This item is only available as the following downloads:
1 FUNDAMENTAL INTERFACIAL STUDIES OF ADVANCED SOLID LUBRICANTS AND THEIR OPERATING ENVIRONMENTS By KEVIN GILLEY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQU IREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Kevin Gilley
3 To Kailey
4 ACKNOWLEDGMENTS I would like to thank first and foremost, my advisor Dr. Scott Perry for witho ut his help and direction my entire graduate career would have not have been possible. I would also like to thank my lab mate Alexander Rudy, for being there to discuss all of my results along the way and helping me to perform measurements. Also deserving of thanks is the entire group of Dr. Greg Sawyer but specifically past members Dr. Brandon Krick and Dr. Rachel Colbert and current members Juan M. Uruea, Kathryn Harris, and Angela Pitenis for all their assistance in taking measurements and knowledge in the area of tribology. Knowing that I could take any result or problem I did not understand and simply ask for help and it would be freely and happily given was an extreme comfort. I would also like to thank my parents, Eddie and Lica, for supporting me th roughout my entire life and raising me to value education and intelligence. I also offer my thanks to all the rest of my family and friends for their support throughout this process. Finally, I would like to express my dearest thanks to my wife Kailey, she has constantly pushed me to excel and has supported me above all others throughout my graduate career.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ 4 LIST OF TABLES ................................ ................................ ................................ ........... 7 LIST OF FIGUR ES ................................ ................................ ................................ ........ 8 ABSTRACT ................................ ................................ ................................ .................. 11 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW ................................ ..................... 13 Solid Lubricants ................................ ................................ ................................ ..... 14 Electrolessly Deposited Nickel Boride Coatings ................................ ..................... 16 Advanced MoS 2 Coatings ................................ ................................ ...................... 23 Cadmium Sulfide and Tribology of Ionic Solids ................................ ...................... 28 Summary Overview ................................ ................................ ............................... 30 2 INSTRUMENTATION AND EXPE RIMENTAL METHODS ................................ ..... 32 X ray Photoelectron Spectroscopy ................................ ................................ ........ 32 XPS Experimental Parameters ................................ ................................ ........ 36 XPS Data Processing Procedures ................................ ................................ ... 36 Pin on Disc Tribometry ................................ ................................ .......................... 41 Pin on Disc Testing Parameters ................................ ................................ ...... 44 Scanning White Light Interferometry ................................ ................................ ...... 45 3 TRIBOLOGICAL AND CHEMICAL EFFECTS OF VARYING ANNEALING TEMPERATURE IN ELECTROLESS NICKEL BORON COA TINGS ..................... 47 Introduction ................................ ................................ ................................ ............ 47 Experimental ................................ ................................ ................................ ......... 49 Results ................................ ................................ ................................ .................. 50 Compositional and Morphological Changes ................................ .................... 50 Friction Behavior ................................ ................................ ............................. 57 Wear Behavio r ................................ ................................ ................................ 59 Discussion ................................ ................................ ................................ ............. 61 Compositional and Morphological Changes ................................ .................... 61 Fr ictional Response ................................ ................................ ........................ 63 Wear Response ................................ ................................ .............................. 65 Summary of Findings ................................ ................................ ............................. 66 4 C OMPOSITIONAL EFECTS OF EXPOSURE TO LOW EARTH ORBIT ON MoS 2 /Sb 2 O 3 /Au COATINGS ................................ ................................ .................. 67
6 Overview ................................ ................................ ................................ ............... 67 Methods and Materials ................................ ................................ .......................... 69 Results ................................ ................................ ................................ .................. 70 Chemical State of Reference Sample ................................ .............................. 70 Chemical State of Flown Sa mple ................................ ................................ .... 74 Discussion ................................ ................................ ................................ ............. 76 Summary ................................ ................................ ................................ ............... 78 5 COMPOSITIONAL EFECTS OF EXPOS URE TO LOW EARTH ORBIT ON MoS 2 /YSZ/Au/DLC COATINGS ................................ ................................ ............. 80 Background Information ................................ ................................ ......................... 80 Methods and Materials ................................ ................................ .......................... 82 Results ................................ ................................ ................................ .................. 83 Chemical State of Reference Sample ................................ .............................. 84 Chemical State of Ram Sample ................................ ................................ ...... 90 Chemical State of Wake Sample ................................ ................................ ..... 93 Discussion ................................ ................................ ................................ ............. 95 Summary of Findings ................................ ................................ ............................. 99 6 TRIBOLOGY OF CADMIUM SULFIDE ................................ ................................ 103 Background Information ................................ ................................ ....................... 103 Me thods and Materials ................................ ................................ ........................ 103 Results ................................ ................................ ................................ ................ 106 Friction of Cadmium Sulfide ................................ ................................ .......... 106 Wear Behavior of Cadmium Sulfide ................................ ............................... 108 Discussion ................................ ................................ ................................ ........... 110 Summary ................................ ................................ ................................ ............. 113 7 CONCLUSIONS ................................ ................................ ................................ .. 114 LIST OF REFERENCES ................................ ................................ ............................ 124 BIOGRAPHICAL SKETCH ................................ ................................ ......................... 135
7 LIST OF TABLES Table pa ge 3 1 Elemental composition, in atomic percent, of coatings after various annealing temperatures. ................................ ................................ ................................ .... 56 3 2 Coefficient o f friction and wear rates of all samples tested. ............................... 61 4 1 Atomic percentages of elements in the near surface region of the flown and reference samples as determined by XPS. ................................ ........................ 71 5 1 Atomic percentages of elements in the near surface region of the reference, ram, and wake samples as determined by XPS. ................................ ............... 84 6 1 Calculated wear rates for CdS under varying loads. ................................ ........ 110
8 LIST OF FIGURES Figure page 1 1 Nickel Boron Phase Diagram, recreated from ASM Intl.  ............................. 18 1 2 Catalytic activity of various metals for diverse reducing agents . ................. 21 2 1 Diagram of the photoelectric effect. ................................ ................................ ... 33 2 2 Differences between the A) linear, B) Shirley, and C) Tougaard backgrounds. The x axis of the plots are binding energy in units of eV and the y axis are arbitrary intensity units. ................................ ................................ ..................... 38 2 3 Spectrum with background subtracted. The x axis of the plots are binding energy in units of eV and the y axis are arbitrary intensity units. ....................... 39 2 4 Final SolidWorks drawing of the novel in vacuo pin on disc tribometer courtesy of Dr. Gregory Dudder. ................................ ................................ ........ 42 3 1 SEM Images of the surface of the NiB coatings a) as received and after annealing at b) 250C, c) 400C, d) 550C, and e) 700C images courtesy of Dr. Wei Qiu and Dr. Juan C. Nino. ................................ ................................ ..... 51 3 2 Ni 2p 3/2 core spectra showing that after higher annealing temperatures no ................................ ......................... 52 3 3 B 1s core spectra showing a) the fitted as received sample with peak assignments and b) the spectra for the 250C, 400C, 550C, and 700C ........ 53 3 4 O 1s core spectra a) showing peak identification for the as received sample and b) remaining samples O 1s spectra ................................ ............................ 55 3 5 Raman spectrum of the electroless Ni3B sample following a 700C anneal in a blanketing oxygen gas. The spectrum is dominated by vibrational peaks attributable to boric ac id, courtesy of Dr. David Hahn. ................................ ....... 57 3 6 Friction coefficient vs. sliding distance for Ni B coatings a) as received and aft er annealing at b) 250C, c) 400C, d) 550C, and e) 700C. ........................ 58 3 7 SWLI image showing the wear track generated on the 550 C sample, courtesy of Dr. Rachel Colbert, and a line plot of SWLI data, averaged over the entire area of the image, indicating the depth of the wear track. .................. 60 3 8 SEM images of wear tracks of NiB coatings for the a) as received sample and after annealing at b) 250C, c) 400C, d) 550C, and e) 700C, images courtesy of Dr. Wei Qiu and Dr. Juan C. Nino. ................................ .................. 61
9 4 1 Schematic design of the MISSE 7 space tribometers from MISSE mission, courtesy of Dr. Gregory Sawyer and Dr. Brandon Krick. ................................ .... 69 4 2 Comparison of O 1s /Sb 3d core spectra for the A) reference sample and B) flown sample ................................ ................................ ................................ ..... 72 4 3 Comparison of Mo 3d core spectra of the A) reference sample an d B) flown sample ................................ ................................ ................................ .............. 73 4 4 S 2p core spectrum of the reference sample. ................................ .................... 73 4 5 Au 4f core spectra of the reference sample. ................................ ...................... 74 4 6 Si 2p core spectra from the flown sample. ................................ ......................... 75 5 1 Comparison of the Mo 3d core spectra of the A) reference and B) ram samples. ................................ ................................ ................................ ............ 85 5 2 Comparison of the Zr 3d spectra for the A) reference and B) ram samples. ...... 86 5 3 Comparison of the S 2p spectra for the A) reference sample and B) ram sample. ................................ ................................ ................................ ............. 87 5 4 Comparison of the Au 4f spectra for the A) reference, B) ram, and C) wake samples. ................................ ................................ ................................ ............ 87 5 5 Comparison of the O 1s spectra for the A) reference, B) ram, and C) wake samples. ................................ ................................ ................................ ............ 88 5 6 Comparison of the C 1s spectra for the A) reference, B) ram, and C) wake samples. ................................ ................................ ................................ ............ 89 5 7 The Si 2 p spectra for the A) ram sample and B) wake sample. No silicon was detected in the surface region of the reference sample. ................................ .... 93 5 8 The F 1s spectrum for the wake sample. ................................ ........................... 95 6 1 Graph of coefficient of friction vs. cycle number for CdS unde r a load of 0.08 N. ................................ ................................ ................................ .................... 107 6 2 Graph of coefficient of friction vs. cycle number for CdS under a load of 0.125 N. ................................ ................................ ................................ .................... 107 6 3 Graph of coefficient of friction vs. cycle number for CdS under a load of 0.25 N. ................................ ................................ ................................ .................... 108 6 4 line profile of the scan (with units of nm on the y axis) courtesy of Alexander Rudy. ................................ ................................ ........... 108
10 6 5 averaged line pro file of the scan (with units of nm on the y the x axis), courtesy of Alexander Rudy. ................................ ......................... 109 6 6 can of a part of the 0.25 N test wear track and the averaged line profile of the scan (with units of nm on the y axis), courtesy of Alexander Rudy. ................................ ................................ ........... 109
11 Abstract of Dissertation Presented to the Graduate School of the Univer sity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FUNDAMENTAL INTERFACIAL STUDIES OF ADVANCED SOLID LUBRICANTS AND THEIR OPERATING ENVIRONMENTS By Kevin Gill ey December 2013 Chair: Scott Pe rry Major: Materials Science and Engineering Solid lubricants are a class of materials that are utilized in applications and environments where traditional lubrication schemes cannot be implemented. A variety of materials display solid lubrication, and i n this study a number of solid lubricants were investigated. Firstly, electrolessly deposited nickel boride alloys were annealed at different temperatures under a flow of oxygen. The surface chemistry, friction, and wear behavior of the coating were then i nvestigated. It was found that when annealed above 550C the coatings had a dramatic change in surface chemistry, where the Ni 3 B had formed a thick layer of B 2 O 3 on the surface. This oxide then reacted at ambient temperatures with moist air to form the lub ricious compound H 3 BO 3 This led to a coefficient of friction below 0.1 and a slight increase of the wear rate from 10 8 mm 3 /Nm to 10 7 mm 3 /Nm. Secondly, the surface chemistry of advanced MoS 2 based coatings that had been exposed to low earth orbit was i nvestigated. It was found that this exposure produced the complete oxidation of the coatings Also exposure to the unique space environments resulted in the deposition of large amounts of contaminant SiO 2 on the surface. Lastly the tribological properties of single crystal cadmium sulfide were
12 investigated. There is nearly no knowledge of the tribological activity of cadmium sulfide in the literature, so the study was performed as an initial investigation into the material. It was discovered that cadmium s ulfide did not show low friction, with a coefficient of friction of approximately 0.25, but did show low wear, with a wear rate of approximately 3x10 7 mm 3 /Nm.
13 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW The most traditional and common approach to lubr ication is the use of a liquid phase material to physically lubricate two separate solid surfaces. However, there are occasions when traditional lubrication schemes cannot be used. In these occasions alternative lubrication methods must be employed. One su ch alternative is the use of solid lubricants, which are solid phase materials that provide tribological benefits to a given system through the provision of a low shear interface not inherently present with the contact of component surfaces. This document focuses on the testing and characterization of a variety of solid lubricant systems. The systems studied include NiB thin films, advanced MoS 2 based coatings, and CdS, a sulfur containing ionic solid. The application of these systems varies, ranging from l ubrication in terrestrial environments to providing lubrication in outer space. Despite their differences in application, solid lubricants typically feature layered structures where low shear forces predominantly arise from the sliding of layers over each other due to low energy bonding or interactions between layers. This class of solid lubricants is widely used in the aerospace industry. The vacuum environment of low earth orbit presents a unique set of problems that prohibit the use of liquid lubricants. Firstly, large temperature variations can occur where depending on the orientation of the sun, components can experience temperatures f rom 200C to 200C. Also, the vacuum environment, where pressures can be as low as 1x10 9 torr, would cause many liquid lubricants to volatilize from the surface of the component  An additional difficulty that comes with the use of sol id lubricants for aerospace applications is that the unique environment experienced in space is very difficult to recreate on Earth. As a result, the materials used must also be
14 able to perform in terrestrial environments or at least resist oxidation or de gradation before use. Solid Lubricants The need for alternative lubrication schemes arose largely from the explosion of engineering and scientific research experienced during WWII and the decade immediately following. As the technology became more advance d, the need for more advanced materials grew. Specifically, the meteoric increase in aerial and aerospace technology necessitated the need for lubricants that could be used at high temperatures and, once the Space Race began, in vacuum environments. As a r esult of this need, many solid systems were evaluated as lubricants. During this time a group of materials evolved to become defined as solid lubricants. Materials in this group are typically soft metals, layered crystalline solids, polymers, or carbon bas ed coatings. Soft metals were among the earliest of materials considered for solid lubrication. Most soft metals exhibit a face centered cubic (FCC) structure, possessing a high number of slip systems  and energetically favoring of slip. The low shear strength of FCC metals provides effective lubrication, however their functionality i s limited by another important part of tribology. Most soft metals have relatively high wear rates, and this limits their applications and functionality  However at higher temperatures, soft metals have been seen to provide tribological benefi ts and are often used under conditions where other solid materials suffer from degradation. Some examples of soft metals that are used as solid lubricants include gold, silver, and lead. Materials with layered crystal structures represent a second class o f solid lubricants. The source of the lubricity of these materials lies in the structural details  These materials show the ability to peel and delaminate easily, displaying the low shear
15 strength between layers which leads to their lubricity. However there is hi gh anisotropy in the properties of the crystal as shearing transverse to the layers is more difficult than shearing along the layers. This anisotropy leads to layered materials such as graphite and molybdenum disulfide (MoS 2 ) to have very high melting poin ts. There are a number of naturally occurring minerals that exhibit layered structures that have been used a lubricants for many years such as talc, mica, graphite, and molybdenite (the naturally occurring form of MoS 2 )  In addition to of these naturally occurring materials, synthetic layered materials such as tungsten diselenide (WSe 2 ) and hexagonal boron nitride (BN) have been engineered to provide solid lubrication [6, 7] Another group of materials that has been successfully emp loyed as solid lubr icants is polymers. Some examples are polytetrafluoroethylene (PTFE), polyetheretherketon e (PEEK), and polyimides (PI). These polymers have both benefits and detriments when compared with other solid lubricants. On the positive side, the y are relatively inert in vacuum; but, being polymers, they are not applicable when high temperatures are involved [1, 8] Also, as most polymers they degrade when exposed to UV light, they are difficult to use in aerospace applications  The final category of materials that have shown to be useful as solid lubricants is comprised of carbon based materials and coatings. Carbon based lubrication is a familiar concept to most as graphite has been used in writing utensils for some time and its natural lubricity is evident. However in recent years a new synthetic form of carbon coating has been made specifically for its solid lubricity. Diamond like carbon (DLC)coatings differ structurally from graphite in the local hybridization of carbon
16 bonding and the inclusion of hydrogen  These DLC films have a low amount of long range ord er and are amorphous in nature. Despite the large number and diverse nature of the classes of materials that exhibit solid lubrication there are only a small number of them that have been extensively studied and are used practically. The development of applications for solid lubricants has been done largely empirically where many materials with the potential for lubrication would be tested. When ma terials with interesting tribological properties would be found, it would be used in applications and studied more extensively to find the best suited environments for it. However the fundamental knowledge of how and why solid lubricants work is an area th at has only minimal knowledge for many materials. Factors such as the influence of environment, contact pressure, effects of annealing etc. can strongly affect the tribological properties of a material, yet for many solid lubricants the effects of these fa ctors is unknown. The aim of this dissertation work is to investigate the fundamentals of solid lubricants in order to provide guiding paradigms for their future implementation in advanced technologies. Electrolessly Deposited Nickel Boride Coatings Nicke l boride coatings can be achieved through a number of processes; one of the most promising is electroless deposition, an aqueous chemical process similar to electrolysis but involving none of the negatives. For example, electroless coating is capable of pr oviding a coating of constant thickness with buildup on edges and can coat complex shapes. Also electroless deposition is possible on a wide range of substrates including non metallic substrates. Electroless deposition was accidently discovered in the 19 th century, but disregarded at the time as a simple curiosity due to the fact that the metal was not a coating but rather formed as powder in the vessel  In 1946, the
17 process was rediscovered when trying to develop a process for electroplating steel tubes with nickel alloys [10, 11] It was observed that when hypophosphate was added to the plating bat h that the inner and outer faces of the tubing were coated with nickel. Transportation Corp and put into production in 1955  These first deposition baths were based on solutions of nickel salts with sodium hypophosphate added a s a reducing agent. In 1954, a bath was proposed using borohydride as a reducing agent and in 1957 the technology was furthered to allow for the deposition of nickel boride coatings  .The deposition of other alloys became possible when other bath compositions were discovered to follow the same deposition process, producing coatings such as pure nickel, nickel phosphide, and the previously mentioned nickel boride alloys  The compositions of nickel boride coatings are specifically tailored by the conditions in the plating bath to deposit a certain phase. There a number of phases that can be deposited and some are desirab le, while others are undesirable for tribological activity. To determine what phases will deposit at certain compositions the Ni B phase diagram must be consulted. Phase diagrams are thermodynamic chart which describes the conditions under which phases of a material will exist at equilibrium. Phase diagrams can be plotted for systems with one or more components. For singular component systems, temperature versus pressure is normally plotted, although any state functions can be plotted against another. For b inary systems, involving components A and B, temperature is often plotted versus weight or atomic percenta ge of either component A or B. All phase diagrams entail phase boundaries, which are
18 lines separating separate phases of the system. The intersection of these phase boundaries are called triple points. Phase boundaries separating a phase from the liquid phase are called liquidus lines while phase boundaries separating a mixture of a liquid and a phase from a completely solid phase are called solidus lin es. Compounds that are formed at only one specific concentration are called line compounds and appear as vertical lines in the phase diagram. Specifically relating to the Ni B phase diagram, the target composition is located at 6 weight % at which is the l ine compound Ni 3 B, which empirically has been shown to exhibit the best mixture of desired properties. Solid solubility of both boron in nickel and nickel in boron is very low, however there are a number of line compounds such as Ni 3 B, Ni 2 B, Ni 4 B 3 and NiB  The nickel boron phase diagram, as seen in Figure 1 1, includes these line compounds and features four eutectics as well as a peritectic. These fea tures highlight the complexity of nickel boron phase behavior and the range of, phases that occur at differing compositions. Figure 1 1. Nickel Boron Phase Diagram recreated from ASM Intl. 
