Anyone who's ever lathered up knows the dilemma.
The same qualities that make surfactants - the chemical compounds in soaps, shampoos and detergents that penetrate grease, dissolve stains and make those satisfying suds in the shower - so effective as cleaners can also act as irritants. When splashed in the eyes, some of these products can cause itching, burning and tears.
"Grease and oils are chemically similar to the membranes in the human body, so accidental exposure of the eyes to these cleaners may also irritate cell membranes," said Dilnoza Amirkulova, a research scientist for consumer products giant Procter & Gamble, or P&G. "We need to understand which surfactants in cleaners can cause eye irritation to ensure our products are safe."
Simulations performed on the Summit supercomputer at the Department of Energy's Oak Ridge National Laboratory could help wash away those risks.
The Oak Ridge Leadership Computing Facility - formerly home to ORNL's Summit, then one of the top 10 fastest supercomputers in the world - routinely offers its leadership-class computing resources to researchers from government, academia, and industry grappling with the largest computing problems in science. Each year, researchers use OLCF resources to achieve breakthroughs in such fields as aerodynamics, biology, chemistry, seismology, engineering, energy, and materials science.
Amirkulova and a team of P&G colleagues used the 200-petaflop IBM AC922 Summit system to create a digital model of the corneal epithelium, the primary outer layer of cells covering the human eye, and test that model against a series of surfactants.
"We take great pride and satisfaction in our many partnerships with industry," said Bronson Messer, the OLCF's director of science. "These collaborations enable leaps in scientific understanding with benefits that people everywhere can see in their daily lives."
P&G, an $82 billion Fortune 100 company, enjoys worldwide recognition for such household mainstays as its shampoos, surface cleaners, and laundry and dishwashing detergents. The company's iconic brands include Tide detergent, Head & Shoulders shampoo, Cascade dish cleaner and Mr. Clean all-purpose cleaner. As the world's largest company for consumer packaged goods, P&G has committed to global sustainability goals through its Ambition 2040 initiative, which seeks to reduce plastic and paper packaging for all products.
The study on Summit grew out of the firm's commitments to product performance, safety and environmental sustainability. To meet its Ambition 2040 goals to reduce packaging waste, P&G needed to concentrate its liquid cleaners to deliver as many or more washes in smaller packages. Concentrated cleaners equal less packaging, fewer rinse cycles, quicker results, less water used and a shrunken carbon footprint. P&G also sought to ensure the concentrated formulas could meet the company's safety standards.
"Our goal was to understand better what causes irritation of the eye and how we can design new surfactants that don't irritate while still cleaning effectively," Amirkulova said. "At P&G, we're committed to delivering the product performance that consumers expect while still meeting safety criteria."
Ethical guidelines ruled out testing on human and animal subjects, which also would have been expensive and time-consuming. The team needed a way to capture the initial contact between the cleaning compounds and the corneal membrane in exacting detail and at every potential angle, starting from the first microsecond, a millionth of a second.
Those criteria made simulation the most promising option.
"Computational chemistry is ideal for such small scales," Amirkulova said. "These surfactants we're testing are really small chemical systems invisible to the naked eye that measure about 30 × 30 nanometers (a millionth of a millimeter), and the membranes we're studying are about 20 nanometers. Due to some experimental limitations, it would be difficult to visualize the process experimentally. We needed those details to capture these mechanisms at work at the molecular level, but we weren't sure our in-house computing resources could handle such calculations. We needed substantially more computing horsepower, the kind that only a machine like Summit could provide."
Eyeing the complexities
The team turned to the Oak Ridge Leadership Computing Facility and received an allocation of time on Summit, which has since been decommissioned.
The computational power of Summit enabled the team to use PACKMOL , a software package for simulating molecular dynamics, to build a scalable model of a liposome - a sac formed by double layers of fatty acids and filled with water molecules - that mimicked the behavior of the corneal membrane. The team then used GROMACS , another software package, to test various amounts of surfactants against the liposome and observe the results for the first three microseconds of contact.
Surfactants in shampoo and other cleaners that irritate the eye do so by disrupting the outer corneal membrane to create microscopic holes that allow chemicals to penetrate to the tissues beneath. The simulations on Summit showed increasing the concentration of surfactants studied didn't necessarily increase damage to the membrane, at least in contact with the liposome alone.
Increased concentration instead caused the surfactant molecules to cluster into micelles - tiny blobs of molecules held together by themselves and surrounded by water. Certain surfactants tended to be more hydrophilic, or attracted to water, and thus more likely to form micelles and less likely to penetrate the membrane. Others tended to be more hydrophobic, or repelled by water, and thus more likely to disrupt the membrane and cause irritation.
"Summit was amazing," Amirkulova said. "We'd never simulated a liposome before, never simulated making holes in a liposome. Not only did we make a stable liposome; we also were able to run the simulation at scale and model a whole liposome instead of just a part of it and to observe the liposome being disrupted by external forces."
The results could point to formulas for gentler, less-irritating surfactants.
"The simulations showed us for the first time that we can potentially double surfactant concentration without doubling the disruption to the outer eye membrane," Amirkulova said. "We think that's a promising route for further study as we develop the next generation of surfactants."
The P&G team plans to publish their findings soon. Next steps could include expanding the simulation by adding proteins and other substances normally found in the eye to the model. Amirkulova plans to draw on insights from the simulations to build a simpler digital tool for predicting liposome disruption and the resulting irritation.
The OLCF, a DOE Office of Science user facility at ORNL, is supported by the DOE's Advanced Scientific Computing Research program.
UT-Battelle manages ORNL for DOE's Office of Science, the single largest supporter of basic research in the physical sciences in the United States. DOE's Office of Science is working to address some of the most pressing challenges of our time. For more information, visit https://energy.gov/science . - Matt Lakin
