Stable isotope-labelled standards and when to use them

What a labelled internal standard buys you

In quantitative LC-MS and GC-MS, the largest sources of error are usually not the calibration curve. They are matrix effects (ion suppression or enhancement from the sample), losses during extraction and clean-up, and injection-to-injection variability.

A stable isotope-labelled analogue of the analyte behaves almost identically to the analyte through extraction, chromatography and ionisation, and is distinguished only by mass. Adding a known amount before sample preparation therefore corrects for all three at once. This is why isotope dilution is the reference approach for methods where accuracy matters.

It works because the labelled compound is a surrogate for the analyte. That gives the first and most important rule below.

Choice of isotope and number of labels

Deuterium is the common and inexpensive choice; carbon-13 is the more robust one.

• Deuterium (D) is cheapest and available for most compounds. The risk is exchange: a deuterium on an oxygen or nitrogen, or on a carbon adjacent to a heteroatom, can exchange with protons in protic solvents or under acidic or basic conditions. Labels placed on aromatic rings or other non-exchangeable positions do not have this problem.
• Carbon-13 does not exchange. It is more expensive and fewer compounds are commercially available, but for methods that run at extreme pH or in strongly protic media it removes an entire failure mode.
• Nitrogen-15 is used where the analyte already contains nitrogen and a 13C label is not available or would be too costly.

Mass shift. The labelled standard should be separated from the analyte by at least three mass units, so that the analyte's natural isotope envelope — the M+1, M+2 contributions from natural 13C, 15N, 34S and so on — does not overlap the labelled signal. For small molecules a shift of three is usually sufficient; for larger molecules with many carbons the envelope is wider and a shift of four or more may be needed. A shift of one or two is generally a design error, not a saving.

Position of the label. For most single-stage quantitative work the position is not critical. Where the method uses a fragment ion for quantitation, the label must survive in that fragment — a label that is lost in the transition contributes nothing to the measurement. This is worth checking against the MS/MS conditions rather than assuming.

What to specify when ordering

Three different purities are in play, and a specification that names only one of them is incomplete:

1. Chemical purity — the percentage of the labelled compound among all material present. Reported by the usual chromatographic methods.
2. Isotopic enrichment — the proportion of molecules carrying the intended number of labels, usually stated as atom % D or atom % 13C. A nominally “d3” standard will normally contain d2, d1 and unlabelled species as well.
3. Isotopic distribution — how the enrichment is spread across the isotopologues. Where the unlabelled species is present at a level that matters relative to the concentration of analyte being measured, it contributes directly to the blank.

Also specify: the labelled positions, the salt or solvate form, whether the material is supplied neat or as a solution (and if a solution, the solvent, concentration, and whether the concentration is gravimetric or verified analytically), and the storage conditions.

A certificate for a labelled standard should state the enrichment and the distribution, not only a chromatographic purity — a 98 % pure material that is only 90 % enriched is a different product from one that is 98 % pure and 99 % enriched.

Traps in quantitative work

• The labelled standard cannot quantify itself. Isotope dilution quantifies the unlabelled analyte against the labelled surrogate, using a response factor determined from standards of known composition. Calibrating the labelled compound against itself proves nothing.
• Label instability in your matrix. Exchange can happen during sample preparation, not only during storage. Where the method uses elevated temperature, extreme pH or prolonged aqueous steps, verify that the label survives the procedure — a recovery experiment with the labelled standard alone will show it.
• Retention time shift. Deuterated analogues can elute slightly earlier or later than the analyte, particularly on reversed-phase systems. The shift is usually small but must be established, especially where the method relies on a narrow retention window or on co-elution.
• Ion suppression difference. Highly deuterated compounds can differ slightly from the analyte in ionisation behaviour. The assumption of identical behaviour is good, not perfect — for a fully validated method the response factor is established experimentally rather than assumed to be one.
• Cross-contamination in the laboratory. Labelled standards are used at low concentrations and are persistent. Glassware, injectors and columns used for them should be kept separate from those used for the unlabelled analyte, or the blank will drift upward over time.

Storage

Labelled standards generally follow the storage requirements of the unlabelled analyte, with two additions. Protect against exchange — keep the material dry, avoid strongly protic or extreme-pH conditions, and prefer non-aqueous stock solutions where the chemistry allows. And limit repeated freeze–thaw cycles of solutions, since each cycle is an opportunity both for degradation and for concentration change through evaporation of the solvent.