Buying nitrosamine reference standards: what the limit actually requires
An NDSRI standard is bought to prove a number. Most of what goes wrong in a nitrosamine project — a purity figure the reviewer questions, a standard that will not dissolve, a lot that degrades in transit — traces back to a specification written before anyone calculated what the limit really is. This guide works backwards from the limit.
The acceptable intake is not one number
A nitrosamine limit is expressed as an acceptable intake (AI) in nanograms per day, not as a concentration in your product. Two frameworks currently dominate, and they do not produce the same figure for the same molecule.
The FDA framework, finalised in August 2023 and maintained on a continuously updated CDER table, assigns a predicted potency category from the structure of the nitrosamine (the Carcinogenic Potency Categorization Approach, CPCA). The EMA framework uses the same five-category structure but publishes its own values in Appendix 1 of the nitrosamine Q&A, and applies a default of 18 ng/day to any nitrosamine that has not yet been assessed.
| Category | Structural feature | FDA AI | EMA AI |
|---|---|---|---|
| 1 | Unhindered α-carbon | 26.5 ng/day | 18 ng/day |
| 2 | Moderately hindered | 100 ng/day | 100 ng/day |
| 3 | α-hydroxyl or branched | 400 ng/day | 400 ng/day |
| 4 | Sterically hindered | 1500 ng/day | 1500 ng/day |
| 5 | Not predicted mutagenic | 1500 ng/day | 1500 ng/day |
Categories 2 to 5 agree. Category 1 does not: 26.5 ng/day in the United States, 18 ng/day in the EU. For a potent NDSRI that is a third more headroom on one side of the Atlantic than the other, and it changes whether your method is fit for purpose.
Three consequences for anyone specifying a standard:
- A compound-specific AI always beats the CPCA value. Both agencies move nitrosamines to a separate table once real carcinogenicity data or a justified read-across becomes available. Check the compound-specific table before you accept a category number — the difference can be an order of magnitude.
- A negative enhanced Ames test changes the route, not just the number. If the mutagenic hazard is ruled out under the enhanced conditions, the substance can be controlled as a non-mutagenic impurity under ICH Q3A/Q3B instead of as a cohort-of-concern nitrosamine. That is a different specification entirely, and it is worth knowing before you buy a standard priced for trace-level work.
- Less-than-lifetime adjustment is not settled. The FDA publishes adjustment factors for short-exposure products; the EU position is more restrictive and the ICH addendum is still in development. If you are buying for a clinical programme rather than a marketed product, do not assume the lifetime limit applies.
Limit values above are taken from the FDA CDER nitrosamine acceptable intake table (CPCA, Table 1) and EMA Appendix 1, both of which are revised periodically. Confirm the current value for your specific nitrosamine before fixing a specification.
Convert to ppm before you talk to a supplier
No supplier can quote against “18 ng/day”. Your method, your specification and your standard all operate in concentration, and the bridge between the two is the maximum daily dose:
limit (ppm) = AI (ng/day) ÷ maximum daily dose (mg/day)
The arithmetic works because 1 ng/mg is 1 µg/g, which is 1 ppm. A worked check against a published FDA interim limit: N-nitroso-sertraline is set at 600 ng/day, which the agency lists as 3 ppm — and 600 ÷ 200 gives 3, where 200 mg is the top of the sertraline dose range. The rows below use the two category 1 values plus the three higher categories from the table above.
| AI (ng/day) | 10 mg/day | 100 mg/day | 1000 mg/day |
|---|---|---|---|
| 18 EMA | 1.8 ppm | 0.18 ppm | 0.018 ppm |
| 26.5 FDA | 2.65 ppm | 0.265 ppm | 0.0265 ppm |
| 100 | 10 ppm | 1.0 ppm | 0.1 ppm |
| 400 | 40 ppm | 4.0 ppm | 0.4 ppm |
| 1500 | 150 ppm | 15 ppm | 1.5 ppm |
Read the table down a column and the real difficulty appears. The same category 1 limit is a routine LC-MS/MS target for a 10 mg tablet and a genuinely hard analytical problem for a 1 g daily dose, where the specification sits at 18 ppb. This is why a high-dose, high-volume API — metformin being the obvious case — generated so much analytical work out of proportion to the chemistry involved.
