xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 002001163
007 cr mnu|||uuuuu
008 090429s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002615
Deva Reddy, Jayadeep.
Mechanical properties of Silicon Carbide (SiC) thin films
h [electronic resource] /
by Jayadeep Deva Reddy.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 62 pages.
Thesis (M.S.M.E.)--University of South Florida, 2008.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Alex A. Volinsky, Ph.D.
ABSTRACT: There is a technological need for hard thin films with high elastic modulus. Silicon Carbide (SiC) fulfills such requirements with a variety of applications in high temperature and MEMS devices. A detailed study of SiC thin films mechanical properties was performed by means of nanoindentation. The report is on the comparative studies of the mechanical properties of epitaxially grown cubic (3C) single crystalline and polycrystalline SiC thin films on Si substrates. The thickness of both the Single and polycrystalline SiC samples were around 1-2 m. Under indentation loads below 500 -Newton both films exhibit Elastic contact without plastic deformation. Based on the nanoindentation results polycrystalline SiC thin films have an elastic modulus and hardness of 422 plus or minus 16 GPa and 32.69 plus or minus 3.218 GPa respectively, while single crystalline SiC films elastic modulus and hardness of 410 plus or minus 3.18 Gpa and 30 plus or minus 2.8 Gpa respectively. Fracture toughness experiments were also carried out using the nanoindentation technique and values were measured to be 1.48 plus or minus 0.6 GPa for polycrystalline SiC and 1.58 plus or minus 0.5 GPa for single crystal SiC, respectively. These results show that both polycrystalline SiC thin films and single crystal SiC more or less have similar properties. Hence both single crystal and polycrystalline SiC thin films have the capability of becoming strong contenders for MEMS applications, as well as hard and protective coatings for cutting tools and coatings for MEMS devices.
Hertzian contact theory
x Mechanical Engineering
t USF Electronic Theses and Dissertations.
Mechanical Properties of Silic on Carbide (SiC) Thin Films by Jayadeep Deva Reddy A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Mechanical Engineering Department of Mechanical Engineering College of Engineering University of South Florida Major Professor: Alex A. Volinsky, Ph.D. Ashok Kumar, Ph.D. Nathan B. Crane, Ph.D. Date of Approval: November 8, 2007 Keywords: nanoindentation, hardness, elastic modulus, fracture toughness, hertzian contact theory Copyright 2008, Jayadeep Deva Reddy
i TABLE OF CONTENTS LIST OF TABLES iii LIST OF FIGURES iv ABSTRACT vi CHAPTER 1 1 1.1 Introduction 1 1.2 An Overview of Silicon Carbide 1 1.3 Hard Coatings 4 1.4 Thin Films for MEMS Devices 5 1.5 Thin Film Deposition 5 1.5.1 Chemical Vapor Deposition (CVD) 6 1.5.2 CVD Mechanism 6 1.5.3 Advantages of CVD 8 1.6 Research Objective 11 CHAPTER 2 12 2.1 Mechanical Characteriza tion of Thin Films 12 2.1.1 Bulge Test 13 2.1.2 Micro-Beam Bending 15 2.1.3 Micro Tensile Test 17 2.1.4 Scratch Test 19 2.1.5 Nanoindentation 21 CHAPTER 3 24 3.1 Nanoindentation 24 3.1.1 Hysitron Triboindenter 24 3.2 Testing of Thin Films 26 3.2.1 Tip Geometry 26 3.2.2 Tip Shape Function 27 3.3 Measurement of Elastic Modulus 31 3.4 Hardness 33 3.5 Fracture Toughness 34 3.6 Hertzian Contact Theory 36 CHAPTER 4 40 4.1 Mechanical Characterization of SiC Using Nanoindentation 40 4.1.1 Sample Preparation 40
ii 4.1.2 Growth of Single Crystal 3C-SiC Films 40 4.1.3 Growth of Polycrysta lline 3C-SiC Films 42 4.2 Experiments and Results 43 4.2.1 Surface Polishing 43 4.2.2 Analysis of Hardness and Elastic Modulus for SiC 48 4.2.3 Fracture Toughness Analysis 54 CHAPTER 5 57 5.1 Conclusions and Recommendations 57 5.1.1 Conclusions 57 5.1.2 Properties of SiC Films 58 5.1.3 Surface Roughness Effect 58 5.1.4 Fracture Toughness of SiC Films 59 5.2 Recommendations and Future Research 59 REFERENCES 60
iii LIST OF TABLES Table 1. Properties of MEMS Materials 10 Table 2. Mechanical Properties of Si ngle Crystal SiC, Single Crystal Si, Polycrystalline SiC and Bulk SiC (Lel y Platelet SiC) 54 Table 3. Fracture Toughness Values for Singl e Crystal and Polycrystalline SiC 56
iv LIST OF FIGURES Figure 1.1 The Tetragonal Bonding of a Carbon Atom With the Four Nearest Silicon Neighbors 2 Figure 1.2 The Stacking Sequence of Double Layers in Most Common SiC Polytypes 3 Figure 2.1 Schematic Representation of Bulge Testing Experimental Setup 14 Figure 2.2 Circular Interference Patter ns Used to Measure Deflection of Bulging Film  14 Figure 2.3 Simple Beam Deflection Schematic 17 Figure 2.4 Schematic of Micro Tensile Testing Machine 18 Figure 2.5 Hysitron Three Plat e Capacitor Transducer 20 Figure 2.6 Scratch Morphology on Gold Film 21 Figure 2.7 Schematic of the Na noindenter (Triboindenter) 23 Figure 3.1 Triboindenter Main Unit (Hysitron Inc) 25 Figure 3.2 Profile of the Film Surf ace Before and After Indentation 27 Figure 3.3 Topographic Image at Various Contact Depths 29 Figure 3.4 Contact Area Plot With Resp ect to the Contact Depth of the Tip 29 Figure 3.5 Multiple Load-Displacement Curves Obtained From Indenting (100) Si 30 Figure 3.6 Schematic of Load-Displacement Curve for Depth Sensing Indentation Experiment 32 Figure 3.7 Schematic of Indentatio n Cross-Section Showing Various Parameters 32 Figure 3.8 Schematic of the Radial Cr acks Induced by Berkovich Indenter 35 Figure 3.9 Comparison of Elastic Lo ad-Displacement Data and the Hertzian Curve Fit 38 Figure 3.10 Elastic Load-Displacement and the Hertzian Curves Obtained From SiC Thin Films 39 Figure 4.1 Rocking Curve From the (200) Planes of 3C-SiC Grown on Si(100) 42 Figure 4.2 RMS Roughness and Aver age Roughness Values of the Unpolished SiC 44 Figure 4.3 Topographic Image of the Polycrystalline SiC Before Polishing 45
v Figure 4.4 Load-Displacement Curves Before Polishing 45 Figure 4.5 RMS Roughness and Average Roughness Values After Polishing Polycrystalline SiC 47 Figure 4.6 Topographic Image of the Polycrystalline SiC After Polishing 47 Figure 4.7 Load-Displacement Curves After Polishing Polycrystalline SiC 48 Figure 4.8 Load-Displacement Curve at a Load of 1 mN (a) Polycrystalline SiC (b) Single Crystal SiC 49 Figure 4.9 Load-Displacement Curve at 10 mN (a) Polycrystalline SiC (b) Single Crystal SiC 50 Figure 4.10 Hardness of Single Crystal and Poly crystalline SiC as a Function of Indentation Depth 51 Figure 4.11 Modulus of Single Crystal a nd Polycrystalline SiC as a Function of Indentation Depth 52 Figure 4.12 Load-Displacement Curves for Bulk SiC, Single Crystal, and Polycrystalline 3C-SiC Films and Bulk Si (100) 53 Figure 4.13 Radial Cracks in Polycrystalline SiC Film 55 Figure 4.14 Radial Cracks in Single Crystal SiC Film 55
vi MECHANICAL PROPERTIES OF SILICO N CARBIDE (SIC) THIN FILMS Jayadeep Deva Reddy ABSTRACT There is a technological need for hard th in films with high el astic modulus. Silicon Carbide (SiC) fulfills such requirements w ith a variety of applications in high temperature and MEMS devices. A detailed study of SiC thin films mechanical properties was performed by means of nanoindentation. The report is on the comp arative studies of the mechanical properties of epitaxially grown cubic (3C) single crystalline and polycrystalline SiC thin films on Si substr ates. The thickness of both the Single and polycrystalline SiC samples were around 1-2 m. Under indentation loads below 500 N both films exhibit Elastic contact wit hout plastic deformation. Based on the nanoindentation results polycry stalline SiC thin films have an elastic modulus and hardness of 422 + 16 GPa and 32.69 + 3.218 GPa respectively, wh ile single crystalline SiC films elastic modulus and hardness of 410 + 3.18 Gpa and 30 + 2.8 Gpa respectively. Fracture toughness experiments were also carried out using the nanoindentation technique and values were measured to be 1.48 0.6 GPa for polycrystalline SiC and 1.58 0.5 GPa for single crystal SiC, resp ectively. These results show that both polycrystalline SiC thin films and single crystal SiC more or less have similar properties. Hence both single crystal and polycrystalline SiC thin films have the capability of
vii becoming strong contenders for MEMS applic ations, as well as hard and protective coatings for cutting tools and coatings for MEMS devices.
