Purity, Identity, and Net Content

A vial can be simultaneously 99% pure, correctly identified, and contain less than three-quarters of the labeled peptide mass. All three statements are true at once and none of them is false. This is the single most consequential source of confusion in research-material documentation, and it exists because three different questions get collapsed into one figure.

• Purity asks: of the UV-detectable, column-eluting material in this vial, what fraction is the main peak? Answered by RP-HPLC/UPLC-UV, reported as area percent at a stated wavelength.
• Identity asks: is that main peak the compound named on the label? Answered by LC-MS, reported as observed mass against theoretical mass with the error in daltons and ppm.
• Content asks: how many milligrams of the named compound are actually present? Answered by amino acid analysis, qNMR, elemental nitrogen, or a measured mass balance, reported as % w/w and mg per unit.

A lyophilized synthetic peptide vial typically contains 70–90% peptide by mass. The balance is counterion (5–20% w/w for a TFA salt, lower for acetate), water (2–10%, higher for hygroscopic sequences), residual solvents, peptide-related impurities, and — where present — bulking agent. None of the non-peptide fraction absorbs at 214 nm or is integrated into the purity calculation. A vial that is 60% mannitol by mass can report 99% area purity, and that is exactly the mechanism by which diluted material passes a purity-only certificate.

Wavelength. Purity is determined at 214 nm, which targets the n→π* transition of the amide backbone. Every peptide carries that chromophore in proportion to its length — the molar absorptivity of the peptide bond at 214 nm is approximately 923 M⁻¹cm⁻¹ — so response is near-proportional to mass across peptide-related impurities. A deletion impurity missing one residue of n has roughly (n−1)/n of the parent response; on a 30-mer that is about 3% relative, and in the conservative direction.

280 nm is reported as a confirmatory second wavelength only. It reports aromatic side chains and disulfides (Trp ≈ 5,500 M⁻¹cm⁻¹, Tyr ≈ 1,490, cystine ≈ 125; Phe effectively invisible), so a sequence containing no Trp and no Tyr is essentially undetectable at 280 nm, and any purity figure claimed there for such a sequence is meaningless.

Column and mobile phase. C18 on high-purity, fully endcapped Type B silica is the default; 100–130 Å pore size below roughly 3–5 kDa, 300 Å above it, since a 100 Å column on a 10 kDa polypeptide produces mass-transfer-limited tailing. C8 and C4 for strongly hydrophobic and lipidated sequences; phenyl-hexyl or pentafluorophenyl where a suspected impurity co-elutes and genuinely orthogonal selectivity is needed; HILIC for short, highly charged sequences that are unretained in reversed phase.

The quantitative purity method uses 0.1% (v/v) TFA in water against 0.085–0.09% TFA in acetonitrile. The deliberate concentration mismatch compensates for TFA's higher molar absorptivity in acetonitrile than in water; matching them produces baseline drift at 214 nm large enough to corrupt integration of 0.05% impurities. Column temperature is controlled and stated, because selectivity changes with it and an uncontrolled oven is a hidden source of between-day disagreement.

Gradient. Peptide retention is extraordinarily sensitive to organic modifier — under linear solvent strength theory the slope S scales roughly with the square root of molecular mass, so a 3–5 kDa peptide can desorb across a window of 1–3% acetonitrile. Methods are therefore developed in two stages: a broad scouting gradient to locate the analyte and reveal late-eluting species, then a focused gradient of roughly 0.3–1.0% B per minute centred on the elution window. Halving the gradient slope roughly doubles resolving power for critical pairs, and deletion sequences, oxidation products, and diastereomers are exactly such pairs.

Every run includes a wash-and-hold at high organic long enough to elute strongly retained species. Truncating a gradient at the main peak is how residual Trt (+242.11), Pbf (+252.06), Boc (+100.05), tBu (+56.06), and dimeric material get missed entirely.

Loading and integration. There is an unavoidable tension between putting enough mass on-column to see a 0.05% impurity at S/N ≥ 10 and keeping the main peak inside detector linearity. We resolve it with paired injections — a low-load injection for main-peak integration, a high-load injection for impurity detection — rather than by choosing one compromise load. Integration parameters, the reporting threshold, and the disregard limit are fixed in the SOP and printed on the report, because baseline placement and skimming decisions can move a reported purity by a full percentage point on a real sample.