Two practical rules follow. First, the method limit of quantification is normally set at or below 30% of the specification limit, so a 0.018 ppm specification implies an LOQ near 5 ppb in the product matrix, not in a clean standard solution. Second, if a nitrosamine is present below 10% of its AI, the FDA does not expect it in routine release testing — but that exemption is about testing frequency, not about the method. You still have to demonstrate at validation that you could have seen it.
Tell your supplier the maximum daily dose. It is the single most useful number you can put in an enquiry, and it is the one most often left out.
What the certificate has to say
A chromatographic purity of 98% tells you what the peak looked like. It does not tell you how much of the material in the vial is the substance you are going to weigh. For a working standard those are different questions, and only the second one lets you calculate a result.
Assigned purity, stated as a value with an uncertainty. Quantitative work needs the content of the analyte, established independently of the chromatography — typically by quantitative NMR against an internal standard, or by a mass balance that accounts for water, residual solvent and inorganic residue. A certificate that reports only “purity 98.5% (HPLC)” leaves you unable to state your own result to three significant figures, which is what a 0.02 ppm specification demands.
The salt form, explicitly. The acceptable intake is expressed against the free base of the parent drug, and residue limits are calculated on that basis. If the reference standard is supplied as a salt or a solvate, every weighing has to be corrected for the difference, and that correction is easy to omit after the first transfer. State the form on the certificate and state it again on the label.
The nitrosation position. An API with more than one nitrosatable amine gives more than one distinct NDSRI, which is why the FDA table carries suffixes such as -1 and -2 against a single drug. These are separate substances with separate limits, separate retention times and, sometimes, enormously different potencies. A standard labelled only “N-nitroso-X” without a position is ambiguous, and an ambiguous standard will be rejected at method validation.
Characterisation data, not just a pass/fail. For a substance with no pharmacopoeial monograph there is nothing to compare against, so identity rests entirely on the data package: high-resolution mass for the molecular formula, NMR assignments that locate the nitroso group, and an elemental or water determination if the purity is established by mass balance.
When there is no CAS number
A large share of NDSRIs have no CAS registry number. They are hypothetical impurities flagged by a risk assessment, or observed in a single product, and nobody has registered them. This is normal for the class and it does not prevent you from buying one.
What it does mean is that the enquiry has to carry the structural definition itself. Four items are enough for a supplier to confirm feasibility:
- Parent drug, with the salt or free-base form you are working against.
- Nitrosation site — which nitrogen carries the nitroso group, described in words if you have no structure file.
- Molecular formula of the intact NDSRI, from which the exact mass follows.
- A structure drawing, or a SMILES string if your system generates one.
“Nitroso impurity of X” alone is not enough, for the reason above: X may support two or three positions, and the supplier will either guess or come back with a question. A mass number from an unknown-peak spectrum is useful supporting evidence, but it does not define the position — two positional isomers are isobaric.
What to put in the enquiry
Most nitrosamine enquiries take two or three exchanges before a supplier can quote. Each of the fields below removes one of those exchanges.
| Field | Why the supplier needs it | Example |
|---|---|---|
| Parent API and form | Fixes the structure and the salt correction | Amlodipine (free base) |
| Nitrosation position | Distinguishes positional isomers | N-nitroso on the piperazine N-1 |
| CAS or structural definition | Confirms identity where no CAS exists | Formula + drawing |
| Maximum daily dose | Converts your AI into the ppm you must meet | 10 mg/day |
| Target market | FDA and EMA category 1 limits differ | EU / US / both |
| Quantity and pack size | Determines whether custom synthesis is viable | 25 mg, two vials |
| Documentation required | COA content, MSDS, TSE/BSE statement | Assigned purity by qNMR |
| Required by date | Stock versus synthesis lead time | Method validation in Q1 |
If the impurity is already catalogued, the enquiry is a stock question and the answer comes back the same day. If it is not, the four structural items above are what turn a “can you make this?” into a feasibility answer with a date.
