A weighed portion of lyophilised peptide is mostly, but not only, peptide. Some of the mass is water, some is residual solvent, some is chromatographic impurity, and for any sequence carrying basic residues a substantial fraction is counterion. In the theoretical mass-balance example set out in a report on the preparation of peptide reference standards, the sum of chromatographic impurities came to 0.92% of total detected area, while acetic acid alone contributed 5.58% w/w, trifluoroacetic acid 0.003% w/w, residual solvents 0.003% w/w and residue on ignition 0.00%; the resulting purity was 0.93 mg of peptide free base per mg of material on an anhydrous basis [1]. The counterion term was roughly six times the chromatographic impurity term. That is the practical case for measuring it: in that example it was the largest of the non-peptide terms standing between a number on a balance and a number in a molarity calculation.
Where the counterion comes from, and why the number is not inferable
Solid-phase peptide synthesis relies on trifluoroacetic acid as the cleavage agent and again as the ion-pairing reagent during reversed-phase purification, with the result that peptides are obtained as trifluoroacetate salts [2]. The trifluoroacetate in the vial is therefore an artefact of how the material was made and purified, not a property of the sequence. It is also not necessarily bound: products of Fmoc solid-phase synthesis carry significant amounts of trifluoroacetate ions in the form of either a counter-ion or an unbound impurity [3]. No assay described below separates those two populations. What you get is a total.
Counting basic residues gives you an upper bound and nothing more. The mass-balance approach used for peptide reference standards requires that all species other than the native peptide — related impurities, counter ion, water, residual solvents, non-combustible residues — be identified and measured and subtracted from 100% [1]. In a candidate certified reference material for angiotensin II, a validated 19F-qNMR method measured the trifluoroacetic acid counter-ion at nearly 25% by mass, and the final assigned purity of that material was 691 ± 9 mg/g (k = 2) [4]. A solid can be clean by peak area and still be close to a third non-peptide by mass.
What suppressed conductivity detection actually measures
Ion chromatography with conductivity detection is probably the most frequently employed method for determining counter-ions in pharmaceutical salts; the alternatives set out in the same review are capillary electrophoresis, which uses indirect UV detection because small ions lack chromophores, and hydrophilic interaction chromatography with evaporative light-scattering or refractive index detection for non-UV-absorbing ions [5].
The suppressor is what makes conductivity usable. In a validated anion method for organic counterions in small-molecule drug substances, an anion-exchange separation was followed by a cation-exchange resin suppressor performing chemical suppression: background conductivity is reduced and all cations from the mobile phase are replaced by H+, so that a high-conductivity eluent is converted to water and carbon dioxide, which have low conductivity [6]. (Method details in this paragraph and in the sections below are from that paper's full text, which is openly accessible.) The detector responds to ions, not chromophores — in that study the free bases of the four drug substances, being non-ionic, gave no response at all in ion chromatography with conductivity detection [6].
The anion run
For trifluoroacetate specifically, the documented difficulty is not sensitivity but a matrix of competing anions. A method using a high-capacity anion-exchange column separated trace trifluoroacetate from an excess of chloride, phosphate and other anions without the need for sample preparation, detected it by suppressed conductivity, and used four potassium hydroxide eluent step changes executed by automated electrolytic eluent generation; method detection limits in phosphate-buffered saline, in an acetate-buffered saline containing protein, and in a commercial peptide were all below 90 ng/mL [7]. In a separate study on a phosphate-buffered saline cell-wash matrix, anion-exchange chromatography with suppressed conductivity reached residual trifluoroacetic acid down to 50 ng/mL in the presence of high concentrations of phosphate and chloride, and improved its sensitivity further by selectively removing halide anions on a silver cartridge before injection [8]. When a halide excess is the limiting interference, taking it out of the sample is the documented fix rather than trying to chromatograph through it.
For synthetic peptides in particular, three techniques were compared head to head for acetate, trifluoroacetate and chloride: ion chromatography gave the best results; capillary electrophoresis was comparable except for limit of detection and limit of quantification; isotachophoresis was judged least useful because of the difficulty of determining chloride, although all three handled trifluoroacetate and acetate satisfactorily [9].
Specificity has to be demonstrated against the anions that could co-elute. The validated small-molecule method above established specificity in the presence of eleven anions — trifluoroacetate, chloride, nitrate, bromide, phosphate, sulphite, succinate, tartrate, sulphate, oxalate and fumarate — with the four target counterions resolved from each other and no interference from the blank, the other anions, or the drug substances themselves [6]. Those analytes were small-molecule active ingredients, not peptides; the list of interferences is what transfers, not the retention times.
