ZHICHENG BIO
L-carnitine: the pharmacopoeial assay is a titration, and a titration cannot see chirality
A 99.5% assay figure on an L-carnitine certificate is a statement about carnitine, not about levocarnitine. Here is what the monograph does and does not lock down.
The short answer
The European Pharmacopoeia assay for levocarnitine is a non-aqueous titration with 0.1 M perchloric acid — and D-carnitine titrates exactly like L-carnitine. So the 98.0–102.0% figure on a compliant certificate constrains how much carnitine is present and says nothing at all about which enantiomer it is. The monograph's only chiral gate is the specific optical rotation, specified as −29.0 to −32.0 on the anhydrous substance. Worked through, that band admits roughly 3% to 5% D-carnitine at its lower limit, depending on which value you take as enantiomerically pure. If you need tighter than that, you have to ask for a chiral method by name.
- The monograph
- Ph. Eur. 1339, Levocarnitine: (3R)-3-hydroxy-4-(trimethylammonio)butanoate. L-carnitine C7H15NO3, Mr 161.20, CAS 541-15-1
- The assay
- "98.0 per cent to 102.0 per cent (anhydrous substance)", by titration with 0.1 M perchloric acid in anhydrous formic acid and acetic acid, crystal violet indicator — an achiral method
- The only chiral gate
- Specific optical rotation −29.0 to −32.0 (anhydrous substance), determined on solution S at 25 °C
- What that band admits
- Taking −32.0 as enantiopure, a reading at the −29.0 limit implies ee = 29.0/32.0 = 90.6%, i.e. 4.7% D-carnitine. Taking −31.0 as enantiopure, it implies ee 93.5%, i.e. 3.2% D. Either way, a few per cent passes
- The impurity that is limited
- Impurity A, (E)- or (Z)-4-(trimethylammonio)but-2-enoate (crotonobetaine): not more than 0.5%
- The derivative arithmetic
- L-carnitine L-tartrate (2:1) is (2 × 161.20) ÷ 472.49 = 68.2% L-carnitine. Acetyl-L-carnitine HCl is 161.20 ÷ 239.70 = 67.3% L-carnitine equivalent
What the monograph actually measures
Levocarnitine is a betaine: a permanently charged quaternary ammonium with a carboxylate. That structure is what makes the Ph. Eur. assay possible in the form it takes — you dissolve the sample in a mixture of anhydrous formic acid and acetic acid, add crystal violet, and titrate with 0.1 M perchloric acid until the colour goes from violet to green. The endpoint counts basic centres.
D-carnitine has exactly the same basic centres. It titrates identically, gram for gram. So the assay result — "98.0 per cent to 102.0 per cent (anhydrous substance)" — is a measurement of total carnitine. A drum of pure racemate would pass it with a perfect number. This is not a flaw in the monograph; a monograph is a system of tests, and the chirality is meant to be caught elsewhere. But it does mean that the headline number on a carnitine COA is not the number a buyer thinks it is.
Identity is likewise partly achiral: infrared absorption compares the spectrum to the reference, and enantiomers give identical IR spectra in solution. The related-substances test does real work — it limits impurity A, crotonobetaine, the (E)/(Z)-4-(trimethylammonio)but-2-enoate that arises from dehydration, to not more than 0.5% — but crotonobetaine is achiral, so this test too is silent on enantiomeric composition.
Which leaves one test carrying the entire chiral load: the specific optical rotation.
How much D-carnitine fits inside the rotation band
The monograph specifies −29.0 to −32.0 on the anhydrous substance, determined on solution S at 25 °C. Optical rotation scales linearly with enantiomeric excess, so the arithmetic is direct once you decide what value corresponds to enantiomerically pure material — and that is the assumption to make explicit, because the monograph does not state it.
Take the top of the band, −32.0, as the enantiopure value. A batch reading exactly at the lower limit, −29.0, has an enantiomeric excess of 29.0 ÷ 32.0 = 0.906. Enantiomeric excess of 90.6% means the minor enantiomer is (100 − 90.6) ÷ 2 = 4.7%. Take a more conservative reference of −31.0 instead, and the same −29.0 reading implies ee 93.5% and 3.2% D-carnitine.
So the honest statement is a range: a batch sitting at the bottom of the pharmacopoeial band carries somewhere in the region of 3% to 5% of the D-enantiomer, and it is fully compliant. On a 1,000 kg order that is 30 to 50 kg of a substance that is not what the monograph is named after. Whether that matters is your call — but it should be a decision rather than a surprise.