19 Since electroless deposition is less understood than traditional deposition techniques, it is important to review how the kinetics and chemistry of the process work. This process is emerging as a desirable way to create new solid lubricants as it is a cheaper an d easier way to make coatings. Electroless deposition is a type of aqueous deposition process in which metal salts are reduced via the oxidation of a chemical agent to produce a resulting solid coating o n a catalytic surface  In its most basic form the proces s can be expressed as a combination of two half reactions given in E quations 1 1 and 1 2  M e z + + ze e (1 1) z + + ze (1 2) In electroless deposition there are two conditions that must be fulfilled for the deposition to occur  The first, called the thermodynamic condition, is that the redox potentials of both the metal and the reducing a gent must be selected so that deposition will be thermodynamically spontaneous, as described in detail below. The second, called the kinetic or catalytic condition, is that the reaction must be controlled in such a way that deposition only occurs on the su bstrate and not in the solution or on the cell walls. The thermodynamic condition can be obtained by choosing a chemistry in which the redox potential of the nickel partial half reaction is greater than the redox potential of the reducing agent reaction. B oth of these redox potentials are calculated from the Nernst e quations, as seen in E quations 1 3 and 1 4 [13, 14] (1 3) (1 4)
20 Once the equilibrium redox po tential of each reaction is known an analytical expression of the thermodynamic condition can be written. Since the reaction must be must be positive. The expression as seen in E quations 1 5 and 1 6 [13, 14] (1 5) (1 6) From E quations 1 5 and 1 6, it can be seen that the equilibrium potentials of both half reactions are positive and the equilibrium potential of the nickel partial reaction must b e more positive than the equilibrium potential of the reducing agent half reaction in order for the thermodynamic condition to be fulfilled. To satisfy the kinetic or catalytic condition, where preferential deposition occurs on the substrate rather than at cell walls or in solution, both the substrate and the deposited metal must serve as catalysts for the oxidation of the reducing agent. To assess whether a given substrate and deposited metal will satisfy this condition, a chart as shown in Figure 1 2 is h elpful. The chart features various reducing agents on the left plotted versus potential on the bottom. For the condition to be fulfilled, the potential of both the substrate and deposited metal must be to the left of the equilibrium potential  In the case of Figure 1 2 nickel is the deposited metal and would work in all reduci ng agents except formaldehyde. Under these conditions, the reaction will only occur on the substrate and then continue to occur o nce the substrate is completely coated with the deposited metal.
21 Figure 1 2 Catalytic activity of various metals for diverse reducing agents  There are a number of components that are included int o the bath used for electroless deposition. The first is a metal salt which provides a source for the metallic ions that when reduced form the coating. A second component is a reducing agent that provides the source of electrons when oxidized, in turn redu cing the metallic ions in solution. Another very crucial component is a complexing agent which has multiple purposes. The complexing agent regulates the concentration of free metallic ions in solution as well as increasing their solubility. The complexing agent also is used to prevent the precipitation of insoluble hydroxides  This is a very important role because the formation of hydroxide precipitates can serve as a potential place for deposition to occur in the solution. Another purpose of the complexing agent is to act as a buffer for the solution bath. A fourth component of the bath is a stabilizer. The stabilizer is used to control the kinetics of the spontaneous deposition reaction by partiall y blocking catalytically reactive sites on the substrate. However, there are some negatives side effects caused by the stabilizer including increased internal stress and
22 porosity as well as decreased corrosion resistance and plating rate  Therefore only minute amounts of stabilizer are added to the bath. The most commonly used stabilizers are sulfur based organic molecules or heavy metal salts. The final components added to the bath are pH regulator s and buffering agents which serve to maintain strict pH conditions in the bath The strict pH requirements arise due to the fact that the most commonly used reducing agents hydrolyze under slightly alkaline, neutral, or acidic conditions  Therefore the pH must be kept above 12 in order to prevent the hydroxyla tion of the reducing agent. As previously mentioned, there are a number of generic components that are involved in electroless deposition. The specific components and conditions of an actual nickel boride bath are as follows. The metal salt that is used is typically nickel chloride hexahydrate (NiCl26H2O) with a concentration of 24g/L  The most commonly used reducing agent is sodium borohydride (NaBH 4 ), with a typical concentration of 0.482g/L  The reducing agent is the main component that must be changed to increase the amount of boron in the film, as it is the exclusive source of boron that will be deposited in the film. There are a number of complexing agents that can be used; one that is commonly used in Ni B deposition is ethylene diamine ( NH 2 CH 2 CH 2 NH 2 ) which is highly concentrated at 59g/L  The typical stabilizer used in the deposition is lead tungstate (PbWO 4 ), which for the reasons previously detailed is kept at a very low concentration, typically 0.021g/L  The final component is t he pH regulator; sodium hydroxide (NaOH) is most typically used, at a concentration of 39g/L  The specific reactions involved in the deposition of nickel boride are given in E quations 1 7, 1 8, and 1 9 
23 (1 7) ( 1 8) (1 9) Many different substrates can be used in electroless deposition of nickel boride films. The only condition required for successful deposition is that the substrate must serve as a catalyst for the N i deposition reaction. Commonly used substrates include most steels, aluminum based alloys, and Inconel alloys. Substrates which cannot initially serve as a suitable catalyst for the deposition can be surface pretreated to allow for the deposition. Also if the substrate is placed into contact with a less noble metal the deposition can occur successfully. Advanced MoS 2 Coatings The earliest reference to the use of MoS 2 as a low friction solid lubricant can be found in a proceeding of the American Physical S ociety from 1941 by Bell and Findlay in which molybdenite was used a lubricant for a rotating anode x ray tube in vacuum  2 as a source of solid lubricity would be performed in which the effects of contact pressure and sliding speed on the coefficient of friction for a MoS 2 coating on steel in air was measured  The number of studies on the MoS 2 rose steadily as its popularity as a solid lubricant played a role in the frictional behavior of MoS 2 and other solids with layered structures In this study it was found that MoS 2 would have a steady state friction coefficient of approximately 0.07 between 40C and 300C. Upon reaching 300C the friction coefficient would rise to a value of approximately 0.5 at 530C. The reason for t his rise
24 in friction after 300C was attributed to oxidation of the MoS 2 to MoO 3 in air at elevated temperatures  Further fundamental studies on MoS 2 found that relative humidity played a large role in the frictional response of the material and that the oxidation of MoS 2 only happened until a monolayer of MoO 3 was formed upon which oxid ation would stop occurring. It was also shown that when exposed to humid air and temperatures as low as 85 100C, MoS 2 would spontaneously form a layer of MoO 3 but as previously stated the layer would only be a monolayer [20, 21] Following the discovery and initial studies of the solid lubricity of MoS 2 the materials became a popular choice for aerospace applications due to its low friction and ability to withstand elevated te mperatures. At that point in time most MoS 2 coatings were manually deposited by burnishing, painting or evaporation from liquid onto the surface. To improve the performance of the coatings, new deposition methods were studied and developed. The first coati ngs deposited by sputter deposition produced increased stability in the friction response as well as an increased wear lifetime on a number of substrates, when tested in vacuum  During this same time studies began investigating the mechanism by which MoS 2 derived its lubricity. Holinski and Gansheimer indicated that there were only weak interactions between the layers in the layered structure of MoS 2 which allowed for the layers to shear easily and slide over underlying layers, thus producing the low friction coefficient. Using scanning electron microscopy (SEM), it was discovered that the layers slid similar to the illustration of a deck of cards where multiple layers moved at once. In the study it was seen that approximately 25 ato mic layers would slide during operation of the test 
25 While shown to operate very well in vacuum, one large problem with MoS 2 coatings was that they operated very poorly in other environments, particularly humid air. New deposition techniques that became available such as RF diode sputtering, DC diode sputtering, and RF magnetron sputtering allowed for the deposition of multiple componen ts in a coating [24, 25] This in turn allowed for the ability to add constituents to MoS 2 coatings to not only increase its base performance and lifetime but also add the ability of the coating to perform in different environments The first study involving co deposited MoS 2 aimed to improve the performance of the sputter d eposited coatings by depositing metals in conjunction with MoS 2 These coatings typically contained nonequiaxed grains which cause high surface area and decreased density in the coating causing increased oxidation and inconsistency in performance. To remed y this problem, metals were co deposited in an attempt to increase the density of the films; however results showed that the overall performance of the coating increased concurrent with the increase in density. The co deposited films exhibited increased li fetimes, lower friction, and uniform friction performance  Different categories of additives have been found to play different roles in the coatings, influencing the tribological effect on the system in a variety of ways. Soft metals such as Au and Pb, already known as solid lubricants, increase the density of the coating, provide increased tribological function in ambient environments and provide lubrication at higher temperatures [24, 26 34] Au and Pb in particular have become widespread in use as constituents in advanced MoS 2 coatings. P b acts not only as a lubricant at high temperatures where MoS 2 suffers, but also acts as an oxidation site, in turn reducing the extent of MoS 2 oxidization and improving tribological function. The
26 ideal concentration of these types of constituents has been determined to be between 5 and 15 at% [31, 3 2, 35] Elements that can form intermetallic compounds with MoS 2 are another category of additives for the coatings. Elements such as nickel, zirconium, chromium, and titanium have been found to increase the hardness of the coating [36 43] This produces an increase in the wear resistance, often displaying wear rates an order of magnitude lower than base MoS 2 coatings. Overloading of intermetallic constituents can produce very brittle films a nd the optimal concentration of this category is also between 5 and 15 at% [44 ] A third constituent category includes elements that have the similar crystal structure s to MoS 2 such as tungsten tantalum, niobium, selenium, and tellurium [45 47] These atoms can substitute into the MoS 2 structure and support the coating by forming complex oxides and sulfides with MoS 2 which prevents the formation of MoO 3 thereby allowing the coating to perform at higher temperatures. However this does not completely eliminate the formation of MoO 3 ; therefore these coatings still display poor performance in high humidity environments and at elevated temperatures. A common problem with all of the previously mentioned constituents is that despite being able to marginally increase the temperature resistance and the ability to perform at high temperatures, the coating suffer performance wise around 450C due to oxidation of th e MoS 2 To address this limitation, the addition of metal oxides such as Sb 2 O 3 MoO 3 and PbO to MoS 2 coatings has proven to be beneficial to the performance of the coating [35, 37, 43, 48 51] When introduced into the system, the oxides have been found to provide oxidation resistance, increased hardness, densification, and inhibition of crack growth. Other mechanically hard oxides,
27 such as Al 2 O 3 and yttria stabilized zirconia (YSZ), have been investi gated and found to also reduce wear at higher temperatures [52 55] Another strategy to increase the range of environments in which MoS 2 coatings can operate is to co deposit a second solid lubricant that can perform well in conditions where MoS 2 is deficien t. The oldest of such mixtures combines MoS 2 with graphite [4, 56] Graphite has a frictional behavior opposite from MoS 2 with respect to environment; the friction coefficient is low in ambient conditions and high in vacuum. By combining graphite and MoS 2 mixed coatings can perform in both ambient and vacuum conditions with increased lifetimes. At elevated temperatures, the graphite acts as an oxidation site so that the MoS 2 will remain un oxidized  However in highly aggressive oxidizing environments the oxidized graphite produces wear debris detrimental to the system, thus limiting the uses of such co deposition [57, 58] While co deposited coatings perform well, specific deficiencies arising from the addition of the const ituents continued to be observed. Therefore tertiary and even more complex MoS 2 based coatings began to be deposited and tested in order to try to find a composition that could perform well in all environments and situations. Zabinkski et al. performed a s tudy in which MoS 2 coatings were co deposited with a wide variety of materials to determine which improved the tribological properties the most  From this study it was conclu ded that decreasing the crystalline grain size of the coating lead to a decrease in the friction coefficient and an improvement of the wear resistance A tertiary composite of MoS 2 /Sb 2 O 3 /Au displayed the smallest grain size  This tertia ry coating had an increased hardness due to the presence of Sb 2 O 3 producing an increased wear lifetime. This also served to limit MoS 2 oxidation. At higher
28 temperatures, Au would contribute tribologically as well as densifying the coating (at all temperat ures). The combination of these constituents lead to a vastly superior coating and this quickly became one of the more popular tertiary coatings used for aerospace applications. Although this particular tertiary coating performed better than base MoS 2 coat ings at elevated temperatures, there was still a need for the coatings to perform at temperatures higher than 450C. Studies by the Air Force Research Labs ( AFRL ) aimed at optimizing performance at increased temperatures yielded a quaternary composite of Y SZ/Au/MoS 2 theoretical ability to perform tribologically in all environments needed for LEO. The graphitic carbon, and a value if 0.02 in a dry nitrogen environment due to MoS 2 in the coating. When increasing the temperature of the testing environment to 500C, the friction coefficient increased only to a value of 0.15, drawing of the Au to the surfa ce of the coating as a source of lubrication  Cadmium Sulfide and Tribology of Ionic Solids metallic cadmium in 1817 as a contaminant in zinc compounds  It would not be until 1820 when CdS would be first produced. The reason behind this bein g that the metal was only found in small quantities as a contaminant and therefore was expensive and hard to obtain in large quantities. CdS was synthesized by heating an acid solution of a cadmium salt in the presence of hydrogen sulfide gas, which would produce powdered CdS. The vivid yellow to orange hue of the CdS powder led to its use as a pigment which came to be known as Cadmium Yellow in dyes and paints; this remains the largest application for CdS.