The cation run
The cations are a second injection on a different column, and skipping them is a real source of unexplained mass. A mixed-mode HPLC method with evaporative light-scattering detection quantified 14 positively and negatively charged counter ions simultaneously within 30 minutes, validated for specificity, accuracy, precision, limit of quantification, sample stability and carry over as proposed by ICH; applied to synthetic peptides, it found differences from the expected concentrations of prepared peptide solutions and an unexpectedly high content of sodium [10]. An anion-only survey cannot see that.
Cation methods have their own selectivity trap. Sodium and ammonium have similar selectivity on stationary phases carrying the commonly used sulfonate or carboxylate functional groups, which makes samples with very different ammonium-to-sodium ratios hard to quantify; a column combining carboxylate and phosphonate groups with a crown ether greatly increased the sodium–ammonium separation, allowing low levels of one to be determined in the presence of high levels of the other under isocratic conditions, with an eluent step change or gradient needed above roughly 4000:1 sodium-to-ammonium [11].
Sample preparation and landing in the linear range
The dilution is the part most often got wrong, because the target is the counterion concentration, not the peptide concentration. In the validated small-molecule method, standards of the four free acids were prepared at 20 µg/mL, and each test solution was weighed out to a different concentration — 134, 100, 66 and 66 µg/mL — set against the theoretical counterion content of each salt (15.13%, 21.24%, 18.48% and 31.27% w/w respectively) so that the counterion arrived at roughly the standard concentration. The diluent was water–acetonitrile 80:20, the injection loop 20 µL, and linearity was established over seven levels from 25 to 200 µg/mL, with accuracy assessed from 50% to 150% of the specification level. Standard and sample solutions and the mobile phase were stable for 48 hours, and the method ran on 10–25 mg of material [6].
Do the arithmetic before you weigh. Take a peptide of free-base molar mass 3,000 g/mol with three basic sites, each assumed to carry one trifluoroacetate: the salt is 3,000 + (3 × 114.02) = 3,342.06 g/mol, so a 1.00 mg/mL solution of the solid is 0.299 mM in salt and 0.898 mM in trifluoroacetate, which at 113.02 g/mol for the anion is about 101 µg/mL. That sits comfortably inside a 25–200 µg/mL calibration window — but only because the assumed occupancy happened to be right. If only half the sites are occupied you are near the bottom of the curve, and a 10 mg/mL stock injected neat is five times over the top of it. This arithmetic is derived here from molar masses and an assumed stoichiometry; it is not a measurement, and its whole point is that you should bracket it with a dilution series until a measured value lands mid-range.
Solubility sets a second limit. Compared with HILIC coupled to charged aerosol detection, ion chromatography imposes much more restriction on the allowable organic percentage of the eluents, which makes HILIC/CAD more appropriate for counterions of poorly water-soluble drug substances [12]. Some organic is tolerated — the anion method above ran 10% acetonitrile in the eluent and 20% in the diluent [6] — but a hydrophobic peptide that needs a high organic fraction or DMSO to dissolve will hit this constraint before it hits any detection limit.
System suitability for a small anion
The robustness data from that validated method give usable acceptance criteria. Flow rate was varied by ±0.1 mL/min, acetonitrile by ±2%, sodium carbonate by ±0.75 mM and sodium bicarbonate by ±0.2 mM; across all of these the elution order was unchanged and all analytes stayed adequately resolved, resolution among the four organic anions remained greater than 1.5, standard relative standard deviation stayed below 2.5%, and the variation in reported content stayed below 0.41% [6]. A suitability test built from that is three checks: resolution against the nearest-eluting anion, replicate-injection RSD on the standard, and retention-time consistency.
For a small anion the nearest neighbour is the whole problem, and which neighbour matters depends on the sample. Trifluoroacetate has to be resolved from chloride and phosphate when those are in large excess [7, 8]; acetate has to be resolved from the other anions a peptide preparation is liable to carry, trifluoroacetate and chloride among them [9]. Carry-over deserves an explicit check rather than an assumption — it was one of the validation parameters in the mixed-mode counterion method [10] — because a strongly retained anion injected at high concentration is exactly the analyte that will reappear in the next blank.
