Modeling, predictive analytics, and smarter data management can help reduce the number of experiments, optimize corporate IP, and save time and money.

Nanoparticle Water 250

A representation of a silica nanoparticle in an aqueous environment. Modeling and simulation have been quite successful in developing new nanomaterials. Image: Accelrys, Inc.

Organizations engaging in materials R&D are caught in the cross hairs of several competing challenges. Product cycle times are shorter than other research-intensive industries like pharmaceuticals, and are shrinking further as organizations seek to discover profitable materials and formulations before the competition.

At the same time, materials science is notoriously complex—millions of elemental combinations may need to be explored to build a new alloy, polymer, or catalyst, and experimental processes require a high level of customization, as well as specialized knowledge and equipment. To speed time-to-innovation, organizations need to be able to reduce the amount of expensive and painstaking trial-and-error experimentation without hampering researchers' abilities to generate quality leads.

Technology that facilitates "predictive" analytics, including modeling and simulation of materials, presents a compelling alternative. Used widely in pharmaceutical research, software-enabled scientific modeling and analytic techniques make it possible for researchers to design and test products in silico. Instead of running multiple experiments in a lab, researchers can use simulations to virtually explore a broad range of ideas and identify promising compound candidates before doing any actual synthesis. This not only significantly reduces the time and expense that goes into materials experimentation and screening, it also allows researchers to investigate more hypotheses than would be possible otherwise.

Two critical conditions need to be met for these techniques to add value for materials researchers. One, the predictive technologies deployed must be both automated and simple enough to use so that materials scientists don't need to be modeling experts. And two, researchers must be able to quickly and easily access, integrate, and share relevant data—both experimental and simulated—that can be leveraged to drive the materials discovery process.

Accelrys kept these conditions in mind while designing the Materials Studio Collection for its scientific informatics platform, Pipeline Pilot. Pipeline Pilot allows project stakeholders to retrieve, merge, analyze, report, and share research data across their enterprises. Based on a services-based open architecture, it supports the plug-and-play integration of multiple sources of information, including solid-state materials data, modeling results, and analytical instrument reports.

Predicting materials properties
The Materials Studio applications operate through this platform and enable users to predict key properties, simulate analytical instrumentation, and identify leads for catalysts, polymers, nanotubes, and other types of materials. These applications use techniques based on quantum mechanics or molecular mechanics to describe materials at the atomic level, yielding properties such as the binding strength of a new adhesive or the tensile strength of a new alloy. Pipeline Pilot's drag-and-drop workflow capabilities enable scientists to customize and automate complex modeling and analysis steps, so that specialized expertise can be captured and re-used by others.

Accelrys Materials Studio applications were recently deployed by specialty chemicals developer Johnson Matthey to accelerate its search for an affordable and high-performing alternative to platinum in fuel cells. With the software, the company was able to investigate thousands of virtual combinations of metals in a fraction of the time that would have been spent on lab experimentation alone. Statistical analysis tools were used to establish correlations between materials composition, stability, and reactivity; re-usable simulations allowed researchers to computationally investigate the expected performance of different ratios of metals. Assisted by automated data integration, the researchers reported saving time and money in narrowing down its list of the most promising alternatives to platinum.

Published in R & D magazine: Vol. 52, No. 6, October, 2010