If you thought a surfactant is some long-haired burnout who spends his days riding the waves at Malibu, you’d be very imaginative, but wrong. A surfactant—short for “surface active agent”—is much smaller than your typical surfer dude. It is, in fact, a molecule. A molecule that looks a bit like a tadpole and is most notable for one particularly interesting characteristic—it lowers the surface tension of the medium in which it’s suspended. When placed in water, for example, a surfactant essentially makes that water wetter.

But the really cool part is how a surfactant operates. The “head” of a surfactant tadpole is hydrophilic (attracted to water) and the “tail” is hydrophobic (repelled by water). Thusly, surfactant molecules added to a mixture of, say, oil and water, will rally to the interface between the two, burying their tails in the former while keeping their heads in the latter. This behavior inevitably breaks the oil into little droplets that mix well with water, a reaction that’s darned near impossible without the presence of surfactants.

Why do we tell you all of this? Because without surfactants, we’d all be a bunch of dirty little people with greasy hair, spotty dishes, and filthy clothes. You see, surfactants are the little gizmos that make the vast majority of our cleaning products do their job. From shampoos to detergents, surfactants remove the offending dirt blobs and then keep them from re-adhering to the surfaces from whence they came.

But not all surfactants are created equal. And concocting a better surfactant—and thereby a better cleanser or shampoo—is precisely why consumer goods manufacturers such as Procter & Gamble devote so much time to surfactant research. Lately, a lot of that research has transpired at Temple University’s Institute for Computational Molecular Science (ICMS), where scientists tirelessly probe the way surfactants act and interact—not only with each other, but also with all the other ridiculously tiny ingredients inside a bottle of laundry detergent or a drop of grease.

But that’s not all they do. The Temple gang also looks at soft water, hard water, and ion-heavy water. They build modified surfactants with longer tails, forked tails, a penchant for grouping together, and other characteristics we only dare imagine. The possibilities are seemingly endless.

In the past, this form of research has occurred in “wet labs,” where test after test is conducted in real-world conditions—often on live subjects, raising other sticky issues. However, it would literally take lifetimes to recreate the near-infinite number of microscopic permutations involved in comprehensive surfactant research in a wet lab. So, Temple’s ICMS researchers conduct the majority of their fundamental studies on computers and supercomputers, turning to the wet lab only when the options have been narrowed down to a much more realistic level.

But even supercomputers have a tough time expediently churning through the simulation mega-numbers and spitting out the necessary data. So the team at Temple recently changed up the routine and added a little something to the equation that didn’t merely reduce the waiting game, but fractionalized it. That something was a $400 graphics card.

An NVIDIA GeForce GTX 285 to be exact, the very same GPU gamers use to blitz through games such as Fallout 3. How is this possible? For starters, modern GPUs are incrementally better than traditional CPUs at “parallel processing,” a multi-thread form of computing that’s ideal for applications involving massive pools of data that require similar instructions. (See this issue’s "Game On" feature for more information.)

 Secondly, researchers all over the world now have HOOMD. That’s right, HOOMD. Short for “Highly Optimized Object-oriented Many-particle Dynamics,” HOOMD is a nifty bit of software that takes advantage of NVIDIA GPUs to perform hyper-speed molecular dynamics simulations.

The result? Stunning, according to ICMS Associate Director, Professor Axel Kohlmeyer. “If you compare a desktop with a single GPU to a BlueGene/L (one of the most powerful of supercomputer architectures), this single GPU is faster than 512 BG/L CPUs. Mind you, BG/L computers are designed to house a huge number of CPUs in a very power-efficient way, and thus, its CPUs run at approximately twenty percent the speed of a (high-end) desktop CPU. Regardless, with current generation CPUs, we get about a 50x speed-up relative to a single processor core by using a GPU.”

Yet better cleaning isn’t the only benefit of Temple’s research. “There are applications in understanding how anesthetics work, finding potential binding sites for drugs to inhibit protein-protein binding (a hot new trend in drug research with applications such as anti-cancer and anti-HIV drugs), and designing ‘nano-capsules’ for drug delivery,” says Kohlmeyer. By encapsulation in surfactant-like molecules that spontaneously self-assemble, Kohlmeyer asserts, such drugs are not only better protected, but also tunable to be more targeted and to deliver fewer unwanted side effects.

Kinda makes you want to treat your Head & Shoulders with a little more respect, doesn’t it?