What if you could perform chemistry on the computer? That’s essentially what the ReaxFF method allows researchers to do — and it’s a tool that more than 1,600 researchers around the world have now requested access to. Developed by Adri van Duin, distinguished professor of mechanical engineering at Penn State, the ReaxFF method can be used in a variety of research fields, including 3-D printing, polymers and biomaterials.
What is the ReaxFF Method?
At the heart of ReaxFF is a combination of equations and parameters used to compute the forces between molecules and atoms. “The ReaxFF, or reactive force field, method allows us to perform simulations on complex materials and molecules including chemical reactions,” explains van Duin.
“ReaxFF enables fully reactive, fully dynamical simulations on systems of greater than 10,000 atoms — and if necessary, well beyond 1 million atoms,” he says. “It can access time-scales of greater than 10 nanoseconds. This is a unique capability. Very few other simulation tools can access these sizes and time scales.”
Wife and cofounder of RxFF Consulting, Inc. Diana van Duin notes, “Essentially, we do chemistry on the computer. What Adri did when he created the ReaxFF method built the framework that allows for chemical simulations. But for each material or for each reaction, you need a particular set of parameters, which are like a piece of software that goes into the larger software.”
ReaxFF Allows Researchers to Use Complex Molecular Dynamics to Optimize Materials
Like the television jingle suggests, “The more you know.” When it comes to research and material development, extensive knowledge is a must-have. That’s exactly what the ReaxFF method promises. Even better, it’s also a more cost-effective and efficient alternative to quantum mechanics, giving researchers and developers another option for investigating materials and developing new concepts.
“ReaxFF works very well for complex materials/molecule interfaces,” says van Duin. “Ideal uses include predicting reactivity, predicting potential synthesis paths and predicting thermal and chemical stability.”
To date, more than 1,600 researchers worldwide have used ReaxFF. Van Duin, along with his team, has used it in research efforts aimed at improving 3-D printing, and Penn State has used it as part of its efforts in developing its graphene membrane — a thin membrane with microscopic holes that may be the future of water filtering, desalination and hydrogen fuel cells. It has been used outside of Penn State in efforts ranging from understanding the reason behind silver tarnishing and the effects that space weather has on satellite materials. ReaxFF can also be connected to other simulation methods and used in biological-type applications.
“Some of the applications of ReaxFF to non-reactive material interfaces have been surprising to me, since it went outside the initial ReaxFF development targets,” notes van Duin.
“ReaxFF has a quite high level of transferability. It uses the same set of equations for all materials and molecules, and as such it has, even for non-reactive materials, become a useful method for simulating interfaces between materials and materials and molecules. For example, it can be used for enzyme interfaces with oxides and metals.”
What’s Ahead for ReaxFF?
Ever the innovator, van Duin envisions a future for the ReaxFF method that expands access and provides an improved user experience. “There is a plan to facilitate access to the ReaxFF parameter sets and ReaxFF input and output files through the Penn State 2DCC center.”
“Furthermore, we are looking to connect ReaxFF with various data science tools – particularly Machine Learning methods – that could enable us to speed up the ReaxFF parameter development considerably.” Along with other researchers, van Duin recently published a paper about the potential for integrating data science with ReaxFF parameter development.
“Within my company, we are looking for stronger connections with SCM Amsterdam, the developer of the AMS/ADF software, that features a ReaxFF module and a parameterization interface. By collaborating with SCM Amsterdam we hope to make these ReaxFF tools within AMS/ADF easier and more efficient to use,” says van Duin.
To obtain published parameter sets, researchers should send a request to the Material Computation Center through the following webform: https://www.mri.psu.edu/materials-computation-center/connect-mcc. Those requests go right to van Duin, who has received more than 1,500 requests over the past five years.