Evaluating a Research Material Supplier

How to Use This Checklist

The purpose of supplier evaluation is to separate what a seller asserts from what a buyer can independently confirm. Those are different categories of information, and most disputes about research material trace back to treating the first as though it were the second. This checklist describes no particular company, confers no endorsement, and is not a certification of anyone.

It helps to keep an explicit hierarchy of evidence in mind. In ascending order of strength: an unsupported statement on a web page; a document the seller wrote; a document a third party wrote that the seller selected and transmitted; a document obtained directly from a third party the buyer selected; and a measurement the buyer made on material from the container that actually arrived. Each step up that ladder costs money and time, and in practice most procurement decisions are made two or three rungs from the top. That is a defensible position provided the buyer knows which rung they are standing on and records it as such.

The failure modes are real rather than hypothetical. In markets where material is sold at distance on the strength of a label, label-to-content agreement is an empirical question, and verification is the rational default posture.

Legal Status Is the Buyer’s Responsibility, and It Is Settled in Advance

The buyer, not the seller, is responsible for determining whether a given material may lawfully be imported into, possessed in, and handled within their own jurisdiction and at their own institution. This determination should be complete before any transaction, because the point at which a shipment crosses a border or arrives at a receiving dock is the point at which any error becomes concrete. A seller’s product description, a disclaimer, or simple willingness to ship is not a legal determination, has no authority, and will not be treated as one by a customs authority, a regulator, or an institutional compliance office.

The categories to check are several and independent of one another. Controlled substance scheduling operates at both the federal level (21 U.S.C. §§ 811–812; 21 CFR Part 1308 in the United States) and at state or provincial level, and the two do not always agree; a compound unscheduled federally may be scheduled locally. Separately, a material may hold medicinal or prescription-only product status, may be subject to import permits or customs restrictions, may fall under precursor or export controls, and may require institutional biosafety, chemical hygiene, or ethics approval before it can be received at all. Classification is compound-specific, jurisdiction-specific, and subject to change, which means a determination made for one material or in one year does not transfer.

One feature of United States law is worth understanding because it is frequently misread. What a product legally is depends on objective intent, which may be shown by expressions, by the design or composition of the article, and by the circumstances surrounding its distribution (21 CFR 201.128). Descriptive labelling does not by itself settle the question in either direction, and the surrounding context is part of the record. The practical consequence is that a phrase appearing on a seller’s site is not a shield, and the question should be routed through counsel or an institutional compliance function rather than resolved by reading marketing copy.

The Document Set to Request, and What Each Document Is Worth

A reasonable request list covers: a written specification stating the tests performed and their acceptance criteria; a lot-specific certificate of analysis, explicitly not a typical or representative analysis; the underlying analytical data rather than a summary, meaning the chromatogram with legible axes, the integration table, and the method parameters (column, mobile phases, gradient programme, flow rate, detection mode and wavelength, injection volume, total run time), plus any mass spectrum with its acquisition parameters; identity data and a statement of how identity was confirmed; a safety data sheet; a statement of whether the seller manufactures the material or purchases and repackages it, and in which country the manufacturing site sits; the retention-sample policy and record-retention period; and the evidentiary basis for any stated storage condition or expiry or retest date, which is either a stability study on this material or an assumption by analogy.

Each of those documents is worth something different. A specification and a result are not interchangeable: a result reported without a specification cannot fail, because nothing has been declared in advance. A typical analysis is a legitimate document type that describes what a seller’s material generally looks like, and it explicitly does not assert correspondence with the container in question; treating one as a lot certificate is a category error, and a seller who substitutes one for the other without saying so has told the buyer something. A safety data sheet is a hazard communication document, not a quality document. In the United States, chemical manufacturers and importers are required to develop safety data sheets and provide them to downstream recipients (29 CFR 1910.1200(g)), so its absence is a basic compliance signal, while its presence says nothing whatever about purity or identity.

How the request is answered is itself evidence, and it is free. Note whether specific questions receive specific answers or general reassurance; whether the seller will put an answer in writing; whether documents arrive as text-bearing files or as flattened images that resist inspection; and whether the manufacture-versus-resale question is answered at all. A reseller is not thereby disqualified, but a reseller cannot vouch for synthesis records it has never seen, and a buyer who does not know which they are dealing with has an unrecorded gap in the chain.

Manufacturer QC Certificate Versus Independent Laboratory Report

A manufacturer’s quality control certificate is a self-declaration issued by the party with a financial interest in the result. The testing behind it may have been performed in-house, or it may have been inherited from an upstream supplier and re-typed onto new letterhead. This does not make it worthless. A well-formed manufacturer certificate names the specification, the methods, the lot, the date, and an authorised signatory, and in doing so it creates a specific, attributable representation that can be checked against subsequent measurement and relied on in a commercial dispute. What it does not carry is any external assessment of the competence of the people who produced it, or any structural separation between the party performing the test and the party who benefits from the outcome.

