At minus 20 °C the question is which route is fastest, not whether fibrillation happens — and the routes do not share an activation energy, so their ranking changes with temperature. minus 20 °C is 25 kelvin below a refrigerator, and below the glass transition of a lyophilised cake the ten-degree rule of thumb stops applying at all — solid-state chemistry is not slow liquid chemistry, it is a different regime, and the failure modes that survive it are mechanical rather than chemical. That multiplier is an average over every route at once, which is exactly why it cannot tell you which one wins. Ordered beta-sheet assembly, effectively irreversible, and its endpoint is opalescence you can see rather than a peak you can integrate. So the way to answer it for your vial is to pick the method that sees fibrillation specifically and run it against a control held cold, rather than to infer a mechanism from a purity number that averages all of them.
The honest answer is that most reported "degradation" is adsorption and dilution error rather than chemistry.
Deamidation converts asparagine or glutamine to the corresponding acid via a succinimide intermediate, adding one dalton. It is base-catalysed, accelerates above neutral pH and is the dominant aqueous pathway for many peptides.
Degradation pathway by condition
| Pathway | Dominant when | Detected by |
|---|
| Deamidation | Solution, neutral to alkaline pH | RP-HPLC, +1 Da on MS |
| Oxidation | Light, trace metals, peroxides | RP-HPLC, +16 Da on MS |
| Hydrolysis | Solution, extremes of pH | RP-HPLC, fragment masses |
| Aggregation | Agitation, interfaces, high concentration | SEC, visual haze; often invisible on RP-HPLC |
| Freeze-concentration damage | Freeze-thaw of buffered solution | SEC, loss of recovered content |
Hydrolysis cleaves the backbone, most readily at aspartate-proline and aspartate-glycine sequences, and is acid-catalysed. In a dry solid it barely proceeds at all.
At dilute concentrations, suspect adsorption before you suspect chemistry.
edited 4 Jan 2026 by Dr_Wren_Halliday — corrected a unit error in the worked example