1 CHAPTER 1 1.1 Introduction This chapter discusses in detail silicon carbide (SiC), hard coatings, thin films in MEMS devices, thin film deposition by Chemical vapor deposition (CVD) and advantages of CVD. 1.2 An Overview of Silicon Carbide Silicon Carbide (SiC) has been used increa singly in electronic devices and MicroElectro-Mechanical Systems (MEMS) due to its capability of operating at high power levels and high temperatures. Another area th at has benefited from the development of thin film technology is in the development of metallurgical a nd protective coatings [1, 2]. One of the challenges in micro level devices is providing corrosion resistance for such environments as biological systems or caustic gases. Silicon Carbide has been recognized as an ideal material for applications th at require superior hardness, high thermal conductivity, low thermal expans ion, chemical and oxidation resistance. Klumpp et. al. were the first to recognize the potential of silicon carbide for use in MEMS devices in 1994 . Since then, it has been used as prot ective coatings in hars h environment [3, 4]. Silicon carbide is a wide band gap se miconductor of choice for high-power, high
2 frequency and high temperature devices, due to its high breakdown field; high electron saturated drift velocity and good thermal conductivity. SiC is a wide band gap semiconductor. It exists in many different polytypes. All polytypes have a hexagonal frame with a carb on atom situated above the center of a triangle of Si atoms and underneath a Si atom belonging to the next layer. The distance, marked as Â‘aÂ’ in Figure 1.1, between neighboring silicon or carbon atoms is approximately 3.08 for all polytypes Â‘C-SiÂ’ is approximately 1.89 Figure 1.1. The Tetragonal Bonding of a Ca rbon Atom With the Four Nearest Silicon Neighbors Silicon atom Carbon C-Si a
3 The carbon atom is the center atom of a tetrahedral structure surrounded by four Si atoms; and the distance between the C atom and each Si atom (marked as C-Si in Figure 1.1) is the same. The stacking sequence is shown in Figure 1.2 for the four most common polytypes, 3C, 2H, 4H and 6H. There are three different layers referred to as A, B, and C in Figure 1.2. If the first layer is A, the next layer according to a closed packed structure will be layer B or C. The different polytype s can be constructed by any combination of these three layers. The 3C-SiC polytype is the only c ubic polytype and it has a stacking sequence ABCABCÂ… or ACBACBÂ… Figure 1.2. The Stacking Sequence of Do uble Layers in Most Common SiC Polytypes A B C A A C B A C A B A A B A A C B A C B A 3C 4H 2H 6H
4 The performance of SiC due to its high-temperature, and hi gh-power capabilities makes it suitable for aircraft, automotive, communica tions, power, and spacecraft industries. These specific industries are starting to take advantage of the benefits of SiC in electronics. SiC films are used as high temperature semiconductors . Thin films have several app lications due to their improv ed mechanical properties, protection against chemical environments, ra diation and mechanical wear. On the other hand thin films have been used due to thei r electrical and optical properties. SiC is suitable for both of these applications. 1.3 Hard Coatings Hardness is an important property for thin films used in electronic, optical, and mechanical applications. Harder coatings also have higher wear resistance, also harder surfaces tend to have lower friction and lubri cation has better results with harder surfaces . Silicon carbide is covalently bonded, wh ich is the reason for its high hardness. Hard coatings have been used successfully fo r two decades to protective materials, and to increase the lifetime and efficiency of cutting t ools. Hard coated surfaces have been used to reduce the problems of chemical diffusion, wear, friction, oxidati on and corrosion and effectively increase the lif e of the lithograp hie-galvanoformung-abformung (LIGA) microdevices [6-8] and other sensitive devices Recently the performa nce and reliability of MEMS components were enhanced dram atically through the incorporation of protective thin-film coatings . SiC hard coatings have helped to increase the efficiency of MEMS devices by protecting them fr om harsh environmental conditions.
5 1.4 Thin Films for MEMS Devices Most MEMS devices are restricted due to low operation temperatures, for example silicon devices are restricted to a maximum temperature of 250 C and can be easily affected chemically. SiC is known for high ther mal conductivity and electrical stability at temperatures higher than 300 C . This has been a vital breakthrough for reliability of MEMS devices in harsh environments. 1.5 Thin Film Deposition There are various techniques for depositing thin films which can have a major affect on the mechanical properties of the film. The most common met hods of depositing thin films are Physical Vapor Deposition (PVD) and Ch emical Vapor Deposition (CVD). The main PVD processes are evaporation and sputte ring. The CVD process involves making a volatile compound react with a material to be de posited with other gase s; in this process a non-volatile solid gets deposited on a suitably placed substrate. SiC thin films can be deposited by CVD. A variety of carbide, nitride, boride film s and coatings can also be deposited by this method . Physical vapor deposition (PVD) is a general term used to describe methods to deposit thin films by the condensation of a vapor ont o the surfaces such as semiconductor wafers. This process involves evaporation at high te mperature in a vacuum, or plasma sputter bombardment. In this chapter we focus mainly on the CVD process.