PLACEHOLDER: Arcadia's default reporting and disregard thresholds for peptide impurity profiling — for example "peaks ≥0.05% area reported, peaks <0.05% disregarded". Confirm the actual SOP values before publishing. Note that ICH Q3A(R2)/Q3B(R2) explicitly exclude peptides from scope, so any threshold used is a laboratory convention and client agreement, not a regulatory inheritance, and should be described as such

Purity is a statement about proportion. It is fully compatible with a 99% pure sample of the wrong compound. A certificate that reports HPLC purity without an identity determination has not established what was purified.

Identity is determined by LC-MS under formic-acid conditions (0.1% FA in both phases). Formic acid is a much weaker ion-pairing agent than TFA — expect earlier elution and reduced resolution of closely related impurities — but ESI response improves substantially, since TFA is a well-documented ion suppressor. Running two methods, TFA-UV for the quantitative number and FA-LC-MS for assignment, is the standard resolution of that trade-off, and the purity figure is always taken from the UV method.

Peptides ionize in positive-mode ESI as a distribution of multiply protonated species, (m/z)ₙ = (M + n × 1.00728)/n, where 1.00728 Da is the proton mass and not the hydrogen atom mass. The observed charge envelope is governed largely by the number of ionizable basic sites, so an envelope inconsistent with the sequence's Arg/Lys/His count is itself a signal worth a second look. Acidic sequences with few basic residues may be run in negative mode.

Mass convention is stated explicitly. Monoisotopic and average masses diverge by roughly 0.05–0.07% of molecular mass — about 0.5 Da at 1 kDa, 2.5 Da at 4 kDa, 6 Da at 10 kDa. Unit-resolution instruments cannot resolve the isotope envelope of a multiply charged peptide ion and report average mass; high-resolution instruments report monoisotopic mass with the observed isotope pattern compared against the pattern computed from the elemental formula. Reporting an observed mass without naming the convention is a defect, and mixing conventions produces both spurious mismatches and — worse — spurious matches.

Intact mass confirms elemental composition, not sequence. Leu and Ile are exactly isobaric and cannot be distinguished by any measurement of the intact species. Gln and Lys differ by 0.03638 Da — 36 ppm at 1 kDa, about 9 ppm at 4 kDa — resolvable on a good high-resolution instrument and invisible on a quadrupole. Two Gly and one Asn are isobaric. Sequence permutations, D/L substitutions, and isoaspartate all preserve mass exactly. A rigorous identity determination therefore uses at least two orthogonal attributes: accurate intact mass with isotope-pattern match, plus MS/MS fragment coverage, plus retention-time agreement with an authentic reference standard where one exists.

For most research peptides no certified reference material exists. In that case the report states plainly that identity is consistent with the stated sequence on the basis of accurate mass and MS/MS fragment coverage, and that no certified reference material was available for retention-time comparison. That sentence is more useful to a technical buyer than an unqualified "Identity: Pass."

MS is never used as the purity detector. Ionization efficiency varies by orders of magnitude between structurally related species, and a hydrophobic protected-peptide impurity can ionize far better or far worse than the parent. Total-ion-chromatogram area percent is not a purity figure and is never reported as one. UV quantifies; MS assigns.

Net peptide content is the mass fraction of vial contents that is actually the target peptide, on a w/w basis. Four approaches, in descending order of rigor:

Amino acid analysis (AAA) is the reference method. Acid hydrolysis (6 N HCl, 110 °C, 20–24 h, evacuated or inert atmosphere) followed by separation and quantification of the liberated free amino acids against a certified amino acid standard, with content back-calculated from recovered residue molar amounts. The known caveats are acknowledged on the report rather than buried: Trp is destroyed by acid hydrolysis; Cys and Met partially oxidize unless performic acid oxidation precedes hydrolysis; Ser and Thr degrade 5–10% over 24 h, corrected by extrapolating 24/48/72 h hydrolysates to zero time; Ile-Ile, Ile-Val, and Val-Val bonds hydrolyze slowly and may be incomplete at 24 h; Asn and Gln convert to Asp and Glu and are reported as Asx and Glx. Quantification is therefore based on the stable, well-recovered residues.

