Effect of column chemistry, temperature, inert flow path, and mobile phase on the separation of GLP-1 peptides and their impurities

Significance of the topic

Chromatographic impurity profiling for GLP-1 receptor agonists (semaglutide, liraglutide, tirzepatide) is critical because these peptide therapeutics often contain low-level impurities that are chemically very similar to the drug substance (diastereomers, single amino-acid deletions/insertions, and site-specific oxidations). High-resolution separations are required for accurate LC-UV quantitation, robust LC–MS detection, and confident identification of challenging isomeric impurities in development, QC and regulatory contexts.

Objectives and study overview

The study systematically evaluated how mobile-phase additives and their concentration, stationary-phase chemistry, column compartment temperature, sample load, and inert flow-path hardware affect chromatographic resolution and peak shape for semaglutide, liraglutide and tirzepatide impurities. The goal was to identify practical conditions that maximize impurity resolution and analytical robustness for routine peptide impurity profiling.

Methodology

• Platform: Agilent 1290 Infinity II Bio Flexible Pump LC systems coupled with DAD and either Agilent Pro iQ Plus MS or 6470 QQQ MS (Agilent JetStream ESI, positive mode).
• Column formats tested: multiple 2.1 × 150 mm columns including Altura Peptide Plus (charged C18, superficially porous 2.7 µm), AdvanceBio Peptide Plus, AdvanceBio Peptide Mapping, InfinityLab Poroshell EC‑C18 (1.9 µm), and ZORBAX Eclipse Plus (1.8 µm).
• Mobile phases screened: water/acetonitrile combinations with acidic additives (0.1% formic acid vs 0.1%–0.5% trifluoroacetic acid) and buffered conditions (50 mM or 40 mM ammonium acetate, pH ~7) to assess alternative selectivity.
• Experimental variables: gradient slope held constant while screening additives and concentrations using a pump Blend Assist feature, column temperatures from 40 °C to 80 °C, a flow rate of 0.5 mL/min, and injection volumes of 1 µL with differing sample loads (sub-microgram up to several µg).
• Detection: DAD at 220 nm or 280 nm (depending on peptide) and targeted SIM/transition monitoring on MS for key peptide charge states.

Used instrumentation

• Agilent 1290 Infinity II Bio Flexible Pump LC with DAD (Bio-inert Max-Light cartridge cell) and Pro iQ Plus MS or 6470 QQQ MS.
• Agilent JetStream ESI source, positive polarity; typical source settings: gas flows and temperatures adjusted to peptide analyses (examples: gas temp ~270 °C, sheath gas temp ~200 °C, capillary ~5500 V).
• Columns evaluated (2.1 × 150 mm): Altura Peptide Plus (p/n 227215-903), AdvanceBio Peptide Plus (p/n 695775-949), AdvanceBio Peptide Mapping (p/n 653750-902), InfinityLab Poroshell EC‑C18 1.9 µm (p/n 693675-902), ZORBAX Eclipse Plus 1.8 µm (p/n 959794-902).
• Mobile phase additives: formic acid (FA), trifluoroacetic acid (TFA) across multiple concentrations, and ammonium acetate buffers (50 mM, 40 mM) to probe pH-driven selectivity changes.

Main results and discussion

• Choice of acidic additive: TFA outperformed FA for semaglutide and liraglutide separations, giving sharper peak shapes, greater number of resolved impurities and cleaner UV baselines. TFA concentration had a pronounced effect on selectivity.
• Optimal TFA concentration differed by peptide: semaglutide separations were optimal near 0.4% TFA, whereas liraglutide showed best impurity resolution at much lower TFA (~0.05%). This illustrates that additive concentration must be optimized per analyte rather than using a single universal value.
• Stationary-phase chemistry: a charged C18 phase with superficially porous particles (Altura Peptide Plus) delivered improved peak shape, higher peak capacity and better selectivity for many semaglutide impurities compared with conventional fully porous or other sub-2 µm C18 phases. The combination of positive surface charge and solid-core particles likely enhances peptide–surface interactions and reduces on-column dispersion.
• Temperature effects: column compartment temperature strongly influenced impurity selectivity. Across experiments, 60 °C provided the best compromise for both semaglutide and liraglutide; higher temperatures (70–80 °C) tended to reduce resolution for some impurities, while lower temperatures altered retention and caused coelution shifts for specific isomers (for example, the Kyn(31) impurity in liraglutide was particularly temperature-sensitive).
• Sample load: increasing injected mass degraded selectivity and peak shape, with liraglutide exhibiting greater sensitivity to overloading under its optimized low‑TFA conditions. Loading capacity advantages were observed for the Altura Peptide Plus charged C18 phase compared with a competitor charged C18, but performance still depended on peptide and mobile phase.
• Buffered mobile phases: while low‑pH TFA conditions were superior for semaglutide and liraglutide, using an ammonium acetate buffer at near-neutral pH (~40 mM NH4OAc, pH ~7) markedly improved resolution and provided different retention order for tirzepatide impurities, indicating that higher pH buffering can be a valuable alternative for certain analogues.

Benefits and practical applications of the method

• Provides a structured approach to optimize peptide impurity separations by systematically varying additive type/concentration, temperature, column chemistry and load.
• Demonstrates that charged C18, superficially porous phases can increase peak capacity and tolerate higher loads, improving robustness for routine QC assays and impurity profiling workflows.
• Shows that pump-based blend and screening tools accelerate method development by removing the need to prepare multiple mobile phases.
• Highlights that alternative mobile phase pH (ammonium acetate buffers) is a practical route to resolve impurities not separable under traditional low-pH TFA conditions, particularly for tirzepatide.

Future trends and potential applications

• Greater use of tailored stationary phases (e.g., charged or mixed-mode peptide phases) combined with superficially porous particles to balance resolution and loading capacity for large peptides and peptide conjugates.
• Integration of automated additive and pH screening (pump Blend Assist or similar) with multivariate optimization tools to accelerate peptide method development and reduce reagent consumption.
• Broader adoption of buffered, higher‑pH mobile phases for certain peptide classes to exploit different selectivity mechanisms and reveal otherwise coeluting stereoisomers or sequence variants.
• Continued optimization of temperature as an orthogonal selectivity parameter; coupling temperature ramps with gradient optimization may further enhance resolution of closely related isomers.

Conclusion

A systematic, multifactor approach identified practical conditions to improve separation of GLP‑1 therapeutic peptides and their subtle impurities. Trifluoroacetic acid at peptide-specific concentrations and a charged, superficially porous C18 phase offered the best general performance for semaglutide and liraglutide, with 60 °C frequently giving optimal selectivity. For tirzepatide, near‑neutral ammonium acetate buffer provided superior alternative selectivity. These findings support targeted method development strategies for peptide impurity profiling in development and QC laboratories.

References

• ASMS 2026 Poster ThP 213, Agilent Technologies: Evaluation of column chemistry, temperature, inert flow path and mobile phase for separation of GLP‑1 peptides and impurities; DE‑015163; Published June 10, 2026.