Assay Above 100%: Why It Happens and When It Is Fine

An assay of 100.4 per cent looks like a data-entry error. A percentage above 100 suggests a measurement that has gone wrong by definition, since a pure substance cannot be more than entirely pure. The instinct to query it is correct. The conclusion that something is broken usually is not.

Content is a ratio, and a ratio is only as meaningful as its denominator. Once the denominator is identified, a figure above 100 per cent stops being puzzling and starts being informative, because the amount by which it exceeds 100 is itself a measurement of something.

The basis is the first thing to find

Content on a bile-acid certificate is normally reported on the dried basis. The weighed sample is not dry; it contains some water and some residual solvent, both of which contribute mass but not content. Reporting on the dried basis removes that volatile fraction from the denominator, so the figure describes the dry substance rather than the substance as it happened to be weighed.

The size of the correction is directly related to the loss on drying. At a loss of 0.4 per cent, the dried-basis correction lifts the reported content by roughly 0.4 percentage points relative to a calculation performed as-is. That is more than enough to carry a result from just below the 100 mark to just above it, without anything having gone wrong. This is the single most common explanation for an assay above 100 per cent, and the one that requires the least alarm.

It follows that the assay line and the loss-on-drying line have to be read together. Loss on drying versus water content sets out why those two measurements are not interchangeable; the point here is narrower. When the assay exceeds 100 per cent, the first place to look is the volatile figure on the same page.

Four sources of a genuine excess

The basis correction explains the ordinary case. Four other mechanisms explain the rest, and they are worth separating because they carry different implications.

The first is the reference standard. Assay is measured against a standard whose own purity is taken as a given. If the assigned purity of the standard is slightly optimistic, every result calculated from it is shifted upward by the same small amount. The effect is systematic rather than random, which is why it shows up as a consistent offset across batches rather than as scatter.

The second is method selectivity. A content method that separates the analyte from related substances measures only the main component. A method that does not separate measures everything that responds and attributes the whole response to the main component, which inflates the figure. Where the impurity panel is measured by a different technique from the content determination, the two can disagree in a way that is entirely explained by selectivity. Stability-indicating assay methods takes up the separation requirement in detail.

The third is the volumetric route. Where content is determined by titration, the titration result and the chromatographic result are measurements of the same quantity by different means, and a bias between them is expected rather than exceptional. A volumetric method responds to the whole acidic function; a chromatographic method responds only to what elutes as the main peak. Small amounts of material that are chemically similar but chromatographically resolved will be counted by one method and excluded by the other.

The fourth is residual moisture that the drying step did not remove uniformly. Where a sample is not homogeneous, the portion weighed for content and the portion weighed for loss on drying can differ in their volatile content, and the two corrections then do not cancel.

A worked illustration

Consider an illustrative case. A batch reports a content of 100.4 per cent on the dried basis and a loss on drying of 0.4 per cent. Read as-is, the volatile correction alone accounts for about 0.4 percentage points, which places the uncorrected figure at approximately 100.0 per cent. The excess is fully explained by the stated basis, and the result needs no further explanation.

Now change one variable. Suppose the same 100.4 per cent is reported alongside a loss on drying of 0.05 per cent. The basis correction now accounts for almost nothing, and the excess is unexplained by the obvious mechanism. That is the case that deserves a question, and the question is not whether the number is wrong but what the method and the standard were.

Pattern	Most likely explanation	Action
Excess roughly equals loss on drying	Dried-basis correction	No action beyond recording
Consistent small offset across batches	Reference standard assignment	Ask for the standard's stated purity
Moderate excess with a flat impurity panel	Method selectivity, non-separating content method	Ask which method produced the figure
Excess that grows batch by batch	Standard degrading, or process drift	Investigate as a trend, not as a point
Excess inconsistent with the volatile figure	Sampling or homogeneity issue	Ask for the calculation sheet

What an assay above 100% can hide

One mechanism in the list above is the only one that should raise a quality question rather than a bookkeeping one. A content method that does not separate related substances will report a high figure precisely because it is counting impurities as though they were the main component. In that case the assay line looks reassuring while the impurity panel carries the real information, and a reviewer who reads only the assay will reach the wrong conclusion.

The defence is straightforward. The content result should be produced by a method whose selectivity has been demonstrated, and the certificate should name that method. Where the method is not named, the content figure is an assertion without a stated basis, and the excess over 100 per cent cannot be interpreted at all. That case is covered in more detail in our note on certificates that omit the method.

How to read the line, and when to ask

Read the content figure together with three neighbours: the basis stated next to it, the loss on drying, and the impurity total. If the stated basis is the dried basis and the excess over 100 per cent is broadly the size of the volatile fraction, the arithmetic is closed and no question is needed. If the excess is materially larger than the volatile correction, the next question is which method produced the number.

Two situations justify sending a query even when the result is inside the specification. The first is a value sitting at the upper limit, because the margin between the result and the limit is what matters for the next batch rather than for this one. The second is a series of results drifting upward with no change in any other line, because a drift in content calculation is usually a symptom of something happening to the standard or to the method rather than to the material.

