Categories
Analytical Methods and Purity

A CAS Registry Number has a check digit you can verify on paper. What it identifies unambiguously, and what it does not tell you at all.

The CAS Registry Number is the most commonly used identifier of a chemical substance and, at the same time, the most often over-interpreted one. It is sometimes read as a guarantee: if the number matches, the material is the right one. In reality, the number answers one very narrow question — which substance is meant — and makes no statement on any other subject.

This text shows what the number settles unambiguously, what it does not tell you at all, and how to check in a dozen or so seconds whether a recorded number contains a typo. The last of these is a surprisingly practical skill, because an incorrect number in a document looks exactly the same as a correct one.

What a CAS Registry Number is, and what it is not

A CAS Registry Number is an entry in a registry — a sequential identifier assigned to a substance at the moment it enters the database. It does not encode the structure of the molecule, does not follow from its formula, and cannot be “derived” from the structure. Substances of similar structure have completely dissimilar numbers, and numerically adjacent numbers have nothing to do with each other.

This distinguishes it from structural identifiers such as InChIKey or SMILES, which are derived from structure and can be generated from the structure alone. A CAS Registry Number is a label assigned from outside, so its only function is to point: “this is the registry entry in question”.

This arbitrariness yields an advantage and a drawback at the same time. The advantage is unambiguity: one number is one registry entry, regardless of how many names the substance goes by. The drawback — the absence of any semantic redundancy: a miscopied number still looks like a number, it just points to something else or to nothing.

The checksum, or how to catch a typo

Fortunately, the CAS Registry Number has a built-in safeguard that few people know about: the last digit is a check digit, calculated from the others. The rule is simple enough to apply on paper.

The number consists of three groups separated by hyphens. We take all the digits from the first two groups, read them from right to left, multiply them successively by 1, 2, 3, 4 and so on, and add them up — and the remainder of dividing this sum by 10 must equal the last digit.

Take modafinil, number 68693-11-8. The digits from the first two groups are 6869311. Read from the right: 1, 1, 3, 9, 6, 8, 6.

  • 1 × 1 = 1
  • 1 × 2 = 2
  • 3 × 3 = 9
  • 9 × 4 = 36
  • 6 × 5 = 30
  • 8 × 6 = 48
  • 6 × 7 = 42

The sum is 168, the remainder of division by 10 is 8 — and that is exactly the last digit of the number. The number is internally consistent.

We checked this rule against all forty CAS Registry Numbers in our catalog: it holds in every case. As a control, we changed the last digit of the modafinil number from 8 to 9 — the rule rejects such a number. That is precisely the point of a checksum: it catches a single typo and most digit transpositions.

It is worth using this with every document copied by hand. A number that fails this test is certainly wrong — without any need to consult a database. A number that passes is correctly written, which still does not mean that it refers to the right substance.

What the number distinguishes

This is the source of many misunderstandings, because a CAS Registry Number distinguishes more than is usually assumed.

  • A salt and the free base have different numbers. In the registry’s sense they are two different substances, with different molar masses and different content of the parent substance.
  • A hydrate and the anhydrous form have different numbers. Water bound in the crystal lattice is part of the substance, so it changes the registry entry.
  • Enantiomers and the racemic mixture have different numbers. A racemate is a separate entry, not “the same thing as an enantiomer”.

The practical consequence: if the number on the order does not match the number on the certificate, this does not necessarily mean a mistake — it may mean that a different form of the same substance was delivered. The difference tends to matter precisely when the content by weight counts.

What the number does not tell you

This list is longer than the list of what it does tell you, and it is worth keeping in mind.

It says nothing about purity. The same number appears on material of 95% and of 99.9% purity. Where the purity figure itself comes from, and why two methods give different results, we describe in the article on what the number on a certificate really means.

It says nothing about crystal form. Different polymorphs of the same substance share a number, yet differ in solubility, melting point, and behavior during storage.

