Loading
Fetching the latest data…
Fetching the latest data…
Chemistry is the science of what a mixture does next, and the law of mass action is the simplest complete answer: every rate is a constant times the concentrations of what reacts. Four real reactions — their species from PubChem, recorded with their ids — run under exactly that law, and the two ideas every introductory course turns on become things you can watch: an equilibrium is not a halt but a balance of two rates, and pushing on it makes it push back. It is also the first module built on the common simulator kit, so what you learn about the integrator here is true in every other room of the platform.
Before you start — None. If you can read a chemical equation — arrows, coefficients, formulas — you can start. The student tier uses the reaction quotient Q and the constant K; the advanced tier states every formula the code computes.
Every rate here is a constant times the concentrations of the species that react, each raised to its coefficient. That is the law of mass action, and it is the entire engine: no temperature, no solvent, no special cases. Watch hydrogen and oxygen: fast at first, then slowing as the reactants thin out — not because anything changed, but because there is less to collide.
Do — Open hydrogen + oxygen and watch the curves. Halve the starting hydrogen and run again: the reaction stops at half the water, because hydrogen was the limiting reagent — the one that runs out first.
Give the reaction a reverse constant and it no longer runs to exhaustion. It stops where forward and reverse rates are equal — Q, the products-over-reactants ratio, has reached K = kf/kr. Nothing has stopped moving; the two directions have matched. That is what every equilibrium is.
Do — Open NO₂ ⇌ N₂O₄ and let it settle. Read the conversion: it stops short of 100 on purpose. Now double kr and run again — K halves, and the equilibrium slides back toward the reactants.
Le Chatelier's principle: disturb an equilibrium and the system shifts to partly undo the disturbance. In mass action it is not a rule but a consequence — inject product, and the reverse rate now exceeds the forward one until the ratio is restored. The module injects product halfway through the run and scores how much came back.
Do — On NO₂ ⇌ N₂O₄, raise the product-injection slider and run. Watch Q spike and relax. Then do the same on hydrogen + oxygen: nothing comes back, because a one-way reaction has no equilibrium to return to.
The simulation moves in fixed steps. Make the step too large for how fast the reaction runs and the arithmetic overshoots: concentrations oscillate and finally go negative — which is not chemistry failing but the method. Stability grades that and only that; it is the same lesson Causal Forge teaches with α, met again in a second science.
Do — Raise dt until stability collapses, then raise kf and watch the safe dt shrink. The edge belongs to the reaction, not to the module. Pace rewards speed — the fastest honest run is the one just inside the edge.
The panel shows element totals and a mass in grams side by side. The atoms are constant to the last digit; the grams drift in the fourth. That is not a bug: the molar masses are PubChem's rounded tabulated values, and a rounded table cannot balance a reaction exactly. The invariant is atoms. This is why conservation is displayed here and deliberately never scored — a metric that cannot fail is a decoration.
Do — Open esterification and read the conservation readout. Carbon, hydrogen and oxygen hold exactly; the gram total moves in the fourth decimal. Then check the CIDs against PubChem yourself — every number on this page has a source.
Has the reaction stopped changing? At the start things move fast; as reactants run low or products build up, the net change slows. This is how close to still the mixture is by the end, compared with its fastest moment.
Try — Halve the number of steps and watch settling fall — the reaction was still moving when you stopped looking. Then double kf: it settles sooner, and stability starts to care.
How much of the reactant that runs out first was actually used up. 100 means it is gone; a reversible reaction stops short of that, because products push back.
Try — On the Haber preset, add hydrogen beyond 3:1 and watch nitrogen conversion rise without touching the rate constants. That is Le Chatelier from the reactant side.
The simulation takes fixed-size steps. If a step is too big for how fast the reaction runs, concentrations overshoot, swing back and forth, and can even go below zero — which is not chemistry, it is the arithmetic failing.
Try — Raise dt until this collapses. The value where it goes is the stability boundary of the method, and it moves when you change kf — the safe step size belongs to the reaction, not to the module.
Halfway through, the module dumps extra product into the mixture. A reversible reaction pushes back — converts some product into reactants again until the ratio is restored. This scores how much of that push-back happened.
Try — Set a perturbation on NO₂ ⇌ N₂O₄ and watch Q spike then fall back toward K. Then try the same on hydrogen + oxygen: nothing comes back, because nothing can.
How quickly the reaction got most of the way to where it ends up. A reaction that reaches 95% of its final conversion early scores high; one still crawling at the end scores low.
Try — Push kf and dt up together. Pace climbs until stability breaks, and the pair of them finds the edge for you — the same edge Causal Forge finds with α.
Surfacing sources you can verify is a deliberate anti-pseudoscience measure, not a bibliography. Nothing on this page asks to be taken on trust.
Glossary — every term used above, defined once.