29 The main use of CdS as a pigment continued until 1925 when Huggins did a theoretical study into the distribution of valence electrons in CdS and other materials having the same crystal structure as diamond  The band gap of CdS was eventually calculated to be 2.42eV. In 1947, single crystal Cd S was first deposited by reacting cadmium vapor with hydrogen sulfide gas by Frerichs and the electrical and optical properties of CdS were studied  Frerichs described CdS and the other materials studied (CdTe and CdSe) incomplete phosphors due to the fact t hat the materials displayed high photo conductivity but showed no phosphorescence Since this discovery the optical and electrical properties of both single crystal and thin films of CdS have been studied extensively [62 74] With the advent of nanotechnology the focus of research on CdS ha s somewhat shifted to the study of its use at the nanoscale. Studies recently have been done on nanoparticles, nanorods, nanowires, nanowhiskers, and nanocrytals of CdS and CdS containing advanced materials as nanoelectronics. Other applications for CdS in clude photoresistors and thin film transistors. The motivation of studying CdS tribologically stems from a number of factors. The first factor relates to similar metal chalcogenides such as MoS 2 iron sulfides, etc. being known solid lubricants. Iron sulf ide has the same crystal structure as CdS and is known to show low friction. Other materials that possess layered structures with alternating sulfur/metal layers are known to show low friction as well. Therefore, investigations of other materials possessin g similar features are warranted by the desire to discover new solid lubricants and interesting tribological behavior. Cadmium sulfide is of particular interest due to its mixed bonding character. Simple calculations into the nature of the bond in CdS show that the bond has only 18% ionic character with the
30 remaining 82% being covalent in nature Another factor for studying the tribological nature of CdS is the dearth of knowledge available in the literature. Searches into the literature of tribology of CdS yielded no results and even in general for ionic solids there is little to no knowledge available. A study performed in 2012 calculated the wear rates of various ionic solids including NaCl, KBr, KCl, BaF, MgF, CaF, MgO, ZnS, and FeS 2 by using scanning w hite light interferometry (SWLI) to do in situ measurements during tribological testing and highlighted the potential of CdS as a tribological material. The study found that as the activation energy of surface defects increased the wear rate of the materia l improved. Of particular interest in this study was the testing done on ZnS. ZnS displayed a wear rate of 1.0x10 7 mm 3 /(Nm) in a test of 38,000 cycles  The similarities in composition and bonding character between ZnS and CdS underscore the motivation behind the Summary Overview The balance of the dissertation addresse s the details of fundamental studies of solid lubricant materials. Chapter 2 discusses the instrumentation used to study three categories of solid lubricants. Chapter 3 presents a study of the effects of annealing electroless nickel boride coatings at a ra nge of temperatures on their tribological properties. The chemical nature of the coating, as well as its tribological properties is discussed in C hapter 3 Chapters 4 and 5 comprise a study of two different advanced MoS 2 coatings that were employed in LEO on the International Space Station. In C hapters 4 and 5 the surface chemistry of the actual coatings that were exposed to LEO is compared to reference samples that were kept in terrestrial environments, providing insight into the contribution of chemical c hange to tribological operation in
31 space environments Finally C hapter 6 discusses the tribological properties of CdS; a material of which the tribological properties are unknown, however based on similar materials may display interesting behavior.
32 CHAPT ER 2 INSTRUMENTATION AND EXPERIMENTAL METHODS X ray Photoelectron Spectroscopy To identify and quantify the elemental and chemical species present in the samples, X ray photoelectron spectroscopy (XPS) was used. For these tests an Omicron Nanotechnology Gh mB Al K (1486.7 eV) source was used with an Omicron Nanotechnology GhmB EAC2000 Sphere hemispherical, 7 channel analyzer. The source features a single Al anode which when excited with electrons emitted from an electrically excited filament, emits non mon ochromatic X rays. These X rays are then focused and diffracted off the <1010> face of a thin crystalline quartz disk. This produces an X ray beam with a spectral energy of 1486.7 eV focused on the sample. The sample is attached to an Omicron GhmB platen v ia Ta strips that are spot welded to the platen. The sample and platen are grounded via attachment to the UHV manipulator arm. The manipulator can be translated in 3 planes and can rotate 360 around the axis of the shaft of the arm. A standard takeoff ang le of 55, relative to the surface normal, was used for the current tests but differing the takeoff angle via angular rotation of the manipulator leads to alteration of the sampling depth in the sample. Lower takeoff angles lead to less surface sensitivity and conversely higher takeoff angles lead to greater surface sensitivity. Once the monochromatic X ray flux strikes the sample, X rays are absorbed by atoms in the sample. This absorption leads to the emission of core level electrons from the atoms, these electrons are known as photoelectrons. These photoelectrons have characteristic energies associated with them, which are influenced by the element they come from as well as the local bonding
33 environment of the element they originate from. This process is known as the photoelectric effect and is diagramed in Figure 2 1 Figure 2 1 Diagram of the photoelectric effect. W hile the X rays penetrate through the entirety of the sample and produce photoelectrons throughout the bulk of the sample; only photoelectrons generated in the top 1 10nm of the sample will be collected with the characteristic binding energy. This is because past this top 1 10nm the photoelectrons will undergo elastic collisions and lose energy thusly losing the c haracteristic energy information. The photoelectrons are collected by the Sphera hemispherical analyzer using a system of electrostatic lenses which can be used to focus collection on microscopic areas of the sample. Once photoelectrons enter the column o f the analyzer they are refocused and filtered utilizing the double hemispherical design of the analyzer. By applying a potential across the inner and outer walls of the analyzer only electrons with a specific energy are allowed to pass through the analyze r to the detector. Electrons
34 with higher and lower energies will not reach the detector. By varying the potential applied across the inner and outer walls and dwelling for a specified time at each voltage, a spectrum can be collected for a range of energie s. For the EAC 2000 Sphera system, photoelectrons that pass through the analyzer are detected and converted into an optical pulse that travels through fiber optic cables from the preamplifier to an amplifier. Here the signal is converted back to an electri cal signal, amplified and then recorded in terms of a plot of counts versus binding energy is generated. The binding energy (BE) is calculated through Equation 2 1 using the kinetic energy (KE) of the electron (which is measured), the energy of the inciden t X ray (hv), and the work (2 1) On the plots of counts versus binding energy, characteristic peaks will be seen at binding energies corresponding to the characterist ic energies of the photoelectrons emitted by the atoms in the sample. XPS is one of the preferred methods of elemental and chemical analysis of surfaces because it can not only identify the elements that are present in the near surface region but it can al so give information on the bonding state of the elements. With the appropriate settings, XPS is sensitive enough to distinguish shifts in binding energy based upon the local bonding environs of the elements. This leads to being able to distinguish between chemical compounds and to detect reactions such as oxidation. This can become useful in trying to establish levels of contamination in samples, as unless a sample is sputter cleaned there will always be some level of C and O contamination from the ambient environment. By distinguishing between compounds, the amount of O or C that is actually in the sample as opposed to that
35 which is contaminants can be determined in some instances. The relative atomic percentage ( at % ) of each element can be calculated as w ell, by integrating the area underneath the BE peaks found within each spectra. In this way the identity of the elements/compounds present is known in addition to the relative amounts of each element/compound. The analyzer was calibrated using a sputter cl eaned, 99.99% pure silver sample by setting the measure silver 3d 5/2 peak binding energy to 368.3 eV. However in working with solid lubricants a consideration that must be taken is the electrical conductance of the sample. Charging of the sample surface ca n be a significant problem. Insulators or poor conductors can accumulate positive charge on the surface due to the fact that as photoelectrons leave they cannot be replenished from a grounding source; this leads to the negatively charged photoelectrons bei ng electrostatically attracted to surface. Thus artificially increases the binding energy measured due to them being harder to remove from the surface. To combat this problem charge neutralization must be utilized. Initially the surface charge on insulatin g or poorly conducting samples is positive. Therefore to neutralize the sample, a flux of low energy electrons (1.5 3 eV) is impinged on the surface. This flux of electrons will neutralize the surface charge on the sample; however, over time the applicatio n of a flux of electrons will induce a negative surface charge. This will cause the photoelectrons to leave the surface with higher kinetic energies, which artificially lowers the binding energies seen in XPS. Therefore, another charged particle must be in troduced to neutralize the now negatively charged surface. To achieve this, a noble gas ion source is utilized, seeking to avoid chemical reactions with the surface of the sample. In the
36 current studies Ar + ions are used at low energies (20 100 eV) to succ essfully neutralize the sample surface without damaging the sample. Using charge neutralization allows for the study of insulators and poor conductors with XPS, but does introduce some negative effects as well. The introduction of both electrons and positi ve ions to the near surface region of the sample will both retard and accelerate the kinetic energy of photoelectrons that interact with the introduced particles. This leads in larger FWHM values for the peaks seen in the XPS spectra. The system used in th is study is fitted with a 15eV CN10 electron gun as well as a 5k eV PHI FIG 5CE ion sputter gun to effectively neutralize poor conductors and insulators XPS Experimental Parameters For the present work, a standard set of parameters was used for all XPS spectra taken, maintaining as many variables and settings constant as possible. For broad survey spectra, a step size of 1 eV was used with the pass energy equal to 50 eV. Only 1 sweep was performed as these large scans are only investigatory and are used to determine the elements present, for which core scans will be performed. The dwell time was set at 0.2 seconds and the energy window for the scan was from binding energies of 1386.7 to 0 eV. For core spectra of individual elements a step size of 0.03 eV was used with the pass energy equal to 20 eV. For these core level scans 5 sweeps were performed over each energy range to improve signal to noise ratios. The dwell time for these spectra was also set at 0.2 seconds; and the energy range differed depending on what core level was being scanned. XPS Data Processing Procedures XPS data is processed using a program capable of processing XPS, Auger electron spectroscopy (AES), and secondary ion mass spectrometry (SIMS) data. The
37 CasaXPS program which was writte n by Neal Fairley and is distributed in the U.S. by RBD Instruments of Bend, OR. Data processing begins with creating a region in the energy window for the particular spectrum. This region should include the entire peak that is located in the particular en ergy window as well as enough space on each side of the peak to establish an appropriate background. Background intensity derives from electrons that when escaping from the sample undergo inelastic collisions with other atoms and lose kinetic energy There are many forms of backgrounds that can be chosen when creating a region. But there are three main backgrounds which are commonly selected: a linear background, the Shirley background, or the Tougaard background. The linear background, seen in F igure 2 2 A is simply a linear line connecting the end points of the region. This method suffers when there is a decrease in the background signal after a peak, which can lead to cutting off intensity from the peak. The Shirley background is scaled in proportion to t he total intensity below its binding energy position; as a result, it is more sensitive to changes in background intensity as seen in F igure 2 2 B. The Tougaard background, seen in F igure 2 2 C, integrates the intensity of the background at a specific bindin g energy from the spectral intensities at higher kinetic energies, i.e., the approach predicts the probability of an electron undergoing a loss event and therefore contributing to the background at a given energy. While conceptually different, the Shirley and Tougaard backgrounds very similar results.
38 Figure 2 2 Differences between the A) linear, B) Shirley, and C) Tougaard backgrounds. The x axis of the plots are binding energy in units of eV and the y axis are arbitrary inten sity units. Due to the ease of calculation, the Shirley background has become widely used and was used exclusively in the current study. After setting the region and the type of background to be used, the background is subtracted from the spectrum, provid ing the either side of the peak varied as seen in F igure 2 3 After background subtraction peak fitting of the spectra can begin. To do this component peaks or fitt ing peaks must be created to match the collected data. These fitting peaks must account for all of the remaining intensity after background subtraction. There are a number of guidelines to follow when curve fitting. The first guideline that must be followe d is that the number of fitted peaks for a spectrum should not exceed the amount of expected chemical states of the element present. For instance on a spectrum for a piece of silicon wafer, there would only be two expected bonding states
39 nascent Si, and Si O 2 therefore the spectra should only have two fitted peaks as it would be non physical to have a third peak as there would be no chemical state to account for the third peak. While simple in idea, this can be a difficult step in practice; often the sample s being examined may have a number of chemical states that could be possible but how many are present is unknown. Figure 2 3 Spectrum with background subtracted The x axis of the plots are binding energy in units of eV and the y axis are arbitrary intensity units. As with the background shape, there are numerous shapes that can be chosen for the fitted peaks. Overwhelmingly, the most used and therefore the only one discussed in detail is the Gaussian/Lorentzian curve. This lin e shape involves a Gaussian curve mixed with a Lorentzian curve; the Gaussian part is contributed from the spectrometer while the natural core level line shape is Lorentzian therefore a blended line must be used to fit the peaks. For this study a 30% Loren tzian/ 70% Gaussian curves were used for all spectra. In such a case as outlined above a second guideline comes into play: when the exact amount of states is unknown, seek to minimize the number of fitted peaks needed to account for the intensity while a dhering to the remaining guidelines. The third
40 guideline suggests that all fitted peaks in a spectrum must have the same FWHM value. The FWHM value is different for every orbital in every element, although for a single element, the values may only differ s lightly across the different orbitals. However within a single spectrum the fitted peaks should all have the same FWHM value. There are certain instances where in some elements that contain doublets the FWHM values will be different for the each doublet su ch as in the Ni 2p spectrum, but these situations are exceptions to the rule. The next guideline relates to the energy separation between doublets. Doublets, defined as the appearance of two peaks for the same orbital, occur for all orbitals higher in en ergy than the s level orbital (i.e p d, f orbitals). This effect occurs because of spin orbit splitting, the process by which any electron in an orbital with orbital angular momentum experiences, coupling occurs between the magnetic fields of spin (s) a nd the angular momentum (l). This leads to the total angular momentum being equal to E quation 2 2 ( 2 2) As demonstrated by E quation 2 stat es this leads to the doublets seen in XPS. The doublets are labeled as follows: p orbitals are labeled as p 3/2 and p 1/2 d orbitals are labeled as d 5/2 and d 3/2 and f orbitals are labeled as f 7/2 and f 5/2 The two peaks in a doublet have a characteristic energetic separation distance. This means that if a fitted peak is placed in one peak of the doublet, a fitted peak must be added to the other peak in the doublet as well, with the distance between the peaks being the characteristic separation energy. The final guideline is that there is a characteristic ratio of intensities of doublet peaks that must be upheld when fitting doublets. For p orbitals the ratio of the 3/2 to 1/2 peak intensities is
41 2:1, in d orbitals the ratio of the 5/2 to 3/2 peaks is always 3:2, and in the f orbital the ratio of the 7/2 to 5/2 peak is always 4:3. As always it should be considered that these are guidelines not steadfast rules and small adjustments can be made if after following these guidelines the composite fitted signal doe s not match the actual data. Also if there any remaining intensity that has not been accounted for after fitting using these guidelines, additional peak(s) should be added to account for the remainder After peak fitting, the area of each fitted peak is c alculated in the software. The total area of the fitted peaks theoretically equals the area of the actual data for the spectrum. This value is needed for quantification of the relative at% of each element present in the selvage region of the sample. The in tegrated intensity of each spectral region must be normalized by a relative sensitivity factor. This sensitivity factor is related to the X ray flux, mean free path of electron, and the detector efficiency which is element specific. After obtaining the rel ative total intensity for each element, this value is divided by the combined relative intensity for all elements detected. This value is then multiplied by 100 to give the relative at% of the element present in the near surface region. The relative at% fo r all the elements should combine to equal 100%. Pin on Disc Tribometry To measure the coefficient of friction for the materials used in the current study a novel homebuilt pin on disc tribometer was used. The tribometer was built by Dr. Gregory Dudder, b ased on the design of Dr. Brandon Krick and Dr. Gregory Sawyer; the final SolidWorks drawing can be seen in F igure 2 4
42 Figure 2 4 Final SolidWorks drawing of the novel in vacuo pin on disc tribometer courtesy of Dr. Gregory D udder The instrument is based on the archetype design of a set of pin on disc tribometers used in the MISSE 7 mission that were taken to and installed on the International Space Station (ISS) duri ng STS 129 Mission The novel tribometer has the added adv antage of being vacuum compatible. The tribometer consists of five main components all of which are specially adapted to be used in vacuum. The first component was the arm of the tribometer. The main function of the arm is to hold the pin and to undergo st rain in two axes. The arm is constructed of titanium and is fitted with strain gauges from Sensing Systems Corporation of New Bedford, MA. The strain gauges have been calibrated using masses of known value suspended from the head of the tribometer. The sec ond component of the tribometer is the piezo actuator. The actuator applies the normal load to the pin, providing an adequate travel distance to be able to load the sample when no voltage is being applied to the piezo. The instrument employs a model APA 12 0S piezo stack from Cedrat. The model was selected after
43 several dynamic strain simulations were run, showing that this model was the best choice. This specific piezo functions through the application of a positive, DC voltage to a horizontal stack of BiTi O 3 blocks. By applying the voltage to this horizontal stack the head of the arm contracts vertically. The third component of the tribometer is the rotating sample stage. The sample stage represents the Omicron GhmB platen mounting position and acts as the flat disc for the revolving pin on disc test. The stage was made of titanium as well. Two holes are drilled into the stage to allow for the prongs on the transfer arm to enter as required for sample transfer using the existing transfer device in the XPS v acuum system. A square bottom hole was machined on the underside of the sample stage as well for the insertion of the D shaped drive shaft of the motor assembly. The drive shaft secured to the sample stage via a set screw inserted through a threaded hole i n the stage. Platens were secured to the stage via an assembly that mimicked the design of the 4d manipulator stage used in the XPS system. The motor assembly that is attached to the sample stage is the fourth component of the tribometer. The motor assembl y is comprised of two pieces, the motor itself and a planetary gearbox. Both of these were purchased from Maxon Motors of Fall River, MA and are rendered vacuum compatible using specialized grease in the motor and gearbox. The motor is an EC 20 Flat model that measures 10mm in height and 20mm in diameter. The maximum speed this model can reach is 5000 rpm. The planetary gearbox provides a reducing ratio of 84:1 meaning that the maximum sample stage rotational speed allowed was 60 rpm. The reducer gearbox an d motor were assembled together into one piece by Maxon before attachment. All of the previous components are directly attached to the fifth component
44 which is mounting block. The dimensions of the mounting block were determined by the requirements of tran sferring samples, a procedure which utilized three separate transfer arms to transfer from the load lock to the tribometer and from the tribometer to the XPS analysis chamber. The length of the mounting block could not exceed 9cm as it would impede motion of one transfer arm and it could be no more than 5cm tall or it would not align with another arm. The location of the stage had to be far enough away from the front edge of the piezo to allow room for the arm to lock onto the head of the platen. The mounti ng block was attached to a four dimensional manipulator with a inch diameter stainless steel rod held in place with a set screw. All power inputs and signal outputs were wired though UHV feedthroughs on the 4 D manipulator mounting flange. The strain gau ge signal was relayed via a Sensotec in line amplifier to a National Instruments SCB 68 signal accumulation box interfaced with a National Instruments 6221 data acquisition card. A National Instruments Labview V8.5 data acquisition program, written by Ira Hill was used for tribometer operation and data collection. The program allows the user to set the direction and speed of rotation, to customize the experiment time and data acquisition rate, and to monitor and control the applied normal load during operat ion of the tribometer. Data is saved directly to a Microsoft Excel file and includes the lateral force, normal force, time, and friction coefficient. Pin on Disc Testing Parameters The purpose of pin on disc test are to measure the friction produced durin g sliding as well as measuring wear performance of a sample under the tribological conditions during operation. During the current study a standard set of parameters was used. The normal force for every test was set to 1N, the diameter of the wear track wa s
45 a constant 6mm, and the size and type of ball used was a 3mm radius 440C stainless steel ball. Scanning White Light Interferometry There are a variety of techniques that can be used to measure the wear performance of the tribological samples, usually r eported as a wear rate. All of the techniques involve estimating the volume of material removed during testing. One of the simplest ways is to measure the mass of the sample before and after tribological testing. The change in mass is the amount of materia l removed. Multiplying this by the density of the material gives the volume lost during testing (V L ). Knowing the normal force (F n ) used during the test and the total sliding distance (d) of the test, the wear rate can be calculated by using E quation 2 3, with wear rate expressed in units of either m 3 /Nm or mm 3 /Nm. ( 2 3) This technique of measuring mass lost to calculate the wear rate is very reliable but becomes difficult when u sing smaller samples. In addition, when samples that experience low wear, a very fine balance is needed to measure the amount of mass loss as this can often be small fractions of a gram. As a result, alternative methods have been developed to determine the amount of volume lost during tribological testing. These entail measuring the wear scar made during the test and calculating the volume lost directly. There are a number of techniques available to accomplish this including atomic force microscopy (AFM), s canning tunneling microscopy (STM), and white light interferometry. For this study scanning white light interferometry (SWLI) was used as the wear tracks were too large in width to be scanned using AFM. The approach
46 employed a Veeco Wkyo NT9100 Interferome ter and the actual measurements were taken by Dr. Rachel Colbert. SWLI works by using the wave superposition principle to combine waves to form interference rings. In the Veeco Wyko NT9100, white light illuminates a semitransparent mirror. A portion of the light is reflected back onto a second mirror, while some is transmitted through to the sample. After reflecting off the mirror and sample, respectively, the waves of light merge and form an interference image. By changing the z position of the transparent mirror, different interference images are taken at a set sequence of phase differences Based off these interference images at different heights and the movement and intensity of the interference rings on the interference images the software generates a 3 d image of the sample surface. Once the image is recorded, image processing software such as SPIP can be used to produce an average of every possible line scan in the entire image. Once this average line scan has been obtained, a Matlab program was written to determine a zero point on the scan that represents the sample surface and then calculate the area under the zero level that had been removed. By taking SWLI images over multiple areas of the wear track, the volume removed as a result of wear a can be c alculated, by multiplying the average area removed by the total length of the wear track. The wear rate for the sample is then calculated using E quation 2 3 as described above.
