Enabling green conversion of cellulose to reduced graphene oxide for biomedical applications

Abstract

Since its discovery in 2004, graphene has been increasingly explored for biomedical applications due to its large surface area, mechanical strength, and outstanding electronic and physical properties. However, its clinical translation is hindered by cytotoxicity concerns, primarily related to reactive oxygen species (ROS) generation induced by oxygen-containing functional groups. Reduced graphene oxide (rGO) exhibits lower toxicity than graphene oxide and pristine graphene, yet the specific effects of different oxygen functional group types and ratios on ROS generation remain insufficiently understood. To clarify this relationship and improve the biocompatibility of rGO materials, precise control over oxygen functionalities is required, however, current tuning strategies often pose environmental challenges. Biomass has attracted attention as a sustainable carbon source for graphene production, but its inherent heterogeneity across species and batches poses significant challenges for reproducibility and performance control. Among biomass components, cellulose stands out due to its structural uniformity and high abundance. These features make it particularly suitable for controllable deoxygenation and carbonization, enabling the production of biochar with high and tunable carbon-to-oxygen ratios. Thermal conversion of cellulose followed by ultrasonication offers a green route to generate rGO materials. Investigating how oxygen functionalities in cellulose-derived rGO influence biomedical performance is, therefore, critical for advancing safe and sustainable applications.

This thesis aims to enable the green synthesis of rGO from cellulose and to systematically investigate the role of oxygen functional groups in determining the biomedical performance of cellulose-derived rGO, through an integrated theoretical and experimental framework. Three major questions are addressed: 1) What are the reaction mechanisms in the cellulose pyrolysis process? 2) How is cellulose-derived biochar transformed into graphene materials? 3) Why do cellulose-derived graphene materials exhibit enhanced biocompatibility, and how does the adjustment of functional groups influence the performance of graphene materials?

Chapter 4 focuses on understanding and enabling the green conversion of cellulose into rGO. Conventional synthesis routes typically rely on harsh chemical reagents, which not only pose environmental hazards but also introduce impurities that limit biomedical applications. Therefore, a sustainable and chemical-free synthesis pathway is required. In this work, rGO was synthesized via a reduction-then-exfoliation approach using cellulose as a renewable precursor. During the process, cellulose was thermally converted into a graphite-like structure with a high carbon-to-oxygen ratio, while preserving sp² domains prior to exfoliation. The resulting cellulose-derived rGO exhibited a 30% enhancement in near-infrared (NIR) photothermal efficiency, validating its potential for photothermal therapy (PTT) applications. Moreover, life-cycle assessment revealed a 50% reduction in environmental burden and improved economic viability compared with conventionally prepared rGO. To further increase yield, green surfactants were employed to regulate interlayer distance, guided by computational modeling, achieving a 40% improvement in rGO production. Overall, this work presents a high-performance, eco-efficient route for rGO synthesis and establishes a generalizable framework for the design and optimization of biomass-derived carbon materials for potential biomedical applications.

Chapter 5 provides a detailed investigation of the formation mechanism of graphite-like structure, valuable and greenhouse products from cellulose pyrolysis, which underlies the conversion process described in Chapter 4. Understanding cellulose pyrolysis is a critical step in converting it into graphite-like biochar and utilizing the by-product biogas, offering a sustainable pathway for oxygen content control, renewable energy generation, and carbon mitigation. To address the lack of mechanistic insight from experiments, molecular dynamics (MD) simulations based on an accurate deep-learning potential field (DLPF) that >90% agreement with density functional theory (DFT) are developed and applied. An optimized kinetic model is constructed based on MD outputs and validated experimentally, reducing predictive deviation by up to 95% compared to existing models. This enables effective design and operation of cellulose pyrolysis, promoting wider biomass utilization and establishing its role in global carbon neutrality.

Chapter 6 focuses on the molecular modeling of graphene and the theoretical study of how oxygen functional groups affect its performance. Since the thermal conversion process in Chapter 4 adjusts the oxygen functional groups in cellulose-derived graphene materials, this chapter investigates their influence at the atomic level. The elastic and thermal properties of graphene, which are important for biomedical applications such as photothermal therapy and implantable devices, were analyzed. Deep-learning-assisted MD simulations were fitted to X-ray diffraction and Fourier transform infrared data to build experimentally validated graphene models. The results show that epoxy-rich GO/rGO has the largest reduction in elastic modulus, while optimized sp²/sp³ hybrid structures increase the modulus by up to 19% compared with pristine graphene. Non-equilibrium simulations further indicate that higher oxygen content, especially epoxy groups, greatly lowers thermal conductivity, whereas carboxyl groups have a weaker effect. These findings provide molecular-level understanding of how oxygen content affects the properties of graphene materials and form a basis for studying how functional groups influence the biocompatibility of cellulose-derived rGO.

Chapter 7 investigates the effect of oxygen functional groups on cytotoxicity through a combined experimental and theoretical approach, bridging the fundamental understanding from previous chapters to biomedical applications. This chapter proposes a green synthesis strategy that integrates cellulose pyrolysis with ultrasonication to produce few-layer rGO with tunable C/O ratios (8.7-11.3) and well-defined physical properties, without the use of toxic reagents. Cell viability assays reveal that rGO produced at 700 °C exhibits optimal biocompatibility (2.1-fold better biocompatibility than commercial product) and selective tumor cytotoxicity, whereas deviations from this temperature in either direction increase cytotoxicity. As the pyrolysis temperature increases from 600 to 900 °C, the -OH content is significantly reduced, the COOH content slightly decreases, and the -O- groups remain nearly unchanged. These trends are confirmed by material characterization and supported by DFT-based calculations of formation and dissociation energies. DFT and MD simulations further demonstrate that COOH groups facilitate ROS generation, whereas -OH groups suppress it. Moreover, MD simulations reveal that the spatial arrangement of functional groups also plays a critical role in modulating ROS production, consistent with experimental observations. Finally, the PTT efficiency of the cellulose-derived rGO materials was evaluated, showing up to a fourfold increase in tumor-killing efficiency compared with conventionally synthesized rGO reported in the literature. Overall, this chapter offers a framework for the synthesis of biomass-derived graphene materials with enhanced biocompatibility, advancing their applicability in biomedical fields.

Together, these four chapters construct a unified and multiscale roadmap for realizing biocompatible, high-performance graphene materials from sustainable sources. By combining deep-learning-driven molecular simulations, green chemistry, experimental validation, and life-cycle assessment, this thesis provides critical mechanistic insight into how oxygen functional groups dictate the physical, thermal, and biological behavior of rGO. The outcomes lay a solid foundation for the future integration of biomass-derived graphene materials into advanced biomedical applications while contributing to the broader goals of environmental sustainability and carbon-neutral material design.