Transport & Nanostructures Group

Our first publication of 2026 “Accelerated Discovery of Graphene Kirigami with Enhanced Elastocaloric Effect via Machine Learning” has just appeared in Nano Letters, published by the American Chemical Society.

Two-dimensional materials, particularly graphene, have captivated scientific interest due to their extraordinary properties. One promising structural paradigm is graphene kirigami (GK), where precise nanocuts are introduced into the graphene lattice to form complex architectures inspired by the Japanese art of paper cutting. Recent experimental advances have demonstrated that GK can be fabricated through top-down patterning strategies, enabling precise control over cut geometry. Specific motifs in graphene kirigami structures can display effective elastic constants spanning several orders of magnitude. In addition to altering the mechanical flexibility of graphene, kirigami patterns may also allow control of thermomechanical properties, including the elastocaloric effect, in which a material undergoes a reversible temperature change upon a change in strain due to a compressive or tensile stress, under adiabatic conditions.

Recent studies have examined the elastocaloric response of graphene kirigami (GK) and shown how it may be tailored through geometric design. This tunability makes GK a promising platform for applications in nanoscale solid-state thermal devices. In this work, we combined molecular dynamics (MD) simulations and machine learning (ML) to explore how GK geometries affect the elastocaloric coefficient (ECC), defined as the adiabatic ratio between temperature change and applied tensile stress. A data set of 16,807 GK configurations was generated through systematic cut patterns and evaluated via MD at room temperature. Using this data, both classical and deep-learning models were trained, with a convolutional neural network (CNN) achieving the best performance. Finally, a model-guided optimization identified high-ECC designs 10 times faster than random search, demonstrating the power of ML-assisted strategies for the accelerated discovery of advanced elastocaloric materials.

The results stem from the thesis work of Ph.D student Franklin Ferreira da Silva Filho.

Our latest publication of 2025 “Consensus effects of social media synthetic influence groups on scale-free networks” has just appeared in Chaos, Solitons & Fractals.

Online platforms for social interactions are an essential part of modern society. With the advance of technology and the rise of AI algorithms, content is now filtered systematically, facilitating the formation of filter bubbles. This work investigates the social consensus under limited visibility in a two-state majority-vote model on Barabási–Albert scale-free networks. In the consensus evolution, each individual assimilates the opinion of the majority of their neighbors with probability (1-q) and disagrees with chance q, known as the noise parameter. We define a visibility parameter as the probability of an individual considering the opinion of a neighbor at a given interaction. This parameter enables us to model the limited visibility phenomenon that produces synthetic neighborhoods in online interactions. We employ Monte Carlo simulations and finite-size scaling analysis to obtain the critical noise parameter as a function of the visibility and the network growth parameter. We also find the critical exponents of the system and validate the unitary relation for complex networks. Our analysis shows that installing and manipulating synthetic influence groups critically undermines consensus robustness, leading to serious consequences in the modern society.

Although this is not our main line of work, it is a type of problem I also enjoy working on. The results stem from the work of graduate student Giuliano Porciúncula, who is supervised by my colleague André L. M. Vilela at UPE.