A supplier’s certificate of analysis reports “Not Detected” for a trace metal. Another laboratory reports a measurable concentration of the same impurity. Before rejecting the shipment or questioning either laboratory, the buyer needs to establish one thing: how low a concentration each method could reliably detect and measure.

“Not Detected,” often abbreviated ND, does not mean a chemical is completely free of the substance being tested. It means the laboratory did not establish its presence under the detection criteria used.

This distinction matters when manufacturers compare chemical grades, approve suppliers or assess materials against impurity specifications.

A certificate reporting “ND” without a detection limit, reporting limit or test method may offer less useful quality evidence than one reporting a small numerical concentration with a properly validated method.

For industrial procurement and quality assurance teams, understanding chemical CoA detection limits can prevent unnecessary supplier disputes and, more importantly, the acceptance of materials that have not been tested sensitively enough.

Why “Not Detected” Needs a Number Beside It

Every analytical method has limitations.

A laboratory may use spectroscopy, chromatography, titration or another technique to identify and measure substances within a chemical sample.

At sufficiently low concentrations, the signal produced by the substance becomes difficult to distinguish from background noise and other measurement variation.

That is where detection and quantitation limits become important.

The U.S. Environmental Protection Agency’s August 2025 Detection Limits Best Practices Guide distinguishes three important concepts:

Detection limit (LOD or DL): The lowest concentration at which a substance can be reliably detected under the defined method.

Quantitation limit (LOQ or QL): The lowest concentration at which the substance can be measured quantitatively with acceptable performance.

Reporting limit (RL): A laboratory’s defined lower threshold for reporting a result quantitatively under its reporting system.

These limits serve different purposes.

A laboratory may detect evidence of a substance at a concentration too low to measure it with the accuracy and precision required for routine quantitative reporting.

The result may then be described as detected below the quantitation limit rather than as a fully quantified concentration.

A laboratory’s reporting limit can also depend on its reporting policy or analytical requirements. It should not automatically be assumed identical to the detection or quantitation limit.

EPA’s guidance was developed for water-quality data reporting, not as a universal industrial chemical purchasing regulation. However, its distinction between these measurements provides a useful framework for interpreting laboratory results.

Two Suppliers Can Report “ND” With Very Different Evidence

Consider a hypothetical purchasing specification for an industrial chemical.

The specification establishes a maximum concentration of 0.50 milligrams per kilogram (mg/kg) for a particular trace-metal impurity.

Two suppliers submit certificates of analysis.

Supplier A reports:

Trace metal: Not Detected — reporting limit 0.80 mg/kg

Supplier B reports:

Trace metal: Below 0.20 mg/kg — reporting limit 0.20 mg/kg

Both results may appear reassuring.

However, their implications are different.

Supplier A’s laboratory cannot provide a reliable quantitative result below its stated reporting limit of 0.80 mg/kg.

That threshold is higher than the purchasing limit of 0.50 mg/kg.

The reported result therefore cannot independently establish that the material satisfies the specification.

The impurity could be present at a concentration above the purchasing limit while still falling below the laboratory’s quantitative reporting threshold.

Supplier B’s result provides more useful evidence.

Its reporting threshold is below the purchasing limit.

Assuming the method is suitable, the sample is representative and applicable quality controls are satisfactory, the result supports a conclusion that the tested material meets the specified maximum.

This remains a hypothetical comparison. The numbers are illustrative, not actual supplier results or universal impurity limits.

The important difference is not which supplier used more reassuring language.

It is whether the analytical method provides sufficient information to evaluate the applicable specification.

A Reporting Limit Is Not a Guarantee of Compliance

A result below an appropriate reporting limit can support material acceptance, but it does not eliminate every uncertainty.

Sampling errors, contamination, incomplete extraction, interference and unsuitable calibration can affect the reliability of laboratory findings.

The purchasing specification must also establish which measurement basis applies.

For example, a limit expressed in mg/kg of material as received is not automatically equivalent to a result expressed on a dry-mass basis.

Quality teams should therefore verify the analytical method and sample basis before using an ND result as evidence of conformity.

Instrument Detection Limits and Method Detection Limits Are Different

Another common problem occurs when a supplier provides the detection limit of an instrument rather than the sensitivity of the complete analytical method.

An instrument may be capable of detecting extremely small concentrations in a clean calibration solution.

However, a real industrial chemical may require sample preparation before measurement.

That preparation can involve dilution, extraction, digestion or separation, depending on the material and validated procedure.

The final method must account for those operations and their potential effects.

A low instrument detection limit does not prove that the same concentration can be reliably detected in the original product.

The U.S. EPA’s method detection limit procedure in 40 CFR Part 136, Appendix B illustrates this distinction.

The procedure defines a method detection limit using statistical confidence that a measured concentration can be distinguished from method blanks.

Crucially, it requires the determination to account for the complete analytical method, including relevant sample-processing steps.

The procedure applies to specified environmental analytical methods. It is not automatically mandatory for all industrial chemical testing.

Nevertheless, it demonstrates why buyers should ask whether a laboratory’s quoted sensitivity includes the full testing procedure or merely the instrument’s measurement capability.

Sample Dilution Can Make a Sensitive Method Less Useful

Suppose a laboratory normally reports a trace impurity down to 0.10 mg/kg.

A difficult sample requires additional dilution to avoid interference or bring the analyte into the instrument’s suitable operating range.

The effective sample reporting limit may increase.

That increase does not necessarily mean the instrument has become less capable.

Instead, the laboratory’s ability to report low concentrations in the original material has changed because of how the sample was prepared.

A buyer comparing results across suppliers should therefore request sample-specific reporting limits when the method or sample preparation makes them relevant.

A generic instrument sensitivity claim is not enough.

Method Validation Determines Whether the Number Is Trustworthy

Detection and quantitation limits should be supported by evidence that the analytical procedure performs as intended.

That requires more than observing a small signal on an instrument display.

The laboratory must establish suitable performance characteristics for the method’s intended use.

The International Council for Harmonisation’s ICH Q2(R2) analytical procedure validation guideline describes approaches for estimating detection and quantitation limits.

The guideline was adopted in November 2023, with subsequent document corrections.

It includes approaches based on visual evaluation, signal-to-noise measurements, calibration relationships and experimental evaluation of performance near the lower measurement range.

For certain methods, a signal-to-noise ratio of approximately 3:1 can support a detection-limit estimate, while a higher ratio of at least 10:1 may support a quantitation-limit estimate.

These are method-validation conventions, not universal numerical guarantees.

A laboratory still needs to establish that the selected procedure performs adequately for the material and intended measurement.

ICH Q2(R2) specifically addresses analytical procedures within the pharmaceutical regulatory context. Its technical principles can inform other laboratory discussions, but its requirements should not be presented as universally binding on industrial chemical suppliers.

The Chemical Matrix Can Change the Result

A method that performs well with a simple reference solution may behave differently when applied to a complex industrial material.

The term matrix describes the overall material surrounding the substance being measured.

In chemical quality control, the matrix might be a concentrated acid, a mineral powder, an industrial solvent, a salt solution or a mixture containing several components.

These substances can affect sample preparation and analytical performance.

For example, a trace-metal method validated for relatively clean water may require additional evaluation before it is considered suitable for a concentrated industrial salt.

The same chemical element is being measured, but the surrounding material can influence the result.

The 2025 third edition of Eurachem’s analytical method validation guide provides a general framework for assessing whether an analytical method is fit for its intended purpose.

Its scope includes method validation, verification, sampling, performance characteristics and supporting documentation.

The practical lesson is that an analytical method should be judged against the actual measurement task.

A laboratory’s ability to detect an impurity in one sample type does not automatically demonstrate equivalent performance in a different material.

Why This Matters for Ferrous Sulfate

Ferrous sulfate heptahydrate, FeSO₄·7H₂O (CAS 7782-63-0), provides a useful example.

Industrial buyers may evaluate its iron content, impurities, insoluble matter, acidity and physical condition.

The existing guide to ferrous sulfate heptahydrate specifications explains why these attributes need application-specific acceptance requirements.