An independent laboratory report is issued by a party with no stake in the result, and it is stronger for exactly that reason, subject to three qualifications that are easy to overlook. First, independence is not binary. A laboratory owned by the same parent company, or one that derives a large share of its revenue from a single client, or one engaged repeatedly by a seller to run the one test that seller wants run, is independent in form more than in substance. Second, accreditation is scope-limited. Accreditation to ISO/IEC 17025 applies to defined methods and fields listed in a scope document, not to the laboratory as an undifferentiated whole, and certification of a seller’s own management system to ISO 9001 says nothing at all about the accuracy of any test result. Third, and most important, a report describes the sample the laboratory received, under whatever identity the submitting customer supplied. Accredited reports customarily state that results relate only to the items tested. If the seller chose the sample, prepared it, labelled it and shipped it, the chain of custody begins inside the seller’s own premises, and the report is evidence about the seller’s submission rather than about the container in the buyer’s hands.

There is a useful structural test for whether a document is a laboratory report at all. Under ISO/IEC 17025:2017, reporting of results is governed by clause 7.8, which sets a minimum content: the identity of the laboratory, a unique report identifier with pagination and a clear end, the customer’s identity, the method used, identification and condition of the item tested, the date of receipt and date of performance, the results with units, and identification of the person authorising the report. A single-page file bearing a percentage, a logo and no method, no sample identification and no signatory is a summary or a marketing sheet whatever its title says, and should be filed as one.

Verifying the Paper Without the Seller’s Cooperation

Accreditation claims are checkable and should be checked at the source. Obtain the accreditation certificate number, then look it up in the accreditation body’s own public register rather than relying on a logo reproduced in a PDF. Confirm that the certificate is current, and open the scope document to confirm that the specific test method and material type actually appear within the accredited scope. Accreditation bodies that are signatories to the ILAC Mutual Recognition Arrangement maintain searchable public directories for this purpose, and cross-checking takes minutes. A laboratory can be genuinely accredited for one technique and not for the one that generated the number in front of you.

Whether the supplier will name the testing laboratory is close to a threshold question. An unnamed laboratory cannot be contacted, its accreditation cannot be looked up, and its report cannot be confirmed, which reduces a third-party report to the evidentiary weight of a seller’s own assertion. Where the laboratory is named, contact it directly, using contact details taken from the laboratory’s own website rather than the telephone number printed on the report, and ask whether a report bearing that number and date was issued and to whom. Some laboratories operate report-verification portals or issue reports with verification codes. A laboratory may decline to discuss a client’s report, and that response is still informative, because confirming that a client relationship exists is more than the PDF alone establishes.

A published street address, a working telephone number and a named responsible person establish who is accountable and who can be served with correspondence or process. They are a floor, not a quality indicator, and they should be read as such. Look up the trading entity in the relevant public business register, and note the formation date, current status, registered agent, listed officers and any trading-name filings; check whether the entity that appears on the invoice is the entity that appears on the website. Reverse-search the street address to see whether it corresponds to a warehouse, an office, or a suite number in a mail-forwarding or virtual-office building, which is common and not inherently improper but changes what the address tells you. Call the number and observe whether a person answers who can identify the entity and the named individual. None of this speaks to what is inside the container. All of it speaks to what is available to the buyer if something goes wrong, which is a separate and legitimate concern.

Lot Numbers and Traceability

A lot number is supposed to designate a discrete quantity of material produced or packaged under uniform conditions, on a known date, with retained records, so that a document, a physical container and a manufacturing event all refer to the same thing. The questions that test whether that is true are simple and can be asked in one message: what quantity did this lot comprise; what is its manufacture date, and separately its packaging date; is this a manufacturing lot or a repackaging lot; if the material was repackaged, what is the upstream manufacturer’s lot designation; how long are batch records and retention samples kept, and will a retention sample be released to a laboratory the buyer nominates.

Several practical checks follow. Confirm that the lot on the container label, the lot on the certificate, and the sample identification on the analytical report are the same string. Compare lot designations across orders separated by some months, since a catalogue in which every unit ever sold carries a single lot number is not operating a lot system. Watch for numbers that are actually order numbers or date codes wearing a lot number’s clothing. Where more than one lot’s documentation is available, compare the chromatograms directly: identical retention times, identical baseline noise and identical integration values across supposedly different lots indicate one file reused, not two lots that happened to agree. Where PDFs have not been flattened to images, document creation metadata is sometimes visible and sometimes inconsistent with the stated analysis date.

