Clarity and Colour: Two Visual Tests Worth Defending

Clarity and colour of solution are the two lines on a certificate that look like an afterthought. They are visual, they are comparative, they produce a grade rather than a number, and they are the first thing a buyer skips when reading quickly.

They are also the two tests most likely to catch a problem that the chromatogram cannot see: insoluble particulate from a filter that failed, a trace of rust from a reactor, a colour body formed by oxidation during drying. Neither would appear in an assay or an impurity profile.

How the comparison works

Clarity is assessed against a series of reference suspensions, numbered from clear through increasing opalescence. The sample is compared in a tube of specified dimensions, against a black background, under defined lighting. The result is stated as clear, or as not more opalescent than a named reference.

Colour is assessed the same way against a numbered series of colour reference solutions in matching tubes, viewed against a white background. The output is a bracketed grade such as "not more intensely coloured than reference solution Y5".

Both are ordinal. Neither produces a unit. That is not a weakness in a specification context, because the question being asked is whether the material sits inside a bracket, not what its exact optical property is.

Why a comparison beats a number here

A turbidity meter gives a reproducible number and is useful for trend analysis. It does not replace the compendial test, because the reference series is part of how the method defines its own answer. Two laboratories with different meters can produce different numbers for the same sample; two laboratories with the same reference series and the same tubes usually agree.

The comparison also carries an implicit acceptance criterion that a number would have to be translated into. "Not more opalescent than reference suspension II" is a decision. "12 NTU" requires somebody to decide what 12 means, and that decision is not in the monograph.

What the two tests can and cannot see

Finding	Clarity	Colour	Assay or impurities
Insoluble particulate	Yes	Sometimes	No
Trace metal colour body	No	Yes	Usually no
Oxidation products with chromophores	No	Yes	Partly, if the method resolves them
Late-eluting organic impurity	No	No	Yes
Residual moisture	Sometimes, through haze	No	No, unless water is measured

The table is the argument for keeping both tests on a routine basis. They cover a region of the impurity space that chromatography does not enter, and they cover it at a cost of a few minutes and two tubes.

The colour test is not selective, and that is the point

A darkening of one grade can come from oxidation, from a metal catalysed reaction, from a change in particle size distribution, or from residual moisture changing the way light scatters in the solution. The test does not distinguish these, and no honest report should pretend it does.

What the test does is trigger an investigation earlier than the assay would. A batch that reads one grade darker than the previous twenty is telling you something changed. The chromatogram may be unremarkable, the assay may be 99.7 per cent, and the material may still be different from what the customer received last quarter.

This is the same logic that makes batch history more informative than a single certificate. The visual result becomes useful when it is a series rather than a value.

Where the variability really comes from

Not from the sample. From the viewing conditions. Three sources account for most disagreements between sites: the light source, the background, and the observer's technique.

The methods specify a light source with defined characteristics because colour grading under warm or cool artificial light shifts the comparison by a grade in either direction. The background matters because opalescence is far easier to see against black than against white, and colour the reverse. The technique matters because a tube held at an angle shows a different path length than one held vertical.

All three are controllable at no cost. The reason they are neglected is that the test is regarded as trivial, and trivial tests do not get written procedures.

Reading both tests on a certificate

The line usually reads something like "clear, colourless to not more than Y5". The first part is the clarity result, the second the colour bracket. What a buyer should notice is the width of the bracket and whether it matches the previous batches.

A bracket that has widened over time is a quiet signal. It means the process is producing a range of material that the specification still accepts, and the specification is doing more work than it was designed to do. The COA reading guide treats this alongside the numeric lines, because a visual field with a wide bracket tells you as much as a numeric field with a narrow range.

Instrumental support without substitution

A spectrophotometric measurement of absorbance at defined wavelengths, or a turbidimetric reading, is useful in two situations: investigating a deviation, and building a trend that a grade cannot show. Neither replaces the compendial test, and a supplier who reports only an instrument reading where the monograph asks for a comparison has substituted a different method.

The honest arrangement is both, with the compendial result on the certificate and the instrument reading in the investigation file.

Where this leaves a supply-chain partner

We report the visual result as the monograph frames it, with the reference named rather than a vague "passes". Where a customer's specification tightens the bracket, we say what the batch history supports before agreeing to it, because a bracket accepted in a specification becomes an obligation on every future batch.

If a certificate in front of you says only "white to off-white powder" with no reference and no solution test, ask which test was actually run. The identification tests and the visual tests are often confused in translation, and the difference matters when the drum arrives looking slightly different from the sample that was approved.

