Appearance Defects: Judging Colour Against a Reference

Appearance is the first line of the certificate and the first thing a receiving clerk checks. It is also the test with the least method behind it. Two sites can both report "white to off-white powder" about the same material and mean different things, and neither is lying.

The gap matters because appearance is where disputes start. A customer opens a drum, sees a slight cream tone, compares it with the previous batch, and sends a photograph. The supplier replies that the material meets specification. Both statements are true and the conversation has nowhere to go, because the specification was a phrase.

What the three components actually are

Appearance covers three separate observations that are often collapsed into one word.

Form. Powder, crystalline solid, granules, lumps. This is a physical description, and it is the easiest to standardize because the categories are discrete.

Colour. A continuous property described in words, which is where the trouble lives. "White" and "off-white" are separated by a boundary that exists in a standard, not in nature.

Foreign matter. Visible contamination: fibres, dark specks, packaging fragments. This is a presence test rather than a scale, and a single speck is a non-conformance regardless of how the rest of the drum looks.

Methods, and what each one can discriminate

Method	Discriminates	Does not discriminate
Descriptive, by eye	Gross differences, foreign matter	Subtle batch-to-batch shift
Against a colour standard series	Position on a defined scale	Hue differences outside the series
Against a retained reference sample	Change relative to a known batch	Whether the reference itself has aged
Instrumental colour coordinates	Small shifts, trendable over time	Whether the shift matters to the customer

The retained reference sample is the most practical tool and the most neglected. A drum of material from a batch the customer accepted, stored sealed and dark, gives a comparison that is meaningful to both parties in a way that a printed colour chart is not. Its weakness is that it ages, so it needs replacing on a defined interval and requalifying when it does.

Light, observer, background

Three variables decide a visual colour judgement and none of them are usually recorded.

Light. North daylight is the traditional reference because it is broadband and relatively cool. Office fluorescent lighting is not: it is deficient at the red end, which shifts a cream tone toward green and makes two different materials look the same. A daylight-matching lamp with a high colour rendering index removes most of this variable at modest cost.

Observer. Colour discrimination varies between individuals and declines with age, particularly in the blue region. A site where one person makes all appearance judgements has a method with a single instrument and no calibration.

Background and sample depth. The same powder in a thin layer on white paper and in a deep bed in a glass bottle reads differently, because the light path through the material differs. Depth should be specified, or the comparison is between two different experiments.

Yellowing, and why it deserves a test rather than a remark

In bile acid powders, a shift toward yellow is usually a chemical signal. Oxidation of the steroid nucleus, a change in residual solvent, or exposure to light during storage all produce it, and all of them are more informative than the colour itself.

A batch that has yellowed in storage should be tested rather than re-described. If the related substances have moved, the appearance change was a symptom and the batch has a stability question. If they have not, the change may be physical, and the answer is still worth having because it tells you something about the packaging.

This is where appearance connects to the photostability arm and to packaging performance. A material that yellows in a clear bottle and not in an opaque one has told you which pack to use, and it has told you for free.

When the answer is "it meets specification"

That answer is only useful if the specification had a method. A workable appearance specification has four parts: the form, the colour described against a named reference, the acceptance rule for foreign matter, and the viewing conditions. The fourth is what makes the first three reproducible.

Where a customer's own incoming inspection uses different conditions, a disagreement can be predicted in advance and prevented by agreeing the conditions rather than the words. This is the same discipline as the clarity and colour of solution test, where the pharmacopoeial method specifies the reference suspensions and the viewing geometry precisely because otherwise the test is not a test.

How a colour dispute should be handled

Three steps, in order. Compare both parties' reference materials, because a disagreement about colour is frequently a disagreement about the reference. Repeat the observation under controlled lighting with a third party present. And if the difference is real, test the parameters that could explain it rather than continuing to discuss the shade.

The last step resolves most cases, because yellowing almost always has a chemical correlate and the correlate is a number both parties can agree on.

What we put in the record

We describe appearance against a retained reference of the material, under a daylight-matching lamp, with the sample depth stated, and we name the observer. We keep the reference on a replacement schedule. Where a customer reports a difference, we ask for their reference and their lighting before we discuss the colour.

