Cross-Linking
The reaction in which molecules bond not only end to end but across to each other, turning two poured liquids into a single connected solid. Almost every property a coating is sold on is a consequence of how completely this happens.
What happens when a coating cures
Mix the two parts of an epoxy and molecules from each begin bonding to molecules of the other. What makes it cross-linking rather than simply chain building is that the bonds form across as well as along: each molecule ties into several others, and the connections propagate until the whole film is one continuous network rather than a crowd of separate molecules lying together.
That is the difference between a coating curing and a coating drying. Drying is physical, a solvent leaving and the solids staying behind, and nothing new is created. Cross-linking makes something that was not there before, and it does not come apart again. A cured epoxy floor cannot be softened and reworked the way a thermoplastic can, which is worth knowing because it decides what your options are when something has gone wrong: with a cured film, they involve removing it.
Why the mix ratio is not a guideline
The reaction consumes sites on one component using sites on the other, in the proportion the chemistry dictates. Supply too much of either and the surplus has nothing available to react with.
The consequence is less forgiving than intuition suggests. An off ratio does not give you a floor that is uniformly a bit weaker. It gives you a network plus a quantity of unreacted material sitting in it, never having joined, and that shows up as softness, persistent tack or resistance the product was rated for and does not deliver. This is why measuring the ratio properly is not carefulness for its own sake. It is the difference between the reaction happening and not.
What rests on it
Most of what a coating is sold on is downstream of how densely and completely this network forms. Hardness is the network resisting a point pressed into it. Novolac is a resin engineered to make more of these connections, which is exactly why it resists chemical attack and exactly why it gives up some toughness. And our page on hot tire pickup already puts it plainly: resistance comes down to how tightly the cured film is linked.
Recoat windows come from it too, and this is the connection most worth having. While the layer below still has unreacted sites available, a fresh coat can link into it and the two effectively become one material. Once that layer has gone as far as it is going to go, there is nothing left to bond with, and the next coat can only hold on mechanically to whatever texture it finds. That is precisely why the instruction past a recoat maximum is nearly always to abrade: you are manufacturing by hand the grip the chemistry can no longer provide.
You cannot see whether it finished
A film that no longer feels sticky has passed a test of its surface, and told you nothing about the network underneath. Tack free arrives early and full cure arrives much later, and the gap between them is the reaction still running.
Evidence of an incomplete reaction usually turns up later and somewhere unhelpful: a floor that dents under a stand, that stays tacky in patches, that softens beneath a warm tire, or that fails against a chemical it was specified to handle. The pattern of where it happens names the cause more reliably than anything you can observe on the day. That is the practical argument for writing down the ratio, the temperature and the batch size while you are mixing, because those are the facts the diagnosis will want and the only moment to capture them cheaply is before anything looks wrong.
Frequently Asked Questions
How is cross-linking different from something simply drying?
Drying is a physical change: a solvent or water leaves and what was dissolved in it stays behind. Nothing new is made, and in principle the process reverses if the solvent goes back.
Cross-linking is a chemical change. New bonds form between molecules and tie them into one network spanning the whole film. There is no going back from it, which is why a cured epoxy floor cannot be softened and reworked the way a thermoplastic can. What you laid is what you have.
Why is the mix ratio so unforgiving?
Because the reaction needs both components in the proportion the chemistry calls for. Bonds form between sites on one part and sites on the other, so an excess of either leaves material with nothing to react against.
That surplus does not become a slightly weaker floor spread evenly through the film. It stays in the film as unreacted material that never joined the network, which is why an off ratio shows up as softness, tack or lost resistance rather than as a proportionate loss of quality. Measuring accurately is not tidiness, it is the reaction working or not working.
Is more cross-linking always better?
A more completely reacted film is better than an incompletely reacted one, always. Whether a chemistry designed for a denser network is better is a different question, and there the answer is that it trades.
A denser network resists chemical attack and heat better, and gives up some toughness and flexibility doing it. Novolac is the clearest case of that trade on this site, and shore hardness is the same trade seen from the mechanical side.
What does this have to do with recoat windows?
The recoat window exists because of it. Applied while the layer underneath still has unreacted sites available, a new coat can link into it and the two become chemically one thing.
Once the layer below has reacted as far as it is going to, there is nothing left to bond with, and the next coat can only grip mechanically to whatever texture is present. That is why the instruction past the maximum is almost always to abrade first: you are manufacturing the grip the chemistry can no longer supply.
Can I tell by looking whether it finished?
Not reliably, and that is the difficulty. A film that has stopped feeling sticky has passed one surface test and told you nothing about the network below it.
The signs of an incomplete reaction tend to arrive later and elsewhere: a floor that dents, stays tacky in patches, softens under a warm tire, or gives up against a chemical it was specified to handle. By then the diagnosis is about what happened at mixing, which is why a record of ratio, temperature and batch size is worth keeping.