The ability to manipulate domains underpins function in applications of ferroelectrics. While there have been demonstrations of controlled nanoscale manipulation of domain structures to drive emergent properties, such approaches lack an internal feedback loop required for automatic manipulation. Here, using a deep sequence-to-sequence autoencoder we automate the extraction of latent features of nanoscale ferroelectric switching from piezoresponse force spectroscopy of tensile-strained PbZr0.2Ti0.8O3 with a hierarchical domain structure. We identify characteristic behavior in the piezoresponse and cantilever resonance hysteresis loops, which allows for the classification and quantification of nanoscale-switching mechanisms. Specifically, we identify elastic hardening events which are associated with the nucleation and growth of charged domain walls. This work demonstrates the efficacy of unsupervised neural networks in learning features of a material’s physical response from nanoscale multichannel hyperspectral imagery and provides new capabilities in leveraging in operando spectroscopies that could enable the automated manipulation of nanoscale structures in materials.

No abstract

Lithium (Li) dendrite growth poses serious challenges for the development of Li metal batteries. Replacing liquid electrolyte with solid composite electrolyte embedded with nanofiller additives can potentially suppress the Li dendrite growth. However, the underlying mechanism is still not fully understood, and most theoretical works focus on pure liquid electrolyte and ignore the mechanical strain effects. Here we developed a phase-field model to simulate the Li dendrite growth by incorporating the microstructure of the solid composite electrolyte and considering the mechanical effects of the electrolyte. Using aluminum oxide nanofiber embedded poly(vinylidene fluoride-co-hexafluoropropylene) (P(VDF-HFP)) as an example, we discovered two key factors, the elastic modulus and the electrolyte nanochannel width that govern the Li dendrite growth. The difference of the Young's modulus between the Li metal and solid electrolyte acts as the additional mechanical driving force, which partially offsets the electrochemical driving force to either promote or inhibit the dendrite growth. We also discovered that the introduction of the 1D nanofiber arrays could confine the Li ion transport along vertical direction, reduce the concentration gradient across the metal/electrolyte interface, and inhibit the Li dendrite growth. Finally, the dependence of overall Li ion conductivity on the nanofiller is discussed. Our work provides deep understanding and designing strategy for the solid composite electrolyte for improved Li anode stability and Li ion conductivity.

We developed a physical model to fundamentally understand the conductive filament (CF) formation and growth behavior in the switching layer during electroforming process in the metal-oxide-based resistive random-access memories (RRAM). The effects of the electrode and oxide layer properties on the CF morphology evolution, current-voltage characteristic, local temperature, and electrical potential distribution have been systematically explored. It is found that choosing active electrodes with lower oxygen vacancy formation energy and oxides with small Lorenz number (ratio of thermal and electrical conductivity) enables CF formation at a smaller electroforming voltage and creates a CF with more homogeneous morphology. This work advances our understanding of the kinetic behaviors of the CF formation and growth during the electroforming process and could potentially guide the oxide and electrode materials selection to realize a more stable and functional RRAM.

Domain walls and topological defects in ferroelectric materials have emerged as a powerful tool for functional electronic devices including memory and logic. Similarly, wall interactions and dynamics underpin a broad range of mesoscale phenomena ranging from giant electromechanical responses to memory effects. Exploring the functionalities of individual domain walls, their interactions, and controlled modifications of the domain structures is crucial for applications and fundamental physical studies. However, the dynamic nature of these features severely limits studies of their local physics since application of local biases or pressures in piezoresponse force microscopy induce wall displacement as a primary response. Here, we introduce an approach for the control and modification of domain structures based on automated experimentation, whereby real-space image-based feedback is used to control the tip bias during ferroelectric switching, allowing for modification routes conditioned on domain states under the tip. This automated experiment approach is demonstrated for the exploration of domain wall dynamics and creation of metastable phases with large electromechanical response.