A metal catalyst (Zn, Co, or Cr center with a bulky supporting ligand) sits between the two monomers and alternately inserts them into a growing chain, ring-opening each epoxide and inserting CO2 into the metal–alkoxide bond:
–M–O–CH(R)–CH2–O⁻ + CO2 → –M–O–C(=O)–O–CH(R)–CH2–O⁻ (carbonate linkage)
–M–O⁻ + epoxide → –M–O–CH(R)–CH2–O⁻ (ether linkage, no CO2)
Each insertion step is a kinetic competition governed by Arrhenius rates for the two pathways, weighted by CO2 pressure:
k_CO2 = A_CO2 · exp(−Ea_CO2 / RT) · P(CO2)
k_epox = A_epox · exp(−Ea_epox / RT)
P(carbonate insertion) = k_CO2 / (k_CO2 + k_epox)
Higher CO2 pressure and lower temperature favor carbonate insertion, raising CO2 incorporation toward the alternating limit (~50 mol% carbonate). At low pressure or high temperature the chain can "back-bite" — the terminal alkoxide attacks its own carbonate to expel a cyclic carbonate byproduct, truncating the chain (a real side reaction that lowers polyol/polycarbonate molecular weight in industrial CO2-to-polyol processes). Bulky cyclohexene oxide sterically suppresses epoxide homopolymerization and gives higher, more regular carbonate content than propylene oxide, at the cost of a slower turnover rate — both trends match real Zn/Co-catalyzed CO2/epoxide copolymerizations used to make CO2-based polyols and polycarbonates.
- CO2 pressure — raises the CO2-insertion rate relative to epoxide insertion, and (in this 2D version) suppresses back-biting.
- Temperature — speeds up both pathways but disproportionately favors back-biting termination.
- Catalyst turnover rate — scales how fast the chain grows on screen (visualization speed only).
- Epoxide monomer — switches between propylene oxide and cyclohexene oxide, changing selectivity and steric bulk.
- Drag to pan, scroll/pinch to zoom — the chain view is a free 2D camera; it does not reset your framing while the chain grows.
Fix vs. the 3D version: the source engine's back-bite probability used a hard-coded scale factor (kBB · 0.00004) that, verified numerically across the entire slider range, never exceeded ~5×10⁻⁷ per insertion step — meaning the "Restart chain: back-bite" event described in its own theory text was, in practice, unreachable no matter how adverse the conditions. This 2D sibling rescales the same rate constant (factor 12.4, calibrated so the sweep's worst-case kBB maps to ≈0.15 per step) so back-biting stays rare (~10⁻⁵) at favorable conditions but becomes genuinely observable at high temperature / low pressure, matching the described chemistry instead of silently never firing.