Storage, stability and the part nobody plans for
Nitrosamine standards are small, polar and often volatile enough to leave a vial. They are also frequently light-sensitive, and in solution the N–N=O group is not indefinitely stable. The practical consequences are unglamorous and consistently underestimated:
- Ship and store cold. −20 °C, protected from light, desiccated. A standard that spent a week in a warm customs warehouse is a standard of unknown content, whatever the certificate says.
- Equilibrate before opening. Cold vial, warm room, moist air: the contents condense water and the weighing is wrong before you start.
- Prepare solutions fresh. Volatility and adsorption onto glass and septum mean a stored working solution loses analyte, and headspace sampling loses it faster. If a result drifts low across a sequence, suspect the solution before suspecting the method.
- Confirm identity on receipt. A retention time match is a weak identity test for a compound with no monograph. Mass confirmation on arrival costs an hour and protects an entire validation campaign.
None of this is exotic. It is simply the difference between a standard that supports a regulatory submission and one that supports a repeat purchase.
Frequently asked questions
Why does a supplier ask for the maximum daily dose?
Because the acceptable intake is a daily amount, not a concentration. Dividing the AI in ng/day by the maximum daily dose in mg/day gives the limit in ppm, and that number decides whether the method needs LC-MS/MS or can be done by HPLC. Without it a supplier can only guess at the specification you are working to.
Is 95% purity enough for a nitrosamine standard?
For qualitative work such as retention time confirmation, usually yes. For quantitative work the number that matters is the assigned content, not the chromatographic purity, because 5% of unknown composition is 5% of uncertainty in every result you report. Ask for an assigned value with an uncertainty and for the identity of the significant minor peaks; if those peaks include the parent drug or a positional isomer, they may also interfere chromatographically.
My NDSRI has no CAS number. Can it still be supplied?
Yes, and this is routine for the class. Send the parent drug, the nitrosation position, the molecular formula and a structure drawing. Those four items are sufficient to confirm feasibility. A CAS number is an identifier, not a requirement for synthesis. Why some impurity standards have no CAS number covers the underlying reasons and what to send in place of one.
What is the difference between an NDSRI and a small-molecule nitrosamine?
A small-molecule nitrosamine such as NDMA or NDEA does not resemble the drug and turns up across many products; its limit is usually compound-specific and very low. An NDSRI is formed from the drug molecule itself, so it is unique to one API and its limit comes from the potency category of that specific structure. The analytical consequence is that small-molecule nitrosamines can be covered by a multi-analyte panel, while NDSRIs generally need one standard per compound.
Do FDA and EMA limits ever differ for the same impurity?
Yes. Both agencies use the same five potency categories, but the category 1 value differs: 26.5 ng/day at FDA against 18 ng/day at EMA. Compound-specific limits can also diverge. If you file in both markets, specify against the lower of the two, or run two specifications.
Does every pack come with a certificate of analysis?
Yes. Each lot is released with a certificate stating identity, assay and chromatographic purity, together with the characterisation data used to assign the structure. The certificate is issued against the lot shipped, not against a reference lot.
Sources
Regulatory values quoted above are from the following, all of which are revised periodically:
- FDA, CDER Nitrosamine Impurity Acceptable Intake Limits — CPCA table, compound-specific and interim limits, less-than-lifetime adjustment factors.
- FDA, Recommended Acceptable Intake Limits for NDSRIs (August 2023) and Control of Nitrosamine Impurities in Human Drugs (September 2024, Rev. 2).
- EMA, Nitrosamines in human medicines — Q&A document with Appendix 1 (established AI values), Appendix 2 (CPCA) and Appendix 3 (enhanced Ames test conditions).
- EMA, Response of the European Medicines Regulatory Network to nitrosamine impurities.
- ICH M7(R2) and the ongoing addendum on N-nitrosamine risk assessment and control.
Send the structure, get an answer
929 nitrosamine standards are catalogued and in stock. If yours is not among them, send the parent drug and the nitrosation position — we will confirm feasibility before quoting, and we will tell you if the positional isomer is the one you need.