6 1.5.1 Chemical Vapor Deposition (CVD) CVD is a relatively old techni que. The formation of soot du e to incomplete oxidation of firewood since prehisto ric times is probably the oldest example of CVD deposition. The industrial use of CVD could be traced back to a patent literature by de Lodyguine in 1893, who had deposited W onto carbon lamp filaments through the reduction of WCl6 by H2. Around this period, the CVD process was developed as an economically viable industrial process in the field of extraction and pyrome tallurgy for the production of high purity refractory metals such as Ti, Ni, Zr and Ta. One of the important commercial reactions in CVD is: SiCl 4(g) + CH 4(g) SiC(s) + 4HCl (g) (1400 C) for depositing hard wear resistant SiC surface coating. 1.5.2 CVD Mechanism CVD can be performed in a Â‘closedÂ’ or Â‘ope nÂ’ system. In the Â‘closedÂ’ system, both reactants and products are recy cled. This process is normally used where reversible chemical reactions can occur with a temperat ure difference. There is no universal CVD or standard CVD. Each piece of CVD equipment is individually tailored for specific coating materials, substrate geometry, etc., whether it is used for R&D or commercial production. In general, the CVD equipment consists of three main components: Chemical vapor precursor supply system, CVD reactor,
7 Affluent gas handling system. The CVD equipment is designed and operate d using optimum processing conditions to provide coating with uniform thickness, surface morphology, structure and composition. Suitable designs have taken into consideration the temperature control, reactant depletion, fluid dynamics and heat tran sfer in the system . In general, the CVD process involves the following key steps: Generation of active gaseous reactant species; Transport of the gaseous species into the reaction chamber; Gaseous reactants undergo gas phase reac tions forming intermediate species: at a high temperature above the decom position temperatures of intermediate species inside the reactor. Homogene ous gas phase reaction can occur where the intermediate species undergo subse quent decomposition and/or chemical reaction, forming powders and volatile by-products in the gas phase. The powder will be collected on the substrate surface and may act as crystallization centers, a nd the by-products are tran sported away from the deposition chamber. at temperatures below the dissociation of the intermediate phase, diffusion/convection of the intermediate species across the boundary layer (a thin layer close to the substrate surf ace) occurs. These intermediate species subsequently undergo the following steps: Absorption of gaseous reactants onto the heated substrate, followed by heterogeneous reaction at the gasÂ–solid interface (i.e. heated substrate) which produces the deposit and by-product species. The
8 deposits will diffuse along the heated substrate surface forming crystallization centers followed by film growth. Gaseous by-products are removed from the boundary layer through diffusion or convection. The unre acted gaseous precursors and byproducts will be transported away from the deposition chamber . 1.5.3 Advantages of CVD CVD has the following distinctive a dvantages over other methods: It has the capability of producing highly dense and pure materials. CVD method has high reproducibility and deposits films uniformly at a reasonable deposition rates. It can be us ed to uniformly coat complex shaped components and deposit films with good conformal coverage. Such distinctive feature outweighs the PVD process. It has the ability to control crystal st ructure, surface morphology and orientation of the CVD products by controlling the CVD process parameters like temperature of the system, flow of precursor ga s, and flow of carrier gas. Rate of deposition can be easily controll ed. CVD at lower deposition rates yields epitaxial thin films for MEMS applicati ons. For thick protective coatings, a high deposition rate is preferred and th e deposition rates can be tens of m per hour. Many techniques cannot achieve higher depos ition rates, except plasma spraying. CVD is more economical in the field of thin film technology compared to other techniques.
9 CVD allows the deposition of a large spectrum of materials including, metals, carbides, nitrides, oxides, sulfides, IIIÂ–V and IIÂ–VI materials by using a wide range of chemical precursors such as halides, hydrides, and organometallics. Relatively low deposition temperatures a nd the desired phases can be deposited in-situ at low energies through vapor pha se reactions, and nucleation and growth on the substrate surface. This enables the deposition of refractory materials at a small fraction of their melting temperatures Silicon carbide can be deposited at a lower temperature of 1000 C using chemical reactions rather than doing it at higher temperatures. Relatively low deposition temperatures and energies. Using CVD the desired phases can be deposited in-situ at low en ergies through vapor phase reactions, and nucleation and growth on the substrate surface. This enables the deposition of refractory materials at a small fraction of their melting temperatures. Silicon carbide for example can be deposited at a temperature of 1000 C using chemical reactions; this is less than temper atures used on other processes. Table 1, shows important proper ties of SiC and compares them to other MEMS materials. Among the materials of interest SiC has a better thermal conductivity, Hardness, YoungÂ’s Modulus and physical stability compared to Silicon or Gallium Arsenide. These properties give an advantage to SiC thin films. It can be noted from the table that SiC has a reasonably high electron mobility and large bandgap. SiC devices have been effectively operating at higher temperatures up to 600 C, while the most commonly used and available semiconductor Si could operate onl y at temperatures of around 200 C [1, 2]. These properties make SiC more appealing in the field of MEMS, hard coatings, medical
10 applications, biotechnology, chemical sensor s and other electroni c applications. The advantages of SiC over the other MEMS material has led to more research in the recent years. Table 1. Properties of MEMS Materials Properties/ Material 3C-SiC Si GaAs Diamond Bandgap Eg (eV) 2.4 1.1 1.4 5.5 Thermal Conductivity (W/cm C) 5 1.5 0.5 20 Thermal Expansion ( 10 -5 / C) 4.2 2.6 6.88 1 Hardness (GPa) 35-45 12 7 70-80 Electron Mobility (cm2/ Vs) 1000 1400 8500 2200 Young's Modulus (GPa) 448 190 75 1041 Physical Stability ExcellentGood Fair Excellent Breakdown Voltage (105 V/ cm) 4 0.3 0.4 10
11 1.6 Research Objective The main objective and goal of the present re search is to determine the mechanical properties of the single crystal and polycrystalline SiC thin films.
12 CHAPTER 2 2.1 Mechanical Characterization of Thin Films The main objective of this ch apter is to review various methods used for thin film mechanical characterization. There is a large variation in the mech anical properties of thin films due to various conditions in th e deposition process among other factors. To increase the lifetime and reliability while main taining cost effectiveness, characterization of mechanical properties is necessary. Mechanical properties measurements play an important role in thin film industries because the properties of thin films can di ffer substantially from the bulk mechanical properties [13, 14]. In recent years there has been considerable interest in the mechanical properties of materials at the micro and nano scales. This is motivated partly by interest in inherently small structures such as thin film systems, Micro Electro Mechanical Systems (MEMS), and small-scale composites, and partly by newly available methods of measuring local mechanical properties in small volumes . The test techniques that were used to determine the mechanical propert ies of the bulk materials cannot be directly applied to measure the mechanical propertie s of thin films. Therefore, several new methods have been developed to study the mech anical properties of thin films, which include the bulge test, micro tensile testing, beam deflec tion techniques, nanoindentation or depth sensing technique and re sonance testing to name a few.
13 2.1.1 Bulge Test The Bulge test is one of the most versatile test s. It can be used to characterize the residual stress, elastic modulus, and other important para meters such as yield strength and fracture toughness. Bulge testing is one of the most promising testing methods to determine the YoungÂ’s modulus, residual stress and PoisonÂ’s ratio. In Bulge testing the substrate is locally removed by etching, and a thin film di aphragm is left behind. The basic principle of the bulge test is to pre ssurize the diaphragm up to the desired maximum pressure, and observe interference patterns on the bulging film . Figure 2.1, shows a schematic for the experiment al setup of the bulge test. Pressure is applied to obtain the load-deflection respons e. The film whose properties are to be measured is placed on top of the chuck and adhered strongly using hard wax or epoxy. A pressure manifold is attached to the chuc k through the minute hole provided. As pressure is applied the film deflects a nd fringes are formed as show n in Figure 2.2. These fringes are observed through the lens of the microscope/interferometer placed on top of the bulging film. Now, the pressure is gradually decreased to atmospheric level, during the decrease in the pressure the number of fringes that were fo rmed also decreases. With each decreasing fringe, the pressure is record ed, and load versus deflection graphs are plotted.
14 Figure 2.1. Schematic Representation of Bulge Testing Experimental Setup Figure 2.2. Circular Interference Patterns Used to Measure Deflection of Bulging Film  Thin film CCD Camera 540 nm Light source TV monitor Computer Microscope Interferometer Senso r Pressure Manifol d Air Table Chuck
15 Using the bulge test, the deflection of the thin film is measured as a function of applied pressure. The residual stress and YoungÂ’s modulus E values are then extracted from the linear and cubic coefficients of equation (1). Using a least-sq uare fit, equation (1) can be used to extract the residua l stress and YoungÂ’s modulus for circular, square, and rectangular diaphragms : 2 0 2 0 1 0 21 W E a f C W a t P (1), where 0 = residual stress; P = pressure; t= film thickne ss; 2a = diaphragm width or diameter; W0 = maximum center deflection; C1= constant; v = PoissonÂ’s ratio; and f ( ) = function of Poiss onÂ’s ratio. The bulge test is a potentially powerful tool for characte rizing thin film mechanical properties but is not utilized that much because of its sensitivity to experimental error and tedious sample preparation. 2.1.2 Micro-Beam Bending In a bending test the force required to defl ect the beam is much smaller than the force required for a tensile test or an indentation test In a tensile test the force does not result in a visible displacement whereas in bending the same force yields a displacement that is large enough to be measured optically (e.g. laser interferometer) or mechanically (e.g.
