Capillary electrophoresis as an orthogonal purity method for synthetic peptides

Reversed-phase HPLC is the default purity method on nearly every synthetic peptide certificate of analysis. It is robust, well understood, and it resolves the impurity classes that dominate solid-phase synthesis — deletion sequences, truncations, incompletely deprotected side chains. What it does not do reliably is separate impurities that differ from the parent peptide in charge but not in hydrophobicity. A deamidated asparagine adds a negative charge at neutral pH while changing the hydrophobic surface almost imperceptibly. A succinimide intermediate, an isoaspartate rearrangement product, a scrambled disulfide isomer — each of these can and frequently does co-elute with the main peak under standard gradient conditions.

Capillary electrophoresis separates on an unrelated physical basis: the ratio of net charge to hydrodynamic size. Two methods operating on unrelated mechanisms will not, in general, conceal the same impurities. That is the whole argument for running CE alongside RP-HPLC, and it is why characterization expectations for peptides increasingly reference orthogonal purity assessment rather than a single chromatographic number.

The separation mechanism, and why it differs

In capillary zone electrophoresis, the sample is injected into a narrow fused-silica capillary — typically 25 to 75 µm internal diameter — filled with a background electrolyte, and a potential of 10 to 30 kV is applied across its length. Each solute migrates at a velocity set by its own electrophoretic mobility, which scales roughly as net charge divided by hydrodynamic radius raised to a power between one-half and one, superimposed on the bulk electroosmotic flow generated by the charged capillary wall.

The consequence for peptides is that migration order is governed by intrinsic charge characteristics and Stokes radius, not by interaction with a stationary phase. Published comparisons make the contrast explicit: in RP-HPLC on bonded n-alkylsilica, peptides are retained through hydrophobic and silanophilic solute–sorbent interactions, whereas in capillary electrophoresis selectivity arises from electrophoretic migration according to intrinsic charge and differences in Stokes radii. Because purification of synthetic peptides is itself usually performed by preparative RP-HPLC, an orthogonal technique is not a redundancy — it interrogates precisely the dimension along which the purification step was blind.

A practical corollary is that CE has no stationary phase at all. There is no column to foul, no carryover from a strongly retained impurity, and no bonded-phase batch variability. There is instead a capillary surface whose charge state must be controlled, which turns out to be the dominant method-development problem.

Buffer pH is the primary selectivity lever

The most important method variable in peptide CZE is the pH of the background electrolyte, because pH sets both the net charge of the analyte and the magnitude of electroosmotic flow. Moving the buffer pH from above to below a peptide’s isoelectric point reverses the sign of its net charge — a principle codified in the harmonized pharmacopeial chapters on the technique.

Two operating regimes dominate synthetic peptide work.

Low-pH CZE, roughly pH 2.0 to 3.0. Phosphate or citrate buffers in this range protonate the silanol groups on the capillary wall, suppressing electroosmotic flow toward zero, while the peptide carries net positive charge from a protonated N-terminus plus lysine, arginine, and histidine side chains. Separation becomes almost purely electrophoretic. This is the workhorse regime for basic and neutral peptides, and it is where deamidation products resolve well, since the introduced carboxylate still registers through the loss of a positive contribution.

Higher-pH CZE, roughly pH 7.0 to 9.0. Borate or Tris buffers with substantial electroosmotic flow. More useful for acidic peptides and for cases where the analytically relevant modification changes charge near neutrality. Wall adsorption of basic peptides becomes a serious problem in this regime, which is why coated capillaries — polyacrylamide, polyvinyl alcohol, or dynamic coatings applied through the buffer itself — are common at higher pH.

For peptides with poor aqueous solubility or a tendency to associate, organic modifiers in the background electrolyte are well established. Work on synthetic peptides in organic/aqueous buffer systems has shown that acetonitrile, methanol, and trifluoroethanol additions change both bulk viscosity and analyte solvation, shifting selectivity in ways not always predictable from the aqueous separation. Trifluoroethanol in particular can stabilize helical conformation and thereby alter hydrodynamic radius — a selectivity handle with no clean RP-HPLC analogue.

What CE detects that chromatography tends to miss

Several impurity classes make the orthogonality argument concrete.

Deamidation and isoaspartate formation. Asparagine deamidation proceeds through a succinimide intermediate to a mixture of aspartate and isoaspartate. All three products differ from the parent in charge. The isoaspartate isomer is notoriously difficult to resolve chromatographically because it is isobaric with aspartate and nearly identical in hydrophobicity, yet CZE at low pH frequently separates what a shallow acetonitrile gradient does not.

