Vacuum Plasma Sprayed (VPS) Coating – Substrate System for Fundamental Study of Adhesion

You don't have the latest version of Adobe Flash Player.

Please update your flash player.

Please login or register to make a comment!

...Description...... more. less.

process ........................................................................................ ....40 2.2. Investigation methods .......................................................................................... ...43 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 5 2.2.1. Morphology and chemical composition by SEM-EDX analysis...........43 2.2.2. Crystal structure by XRD analysis .................................................... ...48 2.2.3. Topography and bending curvature by profilometry............................50 2.2.4. Adhesion between the coating and substrate......................................54 Chapter 4 RESULTS and DISCUSSION 1. Al 2 O 3 on Si wafer................................................................................................................... .....64 1.1. SEM-EDX analysis ............................................................................................... ...64 1.1.1. Deposition of Al 2 O 3 in one passage .............................................................. ...66 1.1.2. Deposition of Al 2 O 3 in multiple passages ....................................................... ...67 1.2. SEM In-situ scratch test ....................................................................................... ...69 1.3. XRD analysis ........................................................................................................ ...70 1.4. Topography .......................................................................................................... ...72 1.5. Mechanism of crater formation ............................................................................. ...77 1.5.1. Erosion ........................................................................................................... ...77 1.5.2. Corrosion ....................................................................................................... ...78 2. Al 2 O 3 on metallic substrate......................................................................................................... .81 2.1. SEM-EDX analysis ............................................................................................... ...81 2.2. XRD analysis ........................................................................................................ ...86 2.3. Topography .......................................................................................................... ...87 2.4. Interfacial indentation ........................................................................................... ...88 2.5. Rockwell indentation...............................................................................................88 3. Al 2 O 3 on sapphire................................................................................................................... .....89 3.1. SEM-EDX analysis. ............................................................................................... ..89 4. NiCr(80-20) on metal substrate................................................................................................. ..92 5. NiCr(80-20) on sapphire........................................................................................................ ......92 Chapter 5 SUMMARY.............................................................................................................. ...............94 Literature..................................................................................................................... ............................97 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 6 ABSTRACT Al 2 O 3 sprayed on Si (001) substrate has been selected as a novel fundamental approach of studying adhesion of Vacuum Plasma Sprayed (VPS) coatings. The choice of substrate had two reasons: i) perfectly flat surface of wafers allows to avoid the mechanical anchoring effect ii) mechanical and chemical properties of Si are very well known. In result Si wafers seem not to allow thermal spraying of alumina under whatever conditions. Moreover the opposite effect 3 substrate removal has been observed. Two possible failure mechanisms of the VPS alumina and Si wafer are investigated: i) erosion due to impact of molten alumina particles, ii) corrosion due to evaporation and melting of Si under vacuum and plasma conditions, and formation of SiO 2 at the interface, which increased 5 times the cooling stress in Si. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 7 Chapter 1 INTRODUCTION 4. General information on Thermal Spray (TS) coating Thermal spray (TS) is the most common technology for deposition of advanced thick film coatings. The process resembles spray-painting with molten coating material. Typical steps undertaken to process TS coatings are discussed below (see Figure 1.1): Figure 1.1: Principle of thermal spray coating; ref 1: R.C. Dykhuizen in Journ. of Therm. Spr. Techn. Vol 3, No. 4 (94); ref 2: M. Vardelle et al. in Journ. of Therm. Spr. Techn. Vol 4, No. 1 (95) In first step a suitable chemistry is selected for any given application, e.g. typically wear, corrosion or thermal resistance, or specific functional property such as electromagnetic, chemical or biological function. The composition of the feedstock (powder, wire, or rod) used in deposition is identical to the desired coating. The most popular are powders, which average particle size is broadly in range of 10-100 [ µ m]. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 8 However preferably, they are of narrow size distribution and equi-axial (spherical) shape as they are fed to the hot zone by compressed gas. In second step, the feedstock is melted so the initial temperature of the process must exceed its melting point, which in the case of ceramic is often over 2000 [°C]. Commonly the feedstock is fed into the hot zone of TS torch, where temperature ranges from ~ 2,000 to 15,000[°C] depending on the TS technique used. The melting process is controlled by heat transfer from hot gas (plasma) to the particles moving rapidly (more than 100[m/s]) through the hot zone with large gradient of temperature. For the hottest (plasma) torches some evaporation and loss of powder may take place. The process gasses expand in the hot zone and thus accelerate towards the nozzle. The hot process gas transfers momentum (and heat) to the particles of the feedstock, melting and in next step accelerating them towards the substrate, at the velocity of 100-1000 [m/s]. During this process some (larger) particles may melt only partially, i.e. retain solid core; alternatively, some smaller particles may evaporate. The desired first splat is round shaped and flat, which indicates strong adhesion to the surface, and consequently dense and good quality coating. Prior to deposition, the substrate, typically an alloy, is sand blasted to induce roughness and adhesion through interlocking