The compendial methods for colour and clarity are maintained by EDQM and in the USP general chapters, and the colorimetric conventions in the ISO catalogue.

The question worth asking a supplier is short: what do you compare it against, under what light, and who decides?

Writing an appearance specification that can be enforced

An enforceable appearance specification has four parts: form, colour against a named reference, the acceptance rule for foreign matter, and the viewing conditions. Without the fourth, the first three are opinions with a document number.

The viewing conditions are cheap to write and rarely written. Lamp type and colour rendering, the distance from the eye, the background, the sample depth, and whether the sample is in its original pack or spread on a plate. Each of them changes what the observer sees, and none of them appears in a typical specification.

The reference is the harder part. A pharmacopoeial colour series gives a scale, but it is designed for solutions and is awkward for powders. A retained sample of the material is more practical and requires a lifecycle: a defined replacement interval, storage in the dark, and a requalification when the reference is renewed. A site that keeps a reference drum for eight years has a reference that describes a material it no longer makes.

Photographs as evidence, and why they fail

A customer who reports a colour concern almost always sends a photograph. The photograph is useful as an indication and useless as evidence, because the camera applies a white balance, the phone applies a tone curve, and the screen applies a profile. Two photographs of the same powder taken in two rooms can differ more than the two batches do.

If photographs are to be used at all, they should be taken under a stated lamp, with a white reference card in frame, at a stated distance, and with the camera's automatic adjustments disabled. That is a procedure, and a procedure is what turns a picture into an observation.

The commercial cost of an unenforceable specification

A phrase like "white to off-white powder" costs nothing to write and a great deal to apply. Every shipment is arguable. Every argument is settled by relationship rather than by measurement, which means the party with more to lose concedes. Over a few years, that produces a supplier who accepts returns for appearance and a customer who expects them.

The alternative is not more expensive. It is a reference sample, a lamp, a written condition and a named observer, and it converts a recurring conversation into a check that takes two minutes. The compendial methods for colour and clarity are maintained by EDQM and in the USP general chapters, and the colorimetric conventions in the ISO catalogue give the instrumental route if a site wants coordinates rather than words.

What a receiving inspection can realistically do

A warehouse cannot run a colorimetric method, and it should not try. What it can do is consistent: the same lamp, the same reference, the same depth, the same person where possible, and a record of the outcome in the terms of the specification. That is enough to detect a change, which is the purpose of the check.

The useful discipline is to keep the reference near the inspection point rather than in a quality office, because a reference that requires a walk to another building will not be used. The same logic applies to the lamp: a daylight-matching fitting over the inspection bench costs less than one rejected shipment and removes the most common source of disagreement between two sites.

Where a difference is found, the escalation should be defined rather than improvised. Hold the drum, photograph it under the lamp with a white card, notify the supplier with the reference and the conditions, and ask for their observation. Four steps, written down, and the conversation stays technical.

Frequently asked questions

Is "white to off-white powder" a specification?

It is a description with a range, and it becomes a specification only when the range is anchored to something. Without a reference standard, a light source and a recorded observation, two competent people can disagree about the same drum and both be honest.

What reference should be used for colour?

Either a pharmacopoeial colour standard series, where one exists for the method, or a retained reference sample of the material itself, qualified and stored under conditions that keep it representative. The second is more useful in practice and more often missing.

Does the light source matter?

Substantially. The same powder looks different under north daylight, under a daylight-matching lamp and under office fluorescent tubes, which are deficient in the red end and push yellow toward green. A colour judgement made under unspecified lighting is not reproducible.

What about instrumental colour measurement?

It is more reproducible and more expensive. A colorimeter gives coordinates in a defined space, which can be trended and compared across sites. It does not remove the need for a decision about which coordinates are acceptable, and that decision still has to be made by someone.

Is yellowing a stability signal?

Often. Yellowing in a bile acid powder usually indicates oxidation or a change in residual solvent, and it can precede any movement in the assay. A batch that has yellowed in storage is worth an impurity check rather than a re-description.

What should the record contain?

The sample, the reference used, the light source, the observer, the date, and the result in the terms of the specification. Four of those five are usually missing, and their absence is what makes appearance the easiest test to dispute.