Quantitative NMR (qNMR) — ¹H qNMR against a certified internal standard such as maleic acid, dimethyl sulfone, or benzoic acid, in DMSO-d₆ or D₂O. Gives an absolute, SI-traceable w/w assay without a compound-specific reference standard, and simultaneously quantifies residual solvents and, via ¹⁹F, trifluoroacetate. It requires a well-resolved non-exchangeable analyte signal, a 90° pulse, a relaxation delay of at least 5×T₁ (commonly 30–60 s), and accurate gravimetry.

Elemental nitrogen (Kjeldahl or combustion) converts total N to peptide via the theoretical nitrogen content of the sequence. Fast and inexpensive; inflated by any nitrogen-containing impurity such as residual DMF, NMP, or piperidine. Its virtue here is that TFA and acetate are nitrogen-free.

Measured mass balance — net content ≈ (100% − water − counterion − residual solvent − ash) × (HPLC area purity / 100), with water by coulometric Karl Fischer per USP ⟨921⟩ Method Ic, counterion by ion chromatography or ¹⁹F qNMR, residual solvents by headspace GC, and ash by residue on ignition. This is a defensible and considerably cheaper approximation to AAA — but only if all four subtractions are actually measured. If any term is assumed rather than measured, the result is not a net content value and is not reported as one. Where a mass-balance estimate is reported, the arithmetic is shown and it is labeled an estimate.

Counterion arithmetic. Synthetic peptides are isolated as salts of the purification mobile-phase acid, with one counterion per basic site. Trifluoroacetate has a formula mass of 113.0 Da, acetate 59.0 Da. A 3,000 Da peptide with three basic sites fully loaded as the TFA salt carries about 10.2% w/w TFA; as the acetate salt, about 5.6%. Arg- and Lys-rich sequences can exceed 20% TFA by mass. Counterion identity is reported alongside the level, since a product sold as the acetate salt that is in fact the TFA salt is a real and detectable mislabeling.

Area-percent purity is a relative measure of UV-absorbing, column-eluting species under one specific set of conditions. It is not a mass fraction, and it carries no information about how much material is in the vial. Its documented limits:

• Invisible species. Water, inorganic salts, counterions, bulking agents such as mannitol, sucrose, trehalose, and glycine, and most residual solvents either do not absorb at 214 nm or elute in the void and are excluded by the integration rule. They are absent from both numerator and denominator.
• Non-eluting species. Aggregates, precipitates, and strongly adsorbed material never reach the detector. Poor mass recovery is the only tell, and it is only visible if recovery is actually measured. A longer reversed-phase gradient does not fix this — SEC or DLS is the correct orthogonal check for aggregate.
• Co-elution. Enantiomers, most diastereomers, isobaric variants, Leu↔Ile substitutions, and disulfide-bond isomers can all co-elute. One peak is evidence of one chromatographic behavior, not one compound. Diode-array peak-purity analysis detects only co-eluting species with different UV spectra; a deletion analog or a diastereomer has an essentially identical spectrum and will pass it.
• Unequal response factors. Area percent assumes equal absorptivity per unit mass. At 214 nm that approximation is good for peptide-related impurities and poor for anything else.
• Sample preparation bias. Filtering an incompletely dissolved sample removes insolubles from both numerator and denominator and biases purity high. Dissolution behavior is observed and recorded, and any filtration is stated.

Accurate intact mass confirms elemental composition within tolerance. It does not confirm sequence. Where sequence confirmation is required, MS/MS fragment coverage must be requested and will be reported with the coverage achieved.

A net content result applies to the unit sampled. Where the laboratory did not perform the sampling, no statement about the lot is possible and none is made.

No result on this page is a safety determination. Purity, identity, and content characterize a material as an article of laboratory research. They are not a determination of sterility, endotoxin content, elemental impurity content, biological activity, or fitness for administration to humans or animals.