Kinetic processes in the high-temperature pressure-infiltration of Al into Al2O3

We explore the influence (i) of the interaction between aluminium and alumina, and (ii) of sodium impurities present in Bayer alumina, on the pressure infiltration of alumina particle preforms with molten aluminium. At 1000°C or above, although the aluminium/alumina system is non-reactive, capillarity-driven solution-reprecipitation processes cause the liquid-solid interface to become mobile. Data show that this can result in infiltration kinetics that resemble those observed with reaction-driven pressure infiltration, namely a continuously increasing melt saturation under fixed infiltration pressure. Resulting isobaric saturation velocities are measured at 1000°C, 1050°C and 1100°C. The role of alumina particle shape and of Na-containing inclusions is investigated. It is found that the main factors affecting the rate of high-temperature isobaric infiltration in this system is the particle geometry. Measured steady infiltration rates give an activation volume on the order of ≈ 200 nm3 and an activation energy in the range of 300-500 suggesting that isobaric infiltration kinetics are governed by diffusion through the solid alumina. Sodium impurities of Bayer alumina are present within β″-Al2O3. They do not influence steady pressure infiltration but ease initial melt penetration into the preform, possibly because evaporated Na2O alters the oxide skin layer that lines the surface of molten aluminium.

Dynamic wetting in pressure-infiltration of ceramic particle preforms by molten metal

Gionata Schneider

This thesis investigates the interaction between mechanically induced flow of a liquid metal into a porous solid, and kinetic effects at the triple line. The approach is experimental and focuses on metal-ceramic systems for which the literature gives values of wetting and local kinetic parameters (Simga, SimgaLV, and interfacial phenomena near the triple line) at high temperature, gleaned by means of the sessile drop experiment. We investigate the infiltration with Al of Al2O3 preforms at 1000°C, 1050°C and 1100°C, and the infiltration with Cu-x (x=Si, Ti or Cr) of porous CGr between 1050°C and 1200°C. Those three alloys are known to form carbides better wetted by the liquid than is graphite. If measurements are taken at various values of applied metal overpressure, in the presence of thermally activated processes at the triple line, time intervenes in capillarity and the front does not stabilise. We observe that the metal continues to invade slowly the porous medium, gradually ¿creeping¿ into its pores under the action of local phenomena that take place at the triple line. We characterise such dynamic infiltration by measuring rates of steady isobaric infiltration, measuring also the variation of infiltration rates at Pth and Pth+0.05MPa. Saturation rates are then computed by means of a linear regression over 100s of the measured saturation for each pressure step. Data are interpreted using an analogy with high-temperature plasticity. This leads to assume that the applied pressure contributes a proportional reduction in the activation energy of the process that governs the rate of preform infiltration, the constant of proportionality defining an activation volume that characterises the thermally activated process. Cu-46at.pct.Si infiltrating carbon is a mildly reactive system that forms SiC at the interface. Isobaric saturation rates of infiltration and their dependence on variations in the applied pressure were conducted between 1050°C and 1200°C. It is found that an activation volume on the order of ~200nm3 makes the data compatible with the Arrhenius law, with an estimated activation energy value of ~400kJ/mol, which is realistic for a process limited by the rate of SiC growth along the interface. Aluminium infiltrating three different Al2O3 preforms shows similar behaviour as does the mildly reactive Cu-Si/C system if particles in the preform are highly angular, regardless of the presence or absence of Na impurities. Isobaric saturation rates measured between 1000°C and 1100°C give a similar activation volume as the Cu-Si/C system, on the order of ~200nm3, and an activation energy ~300kJ/mol, suggesting that the kinetics are likely limited by solid phase diffusion. It is also shown that sodium impurities in alumina serve to facilitate its infiltration by aluminium, likely because sodium alters the oxide skin that initially covers the melt. Finally, in the infiltration of Cu-1at.pct.Cr or Cu-10at.pct.Ti into CGr, one finds that interfacial reaction does not aid pressure infiltration. In the former, it is because initial reaction between melt and preform depletes the alloy in reactive Cr, leading to similar infiltration as with pure Cu. In the latter, initial carbide formation is so extensive that it blocks ingress of the metal into the preform. For alloying elements forming a better wetted interface to aid pressure infiltration, reaction must thus be sufficiently slow compared to the rate of metal flow.

