Does lyophilised powder actually need −20 °C, or is 4 °C fine? What is the real mechanism?

Every supplier sheet I have says "store lyophilised powder at −20 °C for long-term, 2–8 °C for short-term, room temperature acceptable for shipping". Nobody explains what "long-term" and "short-term" mean in months, and nobody explains what physically changes between those temperatures for a dry solid.

What I actually want to understand is the mechanism. A dry powder has no bulk water, so hydrolysis should be slow to the point of irrelevance. If that is true, why does anyone bother with a freezer? And if it is not true, what is the water doing?

Specific things I cannot resolve:

• Is the freezer buying real months, or is it a cheap default recommendation because nobody has run the study?
• Does a freeze-thaw cycle on the dry powder matter? I take a vial out, it warms, I put it back. That feels different from freeze-thawing a solution but I do not know that it is.
• Where does condensation fit in? If I take a −20 °C vial straight out into a humid room, something must be happening at the glass.
• Do those little desiccant sachets in the shipping box do anything for a sealed, crimped vial?

Research-use material. I would rather understand the physics than collect more rules of thumb.

The freezer is buying real time, and the mechanism is residual moisture plus molecular mobility in an amorphous solid, not bulk hydrolysis. Once you see that, all four of your sub-questions answer themselves.

A lyophilised cake is not dry

Freeze-drying removes bulk water by sublimation and then removes some bound water by secondary drying, but it does not remove all of it. A well-made pharmaceutical cake typically retains on the order of 1 to 3 % water by mass, measured by Karl Fischer titration or loss on drying under the compendial water-determination chapter. That residual water is not inert. It is a plasticiser.

Most peptide cakes are amorphous glasses, not crystals. An amorphous solid has a glass transition temperature, Tg, below which molecular motion is extremely slow and above which the material becomes a rubbery state with mobility orders of magnitude higher. Degradation chemistry in a solid — deamidation, oxidation, aggregation via solid-state contact — requires molecular mobility. So:

• Store well below Tg and reactions are kinetically arrested. Not zero, but slow enough that shelf life is measured in years.
• Approach or exceed Tg and mobility rises steeply, and the same chemistry that was frozen out starts to proceed.
• Water lowers Tg, hard. This is the crux. A cake at 1 % water might have a Tg comfortably above room temperature; the same formulation at 4 % water can have a Tg that has dropped by tens of degrees. Moisture ingress does not just add a reagent, it moves the whole stability regime.

That is why the recommendation ladder exists. −20 °C puts you far below Tg for essentially any residual moisture level you are likely to have. 2–8 °C puts you below Tg for a well-dried cake but with much less margin if the cake picked up water. Room temperature relies entirely on the cake being properly dried and staying that way.

So how long, in months?

Honestly: for research-grade material, nobody has run the study on your lot, and that is the real answer to your first question. What exists is the general framework and the approved-product analogy. The regulatory stability framework is the ICH stability guideline — long-term storage at 25 °C / 60 % relative humidity, accelerated at 40 °C / 75 % RH for six months, with intermediate conditions where accelerated data shows significant change. Approved lyophilised peptide products routinely carry 24 to 36 month expiries at refrigerated storage on the strength of that testing.

For an unstudied research cake the defensible position is: −20 °C is a low-cost way to buy a large kinetic margin against an unknown, and the marginal cost of using a freezer you already own is zero. That is a better argument than any number someone will quote you.

Dry-state temperature cycling

Materially less harmful than cycling a solution, for exactly the mobility reason above — there is no ice to form, no ice-water interface, no freeze-concentration of solutes. Taking a vial from −20 °C to room temperature and back is not the same event as thawing and refreezing 2 mL of reconstituted solution, and it is not close.

But it is not free, and the reason is your third question.

Condensation — the actual failure mode

This is where people damage cakes. A vial at −20 °C brought into a room at 22 °C and 55 % RH has an exterior surface far below the dew point. Water condenses on the glass, and if you break the crimp or unstopper while the vial is cold, that water goes straight to where the cake is. Even without opening it, repeated condensation cycles wet the label, wet the crimp, and wet the septum, and a septum that has been repeatedly wetted is a worse barrier than a dry one.

The fix is one line: let a frozen vial equilibrate to room temperature, fully, sealed, before you open it. For a small vial that is fifteen to thirty minutes on the bench. Do not accelerate it in warm water, and do not open it "just to check". If you want to reduce cycling further, store working stock at 2–8 °C and reserve the freezer for material you are not touching, so the freezer vials get exactly two temperature transitions in their life.

