High Pressure (HP) in Spark Plasma Sintering (SPS) Processes: Application to the Polycrystalline Diamond

Abstract

High-Pressure (HP) technology allows new possibilities of processing by Spark Plasma Synthesis (SPS). This process is mainly involved in the sintering process and for bonding, growing and reaction. High-Pressure tools combined with SPS is applied for processing polycrystalline diamond without binder (binderless PCD) in this current work. Our described innovative Ultra High Pressure Spark Plasma Sintering (UHP-SPS) equipment shows the combination of our high-pressure apparatus (Belt-type) with conventional pulse electric current generator (Fuji). Our UHP-SPS equipment allows the processing up to 6 GPa, higher pressure than HP-SPS equipment, based on a conventional SPS equipment in which a non-graphite mold (metals, ceramics, composite and hybrid) with better mechanical properties (capable of 1 GPa) than graphite. The equipment of UHP-SPS and HP-SPS elements (pistons + die) conductivity of the non-graphite mold define a Hot-Pressing process. This study presents the results showing the ability of sintering diamond powder without additives at 4-5 GPa and 1300-1400 °C for duration between 5 and 30 min. Our described UHP-SPS innovative cell design allows the consolidation of diamond particles validated by the formation of grain boundaries on two different grain size powders, i.e., 0.75-1.25 μm and 8-12 μm. The phenomena explanation is proposed by comparison with the High Pressure High Temperature (HP-HT) (Belt, toroidal-Bridgman, multi-anvils (cubic)) process conventionally used for processing binderless polycrystalline diamond (binderless PCD). It is shown that using UHP-SPS, binderless diamond can be sintered at very unexpected P-T conditions, typically ~10 GPa and 500-1000 °C lower in typical HP-HT setups. This makes UHP-SPS a promising tool for the sintering of other high-pressure materials at non-equilibrium conditions and a potential industrial transfer with low environmental fingerprints could be considered.

Keywords: Spark Plasma Sintering; binderless diamonds; high pressure.

Figure 1

Drawing representing the different SPS assemblies and pathways of the injected current (red arrows). Conv-SPS assemblies (left columns) showing the mold contents, the sample and the furnace. These assemblies consist of mold mostly in graphite (Gr) to reach P_max = 0.1 GPa and T_max ~ 2500 °C and for HP-SPS in WC/Co to reach P_max = 1 GPa at low temperature (100–200 °C) (higher temperature for lower pressure is also possible). HP-SPS special assemblies (middle columns) in conventional equipment are more sophisticated and consist of double stage mold with an outer shell (acting as furnace) and inner die in graphite with hard materials discs/pistons (SiC and or binderless WC). In that case, P_max = 1 GPa and T_max = 1100 °C can be reached but the current pathway depends on the composition of pistons and discs. Therefore, the SPS process does not necessarily occur. In UHP-SPS setups (right column), assemblies are much more complicated to ensure good hydrostaticity. Typically, a ceramic cell is used to receive a thin wall graphite furnace in which powder is loaded. Powder is then packed between graphite punches inside the graphite tube furnace that allow current injection inside the sample (SPS principle). Hence, P_max = 6–8 GPa and T_max = 2000 °C can be achieved.

Figure 2

Graphic illustration of Equation (1) showing the efficiency of pressure (colored curves) to activate a physical or chemical process A as a function of temperature. Starting parameters A₀, Q and V have been chosen arbitrary for illustration purposes. Sintering can be activated at several hundred degrees lower when applying high pressure of few GPa than at 1 bar (10⁻⁴ GPa).

Figure 3

Pressure–temperature diagram showing filed of applications of the different SPS setups. Conv-SPS and HP-SPS are both used in conventional equipment and never exceed 1 GPa. Maximum temperature depends a lot on materials constituting the mold, i.e., graphite, WC/Co, WC or SiC. In UHP-SPS equipment, a wide range of pressure–temperature conditions are covered offering the possibility to study sintering behavior of high-pressure phases such as c-C (red-curve) and c-BN (black curve) without any additions of binder.

Figure 4

XRD patterns of starting materials (black) and run products after experiments (red): (a) For the experiments conducted with 0.75–1.25 μm grain size powder and (b) with 8–12 μm grain size powder. The starting powders are very pure, and grains are well crystalized (sharp peaks). After experiment, diamond is fully conserved with no evidence of graphite back transformation. It is of note that diffraction peaks in experimental products are wider than in the starting powder.

Figure 5

SEM images (SEI mode, 10 kV, WD = 10 mm) of the starting powders: (a) 0.75–1.25 μm grain size and (b) 8–12 μm grain size.

Figure 6

Macroscopic view of starting materials (a,b), with grain size 0.75–1.25 μm (a) and 8–12 μm (b), respectively. Starting powders are yellowish. Bottom row shows recovered samples after experiments at HP-HT using our UHP-SPS setup and powders described above (c,d). Recovered samples are solid discs of 11 mm diameter and 2–3 mm thickness.

Figure 7

(a) Load (tons, in blue) and power (W, in red) parameters as programmed and recorded during experiments and associated displacement in mm (b). During dwell, both load and power, i.e., pressure and temperature, are very stable. Most of the displacement occurs during cold compression (few mm) whereas it is only of few tens to hundreds of microns during heating and dwell.

Figure 8

Microstructure of HP22-09-MS1 recovered sample observed in SEM (SEI, 15 kV, WD = 10 mm) at different magnification ((a) ×2000, (b) ×5000 and (c) ×10,000). First, there is no evidence for grain fracturation, nor the presence of other phases. There are two types of regions inside the samples (a,b), typically more or less sintered suggesting that densification is heterogeneous and not total (a–c). In more densified areas, a lot of grain boundaries are observed (b) whereas grains are only packed in less densified areas with only few grains that can be sintered (center of (c) for example).

Figure 9

Microstructure of HP22-40-MS10 recovered sample observed in SEM (SEI, 10 kV, WD = 10 mm) at different magnification ((a) ×2000, (b) ×2500, (c,d) ×5000 and (e) ×10,000). There is no evidence for grain fracturation, nor the presence of other phases. Due to larger grain size, there are less areas sintered (a,b). However, in these areas, it is evident that grains boundaries are formed or in the process of formation (c–e).