At minus 80 °C the question is which route is fastest, not whether dimerisation happens — and the routes do not share an activation energy, so their ranking changes with temperature. minus 80 °C is 85 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. Two chains join, usually through a disulfide, so the product is roughly twice the mass and shows up as a late peak — or as nothing, if it never comes off the column. So the way to answer it for your vial is to pick the method that sees dimerisation specifically and run it against a control held cold, rather than to infer a mechanism from a purity number that averages all of them.
Start with the sequence, because which pathways are available depends on which residues are present.
Aggregation is physical: peptides unfold at air-liquid interfaces and associate. Shaking maximises that interface, which is why swirling and shaking produce visibly different outcomes on the same vial.
Freeze-thaw cycling drives aggregation through concentration at the ice interface and pH shifts as buffer components crystallise out at different rates. Each cycle costs something.
Aggregation at air-liquid interfaces is established from surface-tension and particle-count studies and is the basis for anti-agitation handling guidance.
Swirl, never shake. Aggregation is a handling problem more than a time problem.
edited 13 May 2024 by Dr_Ingrid_Baumgartner — added the citation requested in comments