Loss on Ignition: Meaning, Evidence and Controls
Understand what ignition changes in botanical and mineral materials, how to calculate the result, and what evidence makes it interpretable.

On this page · 11 sections
- 1What loss on ignition measures—and what it cannot prove
- 2LOI, sulfated ash, total ash, acid-insoluble ash and LOD
- 3Why botanical and mineral materials change mass during ignition
- 4Binding US supplement rules, WHO guidance and geological practice
- 5From representative test portion to a defensible constant-mass endpoint
- 6Calculate LOI using the sample denominator and an explicit blank rule
- 7Worked cases: constant mass, crucible drift and a negative result
- 8The records needed to reconstruct an LOI result
- 9Move results between systems without losing their analytical meaning
- 10Recognize procedural failures before interpreting a material difference
- 11A final checklist for method adoption and result review
How does Loss on Ignition (LOI) — Botanicals & Minerals apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What loss on ignition measures—and what it cannot prove
Loss on ignition is the net mass loss of a test portion after ignition under a specified method, expressed relative to its initial mass on the stated reporting basis. Often abbreviated LOI, it describes the portion actually tested under those conditions. It is not an intrinsic material constant independent of sample preparation, ignition procedure or the basis used to report the result.
LOI alone does not prove identity or purity: different materials can produce similar net mass changes. For botanicals, adhering soil can warrant a separate contamination assessment without automatically establishing that the plant material has failed identity testing. Evidence addressing what the botanical is belongs in an appropriate identity assessment, potentially using complementary identification methods, rather than being inferred from a gravimetric result.
Botanical and mineral testing are different analytical contexts. A procedure useful for characterizing a mineral material is not automatically suitable for assessing an herbal ingredient. Method selection depends on the material, the applicable specification and the intended decision; LOI is not a universal testing recommendation for either group.
Keep the measured result, specification comparison and human disposition decision separate. The result records an observation; comparison asks whether it meets the applicable criterion on a matching method and reporting basis. An unexpected or failing comparison may lead to out-of-specification handling. Disposition then considers the reviewed analytical evidence and any investigation, rather than treating the LOI value itself as a release instruction.
03Why botanical and mineral materials change mass during ignition
Heating can remove water, consume organic matter through combustion, or volatilize organic constituents and their decomposition products. Susceptible mineral constituents can also release carbon dioxide during carbonate decomposition. Each process can reduce the mass retained in the crucible, but the balance records their combined effect rather than identifying the substances that left.
A botanical matrix may contain organic material alongside inorganic constituents. Its loss can therefore reflect more than organic matter alone. A mineral matrix may contain hydrated, carbonate-bearing or oxidizable constituents, allowing several reactions to contribute to the same measurement. These are possible compositions, not assumptions about every botanical or mineral sample; the material’s composition informs which explanations are plausible.
Oxidation can act in the opposite direction: oxygen incorporated into the retained material increases its mass. If that gain exceeds simultaneous losses, the signed LOI result is negative. Offsetting gains and losses can also produce a modest net result despite substantial chemical change. Similar LOI values can consequently arise from different combinations of reactions, rather than from chemically equivalent materials.
Interpretation depends on the atmosphere, sample preparation, prior thermal history and exposure during cooling. The atmosphere affects reaction pathways; preparation affects exposure and reaction extent; previous heating may already have altered the material. Cooling exposure can permit moisture uptake or other interaction with the surroundings before weighing. Operational controls for these variables belong in the method lifecycle.
On the stated sample and blank basis, a negative result indicates net mass gain under the method—not oxidation by itself. Interpreting it first involves separating the sample change from the weighing assembly, as in net-versus-gross reconciliation. The worked negative-loss case applies this distinction to the mass ledger.
04Binding US supplement rules, WHO guidance and geological practice
For US dietary supplements, 21 CFR 111.75(a)(1)(i) requires at least one appropriate test or examination to verify the identity of a dietary-ingredient component before use, unless FDA grants the petition-based exemption identified in that provision. Apply the analytical evidence boundary established above when deciding what an LOI result contributes to component identity confirmation.
| Authority or reference | Product and jurisdictional scope | Status | Appropriate use |
|---|---|---|---|
| 21 CFR 111.75(a)(1)(i) | US dietary supplements; dietary-ingredient components | Binding US regulation | Identify the component-identity obligation described above. |
| WHO: Quality control methods for herbal materials | Herbal materials; international reference | Official guidance reference | Consult for its general subject: quality control of herbal materials. |
| USGS analytical-methods report for geologic and other materials | Geologic and other materials; US geological technical context | Technical reference, not a manufacturing regulation | Consult as an analytical reference, separately from manufacturing obligations. |
Keep the legal obligation separate from method selection. In laboratory practice, selecting LOI starts with the analytical question and the applicable material specification—not simply with a reference appearing in this map. Document the intended use of the result alongside the relevant component specifications; do not treat this map as a universal LOI testing instruction.
