Flash Cure

If you apply a blast of heat energy to the surface of a coating, it seems intuitive that the surface should get very hot, very quickly, with the rest of the substrate remaining relatively cool. This is the principle of flash curing or, as it is commonly known, photonic curing.

Our intuitions of what constitutes a "blast" are probably off by a few orders of magnitude or, to put it another way, our estimation of the effects of the thermal conductivity of the substrate are off by a few orders of magnitude.

Absorbing 100W/cm² within 1ms into the top 1μm of a coating sounds as though the coating should get very, very hot and most of the substrate would remain cool. Yet if you try this, you find that the top 1μm has reached just 70°C with the layer at 10μm at 45°C. Even in that 1μs, a large amount of heat has gone into the substrate. Thermal conductivity is very efficient at taking heat away.

So for those who want to, say, sinter silver inks, with a (xenon) flash, the starting power density has to be at least 1kW/cm². That is a lot of power and modern lamps can deliver up to ~8 kW/cm² for less than 1ms. With these very large powers, then flash curing can take place - provided your 1μm coating can absorb all that power.

It is well known that xenon flashed can achieve this sort of curing. It is also said that powerful near-IR lamps can achieve similar results. With powers quoted as being in the 1MW/m² that sounds OK, till we see that that is "only" 0.1kW/m². To get to a temperature of 200°C in the top 1μm requires some rather unlikely conditions, as discussed below.

First, play with the settings to get the general idea. The maximum temperature reached in the coating is shown. The graph shows depth and temperature with time being colour-coded via the rainbow colours¹ - from short times in blues to long times in red. Moving the mouse over the graph gives you a readout of all the information. Details are provided below.

The two obvious settings are the flash power and the time of the flash. Then we have the heat capacity. For a typical polymer this is ~2000 J/kg.K (i.e. 2J/g.K when water is 4J/g.K). The higher it is, the more heat is absorbed from the surface. To avoid having an extra input for density, a value of 1200 kg/m³ is assumed. If your density is higher, then increase your heat capacity proportionally. The most critical value is the thermal conductivity. For many polymers this is around 0.2 W/m.K which is surprisingly large. For paper it is around 0.05 so it is very much easier to flash cure on paper. The final parameter is whether the rear surface temperature can float or not. If the substrate is in good thermal contact with a bottom plate then we can assume that the rear temperature is constant at 25°C. If the rear side is free in air then we can assume no heat loss from that surface within these timescales, so the temperature can float upwards.

Those who are wondering about the absence of inputs for the thickness, heat capacity and thermal conductivity of the absorbing layer, the answer is that they make so little difference to the end results that it is not worth the extra complication on the inputs. Whatever their values (within reason), the temperature is limited not by those values but by the heat loss into the substrate.

What happens to the temperatures after the flash? Slide the t_extra value to the desired time and the graph starts to curl back on itself, but you can still read out temperatures and times. The time steps are larger to reduce calculation overheads.

As mentioned above, to get a flash cure temperature of 200°C when you "only" have 1MW/m², it requires a thin coating that absorbs all the IR energy, a thin, floating substrate, a thermal conductivity less than 0.1 and a low heat capacity and a 5μs flash. The problem is that under those conditions the top 10μm of the substrate has already risen to greater than 100°C, so the flash is now in some danger of damaging the substrate.

As the near-IR flash cure systems are successful for many users, the point is not to denigrate them. Rather it is to show that one has to attend to the details of the whole system in order to get the maximum cure with the minimum damage to the rest of the system. These systems deliver their power over a large sample area so they are "real world" systems

The problem with the xenon flash is that their awesome power density is delivered over a rather small area, with a recovery time required between flashes. If the distance between lamp and substrate increases from, say, 10mm to 40mm, the power density drops by a factor of 2 and systems able to work over large areas at high speeds become very expensive.