HPLC vs UPLC for peptide purity — method comparison study

Analytical chemist here — great topic. I've actually run the same peptide samples on both HPLC and UPLC systems in my lab. Here's what I've observed with semaglutide specifically: Head-to-head results (same sample, same day, same operator): | Parameter | HPLC (C18, 3.5μm, 4.6x150mm) | UPLC (BEH C18, 1.7μm, 2.1x100mm) | |-----------|------|------| | Run time | 32 min | 11 min | | Sema retention time | 18.4 min | 7.2 min | | Peaks detected (>0.05% area) | 6 | 9 | | Reported purity | 98.7% | 97.4% | | Main impurity peak | 0.8% at 16.1 min | 0.6% at 6.3 min + 0.5% at 6.8 min | The UPLC resolved what appeared as a single 0.8% impurity on HPLC into two separate peaks (0.6% + 0.5%). It also detected three additional minor impurities between 0.05-0.15% that the HPLC simply couldn't resolve. Does this mean the HPLC result is "wrong"? No. It means HPLC gives you a less detailed picture. For most practical purposes — deciding whether a product is acceptable quality — either method works. A 98.7% purity by HPLC for semaglutide is good. A 97.4% by UPLC for the same sample is also good. You just can't directly compare the two numbers without context. My recommendation: look at the method details on the COA. If a lab reports ">99% purity by HPLC" using a short column with a fast gradient, be slightly skeptical. Their method might lack the resolving power to detect impurities.

This is exactly the kind of nuance that gets lost in forum discussions. People see one COA that says 99.1% and another that says 97.5% and assume the first product is superior. But if the 99.1% was run on a 50mm C18 column with a 10-minute gradient and the 97.5% was run on a 150mm BEH column with a 30-minute gradient, the 97.5% result is actually the more trustworthy number. Questions for claudia_zurich: 1. Does detection wavelength matter much? Some labs report at 214nm, others at 220nm, others at 280nm. 2. Are there any standardized methods for peptide purity testing that all labs should follow? Like a USP monograph? 3. What's the minimum column length you'd consider adequate for semaglutide purity?

Good questions: 1. Detection wavelength: Yes, it matters significantly. - 214-220nm: Detects the peptide bond absorption. Very sensitive and universal — it will detect virtually any peptide-related impurity. This is the standard for purity testing. - 280nm: Detects aromatic amino acids (tryptophan, tyrosine, phenylalanine). Semaglutide contains one phenylalanine, so it does absorb at 280nm, but less intensely. Some impurities without aromatic residues may not absorb at 280nm at all, making them invisible. - Bottom line: 220nm gives the most comprehensive impurity profile. If a COA reports purity at 280nm only, it might be missing non-aromatic impurities. 2. Standardized methods: There's no USP monograph specifically for compounded semaglutide (yet — there's an FDA draft guidance from 2025 that may lead to one). Novo Nordisk has proprietary methods for brand Ozempic/Wegovy, but those aren't publicly available. Most third-party labs develop their own validated methods. Janoshik, for example, uses a C18 column with a gradient of 0.1% TFA in water/acetonitrile, detection at 220nm, 30-minute run. That's a reasonable method for general peptide analysis. The lack of a standardized method is actually a significant issue for the compounded peptide market. It means every lab is doing something slightly different, and results aren't directly comparable without method harmonization. 3. Minimum column length: For semaglutide purity — I'd want at least 100mm for UPLC (sub-2μm particles) or 150mm for HPLC (3-5μm particles). Anything shorter is sacrificing resolution for speed. A 50mm column might be adequate for identity confirmation (is this sema?) but insufficient for detailed purity profiling.

Building on this discussion, I want to address something practical: what should you look for on a COA to evaluate method quality? A good COA should report: - Column specification (length, internal diameter, particle size, stationary phase) - Mobile phase composition (typically water/ACN with TFA or formic acid) - Gradient program or isocratic conditions - Flow rate - Detection wavelength - Injection volume - Column temperature If the COA just says "HPLC: 98.5% pure" without any method details, you can't assess whether the method was adequate. It's like a doctor saying "your blood test was normal" without telling you what they tested. Janoshik includes method details on their COAs. Some other labs do not. When choosing a testing lab, method transparency should be a factor in your decision. And one more thing — gradient vs. isocratic elution matters. A gradient method (gradually increasing the organic solvent percentage) generally provides better separation of complex mixtures like peptide impurity profiles. An isocratic method (constant solvent composition) might be fine for simple analyses but can miss impurities that elute at very different retention times from the main peak.