The alternatives, and what each is blind to
19F NMR is the specific answer for trifluoroacetate. Capillary electrophoresis and 19F NMR methods were both developed for determining the TFA content of drug discovery compounds, which are often isolated as trifluoroacetate salts from preparative HPLC, and both were applied in a high-throughput fashion [13]. A validated 19F-qNMR method supplied the counter-ion term in the angiotensin II mass balance [4]. In a study comparing three routes for displacing trifluoroacetate from the dicationic octapeptide lanreotide, 19F-NMR, 1H-NMR and attenuated total reflectance FT-IR were used together, and ATR-IR was shown to work for monitoring removal as it proceeds [14]. The blindness is structural: a 19F experiment sees fluorine, so it cannot report acetate, chloride or sodium, and cannot close a mass balance on its own.
Capillary electrophoresis with indirect UV detection and a common internal standard determines acetic acid and trifluoroacetic acid in a single assay, with migration times typically 2 to 8 minutes; that method was validated for precision, linearity, accuracy, specificity, limit of detection and ruggedness, and applied to two synthetic opioid peptide samples [15]. Against ion chromatography on peptide counterions, capillary electrophoresis was comparable except in detection and quantification limits [9].
HILIC with charged aerosol detection quantitates anions and cations simultaneously without changing column, eluent or detection mode, showed typical precision below 2.0% for the inorganic counterions tested with determinations within 2.0% of the theoretical amount, and gave better accuracy for chloride in several drug substances than ion chromatography did; its stated drawback is that silica-based HILIC columns dissolve slightly and raise the detector baseline noise [12]. Note the analyte class — that comparison was for Cl−, Br−, SO42−, K+, Ca2+ and Mg2+ in drug substances, not for trifluoroacetate on a peptide, so it does not overturn the head-to-head peptide comparison in which ion chromatography came out best [9].
Titration is on the list of common approaches for anions in drug substances — potentiometric titration, ion-selective electrodes and complexometric methods sit alongside the chromatographic and electrophoretic options — and ion chromatography was selected over them in that work for instrument availability, column availability and sensitivity [6]. A titration returns a total for whatever it is selective for, so it can confirm a stoichiometry when the salt form is already known and single, and cannot tell you that half your trifluoroacetate has become acetate. For a peptide of unknown provenance, that is the wrong shape of answer.
Reading the molar ratio
Convert the mass fraction to moles of counterion per mole of peptide and compare it with the count of protonatable sites: the N-terminal amine plus arginine, lysine and histidine side chains. A recent systematic study used its measured counterion contents specifically to examine how sequence differences, positive charges in particular, relate to the amount and the potential localisation of counterion — localisation being offered there as a possibility, not a determination [2].
A ratio well above the site count means acid that is not charge-paired: unbound trifluoroacetic acid retained in the solid, which is the second population that the same anion measurement counts [3]. A ratio well below it means either that the sites are not all occupied or that another anion is present, and the anion suite is what distinguishes those. A ratio near zero on the anion channel with mass still unaccounted for points to the cation channel — the unexpectedly high sodium found in synthetic peptides is the concrete example [10]. And a ratio that moves after an exchange step is the readout for whether that step worked: exchange onto an ion-exchange resin or by re-running reversed-phase gave partial to almost complete displacement of trifluoroacetate for lanreotide, while a deprotonation–reprotonation cycle of the amino groups permitted complete removal [14].
What the number licenses, and what it does not
It licenses a corrected formula weight and therefore a corrected molarity; it licenses a completeness check on a counterion exchange; and it licenses a stated counterion figure to travel with the material into whatever assay follows, which is the explicit rationale given for quantifying TFA in synthetic peptides — content assessment and the interpretation of functional assay results [10].
It does not license a purity figure. Mass balance needs the chromatographic impurities, the water and the residual solvents as well, and water content in particular is determined at the time the sample is prepared for analysis because it varies with laboratory humidity [1]. It does not license a claim about which residue holds which ion. And it does not, by itself, tell you whether the counterion is doing anything.
On that last point there is a specific published observation to hold in view. Trifluoroacetate at 10−8 to 10−7 M reduced cell numbers and thymidine incorporation in fetal rat osteoblast cultures at 24 hours, with similar effects in articular chondrocyte cultures and in neonatal mouse calvariae; when the trifluoroacetate and hydrochloride salts of the same three peptides were compared in osteoblasts, proliferation was consistently lower with the trifluoroacetate salts, enough that a real proliferative effect could be missed or an antiproliferative effect wrongly attributed [16]. Those authors judged the finding likely to be relevant to purified peptides above 10−9 M in whatever cell or tissue type — an extrapolation from the systems they tested, stated as a likelihood. More recently, differences in measured passive membrane permeability coefficients were found to depend on both peptide sequence and salt form, and the authors of that work make counterion quantification and specification in assays with synthetic peptides their closing recommendation [2]. Neither result tells you what your own counterion is doing. Both are reasons to know the number before you argue about the result.