EPFL2019

Processing and Mechanical Properties of Alumina Reinforced Copper-Aluminum Composites

Carmen Krüger

Copper-based composites reinforced with alumina particles are produced by two different pro- cesses, namely gas pressure driven infiltration of a porous alumina preform and a two-step in-situ process developed in the course of this work. The resulting composites are characterized for their microstructures and mechanical properties. It is shown that during gas-pressure infiltration of angular alumina particles with liquid copper- aluminum alloys at 1300◦C, rounding of the alumina particles occurs. Rounding is quantified by measurements of local curvature, and shown not to be found after infiltration of the same particles with pure copper at 1150◦C. Transport pathways necessary for rounding are examined by analysis of the thermodynamics and kinetics of the copper-aluminum-alumina system, drawing from the literature on wetting of alumina by copper and its alloys. Analysis leads to two different possible explanations for the phenomenon: (i) the presence of aluminum in the melt increases the equilibrium oxygen concentration and thus accelerates volume diffusion or (ii) segregation of aluminum in the liquid alloy to the alloy-alumina interface causes accelerated interfacial diffusion. The pressure infiltration of four different copper-aluminum alloys with varying titanium content (Cu8wt%Alxwt%Ti with x=0, 0.2, 1, 2) into preforms of two types of alumina particle leads to sound, pore-free composites reinforced with 56-59 vol% alumina. Depending on the titanium content of the matrix alloy, a new phase appears at the interface between the alumina particles and the metallic matrix; this phase was identified as the mixed oxide Ti3(Cu,Al)3O. A very thin (5-10 nm) titanium-rich layer between alumina and the ternary oxide is responsible for a mechanically strong interface between the copper-alloy matrix and alumina. Peak mechanical properties in these composites were obtained by infiltration of equiaxed Al2O3 particles with a Cu8wt%0.2wt%Ti alloy. Their measured properties displayed a Young’s modulus near 235 GPa, a density of 5.5 g/cm3, a tensile strength and elongation of 600 MPa and 1% respectively, together with a compressive strength near 1100 MPa reached at 9% strain. Copper matrix alumina composites are also produced via a process combining in-situ formation of alumina and deformation processing. Elemental powders are uniaxially compacted and subjected to a two-step heat treatment that leads to the formation of alumina films in a copper-aluminum solid solution matrix. During deformation processing, the composite is consolidated and the alumina films are broken up into particles of average diameter 1-2 μm. Further hot working in the form of upsetting and extrusion further refines the alumina particles and improves their distribution, which improves the composite mechanical properties. Cu7wt%Al matrix composites reinforced with 26 vol% alumina particles were obtained by this process. In optimum processing conditions of this work, these composites have a tensile strength of 846±44 MPa and a tensile ductility of 2.2±0.8%, with likely room for significant further improvement of these properties after more extensive deformation processing.

EPFL2012

Stability and physical properties of alkylsilane and alkylphosphonate self-assembled monolayer (SAM) films on metal oxide substrates characterized at the micro- and nanoscale