with coating, and cleaned to remove any blasting grit and organics. Molten particles strike the substrate surface and rapidly (more than 10 6 [K/s]) solidify to build-up the coating. cPancake d 3like flow and spread is preferred for the particles building-up the coating. However, splashing or even bouncing-off does also happen, leading to undesired microstructure and powder loss. The proportion of mass solidified as coating to mass fed to the torch rarely exceeds 50% in typical TS process. However, with the optimized torch design it is possible to reach 80%. The substrate heats up during the process, typically to 200-400°[C], which is dependent on whether it is cooled or not, and how close it is to the torch. Thus the conventional TS does not significantly affect microstructure or chemistry of most metallic substrates. However, for heat-sensitive substrates cold-spray technology should be considered. Despite the microstructural defects of the coating, TS is widely used because of many advantages: - the process is relatively fast and cost-effective, Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 9 - there are not many limitations regarding the coated area, coated material, or coating material. 5. The most common application of TS coatings The applications of TS coating are diversified. Some of the most common include - thermal barrier coatings (TBC) for aircraft gas turbines, - abradable coatings for seals in gas turbines, - thermal barrier, abradables, corrosion and wear resistant coatings for industrial gas turbines, - corrosion and wear resistant coatings in pulp and paper production, - bioactive coatings for implants, - thermal barrier, wear and corrosion resistant coatings for automotive industry, - electrical insulations, - corrosion resistant coatings for bridges. 6. Problematic issues of TS technology TS process is controlled by many variables. It is therefore difficult to stabilize (i.e. to obtain coating reproducibility within narrow specifications), which is absolutely necessary in critical, such as aerospace, applications. The diagnostic tools must be supported by extensive experimentation and modeling, which include complex computer programs predicting behaviour of particle during impact on substrate, and splat formation, as a function of velocity, temperature, viscosity, etc. Evaluation and metallography of TS coatings is difficult: Porosity measurements require mercury porosimeter, on the other hand metallography must be very careful due to the easy pull-out as a result of weak cohesive bonds and high density of microcracks. Microstructure and phase composition analysis involves XRD and SEM with EDX. Wear resistance, thermal properties, and corrosion resistance is evaluated using standard (ISO, ASTM) methods. Coatings strength and stiffness is determined in tension of free-standing tubes after mandrel removal. One of the most problematic issue in TS coating evaluation is to clearly determine its the adhesion mechanism. Moreover, adhesion is modified by residual stress in the Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 10 coating, which measurement is also very demanding. Details on those two very important parameters are described below: a. Residual stress According to [1], residual stresses can be defined as those stresses that remain in a material or body after manufacture and processing in the absence of external forces or thermal gradients. Residual stress measurement techniques invariably measure strains rather than stresses, and the residual stresses are then deduced using the appropriate material parameters such as Young's modulus and Poisson's ratio. Often only a single stress value is quoted and the stresses are implicitly assumed to be constant within the measurement volume, both in the surface plane and through the depth. Residual stresses can be defined as either macro or micro stresses and both may be present in a component at any time. Macro residual stresses, which are often referred to as Type I residual stresses, vary within the body of the component over a range much larger than the grain size. Micro residual stresses, which result from differences within the microstructure of a material, can be classified as Type II or III. Type II residual stresses are micro residual stresses that operate at the grain-size level; Type III are generated at the atomic level. Micro residual stresses often result from the presence of different phases or constituents in a material. They can change sign and/or magnitude over distances comparable to the grain size of the material under analysis. The origin of residual stress in thermal spray coating, according to Kuroda and Clyne (1991) [2], is mainly due to: - Quenching stress - cooling of molten particles striking the substrate, - Cooling stress - cooling of the system (coefficient of thermal expansion). Considerable efforts have been made over the recent years to understand and predict the residual stress. In most cases, it can be assumed that stress in the through 3thickness direction are negligible and that the stresses are the same in all directions within the plane of the coating. However establishing the stress state is generally more complex for (thin) films deposited by atomistic processes, since sprayed coatings are considerably thicker and variations in stress level with depth are often significant. Many recent studies have been focused on measurement of residual stresses and modeling of their development during spraying. However, very few studies present reliable correlations established among process conditions, Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 11 measured stress levels, and indicators of coating performance. To get more fundamental understanding of TS coatings residual stress, more investigations at different scales must be performed. According to [1], the methods of measuring residual stresses, most commonly used in industry are: - Hole drilling (30%), - X -ray diffraction (26%), - Neutron diffraction (19%), - Layer removal or curvature (16%), - Other -including magnetic, ultrasonic, Raman (9%) b. Adhesion Adhesion is the tendency of certain dissimilar molecules to cling together. Mechanisms of adhesion of thermally sprayed coating to the substrate involve: - Mechanical interlocking - during deposition the sprayed material flows around the roughness peaks and into the valleys where it is constrained on cooling, - Physical bonding - Van der Waals forces and hydrogen bridge bonding,, - Chemical bonding - ionic, atomic or metallic bonding. Standard adhesion evaluation tests are performed in tension (pull) test according to [3]. However, it is very difficult to determine whether the failure is adhesive, i.e. occurs at the interface between coating and substrate, or cohesive - occurs through the coating. Thus many efforts have been made recently in order to develop measurement methods and to perform fundamental studies of adhesion. This will lead to more efficient control of the spraying process and quality evaluation of the obtained coating. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 12 Chapter 2 MOTIVATION STATE OF THE ART 1. Motivation The purpose of this work was to produce a thermally sprayed coating 3substrate system suitable for scientific investigations of adhesion and residual stress. Three requirements had to be fulfilled: - deposition of a continous film, - smooth (without scale roughness and voids) interface between coating and substrate, - simple interface chemistry. 2. State of the art Fundamental study of adhesion and residual stress of thermal spray coatings will provide new tools for coating development, optimization and control. A lot of efforts have been made in this direction, however interpretation of the results often encountered difficulties due to numerous parameters influencing the measurements. Thus an idea has arisen of using a very simple coating 3substrate system, which would make interpretation easier. The postulated requirements are presented below. 2.1. Deposition of a continuous film Deposition of a continuous film was consider important because most of normalized and recently developed testing techniques of both residual stresses and adhesion are based on continues coatings. The most common of them are described below. However, first splat of TS coating is considered the most contributing to the adhesion to the substrate. Thus knowledge on both residual stresses and adhesion of single splats is recently an intensively investigated subject. Recent evaluation of residual stresses within single splats is presented at the end of this part. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 13 2.1.1. Residual stresses Residual stresses can be evaluated by: - Hole drilling method [1], which is a widely used technique because of several advantages; the test is simple, cheap, quick and versatile. The technique is applied to a wide range of materials. Moreover equipment can be laboratory- based or portable. The test involves drilling of a small hole through investigated material and evaluation of stresses by measurement of the locally relieved surface strains. The calculation of residual stress is standardized [4] and requires Finite Element Analyses. Commonly strains are measured using a special strain gauge rosette. Close to the hole, the strain relief is nearly complete but in case of deeper holes the technique becomes limited in strain sensitivity because of uncertainties related to the dimensions of the hole i.e. its concentricity, depth, profile and diameter. Moreover values obtained by gauge rosette are influenced by surface roughness, flatness, and specimen preparation. Hole drilling test can be done in incremental steps, which improves the versatility of the technique and enables measurement of stress profiles and gradients. - X-ray diffraction method [1], which is commonly used to measure internal stresses due to elastic deformations within a polycrystalline material. The value is calculated based on comparison between the spacing of the lattice planes in stressed and relaxed material. The same spacing will occur in any similarly oriented plane, with respect to the applied stress. The test is relatively straightforward and equipment commonly available. The specimen is irradiated with high energy X-rays that penetrate the surface. The crystal planes diffract X-rays in accordance with Bragg's law. A detector, which moves around the specimen locates X-rays in respect to their diffraction angle and records their intensity. Stresses within the material are evaluated based on the location of the peaks. - neutron diffraction [1] is based on the same principle as X-ray diffraction: stresses are calculated based on comparison between the spacing of the lattice planes in stressed and relaxed material. Neutron diffraction is advantageous when high penetration depth is required. It is possible to evaluate stresses from around 0.2 [mm] down to bulk measurements of up to Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 14 100 [mm] in aluminium or 25 [mm] in steel. Moreover with high spatial resolution, neutron diffraction can provide complete three-dimensional strain maps of investigated material. There are two considerable disadvantages of neutron diffraction in comparison to different X-rays techniques: i) relatively high cost, ii) much lower availability. - curvature and layer removal techniques are generally quick and require just simple calculation. However, a test can be performed only in simple testpiece geometries. In case of bulk material, the layers removal technique is applied. In the case of coatings curvature of the substrate is measured: i) before and after deposition 3 obtained value is a result for the whole coating, ii) after successive deposition of each layer[5] 3 allow obtaining variation of stresses with depth of the coating is obtained. The curvature of can be measured by many methods, such as: optical microscopy, laser scanning, strain gauges, or profilometry. A test is usually performed on narrow strips in order to avoid multiaxial curvature and mechanical instability. (for more details see Chapter 3: Topography and bending curvature by profilometry). 2.1.2. Adhesion Adhesion testing techniques are based on evaluation of the amount of energy which is necessary to induce delamination of the coating. The type of delamination is selected according to ductility/brittleness of the coating- substrate system. Adhesion tests are normalized [3, 6-12]. Drory listed commonly used adhesion testing techniques, involving induced delamination [2] (see Table 2.1). Two relatively novel (developed in 1996), techniques were of interest in this work: - Interfacial indentation 3 applicable for ductile coating - ductile substrate system (for more details see Chapter 4: Interfacial indentation), - Rockwell indentation 3 applicable for brittle coating 3 ductile substrate system (for more details see Chapter 4: Rockwell indentation). Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 15 Table 2.1: Selected adhesion measurement techniques, T denotes tensile stress and C compressive stress [2]. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 16 Table 2.1: Selected adhesion measurement techniques [2] - cont. Investigation within single splats Only few quantitative investigations have been done for single splats. For instance, Matejicek [13] investigated residual stresses in isolated splats deposited on stainless steel substrates using X-ray microdiffraction. Two types of spraying techniques and feedstock were used: i) plasma sprayed molybdenum, ii) cold sprayed copper. The results were discussed with respect to the influence of selected spraying parameters, contribution of quenching and thermal stress component and splat formation. The measured stresses ranged from 50 to 1050 [MPa]. These studies provide quantitative understanding of the residual stress states at the splat level. However they also indicate large-scale local heterogeneities that can exist within the complex microstructure of a thermally sprayed coating. Simpler coating- Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 17 substrate system i.e. not affected by unknown surface chemistry of steel, may help understanding change in residual stress with respect only to spraying parameters only. 2.2. Simple chemistry interface without scale roughness The claim for smooth substrate surface combined with simple interface chemistry should allow two important studies: i) the influence of physical and chemical bonding on adhesion, ii) the influence of sinusoidal interface between coating and substrate on the results of interfacial indentation test. The search for a suitable coating 3substrate system is mainly based on recent investigations on the formation of single splats on different materials in different spraying conditions and by different techniques. Results are listed further in this chapter. 