Two practical cautions on the measurement itself. Rotation is sensitive to concentration, solvent, temperature and water content, which is why the monograph pins all four ("solution S", 25 °C, anhydrous basis); a rotation figure quoted without those conditions is not comparable to anything. And the band is 3 units wide on a value of about 31, so the measurement's own reproducibility is a meaningful fraction of the specification — do not read a single decimal place as a fine discrimination.
The chiral method to ask for, and the derivatives to check
If a few per cent is not acceptable for your application, the rotation band will not get you there and you need a stereospecific method written into the specification. Chiral HPLC after derivatisation with a chiral reagent is the established approach for this molecule; a method using (+)-FLEC derivatisation for the determination of low amounts of d-carnitine in l-carnitine has been published in the Journal of Pharmaceutical and Biomedical Analysis, and it is a reasonable thing to name when you ask a supplier what they can run. Specify a numerical limit for D-carnitine and require the chromatogram.
Then do the salt arithmetic separately, because it compounds with everything above. L-carnitine L-tartrate is the 2:1 salt: two L-carnitine at 161.20 plus one L-tartaric acid at 150.09 gives 472.49, of which 322.40 is L-carnitine — 68.2%, which is exactly the 68% / 32% split a good datasheet declares. Acetyl-L-carnitine hydrochloride is 239.70 (acetyl-L-carnitine 203.24 plus HCl 36.46), of which the L-carnitine moiety is 161.20 — 67.3% L-carnitine equivalent.
Note what happens if you stack the two effects. A tartrate batch that is compliant on rotation but sitting at the bottom of the band delivers 68.2% carnitine of which up to about 5% is the D-enantiomer — so the L-carnitine actually present is around 64.9% of the powder rather than the 68.2% the salt arithmetic promises. Neither number on the certificate is wrong. The product of the two is what you are buying.
Why pharmacopoeias name the levo form at all
It is worth being clear about what this page is and is not saying. Ph. Eur. 1339 is titled Levocarnitine and defines the article as the (3R) enantiomer. The US Pharmacopeia carries a levocarnitine monograph on the same basis. The regulatory framework treats the L-form as the article of commerce, and the D-form as something the specification is supposed to exclude.
That is a statement about what the specified substance is. This page makes no claim about the physiological behaviour of either enantiomer, and a raw-material page is not the place for one. The commercial point stands on its own: you are contracting for levocarnitine, the monograph's own assay cannot verify that you received levocarnitine rather than carnitine, and the one test that can is calibrated loosely enough to let a few per cent through.
The general lesson repeats across this product line. Where a molecule has a defined stereochemistry and the routine assay is achiral, the enantiomeric purity is a separate purchase decision that has to be written down or it does not exist. The same structure applies to ergothioneine, where the regulator has named the enantiomer, and to theanine, where the shortfall has been measured in commercial products.
Frequently asked
Does a 99% assay on an L-carnitine certificate prove it is the L-enantiomer?
No. The Ph. Eur. 1339 assay is a non-aqueous titration with 0.1 M perchloric acid using crystal violet indicator, and D-carnitine titrates identically to L-carnitine. The 98.0-102.0% figure constrains total carnitine content on the anhydrous substance and carries no information about enantiomeric composition.
How much D-carnitine can pass the European Pharmacopoeia specification?
The monograph's only chiral test is specific optical rotation, specified as -29.0 to -32.0 on the anhydrous substance at 25 degrees Celsius. A batch reading at the -29.0 limit corresponds to an enantiomeric excess of 90.6% if -32.0 is taken as enantiopure, which is 4.7% D-carnitine; or 93.5% excess and 3.2% D if -31.0 is taken as the reference. Roughly 3-5% can pass while remaining compliant.
How do I verify L-carnitine enantiomeric purity properly?
Specify a chiral method with a numerical limit for D-carnitine and require the chromatogram. Chiral HPLC after derivatisation with a chiral reagent is the established route for this molecule; a method using (+)-FLEC derivatisation for determining low levels of d-carnitine in l-carnitine has been published in the Journal of Pharmaceutical and Biomedical Analysis. Optical rotation alone is a screening test, not a limit test.
How much L-carnitine is in L-carnitine tartrate and in acetyl-L-carnitine HCl?
L-carnitine L-tartrate is the 2:1 salt with a formula weight of 472.49, of which two L-carnitine units at 161.20 make 322.40 - that is 68.2%, matching the 68%/32% split usually declared. Acetyl-L-carnitine hydrochloride has a formula weight of 239.70, of which the L-carnitine moiety at 161.20 is 67.3% L-carnitine equivalent.
This page describes raw-material sourcing and specification. It is not a health claim and not medical advice. Last updated 2026-08-07.