The viewing station is part of the method

A comparison is only as good as the conditions under which it is made, and the conditions are specified because they change the answer. Three elements carry most of the weight: the light source, the background and the tube.

The light source has to approximate daylight with defined characteristics. Under warm office lighting, a slightly yellow sample reads yellower and a slightly opalescent one reads clearer, because the eye's sensitivity shifts with the spectrum. The background differs between the two tests, black for opalescence and white for colour, because scattered light is seen against darkness and transmitted colour against white. The tube must be matched in diameter and glass, since the path length through the solution determines both the intensity of the colour and the visibility of the haze.

None of this is expensive. A designated viewing cabinet, a matched tube set and a written instruction cost less than a single repeat analysis, and they remove the most common source of disagreement between two sites.

When a grade moves

A colour result that shifts by one grade is a question, not a conclusion. The useful response is to look at four causes in order of likelihood: moisture, oxidation, trace metal and particle size.

Moisture changes the way the powder dissolves and can produce a faint haze that reads as colour. Oxidation produces chromophores directly and usually moves the grade in one direction over time. Trace metals, particularly iron, catalyse oxidation and also form coloured complexes of their own. Particle size changes the rate of dissolution, which affects the reading without changing the substance.

Each cause has a different fix and a different implication for the customer, which is why assigning one before investigating is worse than saying the cause is under investigation. A batch that darkened because of moisture is a drying problem. One that darkened because of iron is a materials-of-construction problem, and it may affect every batch from that reactor.

The pharmacopoeial definitions of both tests sit in the general chapters published by the EDQM and, in their own sphere, by the USP; the British Pharmacopoeia carries equivalent text. The ICH quality guidelines do not define the tests but set the stability context in which a change of appearance becomes a reportable observation.

For a buyer, the practical question is narrow: is the bracket on this certificate the same width as it was three batches ago? A widening bracket is a process drifting inside its own specification, and it is visible on the certificate before it is visible anywhere else.

Trending an ordinal result without pretending it is a number

A grade is not a concentration, and the mistake to avoid is plotting it on a control chart as though it were one. The distance between reference solution Y3 and Y4 is not the same as the distance between Y4 and Y5, and the spacing is unknown. What can be trended legitimately is the frequency distribution: how many batches fell at each grade over the last twelve months, and whether the distribution has shifted.

A distribution that has moved from "mostly clear" to "mostly Y3" has told you something real, and it told you without any instrument. The response is the four-cause check from the section above, applied to the period during which the shift occurred rather than to a single batch.

Clarity trends the same way. A material that has been "clear" for two years and then reads "slightly opalescent" on three consecutive batches is describing a change in the process or in the dissolution, and the fact that all three batches remain within specification does not make the observation less useful. It makes it more useful, because it was caught by a test nobody watches.

The effort to maintain the trend is a spreadsheet with a column per batch. The discipline is to record the grade as observed, including the ones that differ from expectation, and to resist the temptation to re-read a tube until it agrees with the previous batch.

Frequently asked questions

How is clarity actually assessed?

By comparing the sample with a series of reference suspensions under defined lighting against a black background. The result is expressed as clear, or as not more opalescent than a named reference. It is a bracketed comparison rather than a turbidity value, which is why two laboratories can agree without sharing an instrument.

Is the colour test quantitative?

It is comparative against a numbered series of reference solutions, so the output is an ordinal grade rather than a number in a unit. The grade is reproducible when the lighting, the viewing tube and the observer's method are controlled, which the method specifies.

Why not replace both with an instrument?

Instruments exist for turbidity and colour and are useful for trending. They do not replace the compendial comparison, because the reference series is part of the method's definition. A spectrophotometric reading can support a deviation investigation; it does not substitute for the test the monograph describes.

What does a colour change indicate?

Often oxidation or trace metal catalysed degradation, and sometimes nothing more than a change in particle size or residual moisture. Because the test is not selective, the correct response is to look for a cause rather than to assign one. A batch that darkens by one grade has told you to investigate, not what to conclude.

Does light quality affect the reading?

Substantially. The methods specify daylight or an equivalent source with defined characteristics, viewed against a white or black background as appropriate. Reading a colour series under warm office lighting is one of the most common sources of disagreement between two sites.

Can a clear solution still fail another test?

Easily. Clarity says nothing about dissolved impurities, and colour says nothing about potency. Both tests are about what can be seen, which is a narrow window onto a product, and their value is that the window is cheap to open on every batch.