Positional and sequence isomers. Studies on isomeric hybrid peptides — sequences of identical composition and identical mass differing only in residue order or linkage position — have used capillary electrophoresis specifically because the isomers were not separable by other routine means. Mass spectrometry alone cannot distinguish isobaric isomers without fragmentation, and RP-HPLC often cannot separate them at all.

Charge variants generally. Charge-variant characterization has drawn increasing analytical attention because minor modifications generate acidic or basic species that affect batch-to-batch consistency and stability behavior. Methionine or tryptophan oxidation, N-terminal pyroglutamate formation, C-terminal amidation failures, and residual counterion heterogeneity all fall into this category, and all shift electrophoretic mobility.

Aggregation-adjacent species. Soluble dimers and small oligomers migrate with altered charge-to-size ratios. CE does not replace size-exclusion chromatography here, but a mobility shoulder is often the earliest signal that a lot has begun to associate.

Quantitation caveats that matter on a certificate

Two features of the technique complicate direct comparison of a CE area-percent number to an HPLC area-percent number, and both are frequently misread.

First, CE peak areas must be corrected for migration time. In chromatography every solute traverses the detector at the same linear velocity, so peak area is proportional to mass. In electrophoresis solutes pass the detection window at different velocities, so a slower-migrating species spends longer in the detection volume and produces a proportionally larger area for the same quantity of material. The standard correction divides raw area by migration time. An uncorrected CE purity value is systematically biased, and the direction of that bias depends on whether the impurity migrates before or after the main peak.

Second, detection sensitivity is limited by path length. UV detection across the internal diameter of a 50 µm capillary gives an optical path two orders of magnitude shorter than a conventional HPLC flow cell, and concentration limits of detection are correspondingly poorer. Bubble cells, Z-cells, and in-capillary preconcentration approaches address this — work on the decapeptide triptorelin demonstrated that CE coupled to mass spectrometry, with multisegment injection and in-capillary preconcentration, reaches sensitivity adequate for both pharmaceutical and biological matrices. A routine CE-UV method, however, will not see a 0.05% impurity that a well-optimized HPLC-UV method reports without difficulty.

The consequence is that CE and RP-HPLC purity figures for the same lot are not expected to agree, and a discrepancy is not by itself evidence that either method is wrong. What the comparison supplies is a coverage argument: an impurity invisible to one mechanism is likely visible to the other. The separate and often conflated question of how area-percent purity relates to actual peptide content is not addressed by either method, and requires amino acid analysis or an equivalent absolute technique.

Compendial status and validation expectations

Capillary electrophoresis is a compendial technique. The general chapter is harmonized across the major pharmacopeias — Ph. Eur. 2.2.47, the corresponding USP general chapter, and the Japanese Pharmacopoeia equivalent — with the USP text having become official on 1 August 2020 following approval by the relevant expert committee. A separate USP general chapter addresses capillary electrophoresis for biotechnology-derived articles.

Compendial status means a CE purity method appearing on a certificate of analysis is subject to the same validation framework as any other analytical procedure: specificity, linearity, range, accuracy, precision, detection and quantitation limits, and robustness, under the ICH Q2 framework. Robustness deserves particular attention here. Capillary surface condition drifts with use; buffer depletion at the electrodes shifts pH across a sequence; ambient temperature affects viscosity and therefore mobility. A CE method not demonstrated robust across capillary lots and conditioning protocols will produce migration times that wander enough to complicate peak identification, which is a specificity problem dressed as a precision problem.

For research-grade material, a CE result on a certificate is worth reading carefully rather than treating as a second opinion on the HPLC figure. Four questions determine whether it carries orthogonal information: was the area corrected for migration time, what was the background electrolyte and its pH, was the capillary bare or coated, and what independent identity evidence supports the main-peak assignment.

Where the two methods leave the picture

Neither technique alone characterizes a synthetic peptide. RP-HPLC resolves the hydrophobicity-differentiated impurities that dominate crude synthesis output. CE resolves charge-differentiated species that survive preparative purification precisely because that purification was hydrophobicity-based. Mass spectrometry establishes identity and catches mass-shifted modifications neither separation resolves, and amino acid analysis fixes absolute content independent of both. The characterization literature on therapeutic peptides has converged on this point: purity framed as a single number from a single method describes one projection of a multidimensional problem.

The most informative disagreement between the two methods tends to be a narrow one — a CE purity a percentage point or two below the chromatographic value, traceable to a defined charge variant. That is orthogonality working as intended. A CE result that matches the HPLC result across many lots is the less reassuring outcome: it more often indicates a method that has not been optimized to see anything the chromatography missed than a lot with nothing left to find.