Tunable Anisotropic Heat Transport in Structurally and Chemically Modified 3D Graphene

Technical Summary: Tunable Anisotropic Heat Transport in Structurally and Chemically Modified 3D Graphene

Problem Statement
Three-dimensional graphene (3DG) networks offer a pathway to translate the high intrinsic thermal conductivity of graphene into macroscopic, freestanding architectures. However, the effective thermal transport in these porous networks is not governed solely by graphene's intrinsic properties. Instead, it is dictated by a complex hierarchy of pores, sheet orientation, network connectivity, and graphene-graphene junctions. While prior studies have established that macroscopic graphene assemblies often conduct heat far below the intrinsic limit due to defects and inter-sheet contacts, the specific mechanisms by which post-synthesis processing—specifically mechanical compression, welding, and chemical doping—alters directional (in-plane vs. cross-plane) heat transport remain incompletely understood. A key challenge is distinguishing between structural reorientation effects and chemical/phonon-scattering effects to determine how to program anisotropic heat flow in freestanding 3DG.

Methodology
The study utilized a single material platform: pristine 3DG synthesized via a Ni-template chemical vapor deposition (CVD) route, resulting in a continuous, porous, freestanding network. This pristine material was processed into four distinct states:

1. Pristine: As-synthesized porous 3DG.
2. Cold-Rolled: Samples compressed to varying final thicknesses (109, 53, and 6 µm) to induce densification and structural reorientation.
3. Welded: Stacks of pristine sheets (two or three layers) cold-rolled to a final thickness of ~109 µm, creating internal "weld planes" (mechanical junctions) without metallurgical fusion.
4. Chemically Modified: Samples treated with various n-type (PVP, PEI-A, N-DMBI, TBD) and p-type (H₂SO₄, HNO₃) dopants at varying concentrations and exposure times.

Thermal transport was characterized using paired, room-temperature measurements for both directions:

• Cross-plane ($\kappa_\perp$): Measured using a differential 3$\omega$ method with transferable polyimide-supported metal heater lines to avoid direct contact issues on the porous, conductive surface.
• In-plane ($\kappa_\parallel$): Measured using a suspended four-probe electrothermal method, utilizing the 3DG strip as both the Joule heater and resistance thermometer, with corrections for distributed heat losses.

Key Results

• Mechanical Compression (Cold-Rolling): Compression induced a dramatic, opposing shift in directional thermal conductivity. As thickness decreased from 170 µm to 6 µm, in-plane conductivity ($\kappa_\parallel$) increased from ~4.9 to ~24.9 W m⁻¹ K⁻¹, while cross-plane conductivity ($\kappa_\perp$) decreased from ~7.2 to ~0.21 W m⁻¹ K⁻¹. This resulted in a massive amplification of thermal anisotropy ($A_\kappa = \kappa_\parallel / \kappa_\perp$), rising from ~0.7 in the pristine state to ~120.5 in the most compressed state. The authors attribute this to the reorientation of the graphene network, which aligns segments laterally (increasing parallel pathways) while making cross-plane transport more tortuous and junction-limited.
• Weld Interfaces: Introducing internal weld planes (mechanical junctions between stacked sheets) acted as discrete thermal resistors. While $\kappa_\parallel$ remained high (comparable to rolled samples), $\kappa_\perp$ was significantly suppressed. By comparing welded samples to an "ideal-contact" baseline (summing the resistance of compressed layers without interface resistance), the study quantified an excess weld-associated resistance ($\Delta R_w''$) of approximately $4.5 \times 10^{-5}$ m² K W⁻¹ for one weld and $10.4 \times 10^{-5}$ m² K W⁻¹ for two welds. This confirms that weld planes concentrate interfacial thermal resistance, further enhancing anisotropy.
• Chemical Modification: Chemical doping (both n-type and p-type) suppressed thermal conductivity in both directions without generating the strong anisotropy seen in mechanical rolling. The suppression strength varied by dopant chemistry (e.g., TBD > PEI-A/N-DMBI > PVP for n-type; H₂SO₄ > HNO₃ for p-type). Analysis using the Wiedemann-Franz relation indicated that the electronic contribution to thermal conductivity was negligible ($\approx 0.04$ W m⁻¹ K⁻¹), confirming that the observed suppression is driven by phonon scattering and modified junction conductance rather than electronic transport changes.

Significance and Claims
The paper establishes that structural compression and the insertion of weld planes are practical, post-synthesis routes for "programming" anisotropic heat flow in freestanding 3DG. The study distinguishes between two mechanisms:

1. Mechanical Compression/Welding: Acts as a network reorientation process that amplifies anisotropy by enhancing lateral pathways and creating specific cross-plane bottlenecks.
2. Chemical Modification: Acts primarily as a thermal attenuator that reduces the magnitude of conductivity in all directions without significantly altering the geometric anisotropy.

The authors claim that this work provides a "processing-resolved description" of directional heat transport, demonstrating that a single freestanding 3DG scaffold can be transformed into distinct thermal-transport states. This positions 3DG uniquely between dense thermal conductors and architected porous networks, offering a platform where directional thermal control can be achieved alongside electrical percolation and thermoelectric tuning, which is critical for applications requiring coupled heat-charge transport. The results suggest that while absolute conductivity values may be lower than dense graphite or diamond, the ability to tune anisotropy via accessible mechanical and chemical processing is a distinct advantage for thermal management and thermoelectric applications.