Tin Pest: The Disease That Turns Metal Into Powder

Two Tins, One Element

Tin keeps two completely different solids in its repertoire, switching between them at a temperature a domestic freezer can reach. Above 13.2 °C (286.35 K) at 1 atm the stable form is β-Sn, white tin: body-centred tetragonal, space group I4₁/amd, a = 5.832 Å and c = 3.182 Å, each atom in a distorted six-fold environment — four neighbours at 3.022 Å plus two more at 3.182 Å — with its valence electrons pooled and delocalised. That makes β-Sn a genuine metal, soft and bendable, and a superconductor below 3.72 K.

Below 13.2 °C the winner is α-Sn, grey tin: diamond cubic, space group Fd-3m, a = 6.489 Å — the structure of diamond, silicon and germanium, which tin sits directly below in group 14 and can still imitate. Every atom makes exactly four tetrahedral covalent bonds at 2.810 Å, and the result is not a metal but a brittle, lustreless zero-gap semiconductor.

The counter-intuitive part is the arithmetic. The bonds get shorter — 2.810 Å against 3.022 Å — and the solid still gets dramatically bigger, because four stiff directional bonds hold atoms apart in an open cage while six soft metallic contacts let them nest. Packing fraction collapses 0.535 → 0.340, atomic volume climbs 27.05 → 34.16 ų, and density falls 7.287 → 5.769 g/cm³. Same atoms, a quarter more room. Everything destructive about tin pest follows from that line.

How a Solid Bar Turns Into Grey Powder

The two lattices share no geometric relationship that would let one shear smoothly into the other: α-Sn has to be built atom by atom out of β-Sn — a reconstructive, diffusion-controlled first-order transition — and the product demands +26% more volume, or ~8% on every edge. That misfit is the destructive engine. On a real object:

• Incubation. The metal sits below 13.2 °C, thermodynamically doomed and physically untouched, for months, years or decades — nothing visible.
• Nucleation. Somewhere — a scratch, a grain boundary, an inclusion, or a stray grain of grey tin transferred by contact — one viable α-Sn nucleus survives. That single event is the whole bottleneck.
• Growth into a wart. The nucleus eats into the surrounding metal and, occupying more space than the tin it consumes, raises the characteristic grey blister or wart on a bright surface.
• Spalling. An 8% linear misfit is some four hundred times tin's elastic limit (yield near 10 MPa against a modulus near 50 GPa gives only ~0.02%). The wart cannot stay bonded: it fractures and falls away as loose grey powder.
• Autocatalysis. The powder leaves fresh white tin exposed and in contact with grey tin — a perfect nucleation site. New warts ring the old one and the front spreads until nothing coherent is left.

Hence the impression of infection: it starts at a point, spreads outward, and passes to a sound object by simple contact.

The Thermodynamics: a Driving Force Worth 0.2 RT

Everything hinges on ΔG = ΔH − TΔS. For the α → β direction, ΔH = +2.1 kJ/mol and ΔS ≈ 7.0 J/(mol·K). (ΔH/T₀ = 2100/286.35 = 7.3 J/(mol·K) is what zeroes ΔG exactly at the measured transition; 7.0–7.3 brackets the scatter.) Above 286.35 K the TΔS term wins and white tin is stable; below it, enthalpy wins and grey tin takes over.

Worked example — the pull at −30 °C (243.15 K): ΔG(α→β) = 2100 − (243.15 × 7.0) = 2100 − 1702 = +398 J/mol. So the β → α transformation is downhill by only about 0.4 kJ/mol. Using 7.3 instead gives 0.33 kJ/mol. Compare RT at the same temperature, 8.314 × 243.15 = 2.0 kJ/mol: the entire driving force is ≈0.2 RT, smaller than the thermal jostling of a single atom. Tin's own heat of fusion is 7.03 kJ/mol, and an ordinary chemical reaction runs to hundreds of kJ/mol. Tin pest is one of the feeblest spontaneous processes in metallurgy — which is why nobody sees it in a kitchen freezer over a weekend.

Worked example — pressure: the Clausius–Clapeyron slope is dT/dP = T₀ΔV/ΔH. With molar volumes 20.58 cm³/mol (α) and 16.29 cm³/mol (β), ΔV(α→β) = −4.29 × 10⁻⁶ m³/mol, so dT/dP = 286.35 × (−4.29 × 10⁻⁶)/2100 ≈ −5.8 × 10⁻⁷ K/Pa — about −58 K per kbar. Pressure always favours the denser phase, so squeezing tin pushes the transition down: a genuine, if impractical, cure.

Why It Almost Never Happens: Nucleation and the C-Curve

Classical nucleation theory turns that feeble driving force into a verdict. Per unit volume, ΔG_v ≈ 400 J/mol ÷ 20.6 × 10⁻⁶ m³/mol ≈ 1.9 × 10⁷ J/m³. With a plausible solid–solid interfacial energy γ ~ 0.1 J/m², the critical radius is r* = 2γ/ΔG_v ≈ 10 nm — a cluster of order 10⁵ atoms that must assemble by chance — and the barrier is ΔG* = 16πγ³/3ΔG_v² ≈ 4 × 10⁻¹⁷ J, roughly 10⁴ kT at 243 K. That is not slow, it is never. Homogeneous nucleation of grey tin does not occur; every real case is heterogeneous — an impurity, a strained grain boundary, a cut surface, or a grain of α-Sn pressed against the metal, which skips nucleation entirely.

Growth runs into the opposite problem: cooling raises the driving force but collapses the atomic mobility a reconstructive transformation needs. Driving force → 0 at 13.2 °C, diffusion → 0 far below it, so the rate peaks in between — a classic C-shaped TTT curve with its nose near −30 to −40 °C.

Chemistry supplies the cure. Roughly 0.1 wt% of Sb, Bi or Pb suppresses tin pest, because these solutes dissolve happily in β-Sn but are rejected by the covalent α-Sn lattice; a growing nucleus must expel them, and that partitioning penalty is ruinous against a driving force already worth only 0.2 RT. Traditional 63Sn-37Pb solder carried hundreds of times the needed dose for free. Then RoHS (Directive 2002/95/EC, adopted 2003, applying from 1 July 2006) removed it, and the industry moved to Sn-0.5Cu, SAC305 and pure tin finishes. W. J. Plumbridge reported Sn-0.5Cu fully converted after about 1.5 years at −18 °C: an accidental safeguard, deleted by regulation.

Testing, Measuring and Specifying It

Tin pest is easy to detect and brutally hard to qualify against: the signal is enormous, the clock is long.

• Dilatometry. Cohen's original tool and still the most sensitive: against a +26% volume change, a few per cent conversion registers long before a wart is visible.
• X-ray diffraction. Grey tin announces itself with the α-Sn (111) reflection at 2θ ≈ 23.7° (Cu Kα, d = 3.746 Å), a line with no counterpart in the β-Sn pattern, whose strongest peaks are (200) at ≈30.6° and (101) at ≈32.0°. XRD gives phase fraction non-destructively.
• Calorimetry. The transition enthalpy is 2.1 kJ/mol = 17.7 J/g, about 30% of tin's heat of fusion, so DSC resolves it easily — but only on heating back through the transition, where α-Sn reverts fast once the sample is tens of degrees above 13.2 °C. Cooling scans show nothing: nothing nucleates on a DSC timescale.
• Seeded (inoculated) exposure. Rub α-Sn powder onto the specimen and hold it at −18 °C or −40 °C. Removing the nucleation barrier deliberately makes the test measure growth alone — months instead of decades, a conservative screen rather than a service simulation.
• Long cold soak. Unseeded storage at −18 °C or −40 °C for thousands of hours, with an obvious trap: a clean 1,000-hour result proves nothing when Plumbridge's failure needed roughly 13,000 hours.