47 CHAPTER 3 T RIBOLOGICAL AND C HEMICAL E FFECTS OF V ARYING A NNEALING T EMPERATURE IN E LECTROLESS N ICKEL B ORON C OATINGS Introduction There is a large need in the design of engineered pieces for the protection and enhancement of the performance of surfaces. In 2009 the global coatings market totaled over $90 billion USD with non decorati ve coatings (such as marine, hard coatings, and tribological coatings) making up over 50% of the market  Hard coatings such as diamond like carbon, chrome coatings, and boron nitride are commonly used to increase the hardness of a surface to prolong its operation. By covering the surface with a hard coating the lifetime of the part is increased due to the fact that the piece will experience less wear during operation. Hard coatings can also modify other properties of the surface, potentially providing re sistance to corrosion, improved frictional behavior, and protection from unwanted chemical reactions Nickel boride coatings offer a number of these advantages [13, 77 81] A number of coating processes can be used to deposit nickel boride, including physical vapor deposition, chemical vapor deposition, and electrolytic deposition One of the increasingly used deposition processes is electroless deposition. Electroless deposition is a purely catalytic process in which no external power source is needed to drive the deposition. Also electroless deposition eliminates line of sight is sues as well as edge effects commonly seen in many deposition processes  Advantages of this approach also include the relative simplicity in equipment, low capital expenditure, scalability, and the overall lower workforce skill involved in the deposition [9, 82, 83] The electroless deposition of nickel boron coatings is beneficial in protecting against corrosion and fouling, extending of service lifetimes, and conservation of energy. Nickel
48 boron coatings have been shown in previous studies to have high hardness, low wear, good adherence on a variety of substrates, and provide corrosion resistance  In certain situations, nickel boron coatings have shown low friction but many factors such as de position variables, annealing temperature, and tribological testing environment play a large role in the variance on friction values reported [84, 85] Nickel bo ron coatings are commonly annealed to increase the hardness of the coatings; 400C is commonly as a result of empirical investigations that revealed an enhancement in hardness following this treatment [86, 87] In the present study, the influence of temperature dependent oxygen anneals on the tribological properties of electroless nickel boride coatings have been investigated. Following anneals under a flow of oxygen gas at 250C, 400C, 550C, and 700C, X ray photoelectron spectroscopy (XPS) has been used to document the composition of nickel boride coating surfaces as a function of annealing temperature In addition, Raman spectroscopy has assisted in ascert aining the molecular nature of the boron oxide species detected in the near surface region following high temperature anneals In turn, compositional changes have been correlated with variations in surface morphology and tribological response in order to g enerate a thorough understanding of the origins of favorable performance characteristics Pin on disc tribometry has been used to measure the friction and associated wear properties of the nickel boride coatings as a function of annealing treatments  The pin on disc measurements have provided an average coefficient of friction for a given set of slidi ng conditions while interferometry measurements have been used to follow the amount of material lost
49 during sliding. Using known variables from the pin on disc testing, fundamental wear rates of the variously annealed coatings have been obtained. Experimen tal Numerous 4130 steel substrates measuring approximately 10 mm 14 mm were processed by UCT Coatings Inc. (Stuart, FL) to deposit a boron coating fro m a nickel chloride hexahydrate (NiCl 2 6H 2 O) and sodium borohydride (NaBH 4 ) solution. Lead tungstate (PbWO 4 ) was employed as a s olution stabilizer. The coating thickness was monitored via deposition rate calculations and verifi ed via cross sectional SEM images. A tube furnace allowed for ultrahigh purity O 2 gas to be fl owed over the coated substrates during the various annealing procedures. A steady fl ow of 0.0094 N m 3 /min was e stablished to purge the tube of any ambient air before heating to the specifi ed temperature. Temperature was measured via a K type therm ocouple placed inside the tub e immediately above the sampl e. Once the desired temperature was reached, the samples were he ld at this temperature for 3 hours. Samples were allowed to cool to room temperature under a flow of O 2 gas. XPS was performed using the system and methodology pre viously described. All samples were sputtered for 5 min with the PHI FI G 5CE ion sputter gun with beam energy of 1 kV before general surv ey spectra were taken to remove surface contamination. Processing o f the XPS spectra was done using the guidelines and methods previously described. Raman spectra were collected with a LabRam Infi nity (Horiba Group) micro Raman system, empl oying 1.5 mW continuous helium: objective. Spectra were recorded using 6 s integration times an d a veraged 10 times for a total integration time of 60 s. Backscattered radiation was collected by the same microscope objective, passed through a sharp edge fi lter to reject elastically scattered light, and image d by a CCD
50 detector (1024 9 256 pixels). Tribo logical testing was perform ed via the novel tribometer described previously. A ll samples were run against a 3 mm radius 440C stainless steel ball under ambient conditions with a relative humidity of 40 50%. The tribo logical investigations entailed the application of 1 N normal load and sliding distances in excess of 1 km. Under these conditions, the es timated contact pressure of the sliding interface was 530 MPa, calculat ed assuming a Hertzian contact; higher local pressures likely exist ed in the region of asperities. Scanning white light interferometry (SWLI) scans were taken using a Veeco Wyko NT9100 Interferomet er. Scans over multiple regions were processed and compiled in or der to illustrate broad re gions including the wear track SEM images were obtained using a JEOL NeoScope JCM 5000 benchtop SEM employing a prima ry beam energy of 10 kV. Images were obtained from the region of t he wear track of each sample to illustrate the deformation and damage ca used by the tribological tests. Images were also obtained off of the w ear track to depict the general morphology of the coating and to document changes occurring as a result of the annealing processes. The surface hardness values of the nickel boride coati ngs were evaluate d with a Hysitron Triboindenter, employing a three sided Berkovich diamond indenter. Results Compositional and Morphological Changes Figure 3 1 shows SEM images courtesy of Dr. Wei Qiu and Dr. Juan C. Nino, of the undisturbed portions of the coatings not affected by the tribological testing performed. The as received sample along with the sample annealed at 250C depict s a nodular structure for the nickel boron coatings. In the 400C annealed sample some aggregation can be seen on the sur face where some of the smaller nodules have
51 coalesced, smaller surface cracks have closed, and the deeper cracks are less prevalent. Upon annealing at 550C extensive changes can be seen in the morphology. The nodular structure is completely absent, replac ed by a new phase on the surface. This new structure, dubbed hereafter as flaked, can also be seen in the sample annealed at 700C. Also seen in the 700C image are slight remnants of the nodular structure indicating that the new flaked structure has evolv ed from the previous structure. Figure 3 1. SEM Images of the surface of the NiB coatings a ) as received and after annealing at b) 250C, c) 400C, d) 550C, and e) 700C showing the morphological changes that occur due to the annealing at these temperature images courtesy of Dr. Wei Qiu and Dr. Juan C. Nino Figure 3 2 depicts the Ni 2p 3/2 core spectra for the series of coatings annealed. The Ni 2p spectrum is notoriously difficult to fit due to the number of satellite peaks, denoted in the sp ectra as S1 and S2, which occur due to quantum mechanical effects
52  Despite these complexities, the 2p 3/2 peak can be fitt ed with two smaller peaks. The first peak located at approximately 853 eV is associated with Ni bonded to B in the compound Ni 3 B, the intended stoichiometric compound during deposition. Prior experiments have shown that as nickel boron coatings are anneale d, the films crystallize and the dominant phase seen is Ni 3 B  Figure 3 2. Ni 2p 3/2 core spectra showing that after higher annealing temperatures no The second fitted peak, located at 853 854 eV, is assigned to Ni bonded to O in the compound NiO An additional nickel oxide, Ni 2 O 3 may potentially form with a characteristic Ni binding energy of 856 857 eV. Unfortunately this peak overlaps with one of t he satellite peaks and therefore cannot be fully assessed. Slight variations in the intensity of these spectral features are seen in the as received sample and those annealed to 250C and 400C, however the complete absence of any Ni species is indicated b y the spectra measured from samples annealed to 550C and 700C. Based on these results, lower annealing temperatures are seen to produce a minor oxidation of the nickel boron coating, while higher temperature anneals are seen to induce a drastic
53 compositi onal change to the surface region, consistent with the morphological changes described above. Figure 3 3. B 1s core spectra showing a) the fitted as received sample with peak assignments and b) the spectra for the 250C, 400C, 550C, and 700C samples showing a transition from two peaks in the lower annealed samples to a singular broad peak in the higher annealed samples. Figure 3 3 displays the B 1s core spectra for the samples annealed to different temperatures in the study The relative decrease in S/N ratio compared to the Ni spectra results from the lower photoelectron cross section for boron. As can be seen in the as received, 250C, and 400C samples there are two peaks expressed. The lower binding energy peak is located at approximately 188 eV and corresponds to boron bonded to Ni in the Ni 3 B structure. The higher peak, located at approximately 193 eV can be assigned to B bonded to O. This shows that even in the as deposited coating some degree of surface oxidation has occurred In the samples a nnealed at the higher temperatures, the lower binding energy peak is no longer seen and a single broad peak is observed. This peak is attributed to boron oxide species and/or boric acid resulting from the diffusion of subsurface boron and hydrogen at these temperatures and the extensive oxidation of the surface region The proximity of reported binding energies for
54 B 2 O 3 (193.7 eV) and H 3 BO 3 (192.8 eV )  and the ill defined shape of the spectral peak prevent a quantitative deconvolution into their respective presence The formation of hydrogen containing boric acid is consistent with the inclusion of hydrogen in Ni 3 B films produced through electroles s deposition The breadth of the peak and the shift to higher energies is also observed for both oxygen and adventitious carbon detected in related spectra and is consistent with surface charging of the oxide. These spectral features observed with increasi ng temperature demonstrate that all of the boron present in the near surface region is converted from its deposited form (Ni 3 B). Figure 3 4 displays the correlated changes observed in the O 1s peak as the annealing temperature increases. The lower annealed samples and the as received sample show a peak that can be assigned as arising from two species. Intensity located at approximately 532 and 533 eV and can be attributed to Ni 2 O 3 and B 2 O 3 respectively. For the sample annealed to 250C, deconvol ut ion of th e O 1s peak reveals intensity at 529 eV that can be attributed to NiO in addition to that observed for the as received sample Upon annealing to 400C, a relative shift in intensity located at 533 is consistent with the formation of more boron oxide specie s After annealing above 550C, the O 1s peak shifts to a higher binding energy and widens much like the case in the B 1s peak, consistent with the complete oxidation of the surface region and related surface charging.
55 Figure 3 4. O 1s core spectra a) showing peak identification for the as received sample and b) remaining samples O 1s spectra showing transition in binding energy and broadening of the peak due to the creation of a B 2 O 3 rich surface in the higher annealed samples. From the spectra above the relative percent composition of the surface of the coatings was determined using the methods previously described and are presented as atomic percentages in Table 3 1 As XPS is insensitive to the presence of hydrogen, its relative presence in the ne ar surface region is not reflected. The data for the samples annealed to 250C and 400C indicate that a change in surface composition has occurred as a result of the heat treatment. The presence of C in the as received and lower annealed samples can be at tributed to surface contamination. In the electroless deposition process, PbWO 4 is used, in very minute quantities, as a stabilizer for the bath [17, 86, 87, 91] In some samples, a small amount of Pb has been detected on the surface, usually less than or equal to one atomic percent, as seen in the 400C composition data. The samples annealed to 250C and 400C exhibit a slightly higher B concentration at the cost of lower Ni concentration with respect to the as received coating Significant changes in the composition occur on ce the coatings are annealed at
56 550C and 700C. For these samples, a complete depletion of nickel and higher concentrations of B and O with a much higher concentration of C is measured. The higher concentration of C is likely due to the fact that at these elevated temperatures there is diffusion of C into the actual coating as opposed to contamination resting on the surface. Table 3 1. Elemental composition, in atomic percent, of coatings after various annealing temperatures. Sample Ni (at. %) B (at. %) O (at. %) C (at. %) Pb (at. %) As Received 31 19 42 8 0 250C 22 23 45 10 0 400C 24 24 45 6 1 550C 0 34 49 17 0 700C 0 32 53 15 0 In order to further ascertain the nature of compositional changes occurring with high temperature anneals in oxygen environments, Raman spectra were collected from the Ni 3 B coating annealed to 700C by Dr. David Hahn The resulti ng spectrum is presented in Figure 3 5 and is consistent with the transformation of the near surface region to boric acid, H 3 BO 3 The peaks at 510, 894, and 1182 cm 1 are in agreement with features previously assigned to boric acid [92, 93] In addition, the absence of peaks at 808, 1325, and 1475 cm 1 characteristic of B 2 O 3 i s noted [94, 95] The small feature at 1386 cm 1 in the spectrum reported here has not been assigned, nor were assignments for the features at 3193 or 3265 cm 1 identified throu gh an extensive search of the literature Nonetheless, the Raman spectrum clearly highlights the formation of boric acid and the influence of hydrogen incorporated into the film during deposition.
57 Fig ure 3 5 Raman spectrum of the electroless Ni3B samp le following a 700C anneal in a blanketing oxygen gas The spectrum is dominated by vibrational peaks attributable to boric acid courtesy of Dr. David Hahn Friction Behavior Figure 3 6 ach of the samples tested. As previously stated, the normal force in each test was 1N and the counterface was a 3 mm radius 440C stainless steel ball, resulting in an estimated contact pressure of 530 MPa Again, a separate counterface was employed for eac h unique sample The as received and 250C samples behaved similarly, with each test having a run in period for the first 100 m or so of sliding then holding at a value of
58 around 0.5 for a period of 400 m. Then following ~600 m of sliding distance both sam ples exhibit a steady increase in until the end of the test, by which point the samples have reached coefficient of friction values of ~0.8 0.9. The frictional response of the 400C sample, (c), differs from the previous two samples In the first 30 m of sliding, it begins with a sharp increase of the value from 0.35 to 0.7 and then holds steady at this higher value for the next 170 m of sliding Then, a second, abrupt increase in occurs, increasing to a coefficient value of 1.0 for the remainder of t he test. As will be discussed below, this second increase can be attributed to formation of wear debris within the wear track Figures 3 6(d) and 3 6(e) illustrate a very different frictional response for the coatings annealed in oxygen to 550C and 700C. The 550C sample has a very brief run in period then holds at a value of approx. 0.15 for the duration of the test. The 700C sample starts at a slightly higher friction value then decreases to a steady state value of 0.06 for the remainder of 1 km slidin g test Figure 3 6 Friction coefficient vs. sliding distance for Ni B coatings a) as received and after annealing at b) 250C, c) 400C, d) 550C, and e) 700C.