For pharmaceuticals, foods, devices, chemicals and markets outside the US dietary-supplement context, consult the applicable current product specification, analytical method and jurisdictional framework. Assess their applicability independently rather than extending the supplement provision to those products or markets.
05From representative test portion to a defensible constant-mass endpoint
In routine gravimetric laboratory practice, begin with the analytical question and the approved method revision: which material change is being measured, and on what reporting basis? Treat material preparation, crucible suitability, heating atmosphere, thermal sequence, cooling sequence and endpoint definition as method-controlled choices. Record these choices before testing rather than adjusting them to obtain an expected result. Comparability depends on the sequence, not simply on using the same furnace.
Take the test portion through the sampling and preparation procedure appropriate to the material. Botanical fragments, fines and dense mineral particles can segregate during storage, pouring or subdivision. Mixing and subdivision should preserve the intended representation of those fractions. Grinding, sieving and transfer can selectively remove dust, retain particles on equipment or change moisture exposure; use only the preparation specified by the method and record departures. A conveniently accessible scoop is not necessarily representative.
Identify the crucible so that every weighing remains attributable to the same vessel. Use a crucible suitable for the material and thermal conditions, condition it according to the method, and record its tare. Load the prepared portion without losing material and retain the initial gross mass as well as the calculated sample mass. Keep the tare-verification record connected to that determination, rather than relying on an untraceable balance-zero operation.
Follow the specified heating and cooling sequence. During cooling and transfer, moisture exchange and contamination can change the mass; weighing a vessel before it reaches the method-defined weighing condition can produce unstable readings. Control exposure and handling consistently, with suitable spillage and cross-contamination precautions. Do not substitute a universal cooling time for the selected method’s conditions.
Constant mass means satisfaction of the selected method’s agreement criterion across its prescribed weighing sequence. Apparent stability does not establish complete reaction or analytical specificity. Retain the successive readings used to demonstrate the endpoint, including readings preceding the accepted mass. If the sequence is interrupted or the endpoint remains unmet, record the interruption, observations and sequence position; route the determination for review under the laboratory procedure before accepting, continuing or repeating it.
- Select methodMatch material, specification, reporting basis and intended decision.
- Identify preparationAssociate sample, conditioned crucible and any blank with the run.
- Capture initial massesRecord tare and loaded masses with units and balance identification.
- Ignite and CoolFollow the method's thermal sequence and handling controls.
- Assess endpointRetain successive readings and evaluate the specified agreement criterion.
- Calculate ResultApply the stated denominator and authorized corrections; retain the sign.
- Review and decidePeople assess exceptions, approve results and determine material disposition.
06Calculate LOI using the sample denominator and an explicit blank rule
For this mass balance, express every measured mass in grams. Keep raw balance observations separate from calculated values: the initial loaded gross mass is not the sample denominator. This is the same distinction used in net-versus-gross reconciliation.
| Symbol | Physical quantity | Unit | Role in calculation |
|---|---|---|---|
| C | Initial conditioned sample-crucible tare | g | Raw observation; initial crucible basis |
| T | Initial loaded gross mass | g | Raw observation; crucible plus initial sample |
| R | Final sample-crucible gross mass | g | Raw observation; crucible plus residue |
| B_initial | Matched blank’s initial crucible mass | g | Raw observation; blank starting basis |
| B_final | Matched blank’s final crucible mass | g | Raw observation; blank ending basis |
| S = T − C | Initial sample mass | g | Derived value; reporting denominator |
| d = B_final − B_initial | Signed blank crucible mass change | g | Derived value; conditional correction |
Use blank correction only where the selected method authorizes it and the matched blank represents the relevant sample-crucible change. Under that assumption, the final crucible contribution is C + d. Subtracting it from R gives the corrected residue mass: R − C − d.
