Freeze-Thaw Cycles And What They Do To A Solution

Three separate stresses act during a cycle, and they act at different moments. Cryoconcentration happens during freezing, as pure ice crystals grow and reject solute into the liquid between them. Interface exposure happens across the whole frozen period, because the ice-water boundary is an enormous surface with different properties from bulk water. Mechanical stress happens at the phase change itself, since water expands by about 9 percent in volume when it freezes.

None of these is visible. A solution that has been frozen and thawed four times looks the same as one that has never been frozen, unless aggregation has progressed far enough to scatter light. The measurements that detect the change are chromatographic, and they are run on the solution rather than on the vial.

The word that matters here is cumulative. Cycle damage does not reset on thawing, so the relevant variable is the number of cycles a given volume of solution has experienced, not the total time it spent cold.

Cryoconcentration: the fraction that never froze

Ice crystals accept almost nothing into their lattice. As the ice fraction grows from zero towards its final value, the remaining liquid holds the same absolute quantity of solute in a steadily smaller volume. By the time freezing completes, that unfrozen fraction can be a few percent of the original volume, with solute concentrations 10 to 50 times the starting figure.

Buffer salts do not stay put during this. Sodium phosphate is the standard example: disodium hydrogen phosphate crystallises out of the concentrating fraction before the monosodium form does, which removes base from the remaining liquid. A sodium phosphate buffer at pH 7.0 measured at ambient temperature has been reported to fall below pH 4.0 in the unfrozen fraction during freezing, measured with low-temperature pH indicators. Potassium phosphate shifts in the opposite direction, upward.

The interface, and why crystal size matters

The ice-water interface is not inert. It is a boundary where water molecules are ordered, where the dielectric environment differs from bulk solution, and where amphipathic molecules preferentially locate. Total interfacial area depends on crystal size, so a fast freeze that produces many small crystals creates more interface than a slow freeze that produces few large ones.

This is the reason freezing rate appears in formulation studies as a variable in its own right, and the reason results from one freezer do not transfer cleanly to another. A domestic freezer with an automatic defrost cycle is a different thermal environment from a laboratory unit held at a fixed setpoint, because the defrost cycle warms its contents periodically and each warming episode is a partial thaw.

What a chromatogram shows after cycling

Two signatures appear, and they are separate measurements. Size-exclusion chromatography resolves species by hydrodynamic size, so soluble aggregates appear as area eluting earlier than the monomer peak, reported as high-molecular-weight area percent. A figure moving from 0.3 percent to 1.5 percent across four cycles is the kind of change these studies report, and it is quantitative where an appearance check is not.

Reverse-phase HPLC at 214 nm tells a different story: it separates by hydrophobicity, so it picks up covalent degradation such as deamidation or oxidation products rather than aggregates, which may co-elute with monomer or not elute at all. Purity read by area percent at 214 nm can therefore stay almost flat across cycles while size-exclusion shows clear aggregate growth, and the two results are not in conflict.

The failure mode: insoluble aggregate lost on filtration

The specific thing that goes wrong across repeated cycles is aggregate formation that leaves the solution phase. Soluble aggregate is at least countable by size-exclusion chromatography. Once aggregate grows past solubility it becomes particulate, and a 0.22 micrometre filter step before analysis removes it from the sample entirely. The chromatogram then looks clean, because the damaged material never reached the column, while total recovered peptide has quietly fallen.

It shows up only as a mass balance discrepancy: concentration measured by absorbance at 280 nm, or by amino acid analysis, coming in below the nominal figure while every purity result looks normal. Without a concentration measurement alongside the purity measurement there is nothing in the data to catch it.

What Aurum tests, and what it does not

Every purity and identity figure we publish is measured on lyophilised material as supplied, before any solution exists. Freeze-thaw cycling studies is not among the specifications Aurum publishes, Size-exclusion chromatography is not among the specifications Aurum publishes, and we publish no aggregate content figure for any catalogue item. A published HPLC purity result describes the powder in the vial at the point of testing and carries no information about what happens to a solution afterwards.

That gap is real and it is worth naming rather than implying. Anything a solution does after it leaves the lyophilised state sits outside every measurement on our certificates, and the only way to characterise it is to measure it directly on the solution in question.

Common questions

Is one freeze-thaw cycle equivalent to long cold storage?

No. They are different stresses. Storage at a steady low temperature holds a system in one state; a cycle passes it through the phase change twice, with cryoconcentration and interface exposure on the way through.

Why does the choice of buffer salt change the outcome?

Because buffer components crystallise at different points during freezing, which changes the pH of the unfrozen fraction. Sodium phosphate drops, potassium phosphate rises, and buffers whose components do not crystallise shift far less.

Does aliquoting remove the problem?

It changes the variable rather than removing it. Dividing a solution so that each portion experiences one cycle rather than several reduces cycle count per portion, while each freezing event still involves cryoconcentration and interface exposure.

Can cycle damage be seen in the vial?

Only once particulate matter or visible haze has formed, which is late. Everything before that point is detectable by chromatography and by nothing else.