16 surface profilometer or nanoindenter) . Mi crobeam deflection tests have been used to investigate thin film elastic modulus, and yi eld stress of the beam material. Microbeam deflection tests are done by a nanoindenter usin g its load and displ acement monitoring system. By applying the basic theory of be am deflection one can determine the YoungÂ’s modulus and yield stress. This method can be a pplied to free-standing films as well as to films on substrates . Bilayer beams consisting of a substrate and a thin film are fabricated using conventional deposition techniques. The films are then patterned using sta ndard photolithography processes, followed by anisotropic etching, rinsing, and drying. The final structures consist of cantilever beams extending over an open trench. Beam deflection can be performed by various methods. The most common method for beam deflection measurements is commercial load and depthsensing indentation in strument capable of precise positioning of indentations. Figure 2.3, shows a cantilever beam of length L with a point load F at its end. The deflection of the cantilever beam is: I E FL Y* 3 max3 (2), where F is the applied force, L the length of the cantilever, I = wt3/12 is the relevant second moment of inertia of the beam, w the beam width, t the beam thickness. E* = E where is the anticlastic (saddle-like) correction factor. = 1 when the beam is long enough, and the plane-strain conditions along the beam width apply. This is the basic
17 bending mechanism used in simple beam deflection devices to obtain the YoungÂ’s modulus. Figure 2.3. Simple Beam Deflection Schematic The equation (2) is only valid to extrac t the ideal YoungÂ’s modulus of the beam. However, some other effects such as undercut and anticlastic effects add some additional terms to equation (2). In this case the beam length, L is replaced by ( L + LC) in equation (2), where LC is the required length correction. Thus we get an effective YoungÂ’s modulus by replacing L in equation (2) with the corrected length. 2.1.3 Micro Tensile Test The tensile test is the most efficient method because it directly measures elastic modulus, fracture strength and PoissonÂ’s ratio . A si mple schematic of the micro tensile testing Thin film of thickness Â‘tÂ’ L F Y
18 machine is shown in Figure 2.4. This experimental setup was used by K.M. Jackson et. al.  and W.N. Sharpe Jr. et. al.  to measure the mech anical properties of thin films. The specimen was placed in the grips, aligned and fixed in place with an ultraviolet cured adhesive. The specimen was elongated with a piezoelectric actuator until it failed, and then the strain was measured with a la ser-based direct strain measurement device called the interferometric strain displacement gage (ISDG) . Figure 2.4. Schematic of Micro Tensile Testing Machine This system worked extremely well for polycrystalline silicon and several other materials. But this system was not used fo r SiC because the difficulty in fabricating similar specimens created a greater challenge Ironically, the same characteristics that make thin-film silicon carbide a viable alternative to polycrystalline silicon also make it a Grips Specimen Base Fringe Detector Laser Fringe Detector Air Bearing Piezoelectric Translator Load Cell
19 difficult material to fabricate using convent ional microfabrication tools. Micromolding was used to fabricate the SiC specimen in this case . 2.1.4 Scratch Test Scratch testing is a combinat ion of two operations: a verti cal indentation motion and a horizontal dragging motion. In scratch testin g the tip is dragged horizontally while simultaneously the load is increased in the vert ical direction, which leaves a scratch mark on the thin film. This process ma y also detach the th in film from the s ubstrate. Figure 2.5, shows how the system works based on th e three plate capacitor system. The Triboindenter has sub-nanometer depth resolu tion due to its highly sensitive three plate capacitive transducer. The tip displacement and load are measured by a three plate capacitance system as shown in Figure 2.5. The piezoelectric scanner provides precisely controlled X, Y, and Z indenter tip position. The piezoelectric scanner moves the tip over a specimen, while a feedback loop controls the Zaxis height of the scanner to maintain a constant force between the i ndenter tip and the specimen. The Z-axis movement of the scanner is then calibrated to obtain a three dimensional to pographical image . This technique is used to either do inde ntation or scratch on the specimen.
20 Figure 2.5. Hysitron Three Plate Capacitor Transducer Hardness of the thin film can be calculated by measuring the scratch width b, shown in Figure 2.6. The scratch morphology on a gold film can be seen in Figure 2.6. This width b is used in equation 3 to determin e the hardness of the material: 28 b F HN (3), where FN is the normal applied load and b is the scratch track width. Center plate Tip
21 Figure 2.6. Scratch Morphology on Gold Film The critical normal load measured is used to calculate the practical work of adhesion of the film to the substrate : (4), where r is the contact radius, E is the elastic modulus of the film, WA, P is the practical work of adhesion and h is the film thickness. Equation (4) applies only when the thin film delaminates due to normal applied force, and does not account for the residual stress in the thin film. 2.1.5 Nanoindentation There are various methods used as mentioned in the above sections to determine the mechanical properties of thin films. Bulge test, micro-beam bending, micro tensile test , nanoindentation, etc., are a few methods us ed to determine the mechanical properties 2 1 22 2 h EW r PP A Cr b
22 for the thin films. Nanoindentation is wi dely accepted method used to determine the mechanical properties of thin films. Nanoindentation is a very successful way for measuring the elastic modulus and hardness of thin films , the goal of the majority of nanoindentation tests is to determine the elastic modulus and hardness from load-displ acement measurements . However, it can also be used to measure thin film adhesion, and fracture toughness. General hardness testing machines allow me asuring the size of the residual plastic impression on the specimen as a function of the indenter load. This gi ves the area of the residual imprint for an applied load; which is on the order of few square microns, and can be measured using optical techniques. In th e nanoindentation method, the tip penetration depth is measured as the load is applied to the indenter. Knowing the geometry of the indenter allows the size of the area of c ontact to be determined. The depth of the impression is on the order of tens of nanomet ers. The modulus of the material can be obtained from the measurement of the unloading stiffness i.e., the rate of change of load and depth. The advantage of nanoindentation over other met hods is that it is a re latively simple and direct method. Nanoindentation is the process in which a sharp i ndenter is forced in to the sample of interest and withdrawn. Figure 2.7, gives the schematic of the Nanoindenter used for testing the mechanical properties of th in films.
23 Figure 2.7. Schematic of the Na noindenter (Triboindenter) In nanoindentation, depth sens ing techniques are used where the modulus of the specimen is obtained from the slope of th e initial unloading portion of the loaddisplacement curve. The modulus obtained from the sample is defined as reduced contact, or indentation modulus. Nanoinde ntation and measurement of mechanical properties of thin films using Nanoindentation is further explained in detail, in Chapter 3. Signal Adaptor Lock-in amplifier Transducer controller Scannin g p robe microsco p e 3DPiezo-actuator Sam p le sta g e Indenter tip Transducer
24 CHAPTER 3 3.1 Nanoindentation This chapter provides a brief introduction to the Hysi tron Triboindenter, and various applications of nanoindentati on measurements, including ha rdness, YoungÂ’s modulus and other mechanical properties. 3.1.1 Hysitron Triboindenter The Triboindenter is one of the most a dvanced machines for testing mechanical properties; it is a fully automated multi-load range indentation/scratch testing system designed for measuring hardness, elastic modul us and dynamic viscoelastic properties of thin films. The Hysitron Triboindenter was developed to operate in quasi-static or dynamic loading modes and has optional acousti c emission testing capabilities. The most distinguishing feature of Triboi ndenter is it has a low noise floor, making it possible to do shallow indentations of the order of 10 nm or less and piezoele ctric topography in-situ scanning of pre and post indentati on features on the specimen surface. This high-performance staging system showed in Figure 3.1 offers superior stability and flexibility to accommodate a wide range of applications, sample sizes and types. It uses a patented transducer shown in Figure 2.5. This transducer is the heart of the Hysitron
25 indenter, the force applied is an electrostatic force while the displacement is measured by the change in capacitance. This electrostatic actuation does not use much current, which makes it virtually drift-free due to low heating during actuation. Figure 3.1. Triboindenter Main Unit (Hysitron Inc) This patented three-plate capacitor tec hnology provides simu ltaneous actuation and measurement of force and disp lacement . The Triboindent er can be operated in both open loop or closed loop displacement or forc e control modes. This means that the user chooses the mode of operation and amount of force or displacement that will be applied by the indenter . Hys itron comes with a high load module which can perform indentations up to a load of 2 N.
