Which Analytical Method Proves Identity and Purity?

ICH Q6A, the harmonised guideline on specifications, states it without hedging: “Identification solely by a single chromatographic retention time, for example, is not regarded as being specific”. One peak eluting at the expected minute is not an identity test. It is one piece of circumstantial evidence, and the guideline goes on to say that two chromatographic procedures separating on different principles, or a combination such as HPLC with diode-array detection, HPLC-MS or GC-MS, is generally acceptable instead.

That sentence is the whole subject in miniature. Which analytical method proves identity and purity is the wrong question in the singular; the right question is which method answers which sub-question, and what each one cannot see. This article maps the techniques onto the questions, with the deeper treatment of individual methods handled separately in this category.

Which analytical method proves identity, and which proves purity

Identity and purity are different measurements with different failure modes. Identity asks: is this the substance it claims to be? Purity asks: how much of the material is that substance, and what is the rest? A technique can be excellent at one and useless at the other.

Technique	Question it answers	What it cannot see
HPLC with UV detection	How many separable components, in what proportion of total detector response	Anything without a chromophore; co-eluting species; the true mass ratio
GC-MS	Volatile components, including residual solvents, with a fragmentation fingerprint	Non-volatile and thermally labile material
NMR (structural)	The connectivity and environment of every hydrogen and carbon	Trace-level impurities below its detection capability; anything lacking the observed nucleus
qNMR	Absolute content against a certified internal standard	Impurities whose signals overlap the integrated region
Mass spectrometry	Accurate mass, isotope pattern, elemental composition	Isomers; anything that does not ionise under the chosen conditions
FTIR	Functional groups and a whole-molecule fingerprint against a reference	Quantitative composition; low-level impurities
XRPD	Which crystalline form the solid is in	Chemical identity of the molecule; anything amorphous in a crystalline matrix, below its limits
Karl Fischer titration	Water, specifically	Any other volatile; organic impurities

Read down the third column and the reason for orthogonality becomes obvious. The blind spots do not overlap, which is the point.

Why a single retention time is not an identity test

A retention time is a property of a chromatographic system, not of a molecule. It shifts with column batch, temperature, mobile-phase composition and instrument dead volume. More importantly, it is not unique: many structurally distinct compounds elute at the same time under one set of conditions, and closely related analogues are specifically designed to be similar.

ICH Q6A sets the bar deliberately high. Identification testing “should optimally be able to discriminate between compounds of closely related structure which are likely to be present”, and the guideline names infrared spectroscopy as an example of a test that is specific for the substance. Where the substance is a salt, it adds, identification should be specific for the individual ions.

ICH Q2(R2), adopted on 1 November 2023, formalises how that specificity is demonstrated. It offers three routes: showing absence of interference, comparing results with an orthogonal procedure, or – where the physics itself guarantees discrimination – technology-inherent justification. The guideline gives two examples of the third case: “resolution of isotopes in mass spectrometry” and “chemical shifts in NMR spectroscopy”.

What a chromatographic purity number actually measures

A purity figure from HPLC with UV detection is, by default, an area percentage: the area of the main peak divided by the total integrated area, at one wavelength. Three assumptions are buried in that number, and all three can fail.

That every component absorbs. Compounds without a UV chromophore are invisible. Adnani and colleagues, developing universal quantification for structurally diverse natural products, note that evaporative light scattering detection has historically been used precisely for compounds lacking a UV chromophore – a gap that exists because UV response depends on electronic structure, not on how much material is present.
That everything absorbs equally. It does not. Two impurities at the same mass concentration can differ in response by an order of magnitude, so area percent is not mass percent unless response factors have been determined.
That everything eluted. Material retained on the column, or running in the void, or precipitating in the injector, contributes no area and therefore no impurity.

Mass-sensitive detectors – evaporative light scattering and charged aerosol detection – respond to particle mass rather than absorbance, and so see chromophore-free material. Their response is more uniform than UV, but it is not perfectly uniform either; universal calibration against structurally diverse analytes still carries appreciable error compared with a compound-matched standard.

ICH Q3A(R2), the guideline on impurities in new drug substances, is built on this reality. It sets a reporting threshold, an identification threshold and a qualification threshold, and it defines qualification as “the process of acquiring and evaluating data that establishes the biological safety of an individual impurity or a given impurity profile at the level(s) specified”. Those thresholds scale with the maximum daily dose of the substance; the exact bands are tabulated in Attachment 1 of the guideline and are not reproduced here. The structural point is that an impurity above the identification threshold has to be identified, not merely integrated – which chromatography alone cannot do.