Corrected signed mass loss is initial sample mass minus corrected residue mass: S − (R − C − d) = T − R + d. The signed LOI mass fraction is therefore (T − R + d) / S. Keep this fraction dimensionless during calculation and convert it to percent only at the reporting stage. Preserve the initial sample basis rather than silently switching to a dry-mass or residue denominator.
A positive blank mass change is added back to the apparent loss because crucible mass gain otherwise masks sample loss. A negative d reduces the apparent loss. Do not subtract the entire blank crucible mass: the sample crucible already has its own tare, C. Tare-verified weighing helps keep these separate mass bases explicit.
A blank used to monitor suitability is not automatically authorized for numerical correction. If correction is not authorized, retain the blank observations separately and use the method’s uncorrected calculation. Blank subtraction cannot repair splashing, physical sample loss or an unsuitable analytical procedure.
07Worked cases: constant mass, crucible drift and a negative result
These symbolic examples use a shared illustrative method: report on the loaded as-received sample basis, use a conditioned crucible tare, follow the selected method’s ignition and cooling sequence, and assess its prescribed constant-mass endpoint. Apply a matched blank correction only where that method authorizes it. These are calculation examples, not operating instructions or material acceptance criteria.
All masses are in grams. Let C be the conditioned crucible tare, S the positive initial sample mass, T the initial gross mass, and R the final gross mass. Let L be positive net sample loss, G positive net sample gain, and d the authorized blank mass change, defined as final minus initial blank mass. Thus T = C + S. The tare basis follows the same distinction used in tare-verified weighing.
Case A: evaluate the endpoint before reporting
Let the method’s permitted mass difference be τ grams, and let ε be a signed difference in grams. Take successive readings R_previous = C + S − L + ε and R_final = C + S − L. Their absolute difference is |R_previous − R_final| = |ε|. If the method permits a difference no greater than τ, compare |ε| ≤ τ; also check that its required weighing sequence was completed. A qualifying pair alone does not establish that the full endpoint procedure was followed.
With no blank drift, T − R_final = L, giving the signed mass fraction L / S. That algebra remains true even if the endpoint assessment fails; reportability is a separate decision.
Case B: positive crucible drift masks sample loss
Assume the matched blank represents the crucible’s positive mass drift d. With T = C + S and R = C + S − L + d, the uncorrected fraction is (T − R) / S = (L − d) / S. The authorized correction gives T − R + d = L, hence L / S. The denominator remains S—not the loaded crucible mass or final residue mass. This is a net-versus-gross reconciliation, not a change of reporting basis.
Case C: retain a negative loss
For net sample gain, T = C + S and R = C + S + G + d. Corrected mass loss is T − R + d = −G, so the signed fraction is −G / S. Retain the negative sign. As covered in the thermal interpretation above, it is neither an impossible measurement nor proof of a particular reaction.
| Case | Initial gross mass | Final gross mass | Blank change | Corrected loss | Signed result | Interpretation |
|---|---|---|---|---|---|---|
| A | C + S | C + S − L | None | L | L / S | Net loss; endpoint assessed separately |
| B | C + S | C + S − L + d | +d | L | L / S | Positive drift corrected where authorized |
| C | C + S | C + S + G + d | d | −G | −G / S | Net gain; retain negative loss |
Table results are dimensionless signed fractions, not percentages. They establish calculation logic, not material acceptance. An endpoint failure or compromised preparation remains a separate validity issue even when every subtraction and division is correct.
08The records needed to reconstruct an LOI result
A useful LOI record lets a reviewer reconstruct the result from identified weighings, not just read a percentage. The following illustrative laboratory data model is a practical traceability design, not a jurisdiction-wide legal requirement or a named software product’s feature set.
| Record layer | Linked fields |
|---|---|
| Sample and method | Material lot; test-portion identifier; method revision; preparation record; reporting basis. |
| Run and equipment | Crucible identifier; ignition-run identifier; associated blank where used; balance and furnace identifiers. |
| Recorded observations | Observation timestamps; mass units; tare and loaded-crucible readings; successive endpoint readings; blank readings; relevant run observations. |
| Calculated values | Sample mass; blank change; residue mass; signed LOI; calculation version; input references; rounding decisions. |
| Decisions | Endpoint assessment; deviations; result review; disposition; responsible people, timestamps and rationale for each. |
Keep recorded observations separate from derived values: a corrected residue mass should not overwrite the loaded-crucible reading. Preserve recorded precision and identify where rounding enters the reported result. This separation supports data integrity by design and makes recalculation review traceable to the inputs and calculation version used.
