01 COA & batch testing

What the COA for L-carnitine actually contains

Technical document 01 · Reviewed 2026-11-18 · L-Carnitine

On this page
  1. The certificate as a test report
  2. HPLC, rotation and the anhydrous basis
  3. Release limit versus batch result
  4. What moves between batches
  5. When to order an outside test
  6. What to compare between suppliers

The certificate for an L-carnitine lot reports appearance, infrared identification, assay by HPLC on the anhydrous basis, specific rotation, loss on drying, residue, metals and microbiology. Each line needs a method and a limit beside the result, and the reader needs to know which figure says how much carnitine is present and which says which carnitine it is. The values below are catalog; the certificate confirms one lot.

In one line

The assay says how much carnitine is in the sample, the rotation says which carnitine it is, and the anhydrous basis says what the percentage is a percentage of.

The certificate as a test report

A certificate can be audited only when every row carries the parameter, the method and the limit the lot was released against. A result with no method is an assertion; a method with no limit cannot be judged.

Identity, assay and water · L-carnitine base
ParameterMethodRelease limit
AppearanceVisualWhite crystalline powder
IdentificationInfrared absorptionConforms to reference spectrum
Assay (anhydrous basis)HPLC98.0–102.0%
Specific rotationPolarimetry, 10% solution−29.0° to −32.0°
Loss on dryingGravimetric≤0.5%
Residue on ignitionGravimetric, ashed≤0.1%
Physical, contaminant and microbiological limits
ParameterMethodRelease limit
Particle sizeSieve analysis≥95% through 40 mesh
Heavy metalsICP-MSPb ≤3 · As ≤2 · Cd ≤1 · Hg ≤0.1 mg/kg
MicrobiologyPlate count and absence testsAPC ≤1,000 cfu/g; yeast and mould ≤100 cfu/g; E. coli and Salmonella absent

Read the sheet in blocks: the first table describes the molecule, the second the powder. A certificate that reports the assay and omits the rotation has answered the quantity question and skipped the identity question.

HPLC, rotation and the anhydrous basis

The assay is an HPLC result. Carnitine has no strong chromophore, so a method that reports a percentage has to say how the detector saw the molecule: low-wavelength ultraviolet, a derivatisation step, or an evaporative detector. The method sheet should name the column, the mobile phase, the standard and the calculation; “HPLC 99.4%” alone cannot be reproduced by a second laboratory.

The specific rotation is the identity figure. It is measured on a 10% solution, and the reading depends on solvent, temperature, path length and concentration, so a value quoted without them is not comparable. The window here is −29.0° to −32.0°; the width absorbs normal polarimeter variation and still excludes the racemate, which sits near zero.

The anhydrous basis is a correction, not a second measurement: the assay is divided by the dry matter, so a result on that basis is higher than the same sample reported as-is. With water at ≤0.5%, the correction can move a figure by up to about half a point — enough to carry a lot across the 98.0% line. The first question about an assay figure is therefore not “what is it” but “on which basis”.

Two small lines matter. Loss on drying covers water and volatiles; at ≤0.5% on a hygroscopic base it is a statement about the pack as much as about the dryer. Residue on ignition at ≤0.1% shows that crystallisation removed inorganic material, and it is the line that moves first when a wash is shortened.

Release limit versus batch result

The release limit is the boundary; the batch result is what the laboratory measured. Both belong on the certificate, in separate columns, with the method beside them. A result column filled with the specification range is not a result; it is a copy of the gate, and it tells the buyer nothing about the drum in front of them.

Not every line needs a range in the result column. Limits written as “not more than” — loss on drying, residue, the four metals — are one-sided, and both 0.12% and 0.15% water sit inside the same gate. The rotation and the assay are two-sided, so each should report one figure with its conditions. “Passes specification” is a conclusion, not a measurement, and the reader should be able to reach it from the numbers on the page.

What moves between batches

Water is the first variable. It moves with the drying end-point, the humidity in the packing room and the condition of the liner, and it moves the anhydrous assay with it. A lot can sit inside the assay window and still be wetter than the previous one, which shows up in flow, caking and fill weight long before a rejection.

The rotation moves less, and that is the point. A change of a full degree inside the window is worth investigating, because it is the only line that responds to the enantiomeric composition. Colour and particle size are the practical variables: a cream tint or a shift in the 40-mesh fraction changes how the powder runs on a capsule machine while the chemistry stays in specification.

Related substances are the gap in a short panel. Betaine and crotonobetaine are carnitine's close structural neighbours, and a specification that does not name them is not testing for them; if the risk assessment needs that coverage, add the test rather than assume the assay covers it.

When to order an outside test

Order an independent test for a new source, a first shipment into a regulated market, a rotation at the edge of the window, or a disputed certificate. The useful panel is targeted: rotation on a freshly prepared solution, chiral HPLC for the D enantiomer, loss on drying for water, and ICP-MS for the metals.

Ask the laboratory to state its limit of detection for the chiral method. “D-carnitine not detected” without a limit describes the method's sensitivity, not the sample. The same rule applies to the assay: the report should name the column, the detector and the basis, not only the percentage.

When a result is contested, the sample tested should be the retained sample from the lot, and the lot code on the certificate, the drum and the referee's report should match.

What to compare between suppliers

Put six lines side by side: form and counter-ion; assay with method and basis; rotation with conditions; loss on drying; the contaminant panel; particle size. Two quotations that agree on the assay and diverge on the rotation are not describing the same material; two that agree on both but differ on water will behave differently in the capsule room.

The second layer is traceability. The lot code should be identical on the certificate, the pack and the retained sample, and the certificate should name the manufacturing site and the pack date. A correct certificate can still belong to a drum other than the one on the loading dock.

The catalog specification sits on the L-carnitine product page. The forms are compared on the grade page, the evidence that separates the L form from a racemate is on the identity page, and the release workflow is described on the quality page. The separation itself is explained in the HPLC method note, and loss on drying is defined in the glossary.

Send the specification you intend to use — grade, quantity and destination market — and we will reply with the batch COA, the technical data sheet and a quotation.

Request a quotation