The 3D version watches individual molecules adsorb on a lattice of catalyst sites and reports whatever fraction happens to have reacted at that instant. This 2D companion instead integrates the actual consecutive-reaction rate law for a well-mixed batch — the textbook A→B→C kinetics behind every selective hydrogenation — every frame, alongside an independent population of molecules that flip state stochastically at the same rates, so the two methods can be checked against each other live:
dA/dt = −k1·A
dB/dt = k1·A − k2·B
dC/dt = k2·B
k1, k2 = A0·exp(−Ea/RT)·f(P_H2) (Arrhenius + Langmuir-type H2 coverage)
k2,eff = k2·(1 − p_desorb) (Lindlar poisoning/quinoline attenuate ONLY the 2nd step)
Ea for the first addition (alkyne→alkene) is lower than for the second (alkene→alkane) — a real Horiuti–Polanyi consequence of the twisted, weakly-bound alkene desorbing more easily than the flat alkyne stays put. Poisoning a fraction of the Pd surface with lead raises the probability p_desorb that a freshly-formed alkene lets go before it can react again; that probability is averaged over every site (poisoned sites release almost every time, bare sites rarely do) to give one effective rate constant for the whole reactor, exactly as a chemical engineer would lump it. Quinoline adds a further moderation on top.
- Pb poisoning slider — raises the poisoned-site fraction that discourages the second hydrogenation; watch k2,eff (and the alkane curve) collapse as it rises.
- Quinoline — pushes desorption probability higher still, useful near 100% poisoning to squeeze out the last over-reduction.
- Temperature — raises both k1 and k2 (Arrhenius), but raises k2 relatively more since its activation energy is higher — a real, if secondary, way heat erodes selectivity that a poisoning-only story misses.
- H₂ pressure — speeds the overall turnover (Langmuir-type coverage), without changing which step it's speeding up.
- Best possible alkene yield — the maximum of B(t) for the current rate constants (dB/dt = 0 gives a closed form). This is the real number a chemist chases: quench the reaction at exactly this time and you can never isolate more cis-alkene than this, no matter how careful you are.
The floating dots above the chart are a second, independent simulation: each one is a single molecule that privately rolls dice against the same k1 and k2 every timestep. Their population fractions (translucent dashed lines) should track the solid rate-law curves — small stochastic wobble is expected, a systematic gap would mean a bug.