26 3.2 Testing of Thin Films Elastic modulus E, and hardness, H are th e thin films properties most frequently measured in a nanoindentation experiment Through nanoindentation one can obtain both qualitative and quantitative data from a thin film system. The major misinterpretation of data in nanoindentation arises when the tip indenter unintentionally probes the substrate. In order to avoid this misi nterpretation, the maximum dept h of penetration has to be restricted to 10% of the film thickness . So me of the important fact ors that need to be taken care of while using nanoindentation are described in the following paragraphs. 3.2.1 Tip Geometry In this thesis experiments were performed using Berkovich indenter tip. It has a face angle of = 65.27 which has the same projected ar ea-to-depth ratio as the Vickers indenter. However, the original Berkovich indenter had a diffe rent angle of 65.03 although it gives the sam e actua l area-to-depth ratio as the Vickers indent er as well . For a Berkovich indenter the projected contact area A is given by e quations (6) and (7): 27.65tan3322 chA (6), 25.24chA (7), where hc is the contact depth. Figure 3.3 gives the profile of the film surface before and after indentation. All values obtained from a quasi-static nanoindent ation analysis depend
27 on accurate knowledge of the shap e of the indenter tip. Thus the tip shape function A(hc) must be determined. Figure 3.2. Profile of the Film Sur face Before and After Indentation In nanoindentation, th e contact area at hc is not accurate or may be valid only for the ideal geometry of the indenter tip. In reality the tip geometry varies, so in order to get reliable data one needs to obtain the tip area functi on which takes into consideration the true shape of the indenter tip. This is explained in the fo llowing paragraphs. 3.2.2 Tip Shape Function There are various methods used to find th e tip shape function. The most precise but impractical way is by measuring the residual contact area in the Transmission Emission Microscope using the replica method. Other met hods are to image the tip of the indenter with a scanning force microscope (SFM). Howeve r, in this case the tip shape of the SFM must be accurately known to assess the final shape of the tip which is being imaged. The most prevalent and practical solution to find out the tip shape is by carrying out indentations on various materials with known el astic properties. An iterative procedure is hmax hc hf Pmax
28 used to find the compliance of the machine and the tip area f unction . In this case the experimental data are often fitted using equation (8): 16 1 5 8 1 4 4 1 3 2 1 2 1 2 0) (c c c c c c ch C h C h C h C h C h C h A (8), In case of an ideal Berkovich diamond indenter C0 should be set to 24.5. In general it is preferable to use as few coefficients as pos sible. To calculate both the elastic modulus and hardness the projected contact area is requi red. At large contact depths the ideal tip area function, mentioned in e quation (7) can yield accurate results, whereas at low contact depths the actual tip ge ometry must be taken into acc ount to get accu rate results. Since the indents are small, the area of the indent cannot be measur ed directly, or by optical microscopy. In order to find the contac t area, a number of indents are made in samples with known elastic properties. This c ontact area is then calculated from the empirical formula, A = f (hc), which relates the projected contact area (A) is a function of the contact depth (hc). To determine the tip area function of the indenter tip, multiple indents are made in (100) silicon which has the elastic modulus of 174 GPa. Fi gure 3.3 shows contact area variation with the contact depths. Figure 3.4, shows us the plot betw een contact depth and 0contact area.
29 Figure 3.3. Topographic Image at Various Contact Depths Figure 3.4. Contact Area Plot With Resp ect to the Contact Depth of the Tip Figure 3.5 shows the plots obta ined for the load-displaceme nt curves for Si (100) performed at different loads. The residual impressions of the indenter on the silicon
30 surface can be seen in Figure 3.4. In this pr ocedure, 25 indentations were made starting from 400 N maximum load to 10 mN load with an increment of 400 N to have the full range of required disp lacements. From the known values of Si properties, one can relate contact depths to contact area of the tip fo r various depths from the empirical fit of equation (8). This area func tion is used to determine ha rdness and elastic modulus for unknown materials. Figure 3.5. Multiple Load-Displacement Curves Obtained From Indenting (100) Si 0 4000 8000 0100200Load (N)Depth (nm) P maxS
31 3.3 Measurement of Elastic Modulus The most common method for measuring ha rdness and modulus using nanoindentation methods involves making a small indentation in the film, while c ontinuously recording the indentation load, P, and displacement h, during one complete cycle of loading and unloading. Stiffness of the contact between the indenter and the material being tested is required to determine the mechanical properties of interest. The stiffness, S, is determined from the initial slope of the unloading curve. S = dP/dh where P is described by the power relation given by Oliver and Pharr : P = A( pl)m (9), where A and m are fitting parameters, P and are the load and displacement taken from the top 65% of the unloading curve. The lo ading and unloading portion can be seen in Figure 3.6
32 0 4000 8000 1.2 10404080120Load (N)Depth (nm) S P max Figure 3.6. Schematic of Load-Displacemen t Curve for Depth Sensing Indentation Experiment Figure 3.7. Schematic of Indentation Cro ss-Section Showing Various Parameters Surface Pmax hmaxSurface profile at maximum load Surface profile after load removal Indenter h Loading Unloading hmax
33 Various indent parameters are shown in Figu re 3.7. The indent crosssection explains the residual imprint of the indent after unloading. One can also observe the elastic recovery, which gives a clear picture of surface profile at the maximum load. Once the stiffness is measured using equation S = dP/dh, the reduced modulus can be determined as: A S Er2 (10), where Er is the reduced modulus, which accounts for the measured elastic displacement contributing from both the sample and the inde nter tip. The reduced modulus can be used to calculate the actual modulus of the sample, which is given by: Tip tip sample sample rE E E2 21 1 1 (11), where Etip = 1140 GPa and Vtip = 0.07 are the elastic modu lus and PoissonÂ’s ratio for diamond tip, respectively. From this equation we can calculate the E sample for the given sample. 3.4 Hardness Hardness is the resistance to the plas tic deformation and it is given by: A P Hmax (12), where Pmax is the maximum load and A is the projected area of contact or hardness impression. The effect of indentation depth on hardness measurement has been a real area
34 of concern. When low loads are applied the re sultant area of contact might be very small or sometimes it recovers elas tically with no residual impressi on left behind. This gives an exaggerated hardness value. The most comm on method to determine the hardness of a material is by static indentation Hardness can also be determined using scratch hardness testing using the same nanoinde ntation machine. This method is explained in detail in section 2.1.4. Precautions should be taken in order to avoi d discrepancies in calculating mechanical properties from nanoindentation data. Indents should not be ma de too deep into the thin film as the substrate effects may be noticed [ 29]. In an attempt to avoid substrate effect on thin film elastic modulus and hardness, often excessively shallow indents are made into the thin film. By taki ng indents that are not deep enough, the elastic modulus and hardness measurements will be inaccurate. Th is inaccuracy is due to surface roughness, possible oxidation effects and errors in assessing the tip contact area . 3.5 Fracture Toughness Since a nanoindenter is a versatile machine it can be used to evaluate the fracture toughness of a given bulk material. Small cr acks on the surface of thin films can be induced when higher loads are applied. These patterns of cracks are used to assess the film fracture toughness. Cracks come in different morphology depending on the indentation load, tip indenter geometry a nd material properties The most common kind of cracks are radial cracks for brittle and hard materials [31, 32]. Figure 3.8, shows the schematic of load-induced radial cracks pr opagating from indentat ion using a Berkovich
35 tip. Fracture toughness is calculated from equa tion (13), this is th e most widely used relationship : 2 3 2 1C P H E A Kc (13), where, Kc is the fracture toughness, P is the maximum load, C is the crack length, E is the elastic modulus and H is the hardness, A is an empirical constant, for the Berkovich tip it is 0.016. Figure 3.8. Schematic of the Radial Crac ks Induced by Berkovich Indenter Generally there are three types of cracks, radi al cracks, lateral cracks, and median cracks. Radial cracks occur on the su rface of the specimen at the corners of the indenter edge marks. These cracks are generally formed due to the hoop stress. Figure 3.8, shows a schematic of radial crack propa gation at the edges of the i ndenter contact site. Lateral C Indenter contact area
36 cracks are cracks which occur beneath the surface. These cracks are generated by tensile stress and often extended to the sample surf ace. Median cracks are circular penny shaped cracks that are formed beneath the surface and along the line of symmetry. Fracture mechanics treatment of these types of radial and lateral cracks is useful to provide the fracture toughness based on the leng th of radial cracks . 3.6 Hertzian Contact Theory In nanoindentation the most predominantly used indenter tip is the Berkovich tip, which has the shape of a three-sided pyramid. Berkovi ch indenters are not perfectly sharp. The most common assumption is to describe the Be rkovich indenter as spherical at its tip . However, since the radius R of the tip is not known, mo st users use the tip radius specified by the manufacturer . These assumptions are s ubjected to great uncertainty since the indenters can wear out and change their tip geometry. To avoid such errors some researchers have directly measured th e tip radius by scanning the tip using atomic force microscopy or by scanning electron microscopy . Another popular approach to determine the radius of the tip is to fit the load-displacement curve with the Hertzian equation. The stress es and deflection arising from the contact between two elastic bodies are of particular interest for indentation testing. Hertz found that the radius of the circle of contact Â‘aÂ’ is related to the indenter load P, the indenter radius R, and the elastic properties of cont acting materials by equa tions (14) and (15): Hertz originally derived the equation for two cyli nders in contact; this theory is applied to the spheres in this case.
