Skip to content

Pharmaceutical quality documents

Specified, unspecified and total impurities on a COA

The impurity block on a certificate carries three kinds of limit, and each is read differently. This guide explains specified vs unspecified impurities on a certificate of analysis, and why the total impurities figure is often larger than the sum of the rows printed above it.

Short answer

A specified impurity is listed by name or code with its own limit. An unspecified impurity is any other peak, held to one general limit that applies to each such peak on its own. Total impurities is a separate result: every impurity above the reporting threshold, added together, including peaks that have no row.

What is the difference between specified, unspecified and total impurities on a certificate?

Most certificates of analysis for an active ingredient carry an impurity block, often headed related substances or organic impurities. It usually has three kinds of line, and they answer three different questions.

A specified impurity is one the specification names individually and gives its own acceptance criterion. ICH Q3A(R2) is explicit that a specified impurity can be identified, meaning its structure is known, or unidentified, meaning it is known only by an analytical property such as its retention time. So a row labelled with a relative retention time and no chemical name can still be a specified impurity.

An unspecified impurity is the opposite case. The guideline defines it as an impurity limited by a general acceptance criterion but not individually listed with its own. On a certificate this is the line that reads any other impurity, any unspecified impurity or largest unknown. The limit applies to each unlisted peak separately, which is why the result printed against it is normally the largest one found.

Total impurities is not a sum of the rows. ICH Q3A(R2) says all impurities at a level greater than the reporting threshold should be summed and reported as total impurities. That includes the small unspecified peaks that never get a line of their own.

Line on the certificateWhat the limit applies toTypical wording
Specified identified impurityOne named impurityImpurity A, or a chemical name
Specified unidentified impurityOne impurity known by retention timeUnknown at a stated RRT, or a code
Unspecified impurityEach other peak, one at a timeAny other impurity, largest unknown
Total impuritiesThe sum of everything above the reporting thresholdTotal impurities, total related substances

Where the thresholds come from, and where they do not apply

ICH Q3A(R2) covers impurities in new drug substances, and ICH Q3B(R2) covers degradation products in new drug products. For a drug substance taken at up to 2 g a day, Attachment 1 of Q3A(R2) sets a reporting threshold of 0.05%, an identification threshold of 0.10% or 1.0 mg per day intake, whichever is lower, and a qualification threshold of 0.15% or 1.0 mg per day intake, whichever is lower. Above 2 g a day the three thresholds fall to 0.03%, 0.05% and 0.05%. Q3B(R2) uses its own table: a reporting threshold of 0.1% for a maximum daily dose up to 1 g, and 0.05% above it.

The summary list in section 6 of Q3A(R2) ties these together for organic impurities. A drug substance specification should include, where applicable, each specified identified impurity, each specified unidentified impurity, any unspecified impurity with an acceptance criterion of not more than the identification threshold, and total impurities. Residual solvents and inorganic impurities are listed separately. Q3B(R2) gives the same four lines for degradation products in a drug product.

Two cautions before using these numbers on a real certificate. First, they are written for new substances and products in a registration application; an established material bought against a pharmacopoeial monograph is held to the limits in that monograph, and those can differ from the ICH figures in either direction. Second, the guideline allows lower thresholds where an impurity is unusually toxic and asks for higher reporting thresholds to be scientifically justified. The specification the certificate is issued against is the document that decides which limits apply; the ICH tables explain how such specifications are usually built.

A worked example on an anonymised certificate

The block below is invented for illustration. It describes an active ingredient with a maximum daily dose of 0.5 g, the same dose Q3A(R2) uses in its own illustration, so the reporting threshold is 0.05%, the identification threshold 0.10% and the qualification threshold 0.15%. At that dose the 1.0 mg per day alternative is never the lower figure.

Read down the result column and add the three individual rows: 0.12% plus 0.08% plus 0.08% gives 0.28%. The total impurities line says 0.41%. A receiving team that expects the two to match will think the certificate contains an arithmetic error. It does not.

The chromatogram behind this certificate showed five peaks above the reporting threshold: impurity A at 0.12%, the unidentified peak at relative retention time 1.32 at 0.08%, and three unspecified peaks at 0.08%, 0.07% and 0.06%. Only the largest unspecified peak is printed, because the unspecified limit is checked peak by peak: if the largest passes, every smaller one passes too. All five are added into the total: 0.12 plus 0.08 plus 0.08 plus 0.07 plus 0.06 is 0.41.

There was also a peak at exactly 0.05% and another at 0.03%. Neither is in the total. The guideline sums impurities greater than the reporting threshold, and a peak sitting on the threshold is not greater than it.

So the gap of 0.13% between the rows and the total is the two smaller unspecified peaks. That is the normal pattern. The reverse, a total clearly smaller than the sum of the rows printed above it, is unusual and needs an explanation. There can be a legitimate one: a named impurity printed below the reporting threshold, or a monograph that excludes a particular impurity from the total. The supplier should be able to say which.