It says nothing about origin or batch. The number identifies the substance as a registry entry, not a specific container. That is the role of the batch number and of the document describing it — which we take apart in the article on the certificate of analysis.

And most importantly: it is not proof that the container holds that substance. The number is a declaration, not a measurement. Confirming identity is a separate step and requires spectral data — we return to this in the article on the proton spectrum as proof of identity.

The scale of the identification problem

It is worth realizing how small a fraction of the substances in circulation has actually been studied analytically. A review published in 2023 posed this question directly — how many of the chemicals in commerce have ever been determined in environmental samples — and the answer turned out to be sobering.1 Analyses of compounds present in consumer products showed a similar picture: the registry is far larger than the set of substances actually measured.2

This leads to something important for practice: the mere existence of a number in the registry does not mean that extensive analytical documentation exists for that substance. For rare compounds, the number is sometimes the only reliable piece of information available.

A separate issue is the identification of substances whose number is not known — the literature uses the term “known unknowns” for this. Methods linking accurate mass measurement with searches of chemical databases were developed precisely to assign a structure to a signal that is known to be there but whose identity is unknown.3 These approaches rely on indices and descriptors derived from structure,4 and thus on identifiers that are the opposite of CAS — ones that can be calculated from the structure of the molecule.

Domain databases built around CAS Registry Numbers are, incidentally, created regularly, for example for plant constituents5 or for compounds with a specific application profile.6 In them, the number serves as the key linking records across databases — and that is its strongest, most useful function.

How to use this

A few habits that cost seconds and save hours.

  • Recalculate the checksum for every number copied by hand. A number that fails the test is wrong — there is no need to look anything up.
  • Start your search with the number, not the name. Common, trade, and code names multiply without limit; the number is one.
  • Check that the form matches. A salt, a hydrate, and the free base mean different numbers and different molar masses.
  • Do not treat a matching number as confirmation of the material’s identity. These are two different things, and only the second requires measurement.

A CAS Registry Number is a good key and a poor guarantee. Used as a key — for searching databases, for linking records, for unambiguously indicating what is meant — it is an excellent tool. Used as proof that the material is the right one, it can be a source of errors that come to light only at the first determination that fails.

Sources

Each entry below was retrieved from the PubMed database via the E-utilities interface and is not reproduced from memory. PMID numbers link to the source records.

  1. Muir DCG, Getzinger GJ, McBride M et al. How Many Chemicals in Commerce Have Been Analyzed in Environmental Media? A 50 Year Bibliometric Analysis. Environ Sci Technol. 2023;57:9119-9129. PMID: 37319372. DOI: 10.1021/acs.est.2c09353.
  2. Dionisio KL, Frame AM, Goldsmith MR et al. Exploring consumer exposure pathways and patterns of use for chemicals in the environment. Toxicol Rep. 2015;2:228-237. PMID: 28962356. DOI: 10.1016/j.toxrep.2014.12.009.
  3. Little JL, Cleven CD, Brown SD. Identification of "known unknowns" utilizing accurate mass data and chemical abstracts service databases. J Am Soc Mass Spectrom. 2011;22:348-59. PMID: 21472594. DOI: 10.1007/s13361-010-0034-3.
  4. Zhang Q, Wu C, Zheng F et al. Extension of a Highly Discriminating Topological Index. J Chem Inf Model. 2015;55:1308-15. PMID: 26039156. DOI: 10.1021/acs.jcim.5b00044.
  5. Naeem S, Hylands P, Barlow D. Construction of an Indonesian herbal constituents database and its use in Random Forest modelling in a search for inhibitors of aldose reductase. Bioorg Med Chem. 2012;20:1251-8. PMID: 22261024. DOI: 10.1016/j.bmc.2011.12.033.
  6. Gu J, Gui Y, Chen L et al. CVDHD: a cardiovascular disease herbal database for drug discovery and network pharmacology. J Cheminform. 2013;5:51. PMID: 24344970. DOI: 10.1186/1758-2946-5-51.

Read also