If a seller cannot state what a lot comprised, or cannot connect it to an upstream lot when the material was bought in, then the number functions as an inventory label rather than a traceability record. Where little turns on the outcome, that may be an acceptable trade, but it should be written down as an unresolved unknown rather than quietly counted as a satisfied criterion. The distinction matters later, when a result is anomalous and the first question is whether anything can be traced back at all.

Testing Purity Claims, and the Value of Independent Re-testing

A purity figure detached from a method is not a measurement. The questions to ask are: by what technique; with what detector and, for ultraviolet detection, at what wavelength; on what column and with what gradient; over what total run time; with what integration parameters and threshold; and is the figure an area percentage or a quantified assay against a characterised reference standard. Area percentage by ultraviolet absorbance carries an implicit assumption that impurities respond like the analyte, which is generally false. Species that do not absorb at the chosen wavelength, that do not elute within the gradient window, or that co-elute with the main peak do not appear in the number. A truncated chromatogram, an ordinate scaled so that small peaks vanish into the axis, or a run time shorter than the stated gradient are all worth noticing. Purity is also not identity: an accurate mass confirms a mass, not a sequence or a stereochemical assignment, and isobaric or isomeric impurities require an orthogonal confirmation such as tandem mass spectrometry or sequencing.

A related point is that the mass written on a container is not necessarily the mass of the analyte in it. Lyophilised peptide salts, for example, contain water and counterion in addition to peptide, so chromatographic purity and the fraction of the solid that is peptide are two different quantities determined by two different methods, the latter by quantitative amino acid analysis or elemental analysis, with water by Karl Fischer titration and counterion by ion chromatography. A high chromatographic purity figure is entirely compatible with a substantially lower net peptide content, and a certificate that reports only the first has answered only half the question. It is also fair to ask whether the analytical procedure was validated and against what framework; specificity, linearity, range, accuracy, precision and limits of detection and quantitation are the standard headings, treated in ICH Q2(R2) and in USP General Chapter <1225>.

Independent re-testing is the only evidence that pertains to the container actually in the buyer’s possession. Done properly it means taking material from that container for analysis, assigning it a neutral sample identifier so the laboratory does not know its origin, sending it to a laboratory the buyer selected and pays directly, and keeping the buyer’s own record of custody. Where the stakes justify it, use a second method that is orthogonal to the seller’s, such as a different stationary phase or mobile phase pH, or mass spectrometric detection alongside ultraviolet, because a single method is systematically blind to particular impurity classes. Decide in advance what a discrepancy triggers, and investigate before concluding: sampling variation, conditions in transit, and genuine method differences all produce real disagreements that are not evidence of bad faith.

Re-testing costs money, and on small quantities it can approach the cost of the material, so the sensible policy is periodic rather than universal. A workable rule is to test at qualification of any new source, again on any change worth noticing (a new lot, a new manufacturing site, a change in the appearance of the solid, a change in the documentation package), and at intervals thereafter. Keep the reports. Their value accumulates, because a drift in an impurity profile or a discrepancy against a certificate becomes visible only when there is a locally held baseline to compare against, and no supplier’s documentation can substitute for that.

References

1. ISO/IEC 17025:2017, General requirements for the competence of testing and calibration laboratories. International Organization for Standardization and International Electrotechnical Commission, Geneva, 2017. Reporting of results is addressed in clause 7.8; scope of accreditation is defined per laboratory by the accrediting body.
2. 21 CFR 201.128, “Meaning of ‘intended uses'” (US Food and Drug Administration), as amended by the final rule “Regulations Regarding ‘Intended Uses’,” published in the Federal Register on 2 August 2021.
3. Controlled Substances Act scheduling authority and schedules: 21 U.S.C. §§ 811–812; schedules of controlled substances as published by the Drug Enforcement Administration at 21 CFR Part 1308. State and other national schedules are separate instruments and may differ.
4. Occupational Safety and Health Administration, Hazard Communication Standard, 29 CFR 1910.1200; safety data sheet development and downstream provision requirements at 29 CFR 1910.1200(g). Laboratory workplaces are additionally covered by 29 CFR 1910.1450.
5. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use, ICH Q2(R2), Validation of Analytical Procedures, reached Step 4 on 1 November 2023 and adopted by regional regulators subsequently.
6. United States Pharmacopeia, General Chapter <1225>, Validation of Compendial Procedures. Cited for the standard validation characteristics (specificity, linearity, range, accuracy, precision, detection and quantitation limits); chapter numbering is stable but content is revised across USP editions.
7. International Laboratory Accreditation Cooperation (ILAC), ILAC Mutual Recognition Arrangement. Signatory accreditation bodies maintain public directories of accredited laboratories and their scopes, which can be searched to confirm a certificate number and its scope of accreditation.