Skip to Main Content

Strategic Business Insights (SBI) logo

Nanomaterials July 2018 Viewpoints

Technology Analyst: Marianne Monteforte

Neural Networks for Nanomaterial Design

Why is this topic significant?

Neural networks can dramatically speed up the design of nanoparticle systems with user-defined functionalities and can remove the need for extensive R&D efforts typically associated with traditional nanomaterial design processes.

Description

Recently, a team of researchers from the Massachusetts Institute of Technology (MIT) developed an artificial-intelligence- (AI-) enabled simulation tool—a neural network—that can predict the behavior of engineered spherical nanoparticles. The research team used the neural network to model the relationship between the structural morphology of these nanoparticles and their corresponding scattering behavior (which results in a specific spectrum of colors). These light-emitting nanoparticles are of great importance for applications such as displays, cloaking systems, and biomedical devices.

The tool uses computational neural networks effectively to "learn" how the structure of the nanoparticles affects particle behavior—in this case, how nanoparticles with individual geometries interact with, and scatter, various colors of light. On learning the relationship, the program can effectively run backward to design a particle with a set of user-desired light-scattering properties—the process is inverse design. Tests of the neural network revealed that the tool could predict the way a new nanoparticle scatters colors of light without the need for interpolating with existing examples.

Implications

Nanoparticles are very complex materials systems, and their behavior can rely on a number of properties, including structure, shape, size, and geometry. Typically, to understand nanoparticle systems and relationships better, scientists use computationally expensive simulation processes. The MIT team's development of a neural-network tool reduces the speed and complexity of nanomaterial simulations. However, the process also requires the researchers first to train the neural network and thus requires a large number of dataset examples. As a result, the accuracy of the predictions from these neural-network-simulation tools is highly reliant on good quality and high-volume datasets.

The MIT research team is not only developing these neural networks to simulate and create nanoparticles for practical applications but also seeking to gain a deeper understanding of the technique of creating these neural networks. Thus, the MIT research team is likely to further develop neural-network tools that guide rational nanomaterial design in the near future. Other research teams are also developing AI- enabled tools to speed up the design of nanomaterial systems, such as catalysts with high catalytic activity and stability (see the May 2018 Viewpoints). Players in the nanomaterials industry will likely benefit from forming partnerships with academia to help translate these AI-enabled nanomaterial design techniques to help address industry-specific problems.

Impacts/Disruptions

Materials-by-design research efforts are ongoing in academia, government, and industry. Much of the effort concentrates on developing AI-enabled simulation tools that facilitate the materials-discovery and -design process. However, limitations exist in the range and complexity of nanomaterial systems that today's simulation tools can handle. Nevertheless, as AI-enabled technologies develop, research teams will continue to turn to advanced techniques such as neural networks that will help to accelerate the process of understanding fundamental relationships between nanomaterials and their properties. These tools will enable the future development and large-scale commercialization of materials-by-design tool kits that will have important implications for the future of nanomaterials and the products and services they are made from.

Scale of Impact

  • Low
  • Medium
  • High
The scale of impact for this topic is: Medium

Time of Impact

  • Now
  • 5 Years
  • 10 Years
  • 15 Years
The time of impact for this topic is: Now to 5 Years

Opportunities in the following industry areas:

Nanoparticles, nanophotonics, materials-by-design, simulation tools

Relevant to the following Explorer Technology Areas:

Nanoparticles for Tumor Therapy and Diagnostics

Why is this topic significant?

Recently, a research team from Russia revealed the results of its preclinical trials of a new anticancer drug-nanoparticle system that doubled the lifetime of sick mice.

Description

Researchers from Russia's National University of Science and Technology MISiS and Pirogov Russian National Research Medical University developed and tested a new drug-delivery system for cancer-tumor therapy and diagnostics. The drug-delivery system consists of a cytostatic drug—doxorubicin (to destroy cancerous tumor cells), a vector molecule—vascular endothelial growth factor (to deliver the drug to the infected site), and magnetic iron-oxide nanoparticle (magnetite) carriers.

On introducing the drug-loaded nanoparticles into the cancerous site (in this case, the infected organs of mice) by intravenous injection, the nanoparticles accumulate and—on exposure to an alternating electromagnetic field—heat to 45°C, destroying the cancerous cells. Previous studies have shown that cancer cells are more sensitive to temperature changes than are healthy cells—thus increasing the probability that surrounding healthy tissue will remain intact during the heating process. The results of the research teams' preclinical study on the effectiveness of the novel drug-loaded nanoparticles system on mice show an increase in the median survival rate of mice with tumors by up to 50%.

Implications

The Russian research team's use of magnetite nanoparticles in the design of its novel anticancer drug-nanoparticle system demonstrates the vital role that these nanoparticles can play in advancing medical treatments. Magnetite nanoparticles have approval for clinical use by the US Food and Drug Administration, are readily accepted by the human body, and can break down easily in the body. The magnetic properties of magnetite nanoparticles provide them with the ability to convert their electromagnetic energy to heat and thus destroy cancerous cells.

Overall, nanoparticles show great promise to provide efficient therapeutics and accurate early-stage diagnostics in a range of medical treatments. Yet they still face a number of challenges before translation into clinical use. Some of the hurdles include reproducible manufacturing, efficient scale-up, accurate characterization, in vivo instability and bioavailability, epigenomic impact, potential toxicity, immune response, and regulatory barriers. Nevertheless, many research teams are focusing their efforts on overcoming these challenges and further revealing the benefits of using nanoparticles in a variety of medical applications. As a result, today, about 51 nanoparticulate therapeutics have approval for clinical use, and many more nanoparticulate therapeutics are in phase II and phase III clinical trials. Research in nanomedicine also receives wide support from government-funded organizations. For example, the European Commission is investing huge sums into nanomedicine-related research projects, on topics including targeted nanopharmaceuticals, nanodiagnostics, biomaterials for implants, and regenerative medicines.

Impacts/Disruptions

According to Grand View Research analysts, the global nanomedicine market is growing and by 2025 could reach $350 billion. The ability to control the behavior of nanoparticles—including their composition, size, biodegradability, morphology, and surface functionality—enables research teams to specifically design nanoparticles for medical applications and personalized treatments. These characteristics also greatly influence nanomaterials' potential utility for drug delivery and can influence how they interact with biological systems. Thus, for nanomedicines to reach their full utility, developers urgently need to understand more about these relationships.

Scale of Impact

  • Low
  • Medium
  • High
The scale of impact for this topic is: Medium

Time of Impact

  • Now
  • 5 Years
  • 10 Years
  • 15 Years
The time of impact for this topic is: 5 Years to 10 Years

Opportunities in the following industry areas:

Nanomedicine, cancer treatment, therapeutics, diagnostics

Relevant to the following Explorer Technology Areas: