Reading a Purine Alkaloid COA: Mass Balance and Impurity Limits

Why a Certificate of Analysis Has to Be Reconciled

A certificate of analysis is not a list of independent numbers. It is an accounting statement about one hundred percent of a drum, and the numbers on it have to add up to that hundred percent, one way or another. When they do not add up, the discrepancy is information: it tells you which analytical result is an area percent rather than an absolute content, which limit is a worst case rather than a real value, and which line on the certificate was copied from a template rather than measured.

This article sets out a reconciliation method that works for any solid purine alkaloid, using paraxanthine (CAS 611-59-6) as the worked example, because its specification contains all three of the usual traps in a single page: two named impurity limits, a water figure and a loss on drying figure side by side, and a total heavy-metal limit with four specified metals inside it.

The Mass Balance Equation

For a dry solid the accounting statement is:

100% = assay + specified impurities + unspecified impurities + water + non-aqueous volatiles + residue on ignition + analytical loss

Every term on the right can be measured, and the last one is the honest admission that the sum rarely lands on 100.00%. What matters is that the sign and the size of the gap are known and small.

Term How it is measured Where it hides on a certificate
Assay HPLC against a standard, or titration "Purity" or "Assay (HPLC)"
Specified impurities HPLC peak areas against reference standards "Impurity 1 / Impurity 2"
Unspecified impurities Any peak above the reporting threshold Usually absent, which is the problem
Water Karl Fischer titration "Water"
Non-aqueous volatiles Loss on drying minus water Never printed; has to be calculated
Residue on ignition Gravimetric after ashing "Residue on ignition" or "Sulphated ash"
Analytical loss The remainder Never printed; has to be calculated

Trap 1: Assay and Impurity Limits That Cannot Coexist

A typical paraxanthine specification reads: purity by HPLC above 99.0%, impurity 1 not more than 1.0%, impurity 2 not more than 0.5%. Read literally, the maximum permitted impurity content is 1.5%, which is incompatible with an assay above 99.0%.

The numbers are not wrong; they are doing different jobs. The 1.0% and 0.5% figures are individual worst cases for two specific peaks, the kind of limit written so that a badly made lot is rejected on one impurity without a debate. The 99.0% figure is an area-normalisation result. The only self-consistent reading is that the sum of all peaks must be below 1.0%, while no single specified peak may exceed its own individual ceiling.

Reading Impurity 1 Impurity 2 Total impurities allowed Consistent with assay >99%?
Literal sum of limits 1.0% 0.5% 1.5% No
Correct reading 1.0% max, and total must stay under 1.0% 0.5% max 1.0% Yes

The purchasing fix is simple: put a total-impurity line in the contract, add a limit for unspecified impurities at 0.10% each, and require the individual peak table rather than summary percentages. Then the certificate can be reconciled line by line.

Trap 2: Water and Loss on Drying Both Reported, One Residue Hidden

When a specification lists water not more than 0.5% and loss on drying not more than 0.5%, the two limits look redundant and are in fact a subtraction waiting to happen:

non-aqueous volatiles = loss on drying – water

Lot Water (KF) Loss on drying Implied non-aqueous volatiles Comment
A 0.42% 0.45% 0.03% Consistent: the solid is simply damp
B 0.08% 0.48% 0.40% 0.40% of something that is not water has to be explained
C 0.49% 0.50% 0.01% At the water limit: hygroscopic handling, not a solvent issue

Lot B is the one to ask about. On a product recrystallised from an organic solvent, that 0.40% is residual solvent, and 0.40% of a tonne is four kilograms. If the destination market has a residual-solvent limit for the solvent in question, this is where the compliance question actually sits, not in the assay figure.

Trap 3: A Total Heavy-Metal Limit Is a Budget, Not a Measurement

Paraxanthine is specified at heavy metals below 20 ppm in total, with lead below 1.0, arsenic below 1.5, cadmium below 0.5 and mercury below 1.0 ppm. Add the four specified metals: 4.0 ppm. The total limit therefore leaves 16 ppm of unallocated headroom.

Element Specified limit Share of the 20 ppm budget
Lead <1.0 ppm 1.0
Arsenic <1.5 ppm 1.5
Cadmium <0.5 ppm 0.5
Mercury <1.0 ppm 1.0
Specified subtotal 4.0 ppm 4.0
Unallocated headroom – 16.0

That headroom is where iron, chromium, nickel and residual catalyst metals live. A supplier who reports only the total has reported the least informative number available: a lot can pass at 18 ppm total while carrying 15 ppm of iron picked up in a carbon-steel dryer. If the material is going into an application with its own metal limits, or into a process where iron catalyses degradation, the elemental panel has to be specified element by element.

