The consequence is direct: in the same drum, top and bottom layers can differ by several percentage points in assay. That variation does not exist in a monomer (a crystalline powder is homogeneous) but is normal in an extract. Get sampling wrong and every subsequent data point is void.

  • Stratified sampling: take from the top, middle and bottom layers and combine, rather than scooping from the surface.
  • Sample size: not less than 200 g, mixed and quartered down to the test quantity; a single assay portion not less than 200 mg.
  • Duplicates: run in duplicate; a relative deviation above 2% requires re-sampling. Colorimetry is less precise than HPLC, so this limit is wider than the 1.5% used for monomers — but no wider.
  • Retention sample: not less than 100 g per batch, kept for six months beyond the re-test date.
  • Decision rules: failure of any one of assay, ash, moisture, heavy metals or microbiological limits fails the batch. When the assay sits at the edge of the limit (for example 40.5%), re-sample and re-test before deciding; never release on a single result.

Specification Sheet Fields (Ready to Copy)

Product name: Yam Extract
Botanical source: rhizome of Dioscorea opposita Thunb. (with feed batch correspondence)
Marker: diosgenin (CAS 512-04-9)
Assay: ≥40.0% (vanillin–perchloric acid colorimetry, 540 nm, calculated as
       diosgenin reference standard, after hydrolysis, as aglycone, on the dried basis)
       — if the supplier uses HPLC: ≥40.0% (HPLC-UV 203 nm, external standard, dried basis)
Identification: TLC or HPLC main peak matches the reference standard; starch grains
       seen under the microscope (indicating a carrier)
Appearance: off-white to pale yellow powder
Moisture: ≤5.0%
Total ash: ≤5.0% (≤2.0% may be required for carrier-free grades)
Acid-insoluble ash: ≤1.0%
Sulfur dioxide residue: ≤150 mg/kg (stricter than the 400 mg/kg tier for
       pharmacopoeial yam; recommended as an in-house limit)
Heavy metals: Pb ≤0.5, As ≤1.0, Cd ≤0.5, Hg ≤0.1 (mg/kg)
Microbiological: total aerobic count ≤1000 cfu/g; yeast and mould ≤100 cfu/g;
       E. coli not detectable; Salmonella not detectable
Carrier: □ none  □ maltodextrin, addition rate ____%
Residual solvents: ethanol ≤5000 mg/kg (ICH Q3C class 3)
Particle size: 100% through 80 mesh
Packaging: 25 kg fiber drum with a double PE liner
Storage: sealed, cool and dry, protected from light
Re-test period: 24 months

In a yam extract specification, the three lines most often deleted are precisely the ones to keep: the parenthetical in the assay line (method and basis), the ash line (the denominator of the mass balance), and the tick-box in the carrier line (the legitimate route for dilution).

How to Send a Sample for Third-Party Re-Testing

For third-party re-testing of yam extract, success rests almost entirely on the request form: if method, reference standard and hydrolysis conditions are not stated, the resulting report is not comparable with the original and the re-test is pointless.

  1. Specify the method. Most third-party laboratories default to HPLC and immediately hit the 203 nm integration problem on an extract. The request form must state "vanillin–perchloric acid colorimetry" or "laboratory to select and report the method and its basis".
  2. Ask for the calibration curve and reference-standard lot number. Colorimetric results depend heavily on curve range and reference standard; without both, the numbers are not comparable.
  3. Include a blind sample. Send a sample of known content alongside to assess the laboratory's systematic bias. This screens out some organisations that only produce reports.
  4. State the hydrolysis conditions. Whether hydrolysis is performed, acid concentration, temperature and time directly determine whether the glycoside or the aglycone is measured. Omit this and the re-test is not comparable with the original.
  5. Only historical data from the same laboratory and method are comparable. Most cross-laboratory differences come from the method, not the quality.

FAQ

Q1: The COA says "diosgenin 40%" — how do I know whether that is aglycone or total saponins?

Look at the parenthetical. "After hydrolysis, as aglycone" or "calculated as diosgenin reference standard" means aglycone; "calculated as dioscin" means the glycoside, which converts to only about 19% aglycone. If there is no parenthetical, ask for the full method — this is not pedantry but the only way to pin down the basis in this category.

Q2: Why isn't HPLC used for yam extract?