59 Wear Behavior The wear rates associated with the 1 km sliding runs were determined fr om an analysis of the circular wear tracks measuring 18.85 mm in circumference. Using line scans generated from SWLI and averaged over an entire image, the volume removed from the coating can be determined and from there the wear rate of the coating can be calculated A representative section of an image analyzed in this way is shown in Fig ure 3 7 SWLI images are courtesy of Dr. Rachel Colbert, illustrating the well defined wear track observed on the sample annealed to 550C The wear rates of all 5 sample s are shown in Table 3 2 as well as the average value over the entire test for each sample. The as received sample and 250C sample exhibit similar behavior having both average values and wear rates that are very close together. No wear rate could be calculated for the 400C sample due to debris acc umulating on the surface, which will be discussed in greater detail later. The 550C and 700C samples exhibit a higher wear rate as compared to the lower annealed samples, although having lower coefficients of friction. The difference between the wear rat es of the lower and higher annealed samples can be attributed to the change in both composition and morphology occurring over this temperature range, as described above. Overall, these wear rate values represent a very low rate of material removal from the contact zone as compar ed to many other materials coatings [96, 97] The SWLI measurements further revealed the depth of wear tracks for the lower annealed samples to be between 10 30 nm. In comparison, some of the features on the surface resulting from deposition of the coatings are on the order of 5 15 nm As a result, th ere appears to be more of a smoothing of the nodular structure than material removal. This smoothing effect can be seen in the as received and 250C samples SEM
60 images courtesy of Dr. Wei Qiu and Dr. Juan C. Nino, shown in Fig ure 3 8A and 3 8B Some of th e thinner, deeper lines seen in the images arise from wear debris from the ball. The 400C sample shows a much broader wear track than the other samples as a result of the large flat spot formed on the ball during the testing, which lead to a larger contac t area. The 550C and 700C samples, which displayed higher wear rates, were revealed to have wear track depths between 70 1 20 nm. While this is much more than the lower temperature annealed samples, a removal of 100 nm correlates to 1/1000 of ure 3 8 the higher annealed samples do not display the smoothi ng effect seen in the as received and 250C samples, but exhibit a more typical material removal process within the wear track. The origin of the wider wear track observed on 700C sample (Fig ure 3 8E ) has not been determined in these studies. Figure 3 7 SWLI image showing the wear track generated on the 550C sample courtesy of Dr. Rachel Colbert and a l ine plot of SWLI data, averaged over the entire area of the image, indicating the depth of the wear track.
61 Table 3 2. Coefficient of friction and wear rates of all samples tested. Sample Coefficient of Friction Wear Rate (mm 3 /Nm) As Received 0.55 4.3e 8 250C 0.58 4.8e 8 400C 0.90 N/A 550C 0.17 2.7e 7 700C 0.06 2.1e 7 Figure 3 8 SEM images of wear tracks of NiB coatings for the a) as received sample and after annealing at b) 250C, c) 400C, d) 550C, and e) 700C images courtesy of Dr. Wei Qiu and Dr. Juan C. Nino Discussion Compositional and Morphological Changes As seen in the data presented above, the result of annealing N i B coatings in oxygen at temperatures of 550C and above is a marked change in both morphology and composition These changes in turn lead to notable changes in the measured tribological properties. The typical nodular structure observed on the as receive d
62 sample and following anneals up to 400C is converted to a flaked structure when coatings are subjected to higher annealing temperatures. The resulting coating morphology is smoother than the original coating surface, free of the cracks and gaps between nodules in the nodular structure. This change in surface morphology points to a complete change in the material that makes up the surface region of the coating. This hypothesis is corroborated by the trends seen in the composition of the near surface regi on of the coatings determined via XPS. The as received sample and those annealed at lower temperatures have similar compositions showing both Ni and B in the selvage region. The B 1s core spectra indicate B bonded to Ni in Ni 3 B as well as B bonded to O in B 2 O 3 like species Although the core spectra are consistent with B and Ni bonded together as Ni 3 B, a stoichiometric ratio of 3:1 is not observed In the as received sample, the ratio is 1.6, while being 1:1 in the 250C and 400C samples. These variations likely result from the thermodynamically favorable formation of boron oxides and the resulting diffusion of boron to the surface. This picture is supported by the presence of the large peak in the B 1s spectra that represents B in B 2 O 3 The drastic chan ge in the composition observed for the samples annealed at 550C and 700C are also consistent with the thermodynamically driven formation of boron oxides At elevated temperatures, the diffusion of small hydrogen and boron atoms is enhanced and the surfac e region of the coating is transformed to boron acid. This explains the complete lack of nickel seen in the higher annealed surface as well This surface oxidation mechanism is similar to that observed for the formation on SiO 2 in which Si atoms diffuse t hrough an ever increasing thickness of SiO 2 form the oxide layer For silicon surfaces, higher temperatures are well known to produce thicker oxide
63 layers, consistent with the activated nature of the process  A similar behavior is proposed to be occurring at the surface of nickel boron coatings as hydrogen and boron are drawn to the surface to react with gas phase oxygen  The formation of the boric acid also provides a basis for interpreting the change i n morphology observed following the higher temperature anneals The transformation from the nodular structure to the flaked morphology is seen to result from the significant diffusion of boron and hydrogen through the near surface region and the correspond ing formation of boric acid It should also be note that the Raman data indicate that oxidation is occurring to a substantially greater depth than that detected by XPS, again consistent with observable morphological changes and the friction and wear charac teristics discussed below. Frictional Response The tribological behavior of the coatings exhibits similarly drastic changes as a function of annealing temperature. As seen in Fig ure 3 6 the as received and 250C samples show a period of stability early, a lbeit it at moderately high friction, then begin to increase at approximately 600 m of sliding distance As deposited nickel boride coatings of this composition can thus be categorized as having an intermediate friction with a relatively low wear rate, co nsistent with the known hardness of the compound. The 400C sample showed an immediate increase followed by a small plateau and then another rise to an approximate value of 1. These tribological changes are attributed to an interface within which a high degree of wear is occurring In this case, si gnificant interfacial wear occur s on the stainless steel counterface as well as the coating, producing large quantities of dark red debris in the region of the wear track, consistent with the formation and trans fer of an iron oxide species The mechanism by which this form of wear occurs has not been revealed through the studies performed.
64 The 550C and 700C samples exhibited a completely different tribological response than the samples annealed at lower tempera tures, as would be expected with the significant change in surface chemistry and morphology. These coating samples Boric acid is widely known to form a layered structure  and is further believed to derive its usefulness as a lubricant from this property. In light of this, annealing electroless nickel boron coatings at elevated temperatures in oxygen containing environments is seen to not only produce a given surface finish, but also produce a relatively thick layer of protective, lubricious H 3 BO 3 All of the pin on disk tests presented above were performed at room temperature and pressure in ambient air in which the relative humidity (RH) was ~40% or greater To investigate the role of water in the measured tribological properties, an experiment was performed where sliding began in ambient air with a relative humidity approximately 50% on a sample annealed at 700C in O 2 Following sliding for 9.42 m, a dry N 2 purge was introduced so as to lower the RH to <10% This procedure of alternating testing environments was repeated while continuously sliding, resulting in the tribological response depicted in Fig ure 3 9 Here, regions of sliding in high humidity have been labeled as HH and those in low humidity as LH After reproducing the friction results obtained on a similarly annealed coating, the coefficient of friction was observed to rise slightly upon the reduction in relative humidity The sliding interface was able to recover to the lower value upon returning to HH; this trend repeated for the second LH to HH cycle as well. Upon lowering the humidity a third time however, an almost linear increase in is seen with sliding distance, rising from 0.1 to 0.4 over this distance. This
65 result points to a progressive wear of the protective boric acid layer during low humidity sliding. Following this ramping up of the friction coefficient, a modest recovery is made upon returning to HH, bringing the value down to 0.2. The last region of sliding in LH exhibits an immediate increase back to a value of 0.4, which then holds for the remainder of the 9.42m humidity cycle. The last HH cycle highlights a potential kinetic effect with respect to the parti cipation of ambient water in tribochemical reactions occurring under sliding conditions. Wear Response The wear response of the various coatings reveals further evidence to the fact that annealing temperature has a vast influence on the nature of the surf ace of nickel boron coatings. As previously discussed, no wear data was able to be obtained for the 400C sample as the wear debris generated from the ball during testing accumulated in the wear track, effectively producing a growth in thickness within the wear track Beyond this sample, the results of the SWLI investigation depict a coatings performance of considerable interest. The wear rates calculated for the as received and 250C samples are both very low, in the range of 4.0x10 8 mm 3 /Nm. The 550C and 700C samples exhibit a moderate increase in the wear rate to ~2.0x10 7 mm 3 /Nm; yet, these rates are all fairly low Recognizing that wear rates for many advanced aerospace coatings are in the 10 6 10 7 range demonstrates that these Ni 3 B coatings offer t he potential for substantial wear resistance [96, 97] Although not m easured explicitly, ambient water, in conjunction with the boric layers formed during oxygen anneals, likely plays a strong role in maintaining a low wear rate, as suggested by the low humidity data presented in Fig ure 3 9
66 Fig ure 3 9 me for a 700C sample with 9.42m relative humidity cycles. Summary of Findings Annealing electrolessly deposited nickel boron coatings above a temperature of at least 550C has been shown to transform the surface of the as deposited coating from a nodular structure to a flaked structure. The chemical nature of the coatings shifts from one containing Ni 3 B to a surface that is completely covered with boric acid The presence of these species within the surface layer results in notably reduced coefficients of friction with respect to those of the parent coating when sliding against steel counterfaces in humid environments. Together, the results of this study document correlated changes in surface morphology, composition, and tribological response as result of oxygen annealing and demonstrate the potential for further modification of these useful coatings through post deposition thermal treatments.
67 CHAPTER 4 COMPOSITIONAL EFECTS OF EXPOSURE TO LOW EARTH ORBIT ON MoS 2 /Sb 2 O 3 /Au COATINGS Overview The Materials on the International Space Station Experiments (MISSE) program presents a unique opportunity for researchers dealing in space application materials to have samples exposed directly to low earth orbit (LEO) conditions. Started in 2001, the MISSE program has ha d 8 missions sending samples and experiments to the International Space Station (ISS) with the purpose of characterizing the performance of existing and new space materials during and after exposure to actual space conditions  The materials and experiments selected for testing are mounted in a terrestrial environment into passive experiment containers (PECs). These PECs are then mounted to the outside of the ISS and are left in LEO conditions for a period of time until their retrieval and delivery back to Earth. The MISSE 7 mission contained two PECs, one to be placed on the ram side of the ISS and one to be placed on the wake side of the ISS. The PECs were carried to the ISS on November 2009 a board mission STS 129 and were retrieved during the STS 134 mission on May 2011, leading to an exposure of approximately 18 months  The active experiments consisted of small scale tribometers which would be automated to run after attachment to the outside of the ISS and transmit data on a number of tribologically relevant samples, including the coating reported in the current study  The resulting tribological data is presented elsewhere, with the emphasis of this report focusing on the nature of the chemical changes incurred by exposure to LEO. The commercially available composite coating was comprised of MoS 2 Sb 2 O 3 and Au and designed to act as a solid lubricant in space applications [103, 104] The coating has this unique chemistry to increase the
68 desired properties, ma inly a low friction coating capable of being used in space. The MoS 2 component of the coating acts as the solid lubricant, while the Sb 2 O 3 and Au components are included for structural support and densifying roles [103, 104] The coating was mounted within a tribometer on the ram side PEC, resulting in expos ure to a very unique and chemically aggressive environment. The ram side of the ISS is under ultrahigh vacuum conditions, but experiences collisions with atomic oxygen on the order of 10 14 10 15 atoms/cm 2 s  These atomic oxygen atoms possess a kinetic ener gy of 4.5eV  The ram side also experiences temperature fluctuations from 40 to 60C as well as the potential exposure to various forms of cosmic radiation striking the surface of the sample during orbit  Because the ram side is the forward facing dire ction during orbit, there is finally the danger of debris striking the surface  The unique environment that exists in LEO can potentially produce extreme changes in the composition of the surface of a material. Atomic oxygen is highly reactive and can have severe effects on the tribological response of coating. This is due to the fact that such composite coatings have the ir composition carefully engineered to yield desired tribological properties. Changing this composition can have drastic effects on these properties. Atomic oxygen is known to attack the surfaces, forming oxides with metallic components and eroding others. As a result, the chemical transformation of tribological components potentially may cause the coating to fail prematurely or induce the failure in other components due to wear and high torque. There is great potential for surface analysis techniques such as X ray photoelectron spectroscopy (XPS) in determining how exposure to LEO chemically changes the surface of materials exposed to these environments. XPS offers the ability to determine not only the elements present
69 on the surface but their local chemica l bonding environment. The approach can be used to calculate relative atomic percentages, thus illustrating how the surface of the sample has changed compositionally after exposure to LEO. Methods and Materials The specimen reported in this study is a co mmercially available MoS 2 /Sb 2 O 3 /Au coating. The coating was DC sputtered by Hohman Plating (Dayton, OH) from a composite target source on to a 304 stainless steel substrate disk. The MoS 2 /Sb 2 O 3 /Au coated disk was mounted to the aforementioned tribometer, i n turn mounted on the ra m side assembly as shown in Figure 4 1 image courtesy of Dr Gregory Sawyer and Dr. Brandon Krick Two samples were used for the current study, the first being a coating that was not exposed to LEO and was kept in a clean terrestri al environment; this sample is hereafter referred to as the reference sample. The second sample was the coating exposed to LEO and is hereafter referred to as the flown sample. Figure 4 1 Schematic design of the MISSE 7 space tribometers from MISSE mi ssion courtesy of Dr Gregory Sawyer and Dr. Brandon Krick
70 eV) monochromatic X ray source using the process previously described in Chapter 2. Processing of the XPS spectra was performed using CasaXPS software (Casa Software Ltd.) and was performe d using the methodology that was described in Chapter 2 Results The first aim of the XPS analysis was to identify the species present in the upper 10 strong depth sensitivity and therefore only obtains information fr om the near surface region. After identification of the elements present and their respective binding energies, the values of the peak energies are compared with values in the literature to determine the bonding environment of the elements. From these data the relative atomic percentages of each element present were determined using the methodology described previously To be able to understand the impact of exposure to LEO, a reference sample was compared to the flown sample. The reference sample was a di sc of the same substrate and was coated at the same time as the sample that was flown on the ISS. This reference sample was maintained under pristine conditions during the duration of the MISSE 7 program and was analyzed at the same time as the flown sampl e. By using this reference method, the chemical changes that occurred during exposure could be easily identified through a compositional analysis. Table 4 1 shows the relative atomic percentages of the elements identified in both the flown and reference sa mples. Chemical State of Reference Sample The O 1s / Sb 3d spectra for the reference sample is shown in Figure 4 2A. As can be clearly seen, these two elemental regions extensively overlap and are therefore inherently difficult to differentiate and deconvo lute. To accurately fit this region a number
71 of procedural rules were evoked. First, the area of the Sb 3d 5/2 peaks relative to Sb 3d 3/2 peaks were fixed at a 3:2 ratio, according to the quantum mechanical nature of their origin; second, the separation dis tance between the 3d 5/2 and 3d 3/2 peaks was held at 9.4eV, which corresponds to the reported literature value Finally, the full width half maximum (FWHM) values of the corresponding peaks were kept constant  This general approach of using set peak energy ratios, set separation distances, and consistent FWHM values between peaks was used for fitting all of the regions involved in the study. Using thes e guidelines, two sets of peaks were required to fit the Sb 3d spectrum The first set of peaks located at 529.7 and 539.1 eV are assigned to stoichiometric Sb 2 O 3 While slightly lower than most reported energies [107 109] provided a consistently good fit throughout these data sets. The second set of peaks, located at 528.2 and 537.6 eV, are assigned to sub stoichiome tric Sb x O y sub oxides, which can be produced during the DC sputtering process  After assigning the Sb peaks in the combined spectra, the remaining intensity was attributed to oxygen. The main O 1s peak is located at 530.7 eV and is assigned to O within Sb 2 O 3 and the sub oxides Sb x O y species. The remaining intensity, located at 532.6 eV, is assigned to O bound in C O bonds, a species believed to arise from adventitious carbon located on the surface of the c oating. Table 4 1 Atomic percentages of elements in the near surface region of the flown and reference samples as determined by XPS. Element ( at % ) Sb Mo O C Si Au S Reference 5 15 22 45 0 1 13 Flown sample 2 2 65 9 21 0 0
72 Figure 4 2 Comparison of O 1s /Sb 3d core spectra for the A) reference sample and B) flown sample, indicating that all Sb has been converted from a mixture of Sb 2 O 3 /Sb x O y to Sb 2 O 3 Figure 4 3A displays the molybdenum 3d spectrum for the unflown referenc e sample. Splitting of the Mo 3d 5/2 and 3d 3/2 peaks was held to a constant 3.1eV separation  Metallic Mo peaks were assigned for energies of 228.3 e V and 231.4 eV [106, 108] Peaks with energies of 2 29.4 eV and 232.5 eV were assigned to MoS 2 [107, 109, 110] There is possibility of the presence of other Mo compounds, such as MoO 2 as the energy of these peaks overlaps directly with the peak energies of Mo S 2
73 [107, 111] Yet, their presence cannot be confirmed without further testing, with techniques such as secondary ion mass spectrometry. Figure 4 3 Comparison of Mo 3d core spectra of the A) reference sample and B) flown sample that shows all the available Mo has become MoO3 in the flown sample as opposed to a mixture of Mo/MoS2/MoO2 in the reference sample. Figure 4 4 S 2p core spectrum of the reference sample.