37 rE PR a 4 33 (14), 2 3 2 13 4 h R E Pr (15), where P is the indenter load, h is the displacement, and Er is the reduced modulus,  which can be determined from equation (11). Once the radius is known the stresses in the material can be evaluated from the Hertzian contact mechanics as a function of applied load. In this research we establish the tip radius using the Hertzian curve fit. Figure 3.10, gives the Hertzian fit for the loading curve of a thin film sample which was indented using the Berkovich tip.
38 0 200 400 600 0102030 LoadDisplacement Curve Hertzian FitLoad (N)Depth (nm) Figure 3.9. Comparison of Elastic Load-Dis placement Data and the Hertzian Curve Fit This Hertzian curve fit was obtained from inden titions made in the SiC thin films. To find the radius of the tip, experiments were done at low loads up to 500 N to obtain complete elastic load-displacement curv es as shown in Figure 3.10.
39 0 200 400 600 0102030 L -D curve (Polycrystalline) Hertzian Fit (Polycrystalline) L-D curve (Single Crystal) Hertzian fit (Single crystal)Load (N) Depth (nm) Figure 3.10. Elastic Load-Displacement and the Hertzian Curves Obtained From SiC Thin Films From Figure 3.11, one can observe that the Hertzian curve fit was done on both single crystal and polycrystalline SiC elastic load-d isplacement curves to extract the radius of the Berkovich indenter. The radius of the tip was found to be approximately 100 Â– 110 nm. This indenter tip is us ed in the experiments done, which is explained in chapter 4.
40 CHAPTER 4 4.1 Mechanical Characterization of SiC Using Nanoindentation This chapter deals with growth of single cr ystal and polycrystalline SiC and experiments performed on the single crystal SiC, polycry stalline SiC, bulk SiC (Lely Platelet SiC), and bulk Si (100) films using Hysitron Triboi ndenter. This chapter talks about sample preparation, experimental set up and the obtained results. 4.1.1 Sample Preparation Two samples were studied and their mechanical properties such as elastic modulus, hardness and fracture toughness were compare d. The samples used for the comparative study were single crystal 3C-SiC and polyc rystalline SiC, grown by heteroepitaxy chemical vapor deposition. Film thickness of the samples were around 1-2 m. 4.1.2 Growth of Single Cr ystal 3C-SiC Films The most common technique used to grow cr ystalline films epitaxial ly is CVD. 3C-SiC single crystal films were grow n on 50-mm diameter (100) Si wafers using hot-wall CVD. The design of the CVD reactor can be found elsewhere . The 3C-SiC on Si growth process was developed using the two step carbonization
41 and growth method. C3H8 and SiH4 were used as the precursor gases to provide the carbon and silicon sources, respectively. Ultr a high purity (UHP) hydrogen, purified in a palladium diffusion cell, was employed as the carrier gas. Prior to growth, the samples were prepared using the standard RCA cl eaning method , followed by a 30 second immersion in diluted hydrofluoric acid (HF), to remove surface contaminants and native oxide. The first stage of the process, known as the carbonization step, involved heating the reactor from room temperature to 1140 C at atmospheric pressure with a gas flow of 6 standard cubic centimeters per minute (sccm) of C3H8 and 10 standard liters per minute (slm) of H2. The temperature was then mainta ined at 1140 C for two minutes to carbonize the Si substrate surf ace. After carbonization, SiH4 was introduced into the system at 4 sccm and the temperature incr eased to growth temperature of 1375 C, and gas pressure of 100 Torr was maintained fo r approximately 5 minutes. The temperature and other flow rates were maintained cons tant during the growth process. By this procedure, a sample 2 m thick 3C-SiC was grown. After the growth process was completed, th e wafer was cooled to room temperature in Ar atmosphere . After deposition th e film thickness was measured by Fourier Transform infrared transform (FTIR) a nd confirmed by scanning electron microscopy. The crystal orientation of th e film deposited was determin ed by X-ray diffraction (XRD) using a Philips X-Pert X-ray diffractomer. XR D data proved that the films were single crystal. Figure 4.1, shows the rocking curve obtained from the (200) planes for 3C-SiC grown on (100) Si. This data confirms the film is singl e crystal 3C-SiC.
42 0 2000 4000 -0.500.5Intensity (counts)Omega (degrees) FWHM ~300 arcsec 3C-SiC <200> peak Figure 4.1. Rocking Curve From the (200) Pl anes of 3C-SiC Grown on Si (100) 4.1.3 Growth of Polycrystalline 3C-SiC Films Polycrystalline growth follows the same procedure as single crystal SiC with the exception of a higher gasses flux. The process conditions for the samples studied here were identical to those list ed above except that the SiH4 and C3H8 mass flow rates were 6 sccm and 4 sccm, respectively. This process resulted in a polycrystalline-3C-SiC film.
43 4.2 Experiments and Results This section explains the results obtained from the experiments conducted using the nanoindenter and the analysis of the data to determine the mechanical properties. Samples of same film thickness (2 m) we re used to conduct the experiments. The samples tested were 3C-SiC single crystal grown on Si (100). This sample had a good optical-quality-smooth-surface requiring no fu rther polishing. On the other hand, the deposited polycrystalline sample was rough and needed mechanical polishing. The sample was polished using a 1 m pad with Leco diamond paste to smooth the film surface and reduce the film thickness to match the thickness of the sing le crystal SiC film (2 m). These samples were then cleaved and glued to the sample holders using cyanoacrylate (Super glue). The Berkovich indenter was used for all indentation tests. Th is is the best tip for most bulk samples, unless the RMS roughness is hi gher than 50 nm . Load controlled indentations were performed to determ ine films elastic modulus and hardness. 4.2.1 Surface Polishing Polycrystalline SiC was polished since its as-deposited surface was too rough, and indentation experiments were not giving ma tching load-displacement curves. Figure 4.2, shows the root mean square roughness, av erage roughness, and peak -to-valley height before polishing. Figure 4.3, is the 3D image of the as-deposited poly crystalline specimen before polishing. Figure 4.4, gives an insight as to why polycrystal line SiC had to be polished before doing experiments.