3. Material system 3.1. Substrate 3.1.1. Steel The steel substrate was considered as the most common material for thermal spray. However, its surface chemistry is not well known. Recent works [14-15] revealed complex phenomena on smooth stainless steel surface as a result of spraying NiCr alloy particles by plasma spray technique. They investigated how interface chemistry and splat morphology depend on pre-treatment of substrate surface. In the first step, they formed surface oxides (or hydroxides) and/or induced some modification in surface topology of different samples. In the next step they sprayed NiCr splats and studied their surface with reference to the unsprayed specimens, after the same pre-treatments. As a result of deposition, a diversity of splats ranging from disc-shaped to very fragmented were produced (see Figure 2.1). Description of their diameters, average population and particular characteristics for each sample is given in Table 2.2. The observation of splats microstructure and their interface with the substrate at a micro and nanoscale revealed the occurrence of many forms of oxides, voids, and also jetting of the steel within NiCr. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 18 Figures 2.2 and 2.3 present cross-sections of splats which solidified in different morphologies: round-shaped and splashed. Round shaped splats (see Figure 2.3) showed: 1 3 uniform, columnar grains, 2 3 coincident grain boundaries across the splat-substrate interface, 3,4 3 voids, 5 3 phase identified as NiO (by TEM). Cross section of splashed splat done across the central void (see Figure 2.2) showed: 1 3 regions being presumably a mixture of oxide and NiCr particles, 2 3 shape indicating that the molten metal had been pushed upwards from the substrate. From [14-15] it can be concluded that steel substrate can be not satisfactory material for fundamental study of adhesion and residual stresses because of very sensitive to pre-treatment surface chemistry. Figure 2.1: SEM images of typical splats: (A) disc splat showing a distinct rim and voids at the centre; (B) disc splat with a central void and a distinct rim; (C) flower-shaped splat;(D) fragmented splat; and (E) very fragmented splat [14]. Figure 2.2: FIB cross-sections of a splashed finger across the central void of the NiCr splat [14]. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 19 Figure 2.3: FIB cross-section of a disc-shaped NiCr splat on steel substate (insert shows location of cross-section): [14] Table 2.2: Description of the different types of splat found on each specimen, their diameter, average population and particular characteristic. All samples were polished to the nanoscale. Sample SS_P were sprayed just after polishing. Samples SS_B and SS_BT were placed into boiling distilled water Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 20 for 30 min. SS_PT and SS_BT were thermally treated: subsequently heated at 350 °C for 90 min in air at atmospheric pressure. [14] 3.1.2. Aluminium Aluminium alloy was considered as common substrate for thermal spray. The big advantage of this material is its passive surface due to chemically inert layer of Al 2 O 3 . Recent works [16] revealed occurrence of plastic deformation at the interface as a result of high-velocity air fuel spraying NiCr alloy (see Figure 2.5). This makes that the claim for smooth interface between the substrate and coating is not satisfied. Figure 2.5: Secondary electron image of the aluminum substrate sample with a low density of sprayed NiCr particles [16]. 3.1.3. Silicon Silicon substrate is not used in industry. However, it seemed to be a good choice from chemistry point of view. Its surface chemistry is well known, i.e. it contains a silicon oxide layer, a few nanometers thick. Additionally mechanical properties of this material are well studied. The silicon single crystal coated with passive aluminium oxide coating is well known in electronic industry. During the last few years aluminium oxide films have been widely used in microelectronic devices, optoelectronics, sensors, antireflection Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 21 coatings and surface passivation of solar cells [17]. Regarding its usefulness as an alternative gate insulator, aluminium oxide has many favorable properties [18]. Deposition of aluminium oxide on silicon wafer was successfully performed by chemical vapor deposition (CVD) [19-20], MOCVD, atomic layer deposition (ALD), spray pyrolysis [21], thermal evaporation, sputtering, pulse laser deposition [22], sol 3gel technique [23-24]. Although aluminium oxide is also commonly sprayed by thermal spray technique, silicon wafer has not been so far used as a substrate. A deposition trial has been already carried out for a similar system, i.e. Al 2 O 3 -Al composite cold sprayed on silicon single crystal [25]. The results indicated that a cold spray of hard coating on soft aluminium can be successfully performed (see Figure 2.5). Cratering of the substrate surface affects the integrity of coating/substrate. However, deposition of both pure aluminium oxide and aluminium was not successful. Spraying of pure aluminium oxide caused crater formation in silicon substrate, which is typical of cold spray. In this technique hard, not molten particles are deposited with the highest among TS technologies, kinetic energy, which is spent for flattening and adhesion to the substrate. Therefore a conclusion has been drawn that such coating-substrate system might be successfully obtained by another technique, e.g. Vacuum Plasma Spray, where the particles are molten and require much lower deposition speed. In the case of pure soft metal coating such as aluminium on hard substrate, it is very difficult to obtain proper adhesion due to lack of crater formation. Therefore, a only the composite coating might be successful. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 22 Figure 2.5: Fractured cross-section images (SEM) of Al 3Al 2 O 3 composite coatings using agglomerated Al 2 O 3 powders: (a) 10:1 wt%, (b) 1:1 wt% and (c) higher magnification (1:1 wt%) [25]. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 23 3.2. Coating As was described in first chapter, thermal spray process is controlled and optimized experimentally. The selection of feedstock material has been restricted to previous, successful experience achieved in EMPA (Swiss Federal Laboratory for Material Testing and Research). This experience was related to aluminium oxide coatings, with spraying parameters established for steel and aluminium alloy. However, the coatings were deposited on rough substrates. A very important advantage of spraying aluminium oxide is its thermodynamic stability. Also spraying parameters for Ni-Cr (80-20) alloy have been established for many substrates, in both scientific and industrial applications. 4. Thermal treatment Many independent investigations in recent years have shown that the splat morphology can be significantly modified from the undesired splashed morphology to a more round-shaped morphology with limited or no splashing, by just heating the substrate during spraying up to certain point, called transition temperature, T s . [26-29]. Shape of the splat is important because it implicates microstructure, porosity and properties of deposited coating. Transition temperatures range are relatively low substrate temperature (for most materials 100 3400 [°C]) and are irrespective of the particle melting temperature. Transition occurs over a narrow temperature regime [26,29]. As reviewed in [30], T s is affected by: - Thermal conductivity of the substrate Fukumoto et. al. [21] found T s of nickel splats for variety of materials which include stainless steels, mild steels, copper, aluminium, glass, alumina etc. T s were plotted against thermal conductivity of each substrate. The results indicate that a higher thermal conductivity substrate leads to an increased T s . It was suggested [21] that greater interaction of the splat with the substrate results in a higher probability of splashing (see Figure 2.6). - Surface condition The influence of substrate surface conditions, such as adsorption or condensation of gases, and its effects on the transition temperature has been Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 24 investigated with limited attention. However, this may be a significant factor controlling the splat morphology. Some recent experiments [30] indicate that, adsorbed gases are an important factor responsible for droplet splashing and poor adhesion at low deposition temperature (see Figure 2.7). The significance of this study is that its applicability for all droplet and substrate materials combinations. - Type of deposit particles In the case of alumina, which mainly was of interest in this work, it was found that splat morphology changes from highly fragmented to disk-shaped at T s equals to 100 [°C] [30]. Figure2.6: Relation between thermal conductivity of substrate and transition temperature [32]. Figure 2.7: Morphology of ZrO 2 splats made on mild steel substrate in 250 Torr low pressure chamber with induction plasma (a) polished substrate at room temperature; (b) polished substrate, preheated to Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 25 remove the condensation and adsorption and then cooled down in vacuum (5 Torr) for 17 h to room temperature.[32] 5. TS techniques Several techniques of thermal spray have been developed. Their characteristics are described below [33]: 5.1. High Velocity Oxy-Feul (HVOF) HVOF technique (see Figure 2.8a) is a variant of combustion TS which is a relatively low-temperature 2,000-3,160 [°C] process, with the following characteristics: - special torch/nozzle design allows exit gas velocity of up to 4000 [m/s], - feedstock powder is fed axially to the hot zone, - momentum transfer to the particles accelerates them up to about 500 [m/s], - high velocity impact compacts the coating to near-zero porosity, high adhesion and wear resistance. These positive characteristics of HVOF can be utilized as long as the particles are largely molten despite the short residence in the relatively low temperature hot zone. The HVOF process is therefore limited to cermets (such as WC/Co), whereas pure ceramic material is difficult to melt and deposit through HVOF. 5.2. Wire arc spraying Wire spraying technique (see Figure 2.8b) is suitable for metallic coatings, although cermets may also be deposited. Arc is stricken between two metallic wires or ceramic filled wires. The wires are continuously fed into the hot zone (with velocity about 0.1-1[cm/s]), and melt at ends. Only molten particles are sheared from the two wires ends and accelerated towards substrate by compressed gas or air gun: only molten material leaves the torch. This allows high coating rate, which is about 40 [kg/h], at relatively low cost. Additionally it is possible to achieve a high area coating capability, e.g. for infrastructure 3 bridges (zinc), marine (aluminium), corrosion (chemical industry). However, reactive metals, such as titanium or aluminium must be deposited in an inert gas shroud. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 26 5.3. Plasma Spray (PS) PS coating (see Figure 2.8c) is considered to have the highest performance and versatility among various TS systems. High DC (up to 200 [A]) high voltage (more than 100 [V]) provides medium-high power (20-100 [kW]) arc and high temperature plasma (more than 10,000 [°C]). Plasma - hot ionized gas - is produced out of the mixture of gases fed into the arc zone, which is typically argon (majority inert gas) and hydrogen (minority reducing gas with high enthalpy and high thermal conductivity). Nitrogen can be used to increase the plasma temperature for higher melting point materials. Helium can be used instead of hydrogen but this happens rarely, due to its high cost. PS usually operates with powder, fed into the plasma hot zone by the powder carrier gas. The powder must have ability to cflow d carried by the gas and thus spherical shapes and uniform sizes of powder particles are preferable. The feedstock particle size depends on melting temperature of the material but is generally in the range of 10-50 [ µ m]. The powder injection ports are usually located outside of the torch, and usually inject the powder radially. This is convenient but creates a complex trajectory of the powder particles through the hot zone wherein some particles may under-heat (i.e. remain solidified) and some overheat (evaporate). Axial injection (see Figure 2c) is ideal but creates technical problems for PS (axial injection is easier to achieve in HVOF). Few companies offer axially injected PS torches, such as UBC spin-off Northwest Mettech (N. Vancouver, BC), and Sulzer Metco in Switzerland. PS typically provides medium velocity of particles: 100-500 [m/s]. The velocity and heat transfer decide about particle melting characteristics. The high temperature of plasma allows spraying of almost any non-decomposing material, especially ceramics. PS variants include (i) inert gas shroud to avoid adverse effects of mixing with air, (ii) radial or axial powder feed, (iii) Vacuum Plasma Spray (VPS), where vacuum in chamber is about 0.1 [atm] or Air (APS), (iv) Very High Power (more than 200 [kW]), which allows high deposition rate i.e. more than 5 [kg/h]. This technique needs extensive cooling, e.g. requires that the process is done under water. Recently, a new technology of Liquid PS has been developed, wherein liquid carrying fine (less than [1 µ m]) solid particles is injected into plasma plume. This opens up avenues for nano-coatings. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 27 A large number of process variables controls PS process. Among them best studied are the effects of: - substrate preparation (sandblast) and temperature (100-300 [°C]), - powder and gas characteristics (Ar/He/H 2 /N 2 ), - torch-substrate distance (typically 15-50 [cm]), - voltage/current of arc (typically up to 200 [V] / up to 200 [A]), - nozzle characteristics: may be shaped into csupersonic Laval nozzle d resembling rocket nozzle, - particle velocity (typically 100-500 [m/s]). 5.4. Cold spray Cold spray (see Figure 2.8d) is a novel thermal process, which emerged after 2000. This technique doesn 9t use plasma heating of the powder passing through the nozzle. Instead, the powder particles are accelerated to very high velocity (more than 1 [km/s]) by fast flowing gas, which is typically nitrogen, heated to a relatively low temperature of 300-500 [°C]. The ccold d particles impinge on a substrate and generate local heating sufficient for their amalgamation through cspot-welding d and thus coating build-up. This is why the method is sometimes described as ckinetic compaction d instead of spraying. In fact, as contact pressures of more than 30 [GPa] exceeds yield strength of particle materials up to a factor of 1000 times, extensive plastic deformation without any melting is sufficient to produce dense coatings. The best results of cold spray are achieved for relatively low-yield-strength metals such as aluminium, copper, or nickel. However, many other metals were cold- sprayed, e.g. steel, titanium, tantalum, cobalt etc. Some successful depositions have been carried out with ceramics. The relatively low overall temperature of the process assures little oxidation of the metallic particles, which is very important for reactive metals such as aluminium or titanium, and in oxidation-sensitive applications such as electronics. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 28 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 29 Figure 2.8: Thermal spray techniques; Ref: K.E. Schneider, V. Belashchenko, M. Dratwinski, St. Siegmann, A. Zagorski; Thermal Spraying for Power Generation Components, Wiley-VCH 6. Summary The presented state-of-the-art allows better understanding of the possibilities and shortcomings of different materials systems as well as spraying techniques suitable for fundamental study of adhesion and residual stress. Selection of particular coating 3substrate system will be presented in Chapter 3, in relation to the applicability to particular measurement method. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 30 Chapter 3 MATERIALS EXPERIMENTAL PROCEDURE 3. Coating 3 substrate system Materials from different groups were investigated in order to conclude how formation of thermal spray coating depends on mechanical, thermal and chemical properties of selected materials systems Coatings were thermally sprayed on a smooth (without scale roughness) substrates. Five types of coating-substrate systems were selected (see Table 3.1). System type Coating Substrate a. ceramic-metalloid ± Al 2 O 3 Si(001) b. ceramic-metal 1. ± Al 2 O 3 2. ± Al 2 O 3 1. AlMgSi0.5 2. St37-2 c. ceramic-ceramic ± Al 2 O 3 Sapphire d. metal-metal 1. NiCr (80-20) 2. NiCr (80-20) 1. AlMgSi0.5 2. St37-2 e. metal-ceramic NiCr 80-20 Sapphire Table 3.1: Coating-substrate systems selected for VPS experiments. Detailed reasons of selection of each coating-substrate system are described below: a. Thermally sprayed aluminium oxide on a silicon (001) wafer has been selected as a novel fundamental approach of studying adhesion of VPS coatings. Such investigation requires the simplest possible coating- substrate system, which was expected in this case because of two reasons: i) well known mechanical and chemical properties of silicon, ii) experience in spraying high quality Al 2 O 3 coatings with good adhesion to Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 31 either ceramic or metallic materials [34]. Moreover, Al 2 O 3 -Al composite powders were already successfully sprayed on silicon wafers [25]. b. VPS Al 2 O 3 coatings on polished metallic substrates: aluminium alloy (AlMgSi0.5) and steel (St 37-2), were of interest due to their close to industrial application as wear and corrosion resistant barriers. Successful deposition on flat substrate has been expected because of two reasons: i) presence of natural Al 2 O 3 passive layer on Al alloy which is fully compatible with the deposited particles ii) high ductility of Al alloy, which should compensate stresses generated due to the impact of particles and their fast solidification. However, both systems present constraints for fundamental study of adhesion of VPS coatings due to their complex phase diagrams. Moreover, surface chemistry of the steel substrate is not known. c. VPS Al 2 O 3 on a sapphire has been selected also as a novel fundamental approach of studying adhesion of VPS coatings. A very simple coating - substrate system was anticipated because of three reasons: i) well known mechanical and chemical properties of sapphire, ii) experience in spraying Al 2 O 3 coatings with good adhesion to ceramic materials, iii) chemical identity of the coating and substrate. d. Both VPS NiCr (80-20) on a flat steel (St 37-2) and aluminium alloy (AlMgSi0.5) substrate have been of interest as examples of two ductile materials system and due to potential industrial applications. Successful deposition on flat substrate has been expected, especially in the case of steel, because of similarity of thermal properties of substrate and coating. e. VPS NiCr (80-20) on the sapphire substrate has been selected in order to investigate the adhesion between a ductile coating and a brittle substrate. Tables 2-8 present comparison of basic properties for each selected coating-substrate systems, important for the successful VPS deposition. Values taken from literature have a qualitative meaning only and give an idea what the deposition may result in. If the coating is successfully sprayed, the mechanical properties, important for the particular system and scale of investigation, should be experimentally measured because of: i) size effect [35], ii) dependence on processing method. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 32 VPS alumina on silicon (001) Nomenclature Units Substrate Coating GENERAL Formula Si Al 2 O 3 Density Á g·cm -3 2.33 3.95-4.1 Melting point T m °C 1414 2072 Boiling point T b °C 3265 2977 STRUCTURE Crystal structure diamond cubic ± - rhombohedral (<1000 [°C]), ³ -cubic (>1000 [°C]) THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 0 21675.7 Heat of fusion kJ·mol -1 50.21 - Heat of vaporization kJ·mol -1 359 - MECHANICAL PROPERTIES Elastic modulus E GPa 185 103 Fracture toughness K IC MPa·m 0.5 0.91 3-5 Hardness H kgf·mm -2 Knoop: 1150 Vickers: 1256 Thermal conductivity k W·K 21 ·m 21 149 (25[°C]) 4-3.5 (25-600[°C]) Coef. of thermal ± 10 26 /C° 2.6 5.6-7.0 (100- Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 33 expansion (25[°C]) 300[°C]) Table 3.2: Comparison of selected properties of Al 2 O 3 and Si. Most of data for Al 2 O 3 and Si from Wikipedia, K IC for Si http://design.caltech.edu/Research/MEMS/siliconprop.html K IC , for Al 2 O 3 http://www.accuratus.com/alumox.html H, ± , k, E for Al 2 O 3 Yamasaki R., Physical Characteristics of Alumina Coating Using Atmospheric Plasma Spraying and Low Pressure Plasma Spraying (VPS), 2004 H for Si http://www.ioffe.rssi.ru/SVA/NSM/Semicond/Si/mechanic.html VPS alumina on aluminium alloy (AlMgSi0.5) Nomenclature Units Substrate Coating GENERAL Formula AlMgSi0.5 Al 2 O 3 Density Á g·cm -3 2.7 3.95-4.1 Melting point T m °C 615-654 2072 Boiling point T b °C 2518.85* 2977 STRUCTURE Crystal structure face- centered cubic ± - rhombohedral (<1000 [°C]), ³ -cubic (>1000 [°C]) THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 - 21675.7 Heat of fusion kJ·mol -1 10.79* - Heat of kJ·mol -1 293.4* - Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 34 vaporization MECHANICAL PROPERTIES Elastic modulus E GPa 69 117.2±12** Fracture toughness K IC MPa·m 0.5 14-28* 3-5 Vickers hardness H kgf·mm -2 107 433±68** Thermal conductivity k W·K 21 ·m 21 200 (25[°C)] 4-3.5 (25-600[°C]) Coef. of thermal expansion ± 10 26 /C° 25.6 (25[°C]) 5.6-7.0 (100- 300[°C]) Table 3.3: Comparison of selected properties of Al 2 O 3 and Al alloy (AlMg0.5Si). T m , k, ± for AlMg0.5Si http://asm.matweb.com/search/SpecificMaterial.asp?bassnum=MA6063T832 **E, H for Al 2 O 3 on AlMgSi0.5 - experimental values, measured by indentation method: Indenter: Diamond Pyramid 136,° Load: 0.98 [kgf], Loading/Unloading speed: 0.1 [mm/min] VPS alumina on steel (St 37-2) Nomenclature Units Substrate Coating GENERAL Formula St 37-2 Al 2 O 3 Density Á g·cm -3 7.72-8.0* 3.95-4.1 Melting point T m °C 1370- 1510 2072 Boiling point T b °C 2500 2977 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 35 STRUCTURE Crystal structure *** ± -rhombohedral (<1000 [°C]), ³ -cubic (>1000 [°C]) THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 0 21675.7 Heat of fusion kJ·mol -1 - - Heat of vaporization kJ·mol -1 - - MECHANICAL PROPERTIES Elastic modulus E GPa 190-210* 103 Fracture toughness K IC MPa·m 0.5 50** 3-5 Vickers hardness H kgf·mm -2 213-800* 1256 Thermal conductivity k W·K 21 ·m 21 19.9-48.3 (25[°C])* 4-3.5 (25-600[°C]) Coef. of thermal expansion ± 10 26 /C° 9.4-15.1 (25[°C])* 5.6-7.0 (100- 300[°C]) Table 3.4: Comparison of selected properties of Al 2 O 3 and steel (st37-2). * general properties of tool steel from http://www.efunda.com/Materials/alloys/alloy_home/steels_properties.cfm **for steel 4340 (datafrom Wikipedia) T m for steel http://education.jlab.org/qa/meltingpoint_01.html T b for steel http://www.sapiensman.com/conversion_tables/specific_weights.htm ***Iron-carbon phase diagram http://en.wikipedia.org/wiki/Steel Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 36 VPS alumina on sapphire Nomenclature Units Substrate Coating GENERAL Formula Al 2 O 3 Al 2 O 3 Density Á g·cm -3 3.98 3.95-4.1 Melting point T m °C 2030 2072 Boiling point T b °C 2977 2977 STRUCTURE Crystal structure Hexagonal system, rhomboidal class 3m ± - rhombohedral (<1000 [°C]), ³ -cubic (>1000 [°C]) THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 21675.7 21675.7 Heat of fusion kJ·mol -1 - - Heat of vaporization kJ·mol -1 - - MECHANICAL PROPERTIES Elastic modulus E GPa 345 103 Fracture toughness K IC MPa·m 0.5 2.38 (direction [0001]) 4.54 (direction[11- 20]) 3-5 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 37 Hardness H kgf·mm -2 Knoop: 1835 parallel to C- axis, 2243 perpendicular to C-axis Vickers: 1256 Thermal conductivity k W·K 21 ·m 21 23.1 parallel to optical axis, 25.2 perpendicular to optical axis (73[°C]) 4-3.5 (25-600[°C]) Coef. of thermal expansion ± 10 26 /C° 6.66 parallel to optical axis, 5 perpendicular to optical axis (100[°C]) 5.6-7.0 (100- 300[°C]) Table 3.5: Comparison of selected properties of Al 2 O 3 and sapphire. K IC for sapphire from http://www.ceramics.nist.gov/srd/summary/ftgsaph.htm Crystal structure E, H, k, ± , T m for sapphire from http://www.mt-berlin.com/frames_cryst/descriptions/sapphire.htm VPS NiCr(80-20) on aluminium alloy (AlMg0.5Si) Nomenclature Units Substrate Coating GENERAL Formula AlMgSi0.5 Ni-Cr 80-20 Density Á g·cm -3 2.7 7.75-8.65 Melting point T m °C 615-654 1400-1700 Boiling point T b °C 2518.85* >2000 STRUCTURE Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 38 Crystal structure face- centered cubic * THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 - - Heat of fusion kJ·mol -1 10.79* - Heat of vaporization kJ·mol -1 293.4* - MECHANICAL PROPERTIES Elastic modulus E GPa 117.2±12 220 Fracture toughness K IC MPa·m 0.5 14-28* 65-150 Vickers hardness H kgf·mm -2 433±68 1018 Thermal conductivity (25[°C]) k W·K 21 ·m 21 200 8-17 Coef. of thermal expansion (25[°C]) ± 10 26 /C° 25.6 9-16 Table. 3.6. Comparison of selected properties of NiCr 80-20 and Al alloy (AlMg0.5Si). *Ni-Cr phase diagram http://images.google.pl/imgres?imgurl=http://www.msm.cam.ac.uk/UTC/therm ocouple/images/Chromel2.jpg&imgrefurl=http://www.msm.cam.ac.uk/UTC/ther mocouple/pages/DriftInTypeKBareWiresThermocouples.html&usg=__41sH8d ZY5Phtj0Rtywft- JQS_bw=&h=384&w=517&sz=30&hl=pl&start=6&um=1&itbs=1&tbnid=AeywW 8jaA5d7- M:&tbnh=97&tbnw=131&prev=/images%3Fq%3DNiCr%2B%2Bphase%2Bdia Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 39 gram%26um%3D1%26hl%3Dpl%26client%3Dfirefox- a%26hs%3DAI9%26rls%3Dorg.mozilla:pl:official%26tbs%3Disch:1 *data for NiCr from http://www.nickel-alloys.net/nickel_chrome_alloys.html H for Ni-Cr alloy Sampath R., A structural investigation of a plasma sprayed Ni-Cr based alloy coating, 1992 VPS NiCr(80-20) on steel (St 37-2) Nomenclature Units Substrate Coating GENERAL Formula St 37-2 Ni-Cr 80-20 Density Á g·cm -3 7.72-8.0* 7.75-8.65 Melting point T m °C 1370-1510 1400-1700 Boiling point T b °C 2500 >2000 STRUCTURE Crystal structure *** * THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 0 - Heat of fusion kJ·mol -1 - - Heat of vaporization kJ·mol -1 - - MECHANICAL PROPERTIES Elastic modulus E GPa 190-210* 220 Fracture toughness K IC MPa·m 0.5 50** 65-150 Vickers hardness H kgf·mm -2 213-800* 1018 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 40 Thermal conductivity (25[°C]) k W·K 21 ·m 21 19.9-48.3 * 8-17 Coef. of thermal expansion (25[°C]) ± 10 26 /C° 9.4-15.1 * 9-16 Table. 3.7. Comparison of selected properties of NiCr 80-20 and Al alloy (AlMg0.5Si). VPS NiCr(80-20) on sapphire Nomenclature Units Substrate Coating GENERAL Formula Al 2 O 3 Ni-Cr 80-20 Density Á g·cm -3 3.98 7.75-8.65 Melting point T m °C 2030 1400-1700 Boiling point T b °C 2977 >2000 STRUCTURE Crystal structure Hexagonal system, rhomboidal class 3m * THERMODYNAMIC PROPERTIES Enthalpy of formation f H o 298 kJ·mol -1 21675.7 - Heat of fusion kJ·mol -1 - - Heat of vaporization kJ·mol -1 - - MECHANICAL PROPERTIES Elastic modulus E GPa 345 220 Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 41 Fracture toughness K IC MPa·m 0.5 2.38 (direction [0001]) 4.54 (direction[11- 20]) 65-150 Mohs hardness H kgf·mm -2 Knoop: 1835 parallel to C- axis, 2243 perpendicular to C-axis 1018 Thermal conductivity k W·K 21 ·m 21 23.1 parallel to optical axis, 25.2 perpendicular to optical axis (73[°C]) 8-17 (25[°C]) Coef. of thermal expansion ± 10 26 /C° 6.66 parallel to optical axis, 5 perpendicular to optical axis (100[°C]) 9-16 (25[°C]) Table.3. 