74 The S 2p core level spectrum of the reference sample is shown in Figure 4 4. A 1.1 eV splitting of the S 2p 3/2 and 2p 1/2 was employ ed, consistent with values listed in the literature  The peaks at energies of 161.9 eV and 163 eV were assigned to S associated with MoS 2 and accoun ted for the entirety of the intensity measured in this region [107, 109, 110] Figure 4 5 shows the Au 4f core spectra for the reference sample. The peaks at energies of 84.7 eV and 88.3 eV were determined to be metallic Au and agree with values found in the literature [107, 109] Figure 4 5 Au 4f core spectra of the reference sample. Chemical State of Flown Sample Referring back to Figure 4 2B, the O 1s /Sb 3d core spectrum from the flown sample is seen to differ significantly from the reference sample. As previously mentioned the regions for these two elements overlap and are challenging to assign. However, the flown sample proved to be simpler as fewer species than the reference sample were present. Sb 2 O 3 peaks were assigned energies of 530.3 eV and 539.7 eV,
75 maintaining a splitting of 9.4 eV  The area of the Sb 2 O 3 peak at 530.3 eV was determined by keeping the required 3:2 area ratio between d 5/2 and d 3/2 peaks  After fixing the area of the Sb peaks, the remainder of intensity centered at 530.3 eV. This peak was at tributed to O in both Sb 2 O 3 and MoO 3 which each exhibit an O 1s energy of 530.3 and are present in the material  The largest peak in the spectrum, l ocated at 531.9 eV, was assigned to a combination of O in SiO 2 and C O bonds These assignments correlate well to values found in the literature [107, 109] The presence of Si in the selvag e region of the flown sample is discussed below. The Mo 3d spectrum from the flown sample is shown in Figure 4 3B. As with the Sb spectrum, the Mo spectrum contains only two peaks with energies of 232.3 eV and 235.4 eV. Based on values found in the litera ture, these peaks are assigned to MoO 3 [57, 112] Neither S nor Au were detected in the near surface region of the flown sample (Table 4 1). Figure 4 6 Si 2p core spectra fro m the flown sample.
76 The XPS survey spectrum of the flown sample also contained a large Si peak that was not seen in the XPS of the reference sample. The Si core spectrum, shown in Figure 4 6, could be fit with to a single species by virtue of the single s ymmetric peak at an energy of 102.6 eV, in turn assigned to SiO 2 The SiO 2 has been detected in nearly all of the MISS E 7 tribology samples exposed to LEO, then analyzed using XPS, and is thought to be a contaminant [57, 112 114] The source of the contamination is unconfirmed, though there are a number of likely sources. Potential outgassing of Si based lubricants being used in other parts, contamination from the ISS itself, or space debris striking the surface could account for its presence [106, 113] Discussion As seen in the XPS results shown in Table 4 1, there is a tremendous difference between the surface composition of the flown sample and the reference sample. The most drastic change can be seen in the large increase in the amount of O detected on the surface. In the reference sample, only 22% O is seen and is attributed to C O bonds and Sb 2 O 3 However, after exposure to LEO, the amount of O nearly triples to 65%. W hile the main peak is primarily attributed to SiO 2 conversion of Mo and Sb species to MoO 3 and Sb 2 O 3 is evident and contributes to the increase. In the reference sample, Sb was present in both Sb 2 O 3 and a second form of sub oxide; however, in the flown sa mple only Sb 2 peak was detected, indicating that all Sb had been fully oxidized to the stoichiometric oxide. Further evidence of oxidation of the coating surface could be seen in the Mo 3d spectrum. The reference sample exhibited two peaks for Mo, attribut ed to MoS 2 and metallic Mo. After exposure, the flown sample exhibits only one species, which can only be attributed to MoO 3 As previously mentioned, the large amount of SiO 2 detected, which undoubtedly contributes to the large increase in O signal, stems
77 from contamination coming from some source in LEO. While this contamination partially obscures the true surface of the material, it does not detract from the fact that all of the components elements detected have been oxidized completely from their states in the reference sample. It cannot be ascertained if the Au in the coating has been oxidized, as there was no signal detected for the flown sample. Its absence is most likely a result of it being obscured by the presence of the large amount of contaminant SiO 2 Another noteworthy change following exposure to LEO is the removal of some species from the surface. The amount of C on the surface greatly differed between the two samples, dropping from 45% in the reference sample to 9% in the flown sample. This drastic decrease can be explained by the highly aggressive conditions in LEO. On the ram side of the ISS, where this sample was located, atomic oxygen collides with the surface at a rate of 10 14 10 15 atoms/cm 2 s at an energy of 4.5eV  This leads to what i s essentially a chemical etching of the surface by atomic oxygen. It should be noted that XPS is not sensitive to changes in topography, so it cannot show if any amount of material has been removed as the intensity detected comes from a thin cross section, approximately the top 10nm, of a very broad area, 1.75 x 2.75mm, regardless of the topography. The amount of C present in the reference sample is assumed to be completely adventitious C as there should be no C present in the coating itself. The atomic oxy gen striking the surface would react with adventitious C leading to the formation of CO 2 which would then volatilize leading to a drop in the amount of C detected. Sulfur is another element observed to decrease in atomic % after exposure to LEO. The atomi c percent of S decreased from 13% in the reference sample to 0% in the flown sample. Like C, the S present within the coating in the form of MoS 2 is thought to
78 react with the incident atomic oxygen to form MoO 3 and sulfur oxides. Both species are volatile and likely leave the surface as gaseous products. This reaction mechanism is consistent with the decrease of the concentration of Mo from 15% in the reference sample to just 2% in the flown sample, as well as the complete loss of S  These drastic changes have large implications to the tribological performance of this and similar coatings. The tribological properties of composite coatings have been shown to originate from a variety of surface active species, depending on environment (vacuum versus ambient pressure)  As previously discussed in the case of the MoS 2 /Sb 2 O 3 /Au coating, MoS 2 is intended to serve as a solid state lubricant while Sb 2 O 3 and Au play structural support and densifying roles  However, as seen in the results presented above, these compounds are no longer present in the near surface region probed by XPS, which has become dominated by oxides As a result, the tribological properties of the coating cannot be expected to perform as intended or predicted, based on testing in a terrestrial environment. Summary A MoS 2 /Sb 2 O 3 /Au coating was exposed to LEO for ~18 months dur ing tribological testing on board the MISSE platform on the ISS. This exposure has led to several drastic changes in the chemical nature of the near surface region of the coating. By comparison to a reference sample of identical original composition, it ha s been shown that the original surface, composed of a mixture of components, is transformed primarily to the oxides of each component. The reference sample included a native mixture of Mo, MoS 2 and MoO 2 while the flown sample exhibited only a singular mol ybdenum species MoO 3 after exposure. While the reference sample contained a mixture of Sb and Sb 2 O 3 the flown sample has been transformed completely to Sb 2 O 3 Furthermore,
79 it is noted that the coating composition is altered through the formation and rel ease of volatile reaction products, namely MoO 3 and SO 2 Finally, the aggressive conditions of LEO were seen to drastically reduce the amount of carbon observed on the surface to residual amounts These LEO ram results underscore the need for the protecti on of solid lubricant coatings from the aggressive action of atomic oxygen and highlight the need to design chemically robust alternatives.
80 CHAPTER 5 COMPOSITIONAL EFECTS OF EXPOSURE TO LOW EARTH ORBIT ON MoS 2 /YSZ/Au/DLC COATINGS Background Information A s detailed in C hapter 4 the MISSE program was created to give researchers a chance to expose materials designed for space applications to LEO conditions. make sure both the s amples and mountings can handle the tremendous forces are mounted to the outside of the station. One PEC is mounted on the ram (or leading) side of the station and the othe r is mounted to the wake (or trailing) side of the station. For the MISSE 129 on November 2009 and after an exposure of 18 months were retrieved during the STS 134 mission on May 2011  The MISSE 7 mission is unique in that it involved the first ever active experiment ever conducted on the outside of the ISS. The active experiment involved the use of small scale tribometers used to test the tribological activity of a variety of coatings and samples used as lubricants in space applications  While the tribological data is presented elsewhere, this report will instead focus on the changes seen in the chemical nature of the coatings due to exposure to LEO. The sam ples that are discussed in Chapter 5 represent a unique and complex material coating system. It was developed by the Air Force Research Labs (AFRL) for the purpose of being able to be tribologically active and provide low friction in a variety of environme nts ranging from terrestrial to LEO to pure vacuum and high temperature. There are solid lubricants that perform well in each of these environments individually such as MoS 2 in vacuum, graphitic carbon in terrestrial environments, and Au in high
81 temp. appl ications  But each of these materials performs poorly when introduced to other environments [5, 116] Therefore, AFRL sought to solve this problem of not having a universally beneficial tribological coating by combining a number of solid lubricants together. 2 yttria stabilized zirconia (YSZ), Au, and diamond like carbon (DLC). This coating has been shown to perform very well in differing environments, displaying a friction coefficient of 0.10 in air due to the graphitic C formed from DLC and 0.02 in a dry nitrogen environment due to MoS 2  Upon heating to 500C a friction coefficient of 0.15 was recorded, this was explained in terms of Au segregation to the surface in turn providing low friction  This coa ting was also unique in the MISSE 7 experiments in that a sample was compare how the sample changes chemically depending on its position on the space station. It is well do cumented that the environments in the ram and wake are very different. As discussed in C hapter 4 there is a high collision rate (10 14 10 15 atoms/cm 2 s) of high energy (4.5eV) atomic oxygen, as well as extreme temperature fluctuations ( 40 to 60C), exposure to cosmic radiation, and the danger of debris striking the surface  In contrast the wake side of the ISS is considered to be more like a pure vacuum environment. These differing environments offer the chance to study how the chemical nature of the vari ous constituents of the coating is affected. Like the Hohman coating covered in Chapter 4, this Chameleon coating is carefully engineered to have certain chemical components be tribologically active in these diverse environments. Any reactions that
82 occur w ith any of the components could lead to extreme changes in the tribological response of the coating. This could lead to the coating failing earlier than designed which in turn could lead to the failure of a part that uses the coating as a lubricant. As pre viously discussed, the atomic oxygen that strikes the surface on the ram side is composition by forming surface oxides or reacting to form gaseous species, thus directl y altering the surface composition To analyze how the chemical nature of the surface of these coating changed due to exposure to LEO, X ray photoelectron spectroscopy (XPS) was used. XPS has high energy sensitivity, which gives the ability to distinguish between different bonding environments of the individual species on the surface. This technique can also be used to calculate relative atomic percentages of the elements on the surface. Methods and Materials The sample reported on in this study is a MoS 2 /Y SZ/Au/DLC coating developed by the AFRL. The substrates used for the study were a 304 stainless steel disk. The coating was deposited using magnetron assisted pulse laser deposition (MSPLD), a transition layer of Ti was placed on substrates prior to coatin g and substrates were biased at 150V (with respect to ground) and heated to elevated temperatures to promote crystallization  Four section targets were used for laser ablation for the coatings with the targets consisting of two quarter sections of YSZ, a qua rter section of carbon, and a quarter section of MoS 2 The Au was deposited via magnetron sputtering. After coating, the disks were mounted to the previously mentioned tribometers, which samples were examined, the first was a sample that was not exposed to LEO but instead was kept in a
83 clean environment until XPS was performed after the return of the space samples. This sample is referred to as the reference sample. The second and third s amples are the two which were flown on the ISS, one on the ram side and one on the wake side. They will be referred to as the ram and wake samples respectively. XPS was performed with eV) monochromatic X ray sou rce using the process previously described in Chapter 2. Processing of the XPS spectra was performed using CasaXPS software (Casa Software Ltd.) and was performed using the methodology that was described in Chapter 2. Results One of the main advantages of using XPS to analyze the coatings is that the atomic species can be identified and quantified in the sample. But, a small drawback to XPS is its extreme surface sensitivity. While X rays penetrate the sample a large distance, the signal from XPS only comes from the top 10 surface. Therefore elemental information only pertains to this near surface region and cannot be thought of as representative of the bulk composition which can vary widely from surface composition. As discussed in th e methodology for XPS laid out in Chapter 2, the relative atomic percentages for the elements in the near surface region can be determined through the use of a curve fitting regime and atomic sensitivity factors. Table 5 1 lists the relative atomic percent ages of the elements detected with XPS in the near surface region of the reference, ram, and wake samples. One interesting thing to note is that many of the metals seen in the reference and ram samples are missing in the wake sample. In the wake sample, th e only elements detected were C, O, Si, Au and surprisingly F. The significance of these values and how the value change between samples will be further discussed later in C hapter 5
84 Table 5 1 Atomic percentages of elements in the near surface region of the reference, ram, and wake samples as determined by XPS. Element ( at % ) F Mo Zr O C S Si Au Ram 0 4 2 56 15 2 15 6 Wake 3 0 0 61 7 0 29 <1 Ref. 0 3 5 33 42 4 0 13 Chemical State of Reference Sample The Mo 3d spectrum for the reference sample is sho wn in Figure 5 1A. The spectrum features a doublet with peaks at energies of 232.9 eV and 236 eV. These peaks are attributed to the 3d 5/2 and 3d 3/2 of Mo bonded in MoO 3 [57, 112] The coating is meant to be deposited with MoS 2 being the desired Mo compound due to its frictional properties. However it is seen that the deposited MoS 2 has reacted with the Mo having completely oxidized, existing in its highest oxidation state. This d iffers from many of the coatings that have been made by the AFRL. In their own studies as well as studies done by others, the Mo was seen to be bonded as MoS 2 or MoS x compounds [116, 117] The explanation of th e observation that these samples had oxidized when as deposited coatings were seen to be un oxidized relates to the fact that the samples used in this study had been deposited over 18 months before being analyzed. Even in a clean environment there still is the potential for oxidation and given the long time the samples sat it is likely that the oxidation occurred naturally. YSZ is an important part of the coating as it is included to increase the hardness of the coating and the abrasion resistance [55, 116, 118] These factors help the coating hold up to the operating conditions experienced without having large amounts of wear occurring. Therefore it is important for YSZ to remain in its intended chemistry so as to
85 provide these benefits. The Zr 3d spectrum for the reference sample is seen in Figure 5 2A. There are two prominent peaks displayed, one at 182.5eV and the other at 185.0eV. They have the characteristic separation distance of ~2.5eV and are identifi ed as the 3d5/2 and 3d3/2 peaks for Zr bonded to O in ZrO2  These results coincide well with values seen in the literature for ZrO2 in YSZ [107, 119] Figure 5 1 Comparison of the Mo 3d core spectra of the A) reference and B) ram samples. The S 2p spectrum, shown in Figure 5 3A shows a singular peak at a binding energy of 162.9 eV. The peak is very broad with a FWHM of 3.5eV. The peak position initially would indicate that the peak should be assigned to S bonded to Mo in MoS 2 or some other MoS x compound [107, 109, 110] However, as seen from Figure 5 1A the Mo 3d spectrum shows that Mo is bonded as MoO 3 only wit h no signal seen at the peak position for MoS 2 Therefore since the Mo has obviously been oxidized it is reasonable to assume that the S which broke apart from the MoS 2 that was originally deposited
86 would have formed some new compound as well. Following th is likely scenario the peak seen in Figure 5 3A is attributed to atomic sulfur, with a binding energy of 162 164 eV depending on which literature values are considered and sulfur in the surface that has O atoms adsorbed, which in literature is assigned to a value of 163 eV [107, 120] . Figure 5 2 Comparison of the Zr 3d spectra for the A) reference and B) ram samples Figure 5 4A displays the Au 4f spectrum for the reference coating. The two main peaks detected are located at 84.1eV and 87.8eV. These peaks correspond to the 4f 7/2 and 4f 5/2 peaks for metallic Au. The peak positions agree very well with values found in t he literature for pure Au [107, 109] This is to be expected as it is very difficult to have reactions involving Au, its presence is important to the coating for providing low friction at e levated temperatures.
87 Figure 5 3 Comparison of the S 2p spectra for the A) reference sample and B) ram sample Figure 5 4 Comparison of the Au 4f spectra for the A) reference, B) ram, and C) wake samples.
88 The O 1s spectrum is shown in Figure 5 5A. There is one large peak with a significant shoulder to the higher binding energy side. This singular peak can be broken down into three smaller peaks. The first peak located at 530.6eV is attributed to a combination of the signals of O bonded to Zr in ZrO 2 and O bound to MoO 3 These values agree well with values found in the literature for ZrO 2 and MoO 3 [107, 109, 119] The assignment is further corroborated by the fact that the Mo 3d an d Zr 3d spectra are identified as only arising from as MoO 3 and ZrO 2 respectively. The second peak located at 532.0 eV is attributed to C O, which arises from adventitious carbon that has adsorbed to the surface during the period between deposition and ana lysis  The third peak which has a binding energy of 533.4 is attributed to C=O, also present due to the adsorption of adventitious carbon compounds on the surface of the coating from the ambient environment  Figure 5 5 Comparison of the O 1s spectra for the A) refere nce, B) ram, and C) wake samples.