44 The uneven loading and unloading curves were observed on this specimen mainly because of the coarse surface where the tip sl ips between the peaks a nd the valleys of the film surface. This excessive surface roughness was giving unr epeatable results; in order to obtain repeatable loading and unloading curves the specimen was polished with 1 micron diamond paste. After polishing the topographic scans were taken and the indentation experiment was repeated for the same maximum loads. Figure 4.2. RMS Roughness and Average Roug hness Values of the Unpolished SiC
45 Figure 4.3. Topographic Image of the Po lycrystalline SiC Before Polishing Figure 4.4. Load-Displacement Curves Before Polishing 0 2000 4000 6000 8000 1 1040100200 Indent at position 1 at 10 mN Indent at position 2 at 10 mN Indent at position 3 at 10 mNLoad (N)Depth (nm) Indent 2 Indent 3 Indent 1
46 Figure 4.5, gives the RMS roughness, average r oughness, and peak-to-valley height of the polycrystalline SiC afte r polishing. Figure 4.6, is the topography image of the SiC after polishing and Figure 4.7, shows the lo ad-displacement curves obtained after polishing. Before polishing the average roughness was 44.4 nm, and the RMS roughness was 53.9 nm. After polishing the average roughness was 1.5 nm, and RMS roughness was 2 nm.
47 Figure 4.5. RMS Roughness and Average Roughness Values After Polishing Polycrystalline SiC Figure 4.6. Topographic Image of the Po lycrystalline SiC After Polishing
48 Figure 4.7. Load-Displacement Curves After Polishing Polycrystalline SiC 4.2.2 Analysis of Hardness a nd Elastic Modulus for SiC Standard low load transducer, which can ap ply a maximum load of 10 mN, was used to find the elastic modulus (E) and hardness (H) of the deposited films. The experiment was carried at loads varying between 500 N to 10 mN on both the single crystal and polycrystalline SiC samples. The load-dis placement curves obtained from single and polycrystalline SiC films are compared in Figures 4.8 and 4.9. The hardest materials of the two has less penetration depth for the same load, hence polycrystalline SiC is harder. 0 4000 8000 04080120Polycrystalline SiC After PolishingLoad (N)Depth (nm)
49 From Figure 4.8, it can be inferred that at lower load both the single and polycrystalline samples exhibit similar elastic contact. Also elastic load-displacement curves helps in determining the radius of the indenter tip used in performing the nanoindentation experiments, by using the Hertz theory of elastic contact [16, 24]. Using high loads varying from 5mN to 10 mN we saw the pl astic deformation in the film. Figure 4.10 shows the indentation done at a load of 10 mN, from whic h it can be inferred that the indenter penetrated more into the single crystal SiC than polycrystalline SiC. Figure 4.8. Load-Displacement Curve at a L oad of 1 mN (a) Polycrystalline SiC (b) Single Crystal SiC 0 400 800 1200 01020Load (N)Depth (nm) (a) (b)
50 Figure 4.9. Load-Displacement Curve at 10 mN (a) Polycrystalline SiC (b) Single Crystal SiC Figure 4.10 and Figure 4.11, shows the hardness and modulus values of respective SiC films at various loads obtained from nanoinde ntation tests. Reduced modulus values of the thin film obtained from the nanoindentat ion are calculated us ing the equation (11). 0 4000 8000 04080120Load (N)Depth (nm) (a) (b)
51 Figure 4.10. Hardness of Single Crystal a nd Polycrystalline SiC as a Function of Indentation Depth 26 30 34 38 04080120 Single crystal SiC Polycrystalline SiCHardness (GPa)Depth (nm)
52 Figure 4.11. Modulus of Single Crystal a nd Polycrystalline SiC as a Function of Indentation Depth 350 450 550 04080120 Polycrystalline SiC Single crystal SiCElastic modulus (GPa)Indentation Depth (nm)
53 Figure 4.12, Shows the load disp lacement curves for Lely-Platelet 15RSiC (Bulk SiC), polycrystalline SiC, single crystal SiC and Si (100) at 10 mN. Table 2 gives the hardness and modulus values for 15RSiC (Bulk SiC), polycrystalline SiC, single crystal SiC and Si (100) obtained us ing nanoindentation. 0 4000 8000 0100200300 10 mN load curves Leyl platelet SiC (Bulk) Silicon SIngle crystal SiC Polycrystalline SiC Load (N)Depth (nm) Figure 4.12. Load-Displacement Curves for Bulk SiC, Single Crystal, and Polycrystalline 3C-SiC Films and Bulk Si (100)
54 Table 2. Mechanical Properties of Sing le Crystal SiC, Single Crystal Si, Polycrystalline SiC and Bulk SiC (Lely Platelet SiC) Hardness (GPa) Elastic Modulus (GPa) Silicon (100) 12.46 + 0.78 172.13 + 7.76 Lely platelet 15R-SiC 42.76 + 1.19 442 + 16.34 Single crystal 3C-SiC 30 + 2.8 410 + 3.18 Polycrystalline 3C-SiC 32.69 + 3.218 422 + 16 4.2.3 Fracture Toughness Analysis To determine the fracture toughness (K), low lo ad transducer was replaced with high load transducer. The indentation procedure mentioned in Chapter 3 was followed at higher loads ranging from 100 mN to 500 mN. Figures 4.13 and 4.14 show the microscopic imag es of cracks induced at higher loads in polycrystalline and single 3C-SiC films, resp ectively. Crack length was used to calculate the fracture toughness of the thin films using equation (13) The radial cracks were generated along the sharp corners of the Be rkovich tip used for indentation. Table 3, shows the values of fracture toughness of respective SiC sa mples. The cause for the low values of fracture toughness in this case was due to the tip penetrating into the substrate.
55 Figure 4.13. Radial Cracks in Polycrystalline SiC Film Figure 4.14. Radial Cracks in Single Crystal SiC Film 20 m 20 m
56 Table 3. Fracture Toughness Values for Single Crystal and Polycrystalline SiC From Figure 4.13, we can see that the cracks of the single crystal SiC are propagating along the cubic planes. Same effect is not noticed in the polycry stalline SiC Figure 4.14, since they do not have specific cubic planes to further propagate the crack very easily. Material Fracture Toughness (MPa m) Single Crystal SiC 1.58 0.5 Polycrystalline SiC 1.48 0.6 Bulk SiC 4.6 Bulk SiC ( D.Yang and T.Anderson) 2.18
57 CHAPTER 5 5.1 Conclusions and Recommendations This chapter reviews the experimental results obtained and contains future recommendations for comparing mechanical pr operties nanoindentation data with other methods. 5.1.1 Conclusions In this work, mechanical properties of the single crystal and polycry stalline 3C-SiC thin films were studied using Hysitron Triboindenter in the ambient environment. These films were deposited on silicon substrates by th e chemical vapor deposition technique. Hardness and elastic modulus were measured using nanoind entation tests and compared with bulk SiC (Lely Platelet 15R-SiC) propertie s and the silicon substrate. The effect of surface roughness on polycrystalline 3C-SiC thin film mechanical properties was studied. The properties of rough polycry stalline SiC was compared w ith the smooth or polished sample. Fracture toughness of the films was determined from the indentation experiments.
