Evolution of the structural properties of disordered carbon at high pressure
Abstract
Carbon is important in physics in fields like making novel materials and emerging quantum computing. High-pressure compression is widely used to form new structures. Pressure-induced phase transitions in materials are achieved by exciting the atoms so that they can overcome the energy barrier and form a new structure. Lab synthesized diamond via static compression of graphite within has been reported since 1950s. This shortens the natural formation time of diamond significantly, but usually requires high temperatures. More precursors like polymeric carbon and fullerene, have been reported to successfully synthesize diamond. Some of them were even achieved at ambient temperature. Additional to normal diamond that has a cubic crystal structure, there is another diamond phase called lonsdaleite. It has a hexagonal crystal configuration and is naturally found in meteorite sites. Lonsdaleite has been predicted to be harder than cubic diamond by first-principle calculation, which however needs experimental evidences. Other novel structures like m-carbon have also been found, but more analyses are needed.
The objective of this thesis is to explore pressure-induced phase transformations in a disordered precursor glassy carbon (GC). GC is amorphous and is mostly sp2-bonded. Its structure is described as curly entangled graphene-like sheets. It can form nanocrystalline lonsdaleite and diamonds after room-temperature compression to 80 GPa. This structural evolution without the help of heating has been investigated using electron microscopy, where recovered samples after GC is subjected to different pressures are measured. The formation of an oriented graphitic structure has been confirmed at ~45 GPa. Further nanoindentation measurements showed that after GC is compressed to ~35 GPa, it already showed anisotropic responses, indicating a lower pressure for the oriented graphite. Up to 80 GPa where diamonds are formed, the oriented graphite remains to be the recovered material. An in situ transparent state usually seen at ~60 GPa is referred to as the 'transparent graphite', which is a key transiting state before recoverable diamond.
A shear-driven mechanism has been proposed based on observations that the formed lonsdaleite and diamond have a unique core-shell relationship. The importance of shear was studied by compressing GC under different conditions. No diamonds were formed when it was compressed in a shear-free environment up to 80 GPa. Other factors such as the long-time holding indicate that pressure and shear are critical. Compressions of atomistic models of GC using molecular dynamics simulations supported these experimental observations and provided additional in situ structural information. The oriented graphite has been evident from the aligned graphene-like sheets perpendicular to the compressing axis at ~35 GPa. The Young's modulus that was hard to obtain in experiments has also been calculated. Molecular dynamics simulations have revealed that shear acts as heating spikes to induce phase transformation in room-temperature compression.
The mechanical properties of several diamond and lonsdaleite samples that have similar crystal sizes are measured. The obtained hardness values showed that although lonsdaleite has a hardness similar to that of diamond, it was not significantly harder as proposed. Finite element analysis were done to help understand this based on the geometrical deformation and stress distribution inside the indenter tip. It is seen that when a soft material is indented, a diamond tip can be considered as rigid and give precise measurements. However, a significant underestimation of hardness is observed when measure a superhard material, possibly due to the plastic deformation of the indenter itself. Further electron microscope experiments confirm the plastic deformation of used diamond indentation tip, supporting the conclusion from finite element analysis.