Low-temperature and cryogenic O-rings: brittleness, Tg and material choice from −40 to −60 °C and below

When choosing a sealing O-ring, most engineers focus on the upper temperature limit — how much heat it will survive. In freezer plants, refrigeration loops, outdoor equipment in winter and cryogenic applications, however, the lower limit is the one that bites. In the cold an elastomer does not fail by slow decomposition; it fails by a physical change: it hardens, loses its elasticity and stops pushing back against the sealing surface. The result is a leak that worsens with every thermal cycle. This article explains why that happens, how brittleness is measured, and which materials in our range stay reliable down to −40 °C, −60 °C and below.

Why a seal fails in the cold before it fails in the heat

An O-ring seals because it is compressed on assembly and continuously pushes back against the sealing surface — this residual elastic force closes the leak path. As the assembly cools, two things happen at once. First, the rubber contracts thermally, so both its dimensions and its squeeze in the groove shrink. Second, and more importantly, the material progressively stiffens and loses its ability to spring back to shape. At a low enough temperature the elastomer behaves like a hard plastic that no longer pushes back — and the thin contraction gap it would instantly fill when warm becomes a permanent leak path. This is why, in practice, the dominant cause of cold leakage is loss of stiffness-driven contact force rather than chemical degradation.

Pressure and thermal cycling make it worse. Higher pressure effectively shifts the usable limit upward — a working rule of thumb is roughly 1 °C per about 50 bar — so a ring that still seals at atmospheric pressure may weep under pressure. And every excursion into the cold and back lets micro-cracks and compression set accumulate, so the seal often fails not on the first cold soak but after several cycles. Designing for the cold therefore means designing for the worst-case combination of low temperature, peak pressure and cycle count.

The glass-transition temperature (Tg) and brittleness

The key concept is the glass-transition temperature, written Tg. Above Tg an elastomer is rubbery and elastic: the polymer chains move freely and the material recovers after compression. Below Tg the chain motion is effectively frozen, the rubber enters a glassy state and becomes hard and brittle. Tg is not a sharp line but a transition band — the first signs of lost flexibility appear above it, well before the material is fully glassy. The practical rule is therefore simple: the lower service temperature of a seal should sit safely above the material's Tg, with margin for pressure and cycling on top.

Two distinct mechanisms matter. The glass transition itself is fast and reversible — warm the rubber and it softens again. The second mechanism, crystallisation, is slower: some elastomers gradually stiffen even above their Tg during a prolonged cold soak as part of the polymer orders into a crystalline structure. This is why a seal that passes a short freeze test can still go rigid after a few days inside a cold store. A robust design accounts for both — the instantaneous transition and the time-dependent crystallisation.

How low temperature is measured: ISO 812, ASTM D746 and TR10

To make published figures comparable, standardised tests are used. The low-temperature brittleness test to ISO 812 (its US counterpart is ASTM D2137) finds the lowest temperature at which a specimen does not yet fracture on impact. The related ASTM D746 test reports a brittleness temperature — the temperature at which half of the specimens fracture under a defined impact. Both standards explicitly warn that the measured value is not the lowest usable service temperature: it is a test under specific impact conditions, useful for ranking compounds rather than for setting a guaranteed minimum.

For sealing rings the most informative cold test is the TR10 temperature-retraction test to ISO 2921 (ASTM D1329). A specimen is stretched, frozen and then slowly warmed; TR10 is the temperature at which it has recovered 10 % of the stretch. TR10 correlates well with how a seal behaves in the cold and is routinely quoted on material datasheets, so it is the single number worth asking for when you specify for low temperature. Beware a common confusion: ISO 188 is not about cold at all — it covers accelerated ageing and heat resistance, i.e. the upper end of the range — so do not cite it as evidence of low-temperature capability.

Which elastomers stay flexible in the cold

The best elastomer for low temperatures is silicone (VMQ). In our material database VMQ is listed with a continuous lower limit of −60 °C (and an upper end up to 200 °C), which makes it the first choice wherever oil or fuel is not present and dynamic loading is low — silicone's mechanical strength is comparatively weak. Where fuel or hydrocarbons accompany the cold, fluorosilicone (FVMQ) steps in, also rated to −60 °C: it combines silicone's cold flexibility with FKM-like fuel resistance, at a higher price and in an aerospace-leaning niche.

A dependable all-rounder down to −40 °C is EPDM, which additionally handles brake fluid, hot water and steam, ozone and weathering — ideal for outdoor gear and water circuits in winter. Reaching −40 °C as well are chloroprene (CR, Neoprene), typical of refrigeration and HVAC, polyurethane (PU) for abrasion-resistant dynamic seals, and ethylene acrylate (AEM), which has better cold resistance than its polyacrylate relative. Standard nitrile (NBR) is listed in our database to −30 °C; low-acrylonitrile cold-grade NBR compounds reach lower. By contrast, standard FKM (Viton) only goes to −20 °C and stiffens in the cold — which is why a dedicated low-temperature GLT compound exists (listed among the FKM grades) rated to roughly −40 °C. Even perfluoroelastomer (FFKM), otherwise the most resistant elastomer, is only −20 °C as standard; its special low-temperature grades reach about −42 to −45 °C.

Cryogenic temperatures: where elastomers run out and PTFE takes over

Below roughly −60 °C the elastomer world all but ends. With genuinely cryogenic media — liquid nitrogen (−196 °C), LNG, liquid oxygen — no rubber keeps its elasticity; it hardens, shrinks and cracks under thermal cycling. This is where PTFE (Teflon) takes over; in our database it carries an exceptional lower limit of −200 °C. PTFE, however, is not an elastomer but a rigid polymer — on its own it has almost no elastic recovery (zero compression-set recovery), so it is not used as a plain O-ring but as a PTFE jacket over a metal spring (a spring-energized seal) or as a back-up ring. The spring supplies the contact force PTFE itself lacks, and such assemblies seal down to −250 °C and below.

PTFE in the cold is therefore a backup, not an elastomer: it solves what rubber can no longer do, but at the cost of a different seal construction. For cryogenic projects you are choosing not just a material but a whole assembly — the jacket, the spring and the groove geometry.

A practical selection summary

The procedure is straightforward. First establish the true lowest temperature the seal will see, then add margin for pressure (roughly +1 °C per ~50 bar) and for cycling. Next choose a material whose lower limit sits safely below that figure: down to −40 °C, EPDM, CR, PU, AEM or a cold-grade NBR usually suffices; down to −60 °C, silicone (VMQ), or FVMQ where fuel is present; for cold combined with high temperature or an aggressive medium, reach for the low-temperature grades of FKM (GLT) or FFKM. Below −60 °C and into the cryogenic range, leave elastomers behind and design a PTFE spring-energized seal. Remember that the groove must allow for thermal contraction — enough squeeze when cold matters just as much as the polymer choice. When in doubt, ask for the TR10 figure and talk to us; for cold applications it is the best single indicator you can get.