What mass spectrometry can and cannot settle

High-resolution mass spectrometry gives an accurate mass, and from it a shortlist of possible elemental compositions. It is tempting to treat that as an identification. Kind and Fiehn tested the assumption directly and titled the resulting paper “Mass accuracy is insufficient even at less than 1 ppm”.

Their finding is that accurate mass alone leaves too many candidate formulas standing as molecular mass rises, and that isotopic abundance patterns are the constraint that collapses the list: “applying such an orthogonal filter above 500 u removes always more than 95 % of false candidates”. Their conclusion is blunter still – an instrument capable of 3 ppm mass accuracy and 2 % error on isotope abundance patterns outperforms one with better than 1 ppm accuracy that ignores isotope information.

Two limits follow. First, a molecular formula is not a structure. Second, and more sharply for anyone comparing a spectrum against a database: isomers share a formula exactly. Mass spectrometry cannot distinguish them on mass, only on fragmentation behaviour or chromatographic separation before the inlet.

Why qNMR and mass balance are cross-checked against each other

Quantitative NMR occupies an unusual position. Signal intensity is proportional to the number of nuclei giving rise to it, so purity can be measured against a certified internal standard that need not be the analyte. That removes the circular dependence of chromatography on having a pure sample of the thing you are trying to measure.

An international collaborative study coordinated across thirteen laboratories tested this. Three samples, each certified by a conventional primary method – mass balance, coulometric titration, and freezing point depression combined with qNMR – were measured by proton qNMR with internal calibration. The measured values were equivalent to the reference labelled values, and the authors concluded that qNMR “has the competence to obtain the same quantification performance and accuracy as the conventional primary methods of measurement”.

Equivalence is not redundancy. Davies and co-workers at Australia’s National Measurement Institute compared qNMR against mass balance across organic calibration standards of varying structural complexity, and their argument for running both is specific: demonstrated agreement between two independent purity values confirms accuracy and minimises “the potential for hidden bias”. A hidden bias is precisely the error a single method cannot report, because the method is blind to it by construction.

The tests that chromatography never sees: water, solvents and solid form

Three attributes of a solid reference material are invisible to a purity chromatogram, and each has its own technique.

Water. ICH Q6A treats water content as its own test, important where a substance is hygroscopic, degraded by moisture, or a stoichiometric hydrate. Loss on drying may be adequate in some cases, but the guideline states that “a detection procedure that is specific for water (e.g., Karl Fischer titration) is preferred”. The difference matters: loss on drying reports everything volatile, while Karl Fischer titration reports water and not, for instance, residual ethanol.

Residual solvents. ICH Q3C(R9), current since 24 January 2024, notes that solvents “are not completely removed by practical manufacturing techniques” and that “there is no therapeutic benefit from residual solvents”. It sorts them into three classes by hazard: Class 1 solvents to be avoided, being known or strongly suspected human carcinogens and environmental hazards; Class 2 to be limited, covering non-genotoxic animal carcinogens and agents of other irreversible toxicity; Class 3, of low toxic potential, for which the guideline states no health-based exposure limit is needed. On method, it is direct: residual solvents “are typically determined using chromatographic techniques such as gas chromatography”, with a non-specific method such as loss on drying acceptable when only Class 3 solvents are present and the method is properly validated.

Solid form. Polymorphs are the same molecule in different crystal packings, and they can differ in physical properties while being chemically indistinguishable. ICH Q6A lists the techniques used to establish whether multiple forms exist: melting point including hot-stage microscopy, solid-state infrared, X-ray powder diffraction, thermal methods such as DSC, TGA and DTA, Raman spectroscopy, optical microscopy and solid-state NMR. An HPLC assay dissolves the sample and therefore destroys exactly the information these methods exist to capture.

Orthogonality is a regulatory expectation, not a preference

The word “orthogonal” appears throughout ICH Q2(R2) as a defined tool rather than a stylistic flourish. Specificity can be verified “by demonstrating that the measured result of an analyte is comparable to the measured result of a second, well characterised analytical procedure that ideally applies a different measurement principle”. For accuracy, orthogonal procedures “can be used with quantitative impurity measurements to verify primary measurement values” in cases where spiking studies are impossible because the relevant impurities cannot be obtained.