Trap 4: Microbiological Results Have a Limit of Detection Too

"Escherichia coli: negative in 10 g" is not the same statement as "no E. coli", and it is not the same statement as "zero cfu/g". A 10 g enrichment has a detection limit of one colony-forming unit in 10 g, which is 0.1 cfu/g. It is a sensitive test, and it is a test of a 10 g sample, not of the drum.

The same arithmetic applies to the total plate count. A limit of below 1000 cfu/g and a plate count result of 200 cfu/g are both consistent with a well-made lot and are also both consistent with a sampling plan that happened to draw from a clean corner. Microbiological results are only as good as the sampling plan behind them, which is why a buyer who cares about them should ask how the sample was drawn, not only what the plate showed.

A Reconciled Certificate, Line by Line

Here is what a self-consistent paraxanthine lot looks like once the numbers are put into the equation rather than read as a list.

Line Result Role in the balance
Assay by HPLC 99.4% Main term
Impurity 1 (named) 0.18% Included in the 0.6% non-assay fraction
Impurity 2 (named) 0.09% Included
Unspecified impurities, each <0.10%, three peaks reported Included; this is the line that is usually missing
Water (KF) 0.20% Part of the non-assay fraction
Loss on drying 0.24% Implies 0.04% non-aqueous volatiles
Residue on ignition 0.03% Inorganic fraction
Sum of non-assay terms Approx. 0.6% Consistent with the assay, and with 100%

Nothing on that certificate is unusual. What makes it usable is that the impurity total is stated, the water and drying figures are both present and reconcilable, and the residue is small enough to be plausible for a recrystallised organic solid.

Minimum Receiving Tests for a Purine Alkaloid

A receiving laboratory cannot repeat everything a supplier measured, and it does not need to. The minimum set that catches the errors which actually occur is four tests:

  1. Identity by NMR or by retention time against an authentic standard. This is the test that catches the isomer mix-up, and it is the one that a melting point cannot replace.
  2. Assay plus the individual peak table on one injection. Purity and impurities come from the same chromatogram, and reading them together is what makes the mass balance possible.
  3. Water by Karl Fischer, run against the reported loss on drying. The subtraction is the test.
  4. Residue on ignition. Cheap, fast, and the only routine check on the inorganic fraction.

Everything else, including the full elemental panel and the microbiological set, can be bought as a periodic verification rather than run on every lot, provided the supplier’s own data are on the certificate with methods named.

Writing the Purchase Specification

Most of the disputes in this product family come from a purchase specification that copied a supplier certificate instead of describing a requirement. Five clauses remove most of the ambiguity:

  1. Name the isomers and the identities of the specified impurities, not only their percentages.
  2. Add a total-impurity limit that is consistent with the assay, and a per-peak limit for unspecified impurities.
  3. Require water and loss on drying on the same batch, and require that the difference be reported.
  4. Require the elemental panel to be reported element by element where the application has metal sensitivity.
  5. State the intended regulatory route, so the supplier knows whether microbiological data and residual-solvent declarations are in scope.

FAQ

Is a higher assay always better?

Not by itself. An assay of 99.8% by area normalisation on a method that does not see inorganic residue or water can describe a lot with more total impurity than one at 99.2% that reports a full balance. Compare balances, not headline numbers.

What does "area normalisation" actually mean?

It means every peak in the chromatogram is expressed as a percentage of the total peak area, so impurities that do not absorb at the detection wavelength, or that elute outside the run window, are simply not counted. It is a purity indicator, not an absolute content determination.

Why do some certificates report heavy metals as a single figure?

Because a colourimetric sulphide precipitation test gives a single number and is cheap. It is a legitimate limit test and a poor quantitative tool, and it says nothing about which metal is present. Where the application cares, an ICP panel replaces it.

Should I reject a lot because loss on drying is higher than water?

Not automatically. A small gap, on the order of 0.05%, is normal for a hygroscopic solid handled in ambient air. A gap of 0.3 to 0.5% means something volatile that is not water is present, and that is the point at which the supplier should be asked what the final crystallisation solvent was.

Does this method apply to other product families?

Yes. The same reconciliation works for any solid organic product with an assay, an impurity profile, a water figure and an ash figure, including the phenethylamine alkaloids and food-grade protein ingredients covered elsewhere on this site.

Further Reading

Disclaimer: this article describes analytical and commercial practice for industrial chemicals. It is not a substitute for the buyer’s own quality agreement, pharmacopoeial requirements or regulatory obligations in the destination market.

Contact PolyblueChem for a full certificate of analysis with the impurity table, the water and loss-on-drying pair, and the elemental panel reported element by element.

Scroll to Top