89 Figure 5 6 Comparison of the C 1s spectra for the A) reference, B) ram, and C) wake samples. The C 1s spectrum for the reference sample, shown in Figure 5 6, shows one large peak at 284.8 eV with two smaller shoulders at the higher binding energies of 286.6 eV and 288.5 eV. The large peak is consistent with values in the literature for a carbon single bonded to another carbon  This peak is typically seen in most XPS spectra of samples exposed to ambient atmospheres, regardless of intended composition, due to adventitious adsorption. But this general peak shape has also been seen in DLC co atings with a predominant peak at 284.5 eV [116, 121] Other stud ies
90 have seen that DLC C C peaks tend to be between 284.5 eV, which is typical of a graphitic bonding, and 285.2 eV, which is seen in diamond bonding  While there is some discrepancy on where the main C C peak should be for DLC coatings, it is noteworthy that the main peak does not appear below 284 eV, as this would be indicative of unwanted carbide formation. The two shoulder peaks previously mentioned at binding energies of 286.6 eV and 288.5 eV are attributed to C O and C=O respectively  These two peaks are assigned to adventitious carbon that has adsorbed onto the surface of the coating after deposition. Chemical State of Ram Sample The Mo 3d spectrum for the sample flown on the ram side of the ISS is shown in Figure 5 1B. It shows two peaks with shoulders to lower binding energies. Using the fitting procedures discussed in Chapter 2, the overall intensity can be fitted with two Mo 3d doublets using the characteristic separation distance listed found in the literature  The larger doublet, with peaks at binding energies of 232.7 eV and 235.8 eV, is attributed to Mo bound to O in MoO 3 and similar to the spectrum for the r eference sample; these values agree well with those found in the literature [107, 109] The second smaller doublet, with peaks at energies of 231.6 eV and 234.7 eV, is attributed to sub sto ichiometric oxides (MoO x ) that formed under the high flux of atomic oxygen present in the ram environment. Figure 5 2B shows the Zr 3d spectrum for the ram sample. The spectrum displays two peaks at energies of 182.5 eV and 184.9 eV. The peaks were identi fied as the 3d 5/2 and 3d 3/2 for Zr bound to O in ZrO 2 As with the reference coating the values coincide with the characteristic separation distance of 2.4 eV for Zr doublets as well as literature values for Zr in ZrO 2 [107, 119] The S 2p spectrum for the ram sample is
91 shown in Figure 5 3B. There is one large peak in the spectrum at a binding energy of 168.8 eV. This peak is attributed to S bound in the form SO 2  This differs from the spectrum seen in Figure 5 3A in which the S was bonded as atomic sulfur or sulfur bonded to adsorbed oxygen. This result indicates that exposure to the aggressive ram environment has fully oxidized the S to its highest oxidation state. The Au 4f spectrum for the ram sample is shown in Figure 5 4B. Similar to the reference sample, the spectrum includes peaks at binding energies of 84.1 eV and 87.7 eV. Au is very diffic ult to oxidize and therefore it is not surprising to see that it remains in its pure elemental state. This is desirous as its elemental presence enables the coating to perform tribologically in high temperature environments. The O 1s spectrum seen in Figur e 5 5B, shows a significant change in the coating after exposure to the ram environment. The peak shaped has entirely changed with respect to the peak seen in the reference sample in Figure 5 5A. Instead of a large peak at a lower binding energy with a sho ulder leading to the higher binding energy side there is a larger peak at a higher binding energy with a shoulder to the lower binding energies. The main peak located at a binding energy of 532.2 eV is attributed to a combination of signal from O in SO 2 C O, and SiO 2 All three of these signals are to be expected at this binding energy and are indicated to be present in the surface region of the coating through the S 2p (Figure 5 2B), C 1s (Figure 5 6B), and Si 2p (Figure 5 7A) spectra. The value for this peak agrees well with values seen in the literature for all three species listed [107, 109] The second lower binding energy peak located at 530.4 eV is attributed to a combination of signa l from O bound in ZrO 2 MoO 3 and MoO x All three compounds are expected to contribute signal as they were present in the Mo 3d (Figure 5 1B) and Zr 3d
92 (Figure 5 2B) spectra. The values at which signal is detected coincides well with those in the literatur e  The C 1s spectrum for the ram sample, Figure 5 6B, is seen to have two dominant peaks at binding energies of 284.4 eV and 286.5 eV. The larger of the two peaks, which is located at 284.4 eV, is attributed to C C bonds. The second peak at a binding energy of 286.5 eV is identified as C O. These values, like the values for the 5 6A, coincide with values seen in t he literature [107, 121] It is worth noting that while the location of the peaks has not changed drastically, the ratio of the integrated intensity of the peaks has changed. The ratio of the integrated intensi ty of the C C to C O peak in the reference sample is 3.6 whereas in the ram sample the ratio has decreased to 1.55. This indicates that some portion of the DLC component of the coating has likely oxidized during exposure to the atomic oxygen present in the ram environment Finally, as seen from Table 5 1, the ram and wake samples contained Si which was not found in the near surface region of the reference coating. The possible origins of the Si seen in the two samples flown on the ISS will be discussed fur ther in a later section. The Si 2p spectrum is shown in Figure 5 7A and has one large peak located at a binding energy of 103.0 eV and is attributed to SiO 2 This value is concurrent with values seen in the literature for Si bound in SiO 2 The single peak indicates that all the Si is completely oxidized, existing in a single chemical state on the surface of the coating 
93 Figure 5 7 The Si 2p spectra for the A) ram sample and B) wake sample. No silicon was detected in the surface region of the reference sample. Chemical State of Wake Sample The wake sample varied quite drastically chemically from the reference and ram samp les. The only elements detected were O, C, Si, F, and a very small amount of Au The Au 4f spectrum seen in Figure 5 4C shows two peaks at binding energies of 84.1 eV and 87.8 eV. Au is the only element seen throughout all samples and the only element in w hich the spectrum for each sample was relatively consistent. This can be explained by a couple of factors relating to Au. Firstly the photoelectrons coming from Au will have the highest mean free path when compared to photoelectrons from the other elements This is determined because it has the lowest binding energy of any of the elements, according to E quation 2 1 calculating binding energy, this means it conversely the measured kinetic energy of the electrons is the highest, which gives it the highest mea n free path. The second factor is that the atomic sensitivity factor for Au
94 is the highest at 5.24 it is almost two times higher than the next closest ASF (Mo at 2.867). This means that the signal collected is 5.24 times higher than the actual amount in th e sample. Therefore even a small amount will shows a distinct peak. This means that the other elements deposited in the coating may be present but at atomic percentages that are two low to be detected. As with the reference and ram samples the two peaks de tected were attributed to elemental Au. The O 1s spectrum for the wake sample is shown in Figure 5 5C. The spectrum differs from the O 1s spectra seen in Figure 5 5A and 5 5B in that there only a single peak at a binding energy of 532.5 eV. The reference a nd ram samples both exhibited features representative of O in a variety of chemical environments. This single peak in this spectrum is attributed to O bound in SiO 2 and/or to C O bonds. These compounds are confirmed to be present through corresponding inte nsity appearing in the Si 2p spectrum (Figure 5 7B) and C 1s spectrum (Figure 5 6C). The binding energy values for the signals from these two compounds also agree with those found in the literature  The signal largely derives from SiO 2 as in Table 5 1 there is 29 at% Si in the near surface region; correspondingly 58 at% of the 61 at% detected arises from this compound. The fluorine seen in the wake sa mple arises from contaminants that have adsorbed onto the surface of the sample during flight. The F 1s spectrum, seen in Figure 5 8, shows a large broad peak that requires two species to accurately fit. The molecular identity of these peaks cannot be exa ctly determined due to the fact that the contamination could come from any number of sources. The most likely scenario to occur is that the contamination arises from the volatilization of some kind of fluorinated grease or the degradation of a solid sample
95 fluorinated greases are used in space applications. While it probable that the F seen is bound in some polymer, there is no indication in the C 1s spectrum for the wake sampl e, seen in Figure 5 6C. This is likely due to the fact that there is only a small amount of F detected. A calculation of the intensity of C signal that would be associated with a fluorinated species indicates that the signal would be indistinguishable from the background noise. Another possible situation is that there is some SiO x F y compound that has formed on the surface; yet, like the situation with fluorinated grease, the resulting signal that would be detected in the Si 2p spectrum is indistinguishable from the noise in the background. Figure 5 8 The F 1s spectrum for the wake sample. Discussion From Table 5 1, it is immediately apparent that a significant change in surface composition has occurred due to exposure to LEO. T he largest changes have occurred when comparing the wake sample to the ram and reference samples as the wake sample is seen only to have a few nonmetal constituents. While the ram and reference samples have similar elements in the near surface region there is still a difference in
96 how these elements are bonded that shows how the chemistry of the coating has been changed. The most apparent difference that can be noted is the large increase in the amount of oxygen seen in the ram and wake samples. The referen ce sample contained 33at% oxygen while the ram and wake samples contained 56at% and 61at% respectively. This increase in the amount of oxygen in the ram sample can be explained by two observations. The first is that the atomic oxygen present in the ram env ironment leads to the oxidation of all constituents of the coating to their largest oxidation state. This leads to an increase in the amount of oxygen present but more importantly rids the coating of the desired compounds that were selected for their tribo logical or strengthening properties. This oxidation is less evident than the oxidation seen in the Hohman coating discussed in Chapter 4, because some components of the Chameleon reference coating had already oxidized whereas all of the constituents of the reference Hohman sample showed no initial oxidation. This was indicated through the appearance of Mo being bound exclusively as MoO 3 in the reference sample. Whereas for the reference sample for the Hohman coating, Mo was shown to be bound as MoS 2 and ele mental Mo, then upon exposure was oxidized to MoO 3 The element that was seen to oxidize in the Chameleon coating was S. As previously discussed S was seen in the reference coating as being bound in MoS 2 but after exposure was found to be bound as SO 2 Thi s shows that any MoS 2 found on the surface had undergone chemical reactions to form oxides. The second reason for the increase in the amount of oxygen in the surface of the coating is the introduction of the contaminant SiO 2 In the ram coating there is 15 at% Si found in the surface. This means that 30 at% of the 56 at% O seen is bound in SiO 2 The origins of the SiO 2 will be discussed later in this section. The
97 increase in the amount of O in the wake sample, as previously mentioned can be attributed nearl y exclusively to the introduction of SiO 2 to the coating. As mentioned in the Results section the presence of 29 at% Si indicates that 58 at% of the 61at% of O being bound as SiO 2 The wake environment differs from the ram environment due to the fact that there is less atomic oxygen and less of a chance of collisions with space debris, as it is by definition on the trailing side of the ISS. Another interesting trend seen from the results in Table 5 1 is that some of the elements have decreased in atomic pe rcentage after being exposed to LEO. The amount of C seen in both the ram and wake samples was far less than the amount seen in the reference sample. The reference sample contained 42 at% C whereas the ram and wake samples only had 15 at% and 7 at% respect ively. This decrease could be due to a few different factors. For the ram side the decrease is most likely due to a combination of sputtering and etching by the atomic oxygen that is present. This reaction pathway leading to the decrease of surface C is su pported by the XPS results in Figure 5 6B in which the peak corresponding to C O has risen in intensity when compared to the same peak seen in the reference coating. This shows that either the DLC in the coating or adventitious carbon on the surface is rea cting to form more C O bonds, it is also likely that the formation of CO 2 would produce species leaving the surface as a gas and therefore not be seen in the XPS data. As for the wake sample, the lack of atomic oxygen in this environment means that another mechanism must be occurring to decrease the amount of carbon seen. One possible explanation for this is that the surface is being covered with the contaminant SiO 2 This means that the carbon is still present but is being covered by the large amount of co ntamination present. As
98 previously discussed nearly 90% of the signal coming from the surface of the wake sample is Si and O in SiO 2 therefore only places where the surface has a thinner layer of SiO 2 or adventitious carbon that has settled on the surface after return to the terrestrial environment is detected, producing the very low intensity of carbon detected. Alternatively, adventitious carbon species may have volatilized in the pure ultrahigh vacuum environment of the wake. The atomic percentage of ot her elements was lowered due to exposure to LEO as well. The amounts of Mo and Zr in the reference sample were 3 and 5 at% initially; after exposure the amounts of these two elements had changed to 4 and 2 at% respectively. The removal of these two element s is once again tied to the presence of large amounts of atomic oxygen in the ram environment. While the YSZ in the coating is already oxidized, the Mo present is deposited as MoS 2 Upon exposure to the chemically aggressive ram environment, there is likel y segregation of MoS 2 to the surface where it would undergo oxidation and volatilization. One of the more interesting things that can be noted from the XPS results is that both the ram and wake samples showed large amounts of Si in the near surface region after exposure on the ISS. From Table 5 1, it can be seen that there is no Si present, and from the literature and AFRL press releases, Si is not an intended component of the coatings. Therefore the Si must be coming from some external source and contamin ating the surface [57, 112 114] This Si contamination was seen in all of the MISSE 7 samples that had been exposed to LEO and further all of the Si is seen as b eing bonded as SiO 2 As discussed in Chapter 4 with the Hohman coatings, the exact origin of the Si contamination is unknown but there are several theories. The first being
99 that it comes from a Si based lubricant used in other MISSE samples or parts of the space station. Another theory is that it comes from the ISS itself where Si is a known component of some of the parts on the outside of the space station  A third a nd less likely theory is that the Si comes from space debris striking the surface and depositing small amounts of Si on the surface  This is unlikely though as Si is seen even on the wake samples where no such collisions would occur. Summary of Findings A MoS 2 /YSZ/Au/DLC coating was exposed to LEO orbit via the MISSE 7 mission on the ISS for a period of ~18 months. One sa mple of the coating was exposed on the ram side of the ISS, an aggressive chemical environment in which a high flux of atomic oxygen is present as well as space debris, micrometeorites, and galactic radiation. A second sample of the coating was flown on th e wake side of the ISS, where a more pure vacuum environment is present. Due to these exposures, the chemical nature of the near surface region of the samples was changed dramatically. The ram sample showed that the high flux of atomic oxygen had oxidized all the components of the coating completely, save for Au. The coating was designed for each of the original components to contribute tribologically to the coating under different environments. The breakdown and oxidation of these components could lead to the coating losing the tribological benefits of the constituents. This would lead to the coating failing prematurely and causing the failure of components where this coating is intended to provide solid lubrication. The ram sample contained more atomic pe rcentage of Mo and less atomic percentage of Zr than the reference sample. The reasoning amounts are reversed is due to Mo migrating to the surface to react with the atomic oxygen to form MoO 3 whereas the Zr is already oxidized as ZrO 2
100 The ram sample als o contained less S than the reference sample. This is due to the S oxidizing and forming gaseous compounds that then volatilized from the surface, leaving the coating with less S than was deposited. Furthermore, the amount of C seen in the surface of the r am samples decreased from 42 at% to 15 at%. This is due to adventitious carbon being either sputtered from the surface by the flow of atomic oxygen or reacting and volatilizing from the surface. The atomic percentage of O has also greatly increased. On the ram sample this is partially due to the oxidation of samples from atomic oxygen and also due to the introduction of the contaminant SiO 2 which greatly changed the surface of the samples flown on the ISS. Both the ram and wake samples have been contaminate d from outside sources during exposure. Both the ram and wake samples contained large amounts of Si, 15 at% and 29 at% respectively. The Si was seen by the XPS to be bound as SiO 2 consistent with the large increase in the amount of O detected in the sampl es that had been flown on the ISS. While the source of the Si remains unknown, it is important to note that much of the surface has been covered with this contaminant which could also attribute to further loss of the intended tribological properties of the coating. Furthermore the wake sample also contained a second contaminant, fluorine. The F signal seen in the XPS likely derives from the volatilization of fluorinated grease that is used in aerospace applications or from the degradation of other fluorinat ed polymer samples flown in the wake PEC on the ISS. But the conclusive evidence to take from the wake samples is that nearly the entire surface has become contaminated with SiO 2 and F containing contaminants. Without the knowledge of how fast the contamin ation occurs and how easy it is to remove, it is difficult to predict how the coating would be affected
101 tribologically. But it can be concluded that the sample will not initially perform as intended due to the contaminants masking the species that were int ended to dominate the tribological response of the coating. When comparing the Chameleon samples with the Hohman sample discussed in Chapter 4, a more complete picture of the environments that samples are exposed to in LEO can be seen. The Hohman coating only had one sample flown on the ram side, and similar to the Chameleon coating discussed in Chapter 5 showed oxidation of all of the coating constituents to their highest oxidation state. Also seen in the Hohman ram sample was the same SiO 2 contamination seen in both Chameleon coatings. This SiO 2 contamination has been in seen in the literature for other samples that have been flown on the ISS in other MISSE missions as well [57, 112, 113] The preferential removal of certain species also occurred in the Hohman coatings. The amount of C in the flown sample was far less than the reference coating. This was also the case for the Chameleon coating in which the amount of C in the samples flown on the ISS was far less than the reference coating. As previously mention though, a more complete knowledge of the environment experi enced by materials in LEO was discovered through the study of the Chameleon coatings. Though it was known that the environments differed between the ram and the wake side of the ISS, it was not known how materials would be changed after exposure. With the complete set of Chameleon samples we can see multiple sources of contamination that nearly completely obscure the surface of the experienced in the wake of the ISS.
