Dielectric Properties of Polymer Nanocomposite Interphases Using Electrostatic Force Microscopy and Machine Learning

Knowing the dielectric
properties of the interfacial region in
polymer nanocomposites is critical to predicting and controlling dielectric
properties. They are, however, difficult to characterize due to their
nanoscale dimensions. Electrostatic force microscopy (EFM) provides
a pathway to local dielectric property measurements, but extracting
local dielectric permittivity in complex interphase geometries from
EFM measurements remains a challenge. This paper demonstrates a combined
EFM and machine learning (ML) approach to measuring interfacial permittivity
in 50 nm silica particles in a PMMA matrix. We show that ML models
trained to finite-element simulations of the electric field profile
between the EFM tip and nanocomposite surface can accurately determine
the interface permittivity of functionalized nanoparticles. It was
found that for the particles with a polyaniline brush layer, the interfacial
region was detectable (extrinsic interface). For bare silica particles,
the intrinsic interface was detectable only in terms of having a slightly
higher or lower permittivity. This approach fully accounts for the
complex interplay of filler, matrix, and interface permittivity on
the force gradients measured in EFM that are missed by previous semianalytic
approaches, providing a pathway to quantify and design nanoscale interface
dielectric properties in nanodielectric materials.