Solutions to corrosive problems are often found by looking at both sides of the situation.

It was in that spirit that researchers at the Energy Department’s Argonne National Laboratory looked into the corrosive problem of, well, corrosion. One of the world’s most costly and most common chemical reactions, it typically shows up as rust. And as any automobile owner knows — especially at the spring end of such a bitter winter — it happens at the boundary between water and a metal.

To find solutions, researchers have typically looked at corrosion from the side of the metal. They’ve discovered many ways to prevent, or at least postpone, the problem through rust-proof coatings or metal alloys. Yet the flakes are still flying … and the challenge of corrosion continues.

So a team of researchers, led by Subramanian Sankaranarayanan from Argonne Lab’s Center for Nanoscale Materials, decided to consider corrosion from the other side. They took a peek at the problem from the perspective of the water.

To understand why they did so, it’s worth remembering that corrosion works a bit like a battery — it’s an electrochemical process that involves the transfer of charges. Pure water, composed of just hydrogen and oxygen, doesn’t cut it. Rather, corrosion needs positively and negatively charged particles (called ions) to flow through a solution in order to kick off the rusting process.

Those ions often form when substances known as ionic solids are plopped into water. Sounds complicated, but it can be as simple as table salt (sodium chloride), which dissolves into positive sodium ions and negative chlorine ions when dropped into water. When near a metal boundary (say that of a car door), the ions can move and mix, causing corrosion. This explains why cars tend to develop rusty spots during icy winters, but can drive unscathed through rainy summers.

As published in the journal Physical Review E, Sankaranarayanan’s team — which includes researchers from Harvard University and the University of Missouri — used high-performance computers such as those available at Energy Department Labs to simulate how chloride ions drive corrosion at the atomic level. They found that when relatively few chloride ions are present, water tends to form ordered layers near the metal boundary, and that increasing the amount of chloride (its concentration) didn’t increase the amount of corrosion … up to a point. However, at a critical threshold, the concentrated chloride ions seem to break up the ordered water layers and thereby increase the rate of corrosion.

That transition, which happens in less than a billionth of a second, also decreases the freezing temperature of the solution. This helps explain why snowy roads and sidewalks are slathered with salt in the first place — its ions drive disorder, which makes the frozen stuff melt at a lower temperature. This research is significant because it provides detailed insight into the critical early-stage formation of corrosion conditions on the nanosecond time scale and at the atomic level.

There’s much more to be learned by looking at corrosion through watery lenses. So to that end, Sankaranarayanan’s team has been awarded some 40 million processor-hours on Mira, Argonne Lab’s Blue Gene/Q supercomputer, which they will use to study a variety of corrosion scenarios, and hopefully find ways to improve the durability of materials.

Corrosive challenges remain. But through research at the Energy Department’s National Labs, sophisticated simulations are leading to better solutions.