Every result here is within its limit, and the certificate is internally consistent. Whether the material is acceptable still depends on the specification it was issued against and on the reviewer's judgement, not on the arithmetic alone.

TestAcceptance criterionResult
Impurity ANot more than 0.15%0.12%
Unidentified impurity at RRT 1.32Not more than 0.15%0.08%
Any other individual impurityNot more than 0.10%0.08%
Total impuritiesNot more than 0.50%0.41%

How to check the impurity block on a certificate

Start with the specification, not the certificate. If you hold the agreed specification or the monograph, compare its impurity list with the rows printed. A specified impurity that the specification names but the certificate leaves out is a question for the supplier.

Then work through the block in order.

  • Each specified impurity has its own row, its own limit and, for a quantitative test, a numerical result. ICH Q3A(R2) asks for impurity results in registration batch data to be given as numbers, not as complies or meets limit, so a quantitative result reported only as complies is worth asking about. A limit test, such as some thin-layer chromatography methods, can properly read complies.
  • The unspecified line states whether it is an individual limit. Wording such as any other impurity or largest unknown usually means it is; a line that reads only unknowns is ambiguous and worth asking about.
  • Results carry sensible precision. For drug substances, Q3A(R2) reports results below 1.0% to two decimal places and at or above 1.0% to one decimal place.
  • The total is at least as large as the largest row, and normally at least as large as the sum of the rows. A total smaller than the sum of the printed rows is unusual and needs an explanation.
  • The method reference for the related substances test is named, so the results can be traced to a procedure. The glossary entries on the impurity profile and on acceptance criteria cover both terms in more depth.

Where an automated read helps

Reading one impurity block by hand takes a few minutes. Reading forty from a new supplier's batch history is where rows get skipped and a short total goes unnoticed.

The COA Validator reads the results on an uploaded certificate and lays each impurity result beside the limit printed against it, with the total alongside, so the reviewer starts from a structured table instead of a scan. The judgement stays with the qualified person reviewing the batch. The how it works page describes what the analysis covers, and the guide to checking a certificate of analysis covers the full review beyond the impurity block.

Frequently asked questions

These are the questions receiving teams ask most often about the impurity block. A certificate of conformance is a different document and may carry no results at all; the comparison guide explains the difference.

Definitions

Specified impurity
An impurity individually listed in the specification with its own acceptance criterion. It can be identified or unidentified.
Unspecified impurity
An impurity limited by a general acceptance criterion but not individually listed with its own criterion in the specification.
Reporting threshold
The level above which an impurity is reported. Impurities above it are summed into total impurities.
Total impurities
The sum of all impurities at a level greater than the reporting threshold, whether or not each has its own row on the certificate.

Sources

  1. ICH Q3A(R2), Impurities in New Drug Substances, International Council for Harmonisation (Section 5 on reporting impurity content of batches and section 6 on listing impurities in specifications; glossary definitions and the thresholds at Attachment 1.)
  2. ICH Q3B(R2), Impurities in New Drug Products, International Council for Harmonisation (The summary list of degradation products in a specification, and the thresholds for degradation products at Attachment 1.)

Primary sources are cited in preference to summaries. Where a source is licensed (for example a pharmacopoeial monograph) the reference is given and the text is not reproduced.

About the author

PharmaTrust Team

Written and reviewed by the PharmaTrust team against the primary sources cited above. Corrections are published, never made quietly.

Why we built this

Spotted an error or a newer figure? We correct within a week: info@pharmatrust.tech.

Frequently asked

Does the unspecified impurity limit apply to all unknown peaks added together?

No. The limit for any unspecified impurity applies to each peak on its own. The combined burden of all impurities is controlled by the separate total impurities limit.

Why does total impurities not equal the sum of the rows on my certificate?

Because the total includes every impurity above the reporting threshold, and most of the small unspecified peaks have no row of their own. A total larger than the sum of the rows is normal. A total smaller than that sum is unusual; it can have a legitimate reason, such as a named impurity printed below the reporting threshold, but it needs an explanation from the supplier.

Can a specified impurity be unidentified?

Yes. ICH Q3A(R2) says specified impurities can be identified or unidentified. An unidentified specified impurity is usually labelled by its relative retention time instead of a chemical name, and it still carries its own limit.

Are the ICH thresholds the limits my material must meet?

Not necessarily. The ICH thresholds are written for new drug substances and products in a registration application. A material bought against a pharmacopoeial monograph or an agreed specification is held to the limits in that document, which may be tighter or looser for a given molecule.

What does a result of complies in the impurity block mean?

It depends on the test. For a quantitative method it means the number was not given; ICH Q3A(R2) asks for impurity results in registration batch data to be presented numerically, not in general terms such as complies or meets limit, and a reviewer may reasonably ask the supplier for the figures. For a limit test, such as some thin-layer chromatography methods, complies is a proper result.