58 5.1.2 Properties of SiC Films Modulus and hardness values found from th e nanoindentation tests for polycrystalline 3C-SiC films were 422 + 16 GPa and 32.69 + 3.218 GPa respectively, while single crystalline 3C-SiC films elastic modulus and hardness were 410 + 3.18 GPa and 30 + 2.8 GPa, respectively. Bulk SiC properties were found to be 442 + 16.34 GPa for the elastic modulus and 42.76 + 1.19 GPa for hardness. E and H fo r silicon substrate were found to be 172.13 + 7.76 GPa and 12.46 + 0.78 GPa, respectively. From the results one can observe that the mechanical properties of the Single and pol ycrystalline SiC films are relatively close to the bulk SiC. 5.1.3 Surface Roughness Effect RMS roughness of as-deposited polycrystal line SiC was measured to be 53 nm approximately and topography images show that the peaks-to-valley depth was around 273 nm. This unevenness in the surface affected the measurement of the properties of the thin films. From the experiments, it was observed that H and E values were not comparable with bulk SiC, but did not affect the results. The SiC film surface was then polished to an RMS roughness of 2 nm approximately and the H and E values were compared with rough SiC film and bulk SiC. The H and E values of as-deposited SiC were measured, and found to be 260 + 86.5 GPa and 18.92 + 9.6 GPa respectively, after polishing the film.
59 5.1.4 Fracture Toughness of SiC Films Fracture toughness of SiC films was studied using nanoindentation. Radial cracks were initiated by applying rela tively higher loads using a sharp Berkovich diamond tip. Fracture toughness was calculate d from the crack length and the values are found to be 1.58 0.5 MPa m for polycrystalline SiC and 1.48 0.6 MPa m for single crystal 3CSiC films. These results were compared with fracture toughness for bulk SiC. From the fracture toughness values obtained for both the single and polycrystalline it can be noted that they have relatively low values compared to the bulk material. This low values can be attributed due to various reasons like, s ubstrate effect or residual stress caused during the deposition of the SiC thin films on Si substrate. 5.2 Recommendations and Future Research Additional experiment could be done to compare the hardness values obtained by nanoindentation with a scratch test results. To confirm the values of the fracture toughness, an in-depth analysis of the cubic plan es need to be studied in order to explain the crack propagation in both single crystal and polycrystalline SiC. Nanoindentation tests could be conducted in wet environmen ts to observe the changes in fracture toughness of these films.
60 REFERENCES 1. M Ohring, The Material science of Thin films, Academic press, 1991. 2. www.ifm.liu.se/matephys/new/SiC/ 3. A. Klumpp, U. Schaber, H.L. Offereins, K. KuÂ¨ hl, H. Sandmair, Sensors and Actuators A 41Â–42, 310Â–316(1994). 4. N. Ledermann, J. Baborowski, P. Muralt, N. Xantopoulos, J.-M. Tellenbach, Surface and Coatings technology 125, 246-250 (2000). 5. www.lerc.nasa.gov/WWW/SiC/SiC.html 6. J. Musil, Surface and Coatings Technology 125, 322Â–330 (2000). 7. D. M. Cao, T. Wang, B. Feng, W. J. Meng, and K. W. Kelly, Thin Solid Films 398, 553, References 1(2001). 8. S. Prasad, T. Christenson, and M. Dugger, Adv. Mater. Process. 160, 30, 2002. 9. Y. Tian, A. F. Bastawros, C. C. H. Lo, A. P. Constant and A. M. Russell, B. A. Cook, Applied Physics Letters, volum e 83, number 146 October 2003. 10. Xiao-An Fu, Jeremy L. Dunning, Christian A. Zorman, Mehran Mehregany, Sensors and Actuators A 119, 169Â–176 (2005). 11. K.L. Choy, Progress in Mate rials Science 48, 57Â–170 (2003). 12. S. Venkataraman, D. Kohlstedt, W.W. Gerberich, Metal-Ceramic Interfacial Fracture Resistance using the Continuous Microscratch Technique, Thin Solid Films, 223, pp. 69-275 (1993) 13. P. Benjamin, C. Weaver, Measurement of Adhesion of Thin Films, Proc. Soc. London, 254, pp. 163-176 (1960) 14. D. Beegan, S. Chowdhury, M.T. Laugi er, Surface and Coatings Technology, 201, 5804 (2007). 15. R. Schwaiger and O. Kraft, Script a Materialia, Vol. 41, No. 8, pp. 823Â–829, 1999
61 16. Jay S. Mitchell; Christian A. Zorman; Thomas Kicher; Shuvo Roy4; and Mehran Mehregany, Journal of Aerospace Engineering, 46-54, April (2003). 17. J.H. He, J.K. Luo, H.R. Le, D.F. Moore, Materials Science and Engineering A 423 134Â–142 (2006). 18. H.J. Lee, H.S. Choi C.S. Han, N.K. Lee, G.A. Lee, T.H. Choi, Journal of Materials Processing Technology 187Â–188 ,241Â–244 (2007). 19. Kamili M. Jackson, Jeremy Dunning, Christian A. Zorman, Mehran Mehregany William N. Sharpe, Jr. Journal of MicroElectroMechan ical Systems, Vol. 14, No. 4, August 2005 20. B. Yuan, W.N. Sharpe Jr., Mechanical te sting of polysilicon thin films with the ISDG, Exp. Tech. 21 32Â–35 (1997). 21. Kamili M. Jackson, Sensors and Actuators A 125 34Â–40(2005). 22. W.C. Oliver, G.M. Pharr, An Improved Technique for Determining Hardness and Elastic Modulus using Load and Displaceme nt Sensing Indentation Experiments J. Mat. Res. 7, pp. 1564-1580 (1992) 23. Antony C. Fischer Cripps, Nanoindent ation, Springer-Verlag New York, LLC 24. W.A. Bonin and Hysitron Inc., Multi-dimen sion capacitive transducer, United states Patent, 5869751, 1999. 25. Hysitron Inc., 10025 Valley View Ro ad, Minneapolis, MN, 55344, USA. 26. J.L.Hay, M.E. OÂ’Hern, and W.C. Oliver, Â“The importance of contact radius for substrate-independent property measurement of thin filmsÂ”, Mat. Res. Soc. Symp. Proc. 522, pp 27-32 (1998). 27. E.S. Berkovich, Â“Three-faced diamond pyramid for micro-hardness testingÂ”, Ind. Diamond Rev. 11 127, pp.129-133 (1951). 28. www.nanoindentation.cornell.edu/Testing_I ssues/Nanoindentation-testing_Issues.htm 29. A.A. Volinsky, W.W. Gerberich, Nanoindentation techniques for assessing mechanical reliability at the Nanoscale Mico. Elect. Eng. 69, pp. 519-527 (2003) 30. N. Yu, A. Polycarpou, T. Conry, Tip-radius Effect in Fi nite Element Modeling of Sub-50 nm Shallow nanoindentation Thin Solid Films 450 pp. 295-303 (2003). 31. Anstis GR, Chantikul P, Lawn BR, Ma rshall DB. J Am Ceram Soc; 64:533 (1981).
62 32. Daniel Casellas, Juame Caro, Sivia Molas, Jo se M. Prado, Isaac Valls Acta Materialia 55, 4277-4286 (2007). 33. B.R.Lawn, A.G. Evans, and D.B.Marshall, Â“Elastic/ Plastic indentation damage in ceramics: the median/radial crack system,Â” J. Am. Ceram. Soc. 63, pp. 574581(1980). 34. H.Bei, E.P.George, J.L.Hay, and G.M. Pharr, Â“Influence of Indenter Tip Geometry on Elastic Deformation during NanoindentationÂ” The American Physical Society, July 2005. 35. S.Suresh, T.G. Nieh and B.W. Choi, Scr. Mater. 41, 951(1999). 36. W.W. Gerberich, J.C. Nelson, E.T. Lilleodd en, P. Anderson and J.T. Wyrobek, Acta. Mater. 44, 3585 (1996). 37. K.L. Johnson, Contact Mechanics, Cambridge University press 1985. 38. Thesis S.Harvey, Electrical E ngineering department, USF (2006). 39. W. Kern and D. A. Puotinen, Cleaning solutions based on hydrogen peroxide for use in silicon semiconductor technology RCA Rev. Vol. 31, 187-206 (1970).