Avoid collapsing decisions into a single pass flag. Endpoint acceptance addresses the method’s stopping criterion; a deviation record addresses a departure; result review addresses analytical acceptability; disposition records the authorized material decision. Link these decisions without treating one as evidence that all others occurred. For repeated or superseded determinations, retain the original record and explain which result was selected and why.
Method Version
Defines analytical basis and rules.
- method_id
- version
- sample_basis
- endpoint_rule
- blank_rule
Test portion
Connects preparation to material provenance.
- sample_id
- lot_id
- aliquot_id
- preparation
- method_version
Ignition Run
Associates equipment, crucible and blank.
- run_id
- furnace_id
- balance_id
- crucible_id
- blank_id
Mass Observation
Preserves each original weighing event.
- observation_id
- object_id
- stage
- gross_mass_g
- measured_at
Reviewed Result
Separates calculation from approval.
- result_id
- endpoint_status
- blank_change_g
- signed_loi
- reviewer_id
09Move results between systems without losing their analytical meaning
As laboratory practice, define an interface contract linking balance observations, laboratory calculations and material-status records. Include sample and run identifiers, method revision, measurement stage, units, numerical sign, reporting basis and review status. Distinguish crucible-only, loaded-crucible and post-ignition observations so receiving systems do not treat unlike masses as interchangeable.
Keep the measurement timestamp separate from the import timestamp: arrival does not establish analytical sequence. Preserve the originating record and its identifier whenever a result is transferred or recalculated. This supports data integrity by design without making the receiving system the sole source of evidence.
Check for duplicate imports, missing identifiers, unit mismatches, truncated negative signs and records arriving out of sequence. Route ambiguous records to review rather than silently repairing them. Do not infer a missing sign from an expected positive LOI or overwrite an earlier observation merely because another record arrived later.
Recover by reconciling the transfer with the original observation, documenting the correction, recalculating affected results and obtaining review. Retain the prior record, correction reason and links between versions in the audit trail. Keep analytical approval and material disposition under human control; successful transmission alone is not an approval decision.
10Recognize procedural failures before interpreting a material difference
Treat these as investigation hypotheses, not automatic grounds for invalidation. Use run observations, weighing history, blank behaviour and preparation records to select follow-up work. Apply the negative-result interpretation and constant-mass criteria established above.
| Observation | Possible procedural cause | Discriminating evidence | Control or next action. |
|---|---|---|---|
| Abrupt mass decrease | Sample ejection or splashing | Run notes; deposits outside crucible | Investigate handling and heating sequence. |
| Endpoint not achieved | Incomplete or unstable treatment | Successive masses; treatment records | Apply method-defined endpoint rules. |
| Crucible mass changes | Vessel instability | Blank history; vessel condition | Review tare verification and blank applicability. |
| Mass rises after cooling | Moisture uptake | Exposure history; repeated weighings | Review cooling and transfer conditions. |
| Unexpected deposits or gain | Contamination during handling | Blank behaviour; handling observations | Investigate cross-contamination controls. |
| Reported result contradicts masses | Transcription or calculation error | Original readings versus entered values | Reconcile units, signs and denominator. |
| Preparations disagree | Nonrepresentative test portions | Sampling, mixing and subdivision records | Investigate heterogeneity before replacement preparation. |
Ejected particles represent physical loss, not chemical mass loss through heating. Both reduce measured mass; a blank contains no test portion and therefore cannot distinguish or repair sample ejection. Suspected ejection needs run-specific evidence, not a numerical adjustment.
An unusual result may reflect the material rather than laboratory error. Document the reason for each retest or replacement preparation, retain the original observations, and explain how follow-up work addresses the suspected cause rather than merely seeking a different result.
11A final checklist for method adoption and result review
Use these practice-based checks alongside the method and record controls above; they are not additional regulatory mandates.
Before implementation
- Define the analytical question and assess method suitability for the material and intended decision.
- Identify the method revision, material preparation, reporting basis, endpoint rule and blank role.
- Check calculation controls and agree the reporting format, including units and sign.
- When the method changes, assess comparability before pooling historical results or interpreting them as a single trend in trend review.
Before accepting a result
- Confirm complete observations, a valid endpoint, the correct sample denominator, only authorized corrections and retention of the calculated sign.
