In this webinar, Quantitative AFM Nanomechanics Data for Machine Learning and Materials Developments, we highlight how AFM-based nanomechanical measurements and machine learning can provide new insights into the mechanical properties of polymeric materials.
Watch the webinar recording to:
For material development, it is of interest to understand the relationship between the sample microstructure and its properties at the macroscale. Machine learning can be applied to this problem in several different ways to characterize the microstructure and identify components as well as predict structure-property relationships. Regardless of the application, a large amount of data is generally required to allow the ML model to be developed.
Atomic force microscopy can provide this data, provided an appropriate mode of operation is chosen.
The first part of this webinar is dedicated to the following goals:
In the second part, we discuss computational methods and ML algorithms dealing with data clustering (such as K-Means or Automatic Gaussian Mixture Model) that can be used to detect the different domains and (inter)phases in materials (e.g. polymer blends, hydrogels, nanocomposites, block copolymers, …) by partitioning the recorded data (i.e. the observables) into clusters according to their similarities. Additionally, based on the Tabor coefficient calculation, we also propose some protocols that can be easily implemented to rapidly determine which mechanical model(s) can be applied to obtain the quantitative mapping of the mechanical properties for each local domain or phase.
This algorithmically driven approach enables AFM users to analyze materials with more complex architectures and/or other properties (such as electrical ones), opening new avenues of research on advanced materials with specific functions and desired properties leading to the creation of functional and more reliable structural materials.
This webinar was presented on March 16, 2021
Bede Pittenger, Ph.D.
Bruker Nano Surfaces Sr. Staff Scientist, AFM Applications
Prof. Dr. Philippe LECLERE
Associate Professor, University of Mons