RoHS leaves military and space equipment outside its scope entirely, and medical and monitoring equipment came in only with RoHS 2 (2011/65/EU) from 2014, still carrying Annex IV exemptions for lead in solder. That is why tin-lead solder survives in exactly the hardware that sees −55 °C. Aerospace lead-free control plans (the GEIA-STD-0005 family) require a documented tin-pest risk assessment for sustained sub-zero service, ESA's ECSS-Q-ST-70-71 restricts pure tin finishes, and museum practice for pewter is simply to keep the store above 13.2 °C with a margin.

Zinnpest, Cohen, and the Stories That Are Not True

Northern Europe met tin pest long before it had chemistry for it. Cold cathedrals and hard winters turned tin vessels and organ pipes to grey dust, and because a damaged piece could ruin a sound one stored beside it, the phenomenon was named as an illness: Zinnpest, tin plague. The contagion language was wrong in cause and exactly right in behaviour.

Ernst Cohen, from 1899 in Amsterdam and then for decades at Utrecht, turned the folklore into physical chemistry. Dilatometry and electrochemical cells established that white and grey tin are two allotropes in genuine equilibrium, fixed the transition at 13.2 °C, and showed that a seed of grey tin dramatically accelerates the change — the barrier is nucleation, not thermodynamics.

The best-documented modern victims are high-tin organ pipes — pipe metal runs up to about 98% Sn — in unheated northern European churches, where conservation surveys have found genuine α-Sn conversion and heating is now specified to prevent it.

Two famous stories do not survive scrutiny. Napoleon's buttons — that tin buttons disintegrated during the 1812 retreat from Russia — has no physical evidence behind it: surviving buttons of the period are pewter, alloyed with lead and antimony and therefore immune, and a few weeks of cold is nowhere near the years-long incubation the kinetics demand. Scott's Antarctic fuel cans (1912) certainly leaked, but failed solder seams and perished leather washers explain it without a phase change that the lead in that solder suppresses.

Not Corrosion, Not Whiskers: Telling the Look-Alikes Apart

It is not corrosion. Tin pest needs no oxygen, water, acid or electrolyte; it runs happily in dry nitrogen and in vacuum. Nothing is added or removed — the grey powder is chemically pure tin, not an oxide or salt, and the mass is unchanged. So a conversion coating or a coat of paint that stops rust does nothing here: there is no external reagent to exclude.

It is not tin whiskers. Whiskers are single-crystal filaments of ordinary β-Sn, a micron or two across and up to millimetres long, extruded at room temperature from electroplated finishes to relieve compressive stress. No phase change, no cold requirement, no bulk volume change — and a different failure: a conductive bridge shorting adjacent leads, not a part turning to powder. Different fixes (matte tin, a nickel underlayer, annealing, conformal coating), different standards (JESD22-A121, JESD201). Both are lead-free problems, both were suppressed by lead, and they are constantly confused.

It is not tin cry. The crackling when you bend a white tin bar is mechanical twinning — instant, audible and harmless.

A final misconception: tin pest is reversible as chemistry and irreversible as engineering. Warm grey tin comfortably above 13.2 °C — a modest bake, not a degree or two of margin — and it converts back to white tin within hours to days. But the powder stays powder and the cracks stay cracks. Only remelting and recasting resets the metal.

The same element, two solids: white tin above 13.2 °C, grey tin below it

Property	White tin (β-Sn), stable above 13.2 °C	Grey tin (α-Sn), stable below 13.2 °C
Crystal structure	body-centred tetragonal, I4₁/amd, a = 5.832 Å, c = 3.182 Å	diamond cubic, Fd-3m, a = 6.489 Å — silicon's own structure
Neighbours and bond lengths	6 neighbours: 4 at 3.022 Å plus 2 at 3.182 Å	4 tetrahedral covalent bonds at 2.810 Å
Packing fraction	0.535	0.340
Density and atomic volume	7.287 g/cm³; 27.05 ų per atom (16.29 cm³/mol)	5.769 g/cm³; 34.16 ų per atom (20.58 cm³/mol)
Electrical character	true metal; superconducts below 3.72 K	zero-gap semiconductor; no metallic lustre
Mechanical behaviour	soft, malleable, creeps at room temperature, 'cries' when bent	brittle; disintegrates to grey powder under its own misfit