That last clause describes the ordinary situation for a reference material. If an impurity cannot be isolated, it cannot be spiked, and a single method has no way to demonstrate that it would have detected it. The second method, working on a different physical principle, is the only available check.

The practical form of this is a small set of tests chosen so their blind spots do not coincide: a separation for the profile, a spectroscopic method for identity, a mass-independent method for content, a specific determination of water, a volatiles method for solvents, and – for a crystalline solid where form matters – a diffraction pattern. A certificate of analysis that reports one number from one instrument is not describing the material; it is describing one instrument’s view of it.

Which threads this category develops separately

This page is a hub and stops deliberately short. How to read a certificate of analysis line by line, what area percent and mass balance each actually assert, how relative response factors are determined, how a stability-indicating method is developed and stressed, and the individual techniques in operational detail are each treated on their own page in this category. The chemistry of where the compounds themselves come from belongs to the discovery category, not here. Nothing on this page is a recommendation about use; the terms under which these substances are supplied for laboratory work are set out in the site rules.

Frequently asked questions

Why is a single HPLC retention time not proof of identity?

Because retention time is a property of the chromatographic system, not of the molecule, and it is not unique. ICH Q6A states that identification solely by a single chromatographic retention time is not regarded as being specific. The guideline accepts two chromatographic procedures separating on different principles, or combinations such as HPLC-MS or GC-MS, as generally acceptable alternatives.

What does 99 percent purity by HPLC actually mean?

Usually it means area percent: the main peak’s integrated area as a fraction of total detected area, at one wavelength. It is not a mass fraction. The figure assumes every component absorbs at that wavelength, that all components respond equally, and that everything injected eluted. All three assumptions can fail silently.

Why does UV detection miss impurities without a chromophore?

Ultraviolet absorbance depends on electronic structure, not on how much material is present. A compound with no conjugated system or suitable functional group absorbs negligibly at the analytical wavelength and produces no peak, regardless of concentration. Mass-sensitive detectors such as evaporative light scattering or charged aerosol detection respond to particle mass instead, and so register chromophore-free material.

Can mass spectrometry distinguish between isomers?

Not on mass alone. Isomers share an elemental composition exactly, so they share an exact mass. Kind and Fiehn showed that even sub-part-per-million mass accuracy leaves multiple candidate formulas standing, and that isotope abundance patterns remove more than 95 % of false candidates above 500 u. Separating isomers requires chromatography before the inlet, or fragmentation behaviour after it.

What is an orthogonal analytical procedure?

It is a second, well characterised procedure that ideally applies a different measurement principle from the first. ICH Q2(R2) accepts comparison against such a procedure as a way of demonstrating specificity, and as a way of verifying accuracy for impurity measurements when the relevant impurities cannot be obtained for spiking studies. Different principle means different blind spots.

Why can qNMR measure purity without a standard of the same compound?

Because NMR signal intensity is proportional to the number of nuclei producing it, so a certified internal standard of any suitable compound calibrates the measurement. An international collaborative study across thirteen laboratories found proton qNMR results equivalent to values certified by mass balance, coulometric titration and freezing point depression, and concluded it matches conventional primary methods for quantification.

Which method detects residual solvents in a drug substance?

Gas chromatography. ICH Q3C(R9) states that residual solvents are typically determined using chromatographic techniques such as gas chromatography. A non-specific method such as loss on drying is acceptable only when nothing but Class 3 solvents is expected and the method is properly validated. The guideline classifies solvents into three hazard classes.

What does X-ray powder diffraction tell you that HPLC cannot?

Which crystalline form the solid is in. Polymorphs are the same molecule packed differently, so they are chemically identical and chromatographically indistinguishable once dissolved. ICH Q6A lists X-ray powder diffraction alongside solid-state infrared, thermal analysis, Raman spectroscopy and solid-state NMR as the techniques used to determine whether multiple forms exist.

References

Research use only. Nonsensia Lab supplies analytical reference standards for laboratory and research applications. This article is published for scientific and educational purposes. It is not medical advice, it does not describe any use in humans, and nothing in it should be read as a recommendation to administer any substance to a person or animal.

Filed under: Analytical Methods

Nonsensia Lab supplies the compounds discussed in this guide as analytical reference standards for laboratory and research use.

Identity and Purity: A Complete Guide to Analytical Methods for Small Molecules