Desiccant sachets

For a sealed, crimped vial with an intact butyl closure, a sachet in the shipping box does very little for the contents. Butyl rubber has low but nonzero water vapour permeability, so over years there is a slow ingress path, and a desiccant in a sealed secondary container does reduce the driving gradient. Over a two-week transit it is doing almost nothing chemically.

What it genuinely does: keeps the outside of the vials, the labels and the box dry, which matters for condensation management and for not having a soggy label you cannot read. That is a real if unglamorous benefit. Where desiccant does matter chemically is if you are storing many vials long-term in one sealed box — then a fresh desiccant in that box is cheap insurance against the slow permeation route. Do not put a desiccant sachet inside anything and do not open vials to add one.

edited 8 Sept 2024 by cake_intact — added the method parameters

Adding a practical caveat about domestic freezers, because "store at −20 °C" and "put it in the kitchen freezer" are not the same instruction.

Two problems with a household frost-free freezer:

• Auto-defrost cycles. A frost-free unit periodically warms its evaporator to shed ice, and the compartment temperature swings with it. Depending on the appliance, contents can see excursions of several degrees on a cycle lasting tens of minutes, several times a day. Over a year that is hundreds of small thermal cycles. For a dry cake far below Tg this is tolerable; it is not the stable −20 °C the instruction implies, and it is why laboratory freezers are specified as manual-defrost.
• Door-position gradients. The door shelf of a domestic freezer can run 5 to 10 °C warmer than the back of the cabinet and sees the largest swing every time the door opens. Storing vials in the door is common and is the worst available location.

Mitigations that cost nothing: put vials at the back of the bottom shelf, inside a closed rigid container, ideally with other frozen mass around them for thermal inertia. A sealed food container full of vials plus a fresh desiccant sachet is a reasonable improvised arrangement — the thermal mass damps the defrost swing and the closed container damps humidity cycling on the glass.

If you want to know what your freezer actually does rather than what it claims, a cheap min/max logger left in there for a week is genuinely informative and I would recommend it over any amount of speculation. People are often surprised: some units hold −19 °C within a degree, others swing between −22 and −8 °C routinely, and you cannot tell which you own from the badge.

One thing not to do: store at −20 °C, then move to 2–8 °C, then back to −20 °C as your habits change. Every transition is a condensation event. Pick a location per vial and leave it there.

One correction to a common misreading of the "room temperature acceptable for shipping" line, because it gets over-generalised into "room temperature is fine".

Those two statements are about different quantities. Shipping tolerance is a statement about a short excursion — days to a couple of weeks — assessed against whether the material still meets specification on arrival. Storage recommendation is a statement about the condition under which the assigned shelf life holds. A material can be entirely untroubled by fourteen days at 30 °C and still lose meaningful potency over eighteen months at 25 °C, because the integral of rate over time is what matters and the two integrals are nothing alike.

Practically, the sensible reading of a supplier's three-tier recommendation is:

1. Transit at ambient: expected, designed for, do not panic about it, and the absence of a cold pack is not a defect.
2. 2–8 °C: the right place for anything you are actively working through, on a timescale of weeks to a few months.
3. −20 °C: the right place for anything you will not touch for months, and the place where an unknown lot's unknown degradation rate is most effectively suppressed.

The thing that does not appear on any supplier sheet, and matters more than the difference between tiers two and three: write the receipt date on every vial. Storage temperature only means anything in combination with elapsed time, and a freezer full of undated vials is a freezer of unknown material regardless of how well the freezer performs.

A small mechanistic footnote on why "no bulk water means no hydrolysis" is the wrong intuition, since the question raised it explicitly and it is a good instinct that happens to fail.

Two solid-state routes proceed without bulk water:

• Deamidation via a cyclic imide intermediate. The rate-limiting step is intramolecular attack, which needs conformational mobility rather than a large water reservoir. Residual moisture supplies enough for the chemistry; mobility supplies the rest. This is why the amorphous-glass argument carries the explanation rather than a water-availability argument.
• Oxidation. Methionine and tryptophan residues oxidise in the solid state, driven by residual peroxide from excipients, by trace metals, and by headspace oxygen. Temperature accelerates it and water is not required. This is one genuine argument for headspace control in the fill, and it is not something you can influence after the fact.

There is also physical rather than chemical instability: solid-state aggregation, where neighbouring molecules in an amorphous matrix associate irreversibly. It shows up as incomplete or slow reconstitution — the cake that takes five minutes and a lot of gentle rolling to dissolve when a fresh one took thirty seconds. That symptom is worth noticing, because it is the only solid-state degradation signal visible without instrumentation.