Search

A team at the Harvard MRSEC led by Bertoldi and Rycroft  has  developed  a  framework  to  design mechanical  metamaterials  with  target  nonlinear response. Neural networks were used to accurately learn  the  relationship  between  the  geometry  and nonlinear   mechanical   response   of   these metamaterials.

The  Partnership  for  Research  and  Education  in Materials  between  Navajo Technical  University  and the   MRSEC   based   at   Harvard   focuses   on developing culturally-informed, sustainable pathways into materials science-related careers and advanced studies  for  Navajo  students.

The University of Tennessee, Knoxville's Center for Advanced Materials and Manufacturing has introduced the Element-Weighted Kolmogorov–Arnold Network, a novel interpretable ML architecture that predicts crystal energy landscape properties — formation energy, band gap, and work function — directly from chemical composition. EWKAN achieves state-of-the-art accuracy across large-scale databases, matching or exceeding GNN-based models that require full 3D atomic structure inputs, while using orders of magnitude fewer parameters.

In February 2024, I-MRSEC investigator Daniel Shoemaker, grad student Emily Waite, and outreach coordinator Pamela Pena Martin taught 35 7th and 8th graders at  Franklin STEAM Academy, a Champaign public middle school, about magnetism through a kit they developed, supported by the I-MRSEC and a grant from the APS Group on Magnetism and its Applications. This visit was part of an annual 7-week program that teaches materials science concepts through hands-on activities aimed to build interest and confidence in STEM.

Triangular monomers with positive curvature in one direction and negative curvature in another assemble into trumpet shaped objects predicted to have precise self-limited lengths due to frustration-induced stress.

Northwestern  University  MRSEC  IRG-2  has  developed  an efficient  theoretical  framework  based  on  high-throughput density functional theory calculations and machine learning methods  to  accelerate  the  discovery  of  heteroanionic materials.

An interdisciplinary team of CAMM IRG1 researchers developed a quantitatively reliable helium-benzene potential energy surface with quantum chemical accuracy by combining CCSD(T)/CBS electronic-structure and a multifidelity Gaussian process model that merges sparse high-accuracy data with dense lower-cost DFT data. This is an important result for a weakly bound quantum system in which small errors in the interaction potential lead to materially different many-body predictions.

A new Materials Science From CU program, Light and Energy, enables students to compare the amount of energy needed to illuminate different types light bulbs.  Seen in this photo during the annual STEMapaloozza event held in Denver, Center graduate student Chenhui Zhu encourages a middle student to generate the power to light an array of incandescent light bulbs.

By  employing  redox-active  and  redox-inactive  polymers  in  a mixed-dimensional   heterostructure   architecture,   Northwestern University    MRSEC    IRG-1    has    achieved    vertical    organic electrochemical  transistors  (vOECTs)  for  the  first  time.

New electron microscopy technique developed at UIUC show that magnetic order in antiferromagnetic Fe2As can be resolved at angstrom resolution for the first time.