A supplier reporting an impurity as ND should identify the analytical method and relevant limit.

It would be inappropriate to assume that an ND result automatically means a higher-purity product than another supplier’s material containing a small, reliably quantified impurity concentration.

The comparison must establish whether the laboratories could measure the same substance at sufficiently low concentrations using suitable methods.

What Buyers Should Request When a CoA Reports ND

A complete certificate of analysis should connect the result to the substance being tested, the production lot and the applicable specification.

For an impurity reported as ND, quality teams should establish the following information.

Exact analyte identity: Which element, compound or chemical species was tested?

Analytical method: What standard or validated laboratory procedure was used, including the relevant edition?

Detection and reporting limits: What do ND, below detection limit and below quantitation limit mean under that laboratory’s reporting convention?

Reporting units: Is the result expressed in mg/kg, mg/L, percentage by mass or another unit? What sample basis applies?

Sample preparation: Did dilution, extraction, digestion or another step affect the final reporting limit?

Method suitability: Has the laboratory established that the procedure works for the relevant chemical matrix?

Specification relationship: Is the reporting limit low enough to evaluate the agreed purchasing requirement?

These details become particularly important when results are close to a contractual acceptance limit.

The existing article on supplier CoA verification explains how reported supplier results can be supported by risk-based independent confirmation.

For critical attributes, purchasing teams should also establish how discrepancies will be resolved before the material arrives.

That may involve an agreed reference method, qualified independent laboratory or documented review of the differing analytical procedures.

The correct response depends on the product and the consequence of accepting an unsuitable lot.

Do Not Replace Every Non-Detect Result With Zero

A further problem emerges when companies transfer CoA results into spreadsheets, supplier scorecards or automated quality systems.

An ND result may be entered as zero because the software expects a numerical value.

That shortcut can distort the data.

A non-detect result contains information about the measurement threshold. It does not establish a precisely measured concentration of zero.

Treating such results as exact zeros can artificially reduce calculated averages and obscure differences between supplier testing capabilities.

Replacing each result with half the detection limit is another sometimes-used shortcut, but it is not universally appropriate.

The suitability of any statistical treatment depends on the data, proportion of censored observations and analysis objective.

For procurement records, a better approach is to preserve the reported qualifier, limit type, numerical threshold and units.

For example, a database can retain “below reporting limit” as a distinct result condition rather than silently converting it into a measured concentration.

This also makes historical supplier comparisons more defensible.

If a laboratory improves its detection capability, the apparent frequency of reported impurities may increase even though the material’s true composition has not changed.

Without the associated limits, a quality dashboard might incorrectly interpret that change as deteriorating supplier performance.

A Better Specification Begins With the Required Measurement

Chemical purchasing specifications should be written around the manufacturing requirement rather than around whichever test a supplier happens to offer.

If a trace impurity can affect finished-product quality, catalyst performance, environmental compliance or another critical outcome, the buyer should establish a technically justified acceptance limit.

The required analytical procedure must then be capable of evaluating that limit.

A specification allowing a maximum of 0.50 mg/kg is incomplete if suppliers can demonstrate compliance using methods that cannot reliably assess concentrations below 0.80 mg/kg.

Likewise, a requirement that every impurity must be “not detected” is ambiguous unless the relevant detection criteria are defined.

An effective specification identifies the measured substance, maximum permitted concentration, method requirements, units and reporting conventions.

Where different laboratories are permitted, the buyer should establish how methodological differences will be assessed.

The same document-control principle applies to other supplier records.

A certificate of analysis reports selected lot-specific analytical findings. A product specification defines the agreed acceptance requirements. A safety data sheet communicates chemical hazard and safe-use information, not proof of trace-impurity conformity.

An independent laboratory test can provide additional evidence for the tested sample and analytical procedure, but it does not automatically establish the quality of every other supplier lot.

For industrial chemical buyers, the practical rule is simple: never approve a material solely because its CoA says “Not Detected.”

Approve it when the evidence demonstrates that the material meets the actual specification—and that the analytical method was capable of making that determination.