The specification range we release against is 99.0 to 101.0 per cent calculated on the dried basis, which is the range that appears on EP versus USP limits comparisons in this trade. A result above 100 per cent is inside that interval by construction. The work is not to explain why the number exists but to confirm that the explanation given is the right one, and the arithmetic of the dried basis closes it in most cases.

What the arithmetic looks like when it is written out

The dried-basis correction is easy to describe and easier to misapply, so it is worth writing out once. Suppose a weighed portion reports a loss on drying of 0.4 per cent and a content of 99.7 per cent calculated as-is. The dry matter in that portion is 99.6 parts of every 100 weighed, and the content of the dry matter is the measured content divided by the dry fraction, which comes to approximately 100.1 per cent. The same material read on the dried basis is inside a 99.0 to 101.0 per cent window with room to spare; read as-is it would look marginally low.

Two things follow from that example. The first is that the basis stated on a certificate is not a presentational detail; it changes whether a result looks adequate. The second is that the loss-on-drying figure and the content figure have to travel together, because a content figure alone cannot be interpreted without knowing how large the correction was.

The same arithmetic explains the opposite case. Where the volatile figure is small, the correction is small, and an excess over 100 per cent then requires a different explanation. The size of the correction is therefore itself a diagnostic, which is why the pair of figures is more useful than either one.

Methods and limits as they are described in the general chapters

Content determination is governed by general chapters rather than by a single prescribed procedure. The United States Pharmacopeia and the European Pharmacopoeia both set out the requirements for the chromatographic and volumetric techniques that are used for content, and both allow more than one technique for the same material where the monograph says so. The ICH quality guidelines set out the validation expectations that apply to whichever technique is chosen, including the requirements around range and linearity that decide whether a method is fit for a result close to a limit.

That layering explains a pattern that puzzles buyers. Two certificates for the same material can both be correct and can report content figures that differ by a few tenths of a percentage point, because each was produced under a different permitted technique with its own validation range. The difference is not an error; it is the width of the method landscape that the general chapters permit.

How the figure behaves over a stability programme

Content on the dried basis is one of the few numbers whose behaviour over time is genuinely informative, because it separates chemical change from moisture change. A batch whose content falls while its loss on drying also falls has lost dry substance. A batch whose content rises while its loss on drying falls has gained nothing; the material has simply dried out further, and the dried-basis calculation has removed the effect.

Reading the two together is what makes a stability programme interpretable. A content figure alone, drifting slightly over twelve months, is ambiguous between the two explanations, and the ambiguity is resolved by the volatile line rather than by repeating the content determination.

For buyers who hold material beyond its retest period, the practical point is that the content determination and the volatile determination should be specified as a pair in the stability protocol. Separating them saves a little analytical effort and costs the ability to explain a drift.

Why the basis is stated on the certificate and not assumed

The dried basis is a convention, not a universal standard. Some materials in this family are certified on the anhydrous basis, some as-is, and the choice follows the material and the monograph rather than a general rule. Where a certificate states the basis, the reader can convert; where it does not, the reader has to assume, and assumptions about this particular detail are wrong often enough to matter.

Two consequences follow for anyone comparing documents. The first is that a comparison of content figures across two suppliers is a comparison of two calculations before it is a comparison of two materials. The second is that a specification written without the basis is incomplete in a way that only becomes visible when a borderline batch appears. Adding the basis costs one line on the document and removes an entire class of dispute.

Where the material is held for a long period, the basis also determines how the content figure behaves. A sample that gains moisture in storage will show a falling content on the as-is basis and a stable content on the dried basis. Both are true statements about the same drum, which is why the basis has to travel with the number.

Frequently asked questions

Is an assay above 100% an out-of-specification result?

Not by itself. The specification range for this material is 99.0 to 101.0 per cent calculated on the dried basis, so a result above 100 per cent is inside the specification. Whether it needs investigation depends on how far above, whether it is consistent with previous batches, and whether the method and standard explain it.

What does calculated on the dried basis change?

It removes the volatile fraction from the denominator. The weighed sample contains some water and residual solvent; correcting for the loss on drying lets the result describe the dry substance rather than the substance as weighed. Without that correction the same material would read lower on a damp day.

How much can the correction move a result?

By roughly the size of the loss on drying figure. At a loss of 0.4 per cent, the dried-basis correction adds about 0.4 percentage points relative to an as-is calculation. That is enough to move a result from just below 100 to just above it.

Does an assay above 100% hide an impurity?

It can, if the method does not separate the analyte from related substances. A non-separating method measures the whole response and reports it as the main component, which inflates the figure. That is one reason the method behind the number matters as much as the number.

Should I ask for the raw data when the assay exceeds 100%?

Ask when the value is near the upper limit, when it has drifted upward across batches, or when the certificate does not state the method. In those cases the calculation sheet, the standard's stated purity and the chromatographic or titration parameters answer the question quickly.

Why not simply cap the result at 100%?

Because capping hides information. A result of 100.4 per cent tells a reviewer something about the measurement system; a number truncated to 100.0 per cent tells them nothing and makes the trend impossible to follow.