In the 21st century, chemical engineers are just about as likely to be producing fizzy drinks and pharmaceuticals as they are to be pumping out the polyolefins. It’s the price you pay for versatility. Of course, there’s a bit more to it than simply the versatility of the people. The theories they rely on are highly transferable, too. However, it now turns out that one such theory — nucleation — might not be quite what our college lecturers would have us believe.
For the best part of a century it’s been accepted that nucleation occurs at nucleation sites. If such sites involve suspended particles or minute bubbles you get heterogeneous nucleation, otherwise you get homogeneous nucleation.
Textbooks tell us that a systematic process of nucleation and growth is what turns a boiling liquid into a vapor. First the liquid forms tiny nuclei — microscopic bubbles — that pick up particles and grow accordingly. It’s a fundamental process and of particular importance to anybody who’s ever tried to get a fluid through a pipe or vessel.
However, it appears that the nucleation community has been operating for many generations with something less than a full grasp of the facts — at least as far as homogeneous nucleation is concerned, according to research by chemical engineers at Purdue University, West Lafayette, Ind.
“Our findings indicate that this is not what’'s going on,” says David S. Corti, a Purdue associate professor of chemical engineering. “The bubble grows via a mechanism very different from classical nucleation theory.” Working with doctoral student Mark Uline, Corti found that the molecular mechanism by which a liquid becomes a vapor is nowhere near as strict as the received wisdom would have us believe.
Corti uses the analogy of a mountain pass to explain the process. “In order to get from the liquid to the vapor, you have to go over this mountain pass,” he explains. “If you climb up and you’re not quite at the top, sometimes you can roll back down, but if you get to the top, you can roll down to the other side and get to the vapor phase.”
His research has shown that this mountain pass is broader and flatter than previously thought, meaning there are several possible pathways responsible for the phase transition.
At the same time, very curiously, once you reach the free energy of bubble formation and get over the mountain pass, the surface disappears. “You look at one side and you see the mountain and think everything is okay, but once you climb over, it’s as if the mountain disappears on the other side,” he says.
Textbooks would have us believe that bubbles forming down the pass could reverse direction and head back to the liquid phase. Not so, says Corti: “In our view, as soon as you get over the top of the mountain, the mountain disappears. You have no choice but to plummet to something else, the vapor phase.” The details of Corti and Uline’s work appeared in August’s issue of the journal Physical Review Letters.
Since then, however, Corti’s research has moved on and he’s now tackling the formation of droplets. “So far our findings are very much in line with our bubble work, except that we find that on the other side of the mountain the droplets are channelled into stability areas,” he says. “Now we have to understand what it means if the mountain disappears.”
Corti believes that these apparently strange nucleation processes can be explained in terms of spinodal decomposition. Here, phase separation is delayed until the system enters the unstable region where a small perturbation in composition leads to a decrease in energy and thus spontaneous growth of the perturbation.
To pin down exactly what is going on, Corti is using both density functional theory and dynamical density functional theory. The first should help establish the point of equilibrium while the second should give an insight into how that equilibrium evolves.
“This will help us to get the characteristics of how growth occurs,” he says. He hopes to have a much better grasp of what is really going on in five to six months.
But with such a fundamental piece of accepted wisdom taking a bit of a bashing, one wonders whether Purdue has been overrun with chemical engineers keen to get a better grip on their processes.
“Well, we had our 15 minutes of fame back in August and the nucleation community is quite small really,” concludes Corti. “The point is that although the previous theories don’t work well, they work well enough for the people who use them. I think we’ll garner more interest down the road when the key issues of bubble and droplet formation are hammered out.”
Sean Ottewell, editor at large