It is not that HPLC is not used; the matrix does not allow it. In the 203 nm terminal-absorption region, starch hydrolysates, uronic acids, proteins and pigments all absorb, while saponins are only about 40% of the material, so the main peak is buried in matrix peaks. Colorimetry trades separation power for usability, which is standard practice for extract products.

Q3: Can a 40% from colorimetry be treated as "diosgenin 40%"?

In a contractual sense yes, but understand that it actually means "total steroidal sapogenins, calculated as diosgenin". Related sapogenins and the 25S epimer are included, so the true value may be slightly lower. For regulatory filings, or where downstream use is sensitive to isomers, add an HPLC test and an isomer ratio.

Q4: Is the difference between a 40% extract and a 92% monomer one of process or of purity?

Both require deep refinement; the difference is downstream positioning. The extract keeps a food-compatible process route and a polysaccharide matrix for the formulation end; the monomer must crystallise to the point where a melting point can be measured, for steroid synthesis. Converted to cost per kilogram of saponin, the monomer is usually cheaper — the extract's premium buys food-compliant status and formulating convenience, not purity.

Q5: A mass balance gives a 55% residual — is that normal?

Yes. A yam extract is mostly starch, polysaccharide and protein, so a residual of 30% to 70% is reasonable. Watch for residuals that are too small (under 5%, suggesting abnormal ash or a wrong assay basis) or too large (over 75%, meaning the assay does not fit the sample).

Q6: If a mass balance can detect dilution, why can't it detect the basis?

Because the two problems give different signals. Carrier dilution changes the composition of ash and residual, whereas switching the basis only changes the denominator of the same number, leaving the residual in a plausible range. The first is visible in an addition; the second only through method and reference standard.

Q7: Two tests on the same lot differ by 2% — is that normal?

Common for extracts, for three reasons: stratified sampling, the inherently lower precision of colorimetry compared with HPLC, and inconsistent drying conditions. Standardise sampling (three layers combined) and drying conditions first, then discuss method differences. If two consecutive results exceed 2%, consider batch homogeneity.

Q8: Does yam extract need sulfur dioxide testing?

We recommend it. Chinese yam is a traditionally fumigated species in the pharmacopoeia's 400 mg/kg tier; the extract is not directly subject to that limit, but fumigated raw material carries sulfur dioxide into the finished product. An in-house limit of 150 mg/kg is stricter than the raw-material limit, adds limited cost, and avoids explanations at export.

References and Search Strategies

  1. Chinese Pharmacopoeia Commission. Pharmacopoeia of the People's Republic of China (2025 edition, Part I), monograph "Dioscorea opposita" (botanical source, description, identification and tests; no assay item).
  2. Chinese Pharmacopoeia Commission. Pharmacopoeia of the People's Republic of China (2025 edition, Part IV), General Chapter 0212 General Rules for the Testing of Crude Drugs and Decoction Pieces (sulfur dioxide residue, Chinese yam being in the 400 mg/kg tier), General Chapter 0401 UV-Visible Spectrophotometry, General Chapter 0512 High-Performance Liquid Chromatography.
  3. Ministry of Health. Catalogue of Substances That Are Both Food and Medicine (Annex 1 to Wei Fa Jian Fa [2002] No. 51) — the listing basis for Chinese yam as a food-and-medicine substance.
  4. PubChem: molecular formulae and weights of diosgenin (CID 99474, CAS 512-04-9) and dioscin (CAS 19057-60-4), used to verify the glycoside-to-aglycone conversion factor.
  5. Methods for assaying yam saponins: PubMed search "Dioscorea opposita"[Title/Abstract] AND (saponin OR diosgenin) AND (spectrophotometric OR vanillin OR determination). General search for steroidal sapogenin colorimetry: "steroidal sapogenin" AND "vanillin" AND "sulfuric acid".

Items 4 and 5 are search paths rather than specific citations; use them to locate and verify the original papers.

Disclaimer

This content is technical material on raw-material testing and quality control, intended for R&D, QC and purchasing staff. It does not constitute medical advice or a compliance opinion. Method parameters given here are examples of common practice; actual testing must follow a validated method and the current standard. Regulations and limits change with revisions; for transactions and filings, rely on the current official standard, the agreed contract and third-party test reports.