- Keep exceptions traceable through the deviation record; document the reviewer’s decision and rationale.
- Decide whether this individual test is valid separately from method suitability, then compare the valid result with the applicable specification.
Establish what was measured, establish whether the measurement is valid, then assess what the result means for the material.
Frequently asked questions
Q.How can as-received and dry-mass LOI results be compared without mixing reporting bases?+
First establish whether ignition began with as-received material or a separately dried portion. For sequential drying and ignition, distinguish the drying loss from the subsequent ignition loss; express each against its stated denominator. Merely replacing the denominator with dry mass leaves any moisture contribution in the numerator. Preserve the original masses and calculation sequence when comparing suppliers. Apply the same denominator discipline used for dry-weight reporting, and document any water or LOD compensation separately.
Q.Can a mineral LOI result be converted directly into carbonate content?+
Not without evidence identifying the carbonate phases and establishing which mass changes occurred. Heating may release water, organic decomposition products and carbon dioxide, while oxidation may increase residue mass. A carbonate calculation is defensible only when its chemical assumptions and reaction completeness are supported; mixed minerals complicate the conversion. Use mineralogical identification and suitable carbon or evolved-gas measurements to test those assumptions. An orthogonal identification strategy helps distinguish a carbonate interpretation from an unexplained net mass change.
Q.What follow-up evidence helps distinguish adhering soil from a botanical material's own ash?+
Examine representative unashed material macroscopically and microscopically for adhering particles, and compare findings with authenticated material of the same plant part. Total ash and acid-insoluble ash can contribute different information, but neither alone establishes the origin of every mineral constituent. Washing comparisons are investigative tools only when their effects on the botanical are considered. Pair botanical identity evidence with targeted elemental analysis where appropriate; soil findings warrant contamination assessment, not an automatic conclusion that botanical identity has failed.
Q.When is a loss-on-drying result useful alongside LOI?+
LOD is useful when variable initial moisture or other drying losses could obscure comparisons between ignition results. It can help separate a preliminary drying contribution from later mass change, provided the preparation sequence and reporting bases are explicit. LOD is not necessarily water-specific, and LOI is not simply a second moisture measurement. For separate test portions, assess sample homogeneity before combining results mathematically. A moisture assessment can guide whether a water-specific measurement would better resolve the question.
Q.How should a supplier result be handled when the certificate names LOI but omits the method?+
Treat the result as insufficiently defined for method-to-method comparison. Request the method identifier and revision, sample preparation, reporting basis, heating conditions, endpoint and calculation convention. For US dietary supplement components where supplier-certificate reliance is permitted, 21 CFR 111.75(a)(2)(ii)(B) requires the certificate to include a method description, test or examination limits, and actual results. Review the clarified certificate of analysis against the intended specification before deciding whether it supports acceptance or whether additional testing is appropriate.
Q.Should results from separate preparations be averaged when they disagree?+
An average can conceal a preparation or sampling problem. First compare the disagreement with the method's established precision and replicate-reporting rules. Examine sample segregation, crucible conditioning, blank correction, cooling exposure, transfer losses and the constant-mass sequence. Repeated weighings of one residue assess something different from independent preparations. Use out-of-specification handling when applicable, or document an analytical deviation where warranted. Retain all original results and justify any exclusion; a passing average does not explain discordant measurements.
Q.How can a laboratory maintain meaningful trends after changing its LOI method?+
Mark the method change explicitly rather than merging results into an uninterrupted trend. Where feasible, test representative retained materials using both procedures, covering the material types and result range relevant to routine decisions. Compare systematic shifts and variability, including effects from preparation, atmosphere, endpoint and reporting basis. Preserve method revisions with each result. A documented method-transfer comparison can support interpretation; out-of-trend review should distinguish a measurement change from a genuine change in incoming material.
Q.When is outsourcing LOI testing preferable to establishing the method in-house?+
Outsourcing is often preferable when demand is intermittent, the matrix presents unfamiliar heating hazards, or suitable furnace control, weighing capability and technical expertise are unavailable internally. Evaluate the laboratory's experience with the actual material and requested method, not merely its ability to report LOI. Agree on reporting basis, raw weighing records, blank treatment, endpoint evidence, uncertainty where relevant, and investigation responsibilities. A defined sample handoff helps protect representativeness and traceability during transport, storage and subsampling.
Primary sources
Want to see how Loss on Ignition (LOI) — Botanicals & Minerals could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