Enamul Hoque

Self-assembled monolayer (SAM) films have attracted immense attention for both fundamental and applied research. A SAM is composed of a large number of molecules with a head group that chemisorbs onto a substrate, a tail group that interacts with the outer surface of the film, and a spacer (backbone) chain group that connects the head and tail groups resulting in a coating. Interactions between spacer groups of different molecules, such as van der Waals forces and/or hydrogen bonding, hasten SAM film formation and contribute to its stability. In this dissertation, SAM and thin films have been formed onto copper and aluminum oxide surfaces by reaction with 1H,1H,2H,2H-perfluorodecyldimethylchlorosilane (PFMS), 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFTS), 1H,1H,2H,2H-perfluorodecylphosphonic acid (PFDP), octylphosphonic acid (OP), decylphosphonic acid (DP), and octadecylphosphonic acid (ODP). The properties and stability of the films were investigated employing complementary surface analysis techniques: X-ray photoelectron spectroscopy (XPS), atomic force microscopy (AFM), friction force microscopy (FFM), a derivative of AFM, contact angle measurements (CAMs), and Fourier transform infrared reflection/absorption spectroscopy (FT-IRRAS). The perfluoroalkylsilane SAM on Cu is found to be extremely hydrophobic typically having sessile drop static contact angles of more than 130° for pure water and a surface energy of 14 mJ/m2 (mN/m). FFM showed a significant reduction in the adhesive force and friction coefficient of PFMS modified Cu (PFMS/Cu) compared to unmodified Cu. Treatment by exposure to harsh conditions showed that PFMS/Cu SAM can withstand boiling nitric acid (pH=1.8), boiling water, and warm sodium hydroxide (pH=12, 60 °C) solutions for at least 30 minutes. Furthermore, no SAM degradation was observed when PFMS/Cu was exposed to warm nitric acid solution for up to 70 min at 60 °C or 50 min at 80 °C. XPS and FT-IRRAS data reveal a coordination of the PFMS silicon (Si) atom with a cuprate (CuO) molecule present on the oxidized copper substrate. The data give good evidence that the stability of the SAM film on the PFMS modified oxidized Cu surface is largely due to the formation of a siloxy-copper (-Si-O-Cu-) bond via a condensation reaction between silanol (-Si-OH) and copper hydroxide (CuOH). For a PFTS modified Cu surface (PFTS/Cu), the sessile drop static contact angle of pure water has been measured to be more than 125° and the surface energy to be typically less than 16 mJ/m2. Stability tests show that the PFTS/Cu film can survive in boiling pure water for one hour, boiling nitric acid (pH 1.5 or 1.8) for 30 minutes, sodium hydroxide solution (pH 12, 70 °C) for 30 minutes, and autoclave conditions (steam at 134 °C and 3 atmospheres) for 15 minutes. The more commonly used self-assembled monolayer (SAM) modifications of Cu surfaces, e.g. thiol compounds, are significantly less stable than PFTS/Cu. Extremely hydrophobic (low surface energy) and stable PFMS/Cu SAMs and PFTS/Cu films could be useful as corrosion inhibitors in micro/nanoelectronic devices and/or as promoters for anti-wetting, low adhesion surfaces or drop-wise condensation on heat exchange surfaces. XPS analysis confirmed the presence of perfluorinated and non-perfluorinated alkylphosphonate molecules on the PFDP, DP, and ODP SAMs deposited at the aluminum oxide coated silicon (Al/Si) surfaces. The sessile drop static contact angle of pure water on PFDP SAMs was typically more than 130° and on DP and ODP typically more than 125° indicating that the phosphonic acid SAMs reacted with Al samples were very hydrophobic. The surface roughness for PFDP/Al, DP/Al, ODP/Al, and bare Al was approximately 35 nm, as determined by AFM. The surface energy for PFDP/Al was determined to be approximately 11 mJ/m2 by the Zisman plot method compared to 21 mJ/m2 and 20 mJ/m2 for DP/Al and ODP/Al, respectively. PFDP/Al gave the lowest adhesion and friction force while bare Al gave the highest. The adhesion and friction forces for ODP/Al and DP/Al SAMs were in between. ODP, DP, and OP SAMs have been studied in detail on relatively flat aluminum oxide surfaces. The rms surface roughness for ODP/Al, DP/Al, OP/Al, and bare Al was less than 15 nm, as determined by AFM. The sessile drop static contact angle of pure water on ODP/Al and DP/Al was typically more than 115° and on OP/Al typically less than 105°. The surface energy for ODP/Al and DP/Al was determined to be approximately 21 mJ/m2 and 22 mJ/m2, respectively, compared to 26 mJ/m2 for OP/Al. ODP/Al and OP/Al were studied by FFM to better understand their micro-/nano-tribological properties. ODP/Al gave the lowest coefficient of friction values while bare Al gave the highest. The adhesion forces for ODP/Al and OP/Al were comparable. The chemical stability of ODP/Al, PFDP/Al, DP/Al, OP/Al, and PFMS/Al SAMs has been inspected by exposure to warm nitric acid (pH 1.8, 30 min, 60-95 °C). The XPS data and stability against harsh chemical conditions indicate that a type of bond forms between a phosphonic acid (PA) or silane molecule and the oxidized Al surface. Stability tests using warm nitric acid (pH 1.8, 30 min, 60-95 °C) show ODP/Al SAMs to be most stable followed by PFDP/Al, DP/Al, PFMS/Al, and OP/Al SAMs. For PFTS/Al, stability tests demonstrate that modified aluminum is able to survive exposure to warm nitric acid (pH = 1.8, 60 °C, 30 min) indicating some degree of robustness. Hydrophobic, low adhesion, and robust aluminum surfaces have useful applications for micro/nano-electromechanical systems (MEMS/NEMS), such as digital micro-mirror devices (DMDs). These studies are expected to aid in the design and selection of proper lubricants for MEMS/NEMS.

EPFL2007