Mapping amorphization and crystallization routes in graphene oxide thermal reduction
Reduced graphene oxide (rGO) is a versatile carbon material used in real-world technologies, including conductive coatings and inks, polymer composites for electromagnetic interference (EMI) shielding, electrodes for supercapacitors and batteries, chemical/strain sensors, and membranes for filtration and separation. A widely adopted and scalable route to synthetize rGO starts from graphene oxide (GO), a sheet-like material based on the same carbon honeycomb lattice that makes graphene special but “decorated” with oxygen-containing groups. These functional groups disrupt conductivity, change surface chemistry, and strongly affect how sheets interact.
To make GO more graphene-like, reduction is carried out, often by heating. During thermal reduction, oxygen groups are progressively removed and the carbon network partially recovers order, while the sheets re-stack into layered structures. Because this thermal step is so common, it is tempting to assume a simple rule: reach a given final temperature and you get a predictable rGO. Our study shows why this intuition can fail. GO is not a single, universal material; its chemistry and microstructure depend on its synthesis and processing history. Crucially, the heat treatment kinetics (ramp rate and dwell time) can steer the outcome as much as, or more than, the final temperature.
We monitored GO reduction in real time using X-ray diffraction, with synchrotron powder XRD performed at the MCX beamline of Elettra, complemented by laboratory XRD on oriented thin films. Beyond the usual diffraction signal that mainly tracks interlayer spacing, we also followed features linked to in-plane order (the honeycomb lattice fingerprint) and stacking registry (how well neighboring sheets align). This approach allows us to distinguish predominantly misaligned, turbostratic stacking from locally registered AB pairing (Bernal-like bilayers) and even short ABA motifs.
A central and surprising result is a transient “amorphous-like” window around 140–190 °C. Here, the diffraction fingerprint of the honeycomb lattice nearly vanishes: the graphene-like order becomes temporarily so disordered that its signature fades, see Figure 1. Importantly, this is not just “bad graphene”. Highly disordered graphene-like materials can be functional in their own right. Their conductivity, reactivity, and accessibility can be tuned by controlling the number of defects and active sites.
What happens after the amorphous phase is obtained is not predetermined by temperature alone, but depends strongly on the process kinetics at the atomic scale. Fast ramps can kinetically trap defect-rich states, which remain largely turbostratic even after high-temperature excursions (while still showing local AB pairing). Slower ramps below ~240 °C enable a gentler reorganization that promotes AB ordering and the emergence of short-range ABA motifs. In other words, kinetics acts like a dial that determines which structural “destination” is reached.
Figure 1: In-situ synchrotron PXRD during thermal reduction of graphene oxide (GO). The intensity map (left) shows how the diffraction pattern evolves during rapid heating, two isothermal holds and cooling, revealing a transient loss of structural order (“amorphization”) followed by its recovery (“crystallization”). The temperature-time program used in the experiment is shown on the right.
Why does re-stacking matter for applications? Because it reshapes transport pathways. Stacking registry and interlayer coupling affect electronic transport across rGO films, while packing/disorder influences how ions and molecules move through the material. The process–structure map presented here turns thermal reduction from a black box into a design tool: by choosing ramp rate and dwell time (not only final temperature) rGO can be steered toward the structural state best suited for the targeted device or process.
This research was conducted by the following research team:
Nicolò Galvani1, Jasper R. Plaisier2, Cosimo Anichini3, Alicia Moya4, Paolo Samorì3, Andrea Liscio5 and Fabiola Liscio1
1 Istituto per lo Studio Dei Materiali Nanostrutturati (ISMN), Consiglio Nazionale Delle Ricerche (CNR), Bologna, Italy
2 Elettra - Sincrotrone Trieste S.C.p.A., Trieste, Italy
3 University of Strasbourg, CNRS, ISIS UMR, Strasbourg, France
4 Universidad Autónoma de Madrid, Departament of Applied Physical Chemistry, Madrid, Spain
5 Istituto per la Microelettronica e Microsistemi (IMM) Consiglio Nazionale Delle Ricerche, Roma, Italy
Contact person email:
Local contact email:
Reference
N. Galvani, J.R. Plaisier, C. Anichini, A. Moya, P. Samorì, A. Liscio and F. Liscio, “Kinetically trapped amorphous states and AB pairing in rGO: an in-situ XRD study of process–structure map", Carbon 249, 121247 (2026); DOI: 10.1016/j.carbon.2026.121247.
