Chemical attack on a reactor wall is thermally activated. The corrosion rate follows the Arrhenius equation, scaled by a material–fluid compatibility factor S (0 = essentially inert, 1 = severe attack):
rate(T) = k0 · S · exp[ -(Ea/R) · (1/T − 1/Tref) ]
Tref = 298.15 K, Ea ≈ 66 kJ/mol, k0 = 2 mm/year
wall remaining(t) = t0 − rate(T) · t
time to failure = t0 / rate(T)
- PTFE / PFA are fluoropolymers — near-inert to almost everything except hot oxidizers.
- 316 stainless steel is strong and cheap but chloride/halogen ions attack its passive oxide film, causing localized pitting — the dark spots on the coil.
- Hastelloy C-276 resists a broad range of acids and halides but is not immune to hot oxidizing acids.
- Borosilicate glass resists almost all acids but is etched by strong bases (and by HF).
- Raising temperature 10 °C roughly doubles a thermally-activated attack rate — this is why hot flow-chemistry loops fail materials-compatibility tests that a bench-top run at room temperature would pass.
Real process engineers run accelerated corrosion coupons at elevated temperature and extrapolate with this same equation to size a safe service life before selecting reactor tubing.
Corrected from the 3D version: the source engine used Ea ≈ 30 kJ/mol, which only yields a ×1.35–1.48 rate increase per 10 °C over the sim's own operating range — while its own theory text (and real flow-chemistry practice) claims the rate "roughly doubles" every 10 °C. Verified numerically: exp[-(Ea/R)·(1/(T+10)-1/T)] ≈ 1.4 at Ea = 30 kJ/mol vs ≈ 2.0 at Ea ≈ 66 kJ/mol across 25–70 °C. This 2D version uses Ea = 66 kJ/mol so the slider behaviour actually matches the stated physics.