I've been tracking one batch across three COAs over 6 weeks. Vendor COA says 98.4% by HPLC, area normalization.
Lab A (independent) reports 96.1% with a different column and gradient. Lab B says 97.3% but flags an unknown peak at 0.8%. Same sample? I split it myself, stored dark at -20C. The gap feels too big for just method. Anyone else seen a 2%+ spread on the same lot? What would you ask the lab to clarify first?
Area normalization always flatters. Ask for raw chromatograms.
Agree. That 0.8% unknown peak is doing real work. Is it co-eluting with the main peak on the vendor method?
What column dimensions? A 150 x 4.6 vs 100 x 2.1 can shift resolution and tailing.
The 2.1 mm column point is where I'd start. Method transfer between a 4.6 mm and 2.1 mm column can change selectivity, especially if dead volume isn't adjusted. But a 2.2% gap still feels high for column alone.
I'd ask Lab A and B for system suitability, peak purity, and whether they used the same reference standard. Also check if one lab reports anhydrous vs as-is basis. That can move numbers more than people expect. And if the sample is a salt, the counterion can skew area% if it elutes or doesn't.
Mass balance is the bigger issue. HPLC purity doesn't capture water, residual solvents, or inorganic counterions. Did either lab run KF or GC?
Autosampler degradation is real. Ask for injection order and time stamps. A late-eluting degradation peak can inflate or deflate area% depending on method.
Not convinced it's method alone. I had a lot show 97.8 vs 95.9, turned out to be salt form. NMR plus ICP clarified it. If your two labs prep differently, the gap may be real but not a purity problem.
Asked both labs for injection sequence and time stamps. If the late peak only shows after the standard, that's autosampler drift, not sample. Also want peak purity spectra overlays, not just a yes/no.
Column alone? Doubt it. I've seen 2.1 mm methods change elution order slightly, especially with different pH. But 4.2% is a flag, not a column footnote.
KF and GC or it's just area% theater. If Lab A ran anhydrous and Lab B as-is, the math explains part of the gap before you blame the column.
Salt form again. If one prep hydrolyzes or ion-pairs differently, area% moves. NMR plus ICP settled it for us. Ask for the actual COA basis.
Short quip: 'Same batch' means little if prep differs. Same vial? Same dilution solvent? Same equilibration? No? Then compare methods.
I'd start with system suitability: tailing, resolution, RSD, and whether the 0.8% unknown is integrated the same. Different integration thresholds can eat a tenth or more.
Autosampler degradation—yes. I had a run where injection 1 vs 40 differed by 1.5%. Time stamps and a fresh prep on both instruments would be my next ask.
Mass balance wins. Sum HPLC, KF, GC, counterion, then see what's missing. A 4.2% gap may be methodology plus water/solvent, not a purity argument.
Login to reply to this thread.
Login / Sign up