Machine Learning of Microscopic Ingredients for Graphene Oxide/Cellulose Interaction

Abstract

Understanding the role of microscopic attributes in nanocomposites allows one to control and, therefore, accelerate experimental system designs. In this work, we extracted the relevant parameters controlling the graphene oxide binding strength to cellulose by combining first-principles calculations and machine learning algorithms. We were able to classify the systems among two classes with higher and lower binding energies, which are well defined based on the isolated graphene oxide features. Using theoretical X-ray photoelectron spectroscopy analysis, we show the extraction of these relevant features. In addition, we demonstrate the possibility of refined control within a machine learning regression between the binding energy values and the system’s characteristics. Our work presents a guiding map to control graphene oxide/cellulose interaction.

Cellulose nanofibrils–graphene hybrids: recent advances in fabrication, properties, and applications

“…103 Nevertheless, GO still suffers from some problems such as low conductivity, which can be obviously avoided through the conversion of GO to rGO, leading to high thermal and electrical conductivity by reducing the oxygenated groups and approximating toward the graphene phase, superhydrophobicity, high-quality interface, and even effective catalysis properties. [188][189][190] As reported in subsection 5.3, CNFs/rGO exhibit promising characteristics and can be applied in several fields. Nonetheless, it is worth mentioning that a higher amount of rGO in the hybrid would harshly damage the mechanical features and hence reduce the performance, which would conceivably inhibit its applications in various areas.…”

“…103 Nevertheless, GO still suffers from some problems such as low conductivity, which can be obviously avoided through the conversion of GO to rGO, leading to high thermal and electrical conductivity by reducing the oxygenated groups and approximating toward the graphene phase, superhydrophobicity, high-quality interface, and even effective catalysis properties. 188–190…”

Recent Advances of Graphene and Related Materials in Artificial Intelligence

“…The QSPR used for properties prediction is first built by ML models, plenty of candidates (unexplored) are then screened through the QPSR to perform inverse design. [ 28 ] Once the topological or geometrical features are encoded by a set of structural fingerprints or descriptors, the material property is mapped into structural features. The resulting QSPR is efficient to discriminate potential nanomaterial candidates from virtual graphene libraries or highlight the most influential structural features for structural design.…”

“…The oxygenated group density of GO is the primary attribute ruling the binding energy scale. As a result, the refined control over the binding energy based on the variation of the oxygen density in GO is an effective way [28a] …”

Nanoscale structural and electronic properties of cellulose/graphene interfaces

“…However, they only focus on the global electrical response and ignore the effect of nanocellulose/2D interaction at the nano and atomic levels on the electronic properties. Our group recently used first-principles calculations with a machine learning approach to evaluate relevant chemical and structural parameters that govern the binding energy of graphene oxide/ nanocellulose interfaces, 19 which have been considered promising polymeric composites for gas barriers 20 and water decontamination. 21 In the same direction, Zhu et al 22 employed first-principles methods to study the interface bonding behavior of the composite systems of graphene oxide and cellulose derivatives.…”