8. Comparison of selected properties of NiCr 80-20 and Al alloy (AlMg0.5Si). 2. Experimental procedure 2.1. Thermal spray process Vacuum Plasma Spray (VPS) has been chosen as a deposition technique because of the highest performance and versatility among various TS systems, which were described in Chapter 2. The coatings were obtained in a VPS unit using a Medicoat 50 [kW] with powder feeder (F-4 & V21 nozzle), moved by an industrial robot. Surfaces of metallic substrates were mechanically polished to the roughness R a =1 [ µ m]. Roughness of silicon and sapphire wafers was of a few nm, thus polishing was Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 42 not necessary. Dimensions of substrates used in this work are presented in Figure 3.1. Figure 3.1: Scheme of substrates used in this work: A) silicon wafer, B) sapphire wafer C) polished metals (R a =1[ µ m]): aluminium alloy (AlSi0.5Mg) and steel (St 37-2). Prior to spraying, the samples were washed with ethanol in order to remove any remaining dust or grease from the surface. The material to be deposited was injected in a powder form using argon as carrier gas. Careful control of the substrate temperature while deposition was necessary to optimize adhesion of the coating. Optimal temperature depends on the type of the coating 3 substrate system. Ceramic coatings are generally brittle. Their fracture strain is below 0.1%. The strain due to the rapid solidification is relaxed in multiple cvertical d micro-cracks within the single splat. Thus, in this case, substrate must be kept at low temperature during spraying in order to avoid failure as a result of the cooling stresses. On the other hand, metallic coatings are generally ductile, relax the solidification strain through plastic deformation and rarely crack. In this case preheating of the substrate over a transition temperature, T s before spraying increases the adhesion, which has been clearly indicated by all the experimental studies [32]. Such treatment reduces the surface tension of the coating below the surface tension of the substrate which allows particles to wet the substrate and results in disk- Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 43 shaped, well adhering splats. On the other hand, when the substrate temperature is kept below T s , extensively fingered splats are obtained, easily detaching from the surface. The transition temperature depends on the substrate-coating system and is generally low compared to the melting temperatures of both. The main spraying parameters were the same as in previous successful VPS deposition processes used for industrial applications (see Table 3.9). Spraying parameters constant for each deposition are listed below: Ar Plasma = 40 [sl/min] H 2 Plasma= 6 [sl/min] Vacuum= 100 [mbar] = 10[kPa], which is in range of low vacuum (3 -100 [kPa]) Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 44 Coating/ Substrate Current intensity,A Spray.distance,mm pray.ve oc y,mm s um er o passages Remarks NiCr(80-20) /AlMg0.5Si, St37-2, sapphire 750 275 300 10 Without cooling, preheating 350°C Al 2 O 3 /AlMg0.5Si, St37-2, sapphire 750 275 300 28 Cooling Al 2 O 3 /Si 900 325 150 20 Cooling and 1 9 break between pass. Al 2 O 3 /Si 750 275 150 40 Cooling and 1 9 break between pass. Al 2 O 3 /Si 750 275 150 1 - Al 2 O 3 /Si 750 275 150 40 10 9 break between pass. Al 2 O 3 /Si 750 275 300 30 without cooling Al 2 O 3 /Si 750 275 300 5 without cooling Al 2 O 3 /Si 750 275 300 14 without cooling Table 9: The main spraying parameters. Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 45 2.2. Investigation methods In this work attention was focused mostly on a VPS Al 2 O 3 coating on silicon wafer, which is a novel system, promising for fundamental study of adhesion. For other samples the analyses were performed if first observation after the VPS process indicated successful deposition. Characterization of the obtained coatings comprised: i) morphology and chemical composition by SEM-EDX analysis, ii) crystalline structure by XRD analysis in either conventional Bragg-Brentano geometry in the case of thick layers or grazing incidence X-ray and neutron diffraction (GID) in the case of thin layers, iii) topography of the surface and residual stress expressed as bending curvature, by profilometry. Additionally basic study of adhesion between the deposited material and the substrate was done by: i) in-situ SEM scratch test, ii) interfacial indentation, iii) Rockwell indentation. 2.2.1. Morphology and chemical composition by SEM-EDX analysis a. SEM technique Scanning Electron Microscopy (SEM) is nowadays one of the most popular microscopy techniques for materials imaging. It is mainly due to multiple types of signal emitted from the sample while scanning, which, by use of suitable detector, allows different types of analysis, such as high resolution investigation of morphology delivered by secondary electron (SE) signal as well as chemical composition delivered by, used in this work, X-ray or an Auger electron signal. SEM imaging requires just small preparation investment, and allows obtaining: magnification ranges 10x-10 6 x and high depth of focus. Basic principles and possibilities of SEM imaging described by many authors e.g. in [36]. Image creation in SEM: SEM imaging uses electron wave as a analysed signal instead of optical wave. Electron wave has shorter length and because of it image resolution can be higher. The wave is released and focused onto sample by Vacuum Plasma Sprayed (VPS) Coating 3 Substrate System for Fundamental Study of Adhesion 2010 46 electron beam emitting system consisting of a cathode, the Wehnelt Cylinder (Grid Cap) and an anode. Electrons are created by heating a tungsten hairpin cathode at 2600-2900 [K]. They are accelerated between the cathode and the anode in the optic axis. The Wehnelt Cylinder, which works as s lens, focuses the electrons emitted from the cathode into a small beam diameter, cthe crossover d. Next the condenser lenses within the microscope column, make the crossover smaller, approaching at minimum electron beam diameter of 4 [nm] on the sample surface. Obtained electron beam (primary electrons (PE)) is deflected by Scanning coils so that it scans on the sample surface. (See Figure 3.2.) On the sample surface primary electrons interact with the material and creates multiple type of signal: secondary electron (SE), Auger electron, back scattered electron (BSE) and X-rays. Each type of signal is analyzed by an appropriate detector. Figure 3.2: Scheme of image creation in SEM [36]. Interaction of electrons with material: Primary electrons (PE) penetrate into the sample where they are elastically and inelastically scattered. Elastic scattering means that PE come back again to the surface, and inelastic scattering means that PE lose their Vacuum Plasma Sprayed (VPS) Coating \

less