102 With this ne w knowledge it can be concluded that though these materials may perform as intended in simulated LEO conditions, once the materials are actually sent into LEO the surface chemistry changes. The affects these chemical changes could have on the tribological properties of the coatings is undetermined. But, given the advanced nature of the coatings and the amount of time and care taken to formulate them and deposit them as formulated, any changes could drastically affect how the coating behaves.
103 CHAPTER 6 T RIBOLOGY OF CADMIUM SULFIDE Background Information The metal cadmium was discovered in 1817 as a c ontaminant in zinc compounds due to its rarity and difficulty in purification it would not be until 1820 that cadmium sulfide would be synthesized. As previou sly mentioned cadmium sulfide is primarily used as a pigment. In its crystalline state it can assume three different forms. The first and most common form of CdS found in nature is that of the mineral Greenockite, which exists in the hexagonal Wurtzite cry stal structure. The other natural form of CdS is the rare mineral Hawleyite which was only discovered in 1955. In this mineral CdS is found to be in the cubic zincblende structure. There is another non naturally occurring form of CdS, at high pressures CdS will transform to the cubic NaCl structure [122, 123] Synthetically, there are multiple ways to deposit CdS including chemical bath depos it i on ,sol gel technique, metal organic chemical vapor deposition, sputtering, electrochemical deposition, and screen printing [124 130] In C hapter 6 the frictiona l and wear properties of CdS will be investigated using linear reciprocating tribometry. The properties of CdS relating to tribology are essentially unknown and based on its similarities to other solid lubricants it presents a chance to broaden the knowled ge of the behavior of this material. Methods and Materials As mentioned in Chapter 1, there is virtually no knowledge of the tribological behavior of CdS despite the fact that based on similar materials there is the possibility of it exhibiting desirable t ribological properties such as a low coefficient of friction and
104 low wear rate Single crystal CdS is commercially available but only in smaller quantities as depositing large samples is difficult. Therefore to test this material new methodologies and equi pment had to be used. The sample used in this study was a randomly oriented single crystal CdS sample that was purchased from MTI Corporation. The dimensions of the sample measure 10 mmx10 mmx1 mm. The sample was polished to a surface roughness of R a <10 The single crystal CdS was deposited through physical vapor deposition in the hexagonal Wurtzite structure. The smaller dimensions of the sample did not allow for tribological testing on the home built tribometer described in Chapter 2 as there is no ab ility to change the diameter of the wear track made on the sample, which is larger than the CdS dimensions. Therefore, to test this sample, a linear reciprocating tribometer designed in the lab of Dr. Greg Sawyer at the University of Florida was used. This tribometer is unique in that it is designed for microscale contact and measurements. In this set up a variety of cantilevers of var ying stiffness and accommodating a wide variety of pins and counterfaces is employed. Capacitance probes that measure the di splacement of the cantilever combined with the calibration of the cantilever itself measure the normal and friction forces during testing. The sample is mounted on a motorized stage that provides the reciprocating motion. The smaller size of the tribometer allows for smaller and more between tests was the load. The counterface was a 1.5 mm radius Sapphire ball. Sapphire was chose due to the high modulus of elasticity 345 GPa which would guarantee that wearing of the CdS would occur before wearing of the ball. A new ball
105 was used for every test performed. The magnitude of the linear reciprocation was 600 and the number of cycles in each test was 10000. The samples were all run in under ambient temperature and pressure with the temperature being held at 68F and the relative humidity being approximately 54%. As previously stated the only variable that was changed was the load, normal force, applied during the test. Runs were performed at loads of 0.08 N, 0.125 N and 0.25 N. Using Hertzian contact mechanics, these loads correspond to contact pressures of 330 MPa, 380 MPa, and 481 MPa respectively. Due to the stiffness of the cantilever used to hold the 1.5 mm radius ball the load could not be increased past 0.25 N. Multiple loads were used to discern if any tribological effect could be seen from varying the load. To measure the wear rate of the CdS, atomic fo rce microscopy (AFM) scans were taken using an Asylum MFP 3D AFM imaged the entirety of the wear track width as well as some of the nascent surface on either side After scanning the wear region image processing was p erformed using SPIP software. Processing entailed planarizing and flattening the image and then averaging the line profile perpendicular to the we ar track for the entire scan. The averaged line scan data was then plotted using Origin. After plotting the da ta, a baseline corresponding to the plane of the nascent surface on either side of the wear track was established. Then the area between the data and the baseline, attributed to the wear track, was calculated using the integration tool in Origin. This area equaled the area of the material removed during tribological testing. Multiple scans of each wear track were performed and averaging the values of the area removed from each image gave an
106 average amount of material removed from the wear track. This value multiplied by the length of the wear track gave the volume of material removed by tribological testing. Using E quation 2 3 the wear rate of the material was calculated by dividing the volume of material removed by the normal force used and the total slidi ng distance. Results Friction o f Cadmium Sulfide Figure 6 1 shows the friction response of the CdS using an applied load of 0.08 N which corresponds to a contact pressure of 330 MPa. As previously stated, a sapphire ball of radius 1.5mm was used for the t est to prevent wear of the counterface interfering with the tribological testing. Figure 6 1 shows a brief run value decreases to approximately 0.16 0.17 and remains in this range for 2000 cycles. dily increase over the next 6000 cycles to a value of 0.21 where it remains from cycle number 8000 until the end of the test. As seen from the graph there is only one portion where a steady state is reached and that is the portion from cycle number 8000 10 000 where the value holds steadily around 0.21 0.21. The friction response of sapphire on CdS under a load of 0.125N (resulting in a contact pressure of 380 MPa) is shown in Figure 6 2. This graph shows a steep run in 0.17. Then at cycle ncrease continues for the remainder of the test upon which is finishes at a value near 0.25.
107 Figure 6 1. Graph of coefficient of friction vs. cycle number for CdS under a load of 0.08 N. Figure 6 2. Graph of coefficient of friction vs. cycle numbe r for CdS under a load of 0.125 N. Figure 6 3 shows the graph of coefficient of friction vs. cycle number for sapphire on CdS under a load of 0.25 N, which gives a contact pressure of 481 MPa. After run in the coefficient of friction begins to increase to a value of 0.23 at cycle number 2000. At this point the value oscillates around 0.23 0.25 until around cycle number 9500 at which
108 Figure 6 3. Graph of coefficient of friction vs. cycle number for CdS under a load of 0.25 N. Wear Behavior of Cadmium Sulfide As previously stated the wear rates of the CdS samples were calculated using AFM scans to determine the wear volume removed during tribological testing. For each Repre sentative images of the wear tracks all taken and courtesy of Alexander Rudy, and the averaged line profiles for the 0.08, 0.125 and 0.25 N tests are given in Figure s 6 4 through 6 6. Figure 6 4. N test wear tra ck and the averaged line profile of the scan (with units of nm on the y x axis) courtesy of Alexander Rudy
109 Figure 6 5. N test wear track and the averaged line profile of the scan (with units of nm on the y x axis) courtesy of Alexander Rudy Figure 6 6. N test wear track and the averaged line profile of the scan (with units of nm on the y x axis) courtesy of Al exander Rudy It can be seen from the AFM images that the wear tracks look like they were formed by gouging away of the materials as evidenced by the deep scratches seen. Table 6 1 shows the calculated wear rates for the three tests in the standard units of mm 3 /Nm.
110 Table 6 1. Calculated wear rates for CdS under varying loads. Load During Testing Calculated Wear Rate (mm 3 /Nm) 0.08 N 2.69x10 7 0.125 N 3.01x10 7 0.25 N 3.32x10 7 Discussion The frictional data reveals that despite similarities to other l ow friction solid lubricants, CdS does not display low friction. While the friction coefficient is not exceptionally high, with the highest value only reaching approximately 0.27; it cannot be classified as displaying low friction as this usually requires a coefficient of friction under 0.1. It is interesting that all three tests did not display an exact steady state friction. Even in portions where a plateau in coefficient of friction was observed, there would be strong local oscillations in the value from cycle to cycle. This phenomenon was only observed in the latter portion of all the tests. This could be due to wear debris being trapped in the wear track. As a result, the pressure on the wear debris would be increased as it slides across the surface unt il the debris is worn away and another piece of debris comes into contact and begins sliding. This mechanism would also be consistent with the deep gouges seen in the wear tracks as debris sliding back and forth would cause gouging of the surface. The wea r behavior of CdS is characterized by a desirable low wear rate. As previously discussed the mechanism for wearing appears to be gouging, but this only occurs after the initial formation of wear debris from the surface. The initial wear debris may be quite large as some of the gouges are very large and deep or, alternatively, this could be an effect of wear debris becoming trapped and gouging out more material inside the wear track. This being said a wear rate in the 10 7 mm 3 /Nm is considered to
111 fall in the range of low wear. In work performed by Krick and Sawyer, the wear rate of ionic solids including NaCl, CaF, ZnS, and FeS 2 were studied. From this study it was determined that wear rates of these materials were 1.9x10 5 2.9x10 7 1.0x10 7 and 3.0x10 8 mm 3 /Nm respectively  The results for the wear rates of CdS, the values of which can be seen in Table 6 1 show t hat they are comparable to those seen by Krick. Cadmium sulfide was chosen to be studied as a result of its similarities to other materials that display solid lubrication. First, other metal sulfides such as MoS 2 and FeS have shown to be low friction. ZnS while not displaying low friction has low wear properties. These materials, like CdS have crystal structures that have metal layers bonded to sulfur layers. But, they have several different bonding arraignments and crystal structures. The most common form of CdS is the wurtzite structure and the sample used in this study was the wurtzite form, while the common form of ZnS is the zincblende structure. CdS and Zn S are obviously the closer related of the materials as they have the same oxidation state (+2), s imilar ionic radii (88 pm for Zn and 109 pm for Cd), and very close electronegativit y values (1.65 for Zn and 1.69 for Cd) leading to very similar percent ionic character (19.44% for ZnS and 17.965% for CdS) These factors help to explain why the wear rate s of CdS are very similar to ZnS. As the ionic bond strength would be very similar therefore the difficulty to remove atoms from the surface of both materials would be very close. This would make both crystals wear in similar fashion and display wear rates on the same order of magnitude which was seen experimentally. Both FeS and FeS 2 are used a solid lubricants for brake pads and while they do not display friction under 0.1 they do provide increased lubrication over steel on steel contacts with a coefficie nt of friction around 0.24  FeS has the same crystal
112 structure as ZnS so this is yet another clue that the structure of CdS leads to increased coefficient of friction when comparing to other metal sulfide type materials. MoS 2 however exhibits a differ ent type of structure than CdS, ZnS, FeS, and FeS 2 having a layered structure where the molybdenum is bonded to a layer of sulfur atoms above and below. Adjacent sulfur layers are only weakly bonded to each other, thus providing the low shear surface for s liding. Previous studies have shown that while having low friction, the wear behavior of MoS 2 is interesting. It has been shown that initially the wear rate of MoS 2 is extremely high, where up to 90% of coating thickness will be worn away just in the run i n period, but after this point the wear behavior changes to an extremely low wear rate  At this point very thin films of MoS 2 will provide lubrication for long periods of time showing little wear. The reason for this is expla ined by columnar collapse of the MoS 2 structure until the film is thin enough that the reinforcement of the bonding to the substrate prevents its collapse. From this point on the sliding of the MoS 2 layers begins and low friction occurs. To prevent this in itial wearing away of the film the advanced MoS 2 coatings like those discussed in Chapters 4 and 5 were made and wear rates for those coatings are low throughout the entirety of testing  This behavior by MoS 2 is interesting in relating to CdS in that it shows that having low friction and low wear rate in a pure substance is extremely difficult to find. Often materials that are lower i n friction tend to wear more since it involves the material sliding over itself, whereas materials with lower wear often have higher friction as the material resists sliding. However the fact that a larger number of metal sulfides display lower friction wa rrants continued tribological investigation into other unexplored sulfur compounds.
113 Summary In light of the lack of knowledge of its tribological properties and its similarities to currently known solid lubricants, the friction and wear properties of sing le crystal hexagonal cadmium sulfide were studied using reciprocating tribometry. Using a sapphire pin, the CdS was investigated with loads of 0.08 N, 0.125 N, and 0.25 N. The resulting data showed that CdS did not display low friction, however following r un in, exhibited coefficient of frictions ranging between 0.2 0.27. The wear rate of the material determined from a quantitative assessment of the wear topography. The wear rates, all of which were close to 3x10 7 mm 3 /Nm, show that CdS has low wear propert ies similar to like materials. While the frictional properties of CdS are not low enough for it to be considered a solid lubricant, the wear rate is low where it could be considered for applications where a low wear coating is needed, as this is an applica tion for other ionic solids that exhibit low wear rates.
114 CHAPTER 7 CONCLUSIONS Solid lubricants are a class of materials which are growing in need of both quantity and the breadth of situations in which they can be used. Since the discovery of solid lub rication man y different materials have been used and studied for a wide variety of applications. Solid lubrication is needed in applications where traditional liquid lubrication cannot be used, such as vacuum or space applications, extremely dry and arid e nvironments, and elevated temperatures. As most solid lubricants have been developed empirically, often entailing mere testing many materials in a variety of situations and environments until the material with the best properties in that situation is found A large fundamental base of knowledge on how the lubricants operate or how they are affected by their environment or processing conditions for many lubricants does not exist Also based on performance trends and limitations seen in current solid lubrican ts it is important to discover new lubricants and improve upon older systems to tailor the tribological response for the advanced needs of ic and industrial applications. The preceding dissertation addressed this need for fundamental knowle dge of how environment and processing conditions affect the behavior of solid lubricants for a number of systems. This was done using x ray photoelectron spectroscopy to study the chemical nature of the surface region pin on disc tribometry to determine t he coefficients of friction, and multiple techniques to determine the wear rates of materials. In Chapter 3 a study was presented on an electrolessly deposited nickel boride coatings which were annealed in oxygen at various temperatures and the effects on the chemical and tribological effects of the annealing was investigated.
11 5 C hapters 4 and 5 introduced a study of the chemical effects of exposure to low earth orbit on two different advanced MoS 2 based coatings one being a MoS 2 /Sb 2 O 3 /Au coating and the oth er a MoS 2 /YSZ/DLC/Au coating. Chapter 6 introduced a novel study on the frictional properties of cadmium sulfide an ionic solid that had previously been uninvestigated in the literature for its frictional response. Electroless nickel boride coatings repre sents an emerging class of solid lubricant that are interesting in that they are cheaper and easier to produce and suffer from none of the negative effects seen when using other deposition methods for nickel boride coatings The coatings are annealed at a specific temperature, empirically determined to produce coatings with the desired microstructure while having the highest hardness value and desired tribological properties. In C hapter 3 the effect of changing the annealing temperature, under a flow of oxy gen, on the chemical state of a set of commercially deposited nickel boride coatings, the coefficient of friction, and the wear rate was described It was discovered that little effect was seen on the coating when annealed below temperatures of 400C. The as received, 250C, and 400C samples all showed the nodular structure that is typically seen in electroless NiB coatings and showed similar tribological characteristics. The counter face used were 440C stainless steel balls, and the friction coefficients of as received, 250C, and 400C coatings were all high, above 0.4, which in most applications will provide an improvement over metal on metal contacts which would be present without the coating, however higher than typically desired. The wear rates of the lower temperature annealed samples were also similar being approximately 4x10 8 mm 3 /Nm. However upon annealing at 550C and 700C the properties of the coating changed completely. The nodular microstructure
116 was replaced with a flaked microstructure and co efficients of friction were seen to be 0.16 for 550C and 0.06 for 700C. The wear rates of these two samples increased to values of ~ 2x10 7 mm 3 /Nm. The reason for this transformation in behavior was attributed to the formation of a thick layer of B 2 O 3 w hich upon exposure to humid air underwent a reaction to for the lubricious compound H 3 BO 3 This was confirmed by Raman spectroscopy and a pin on disk test where the relative humidity was modulated between high and low humidity This knowledge of that fact that annealing at higher temperatures causes the surface to be fundamentally changed from the as deposited NiB 3 to B 2 O 3 which in humid air will spontaneously form H 3 BO 4 can shape how this coating could potentially be used and processed. If the coating is going to be used for tribological purposes in a humid environment it would actually be beneficial to anneal above the temperature needed to form B 2 O 3 so that the lubricious boric acid can form and provide a low friction coefficient. Also this can effect p rocessing in that knowing how annealing at a certain temperature effects the properties of the coating. In C hapters 4 and 5 an extremely unique experiment was performed. MoS 2 is a solid lubricant that has long been known to perform extremely well in vacuum This has led to it widely being used in aerospace applications, and especially in parts put into low earth orbit However, MoS 2 suffers in terrestrial environments and at elevated temperatures. Therefore, to try to improve the performance and application s in which MoS 2 can be utilized, advanced coatings are produced in which MoS 2 is co deposited with other materials that improve various properties of the coating or provide for lubrication in environments where MoS 2 is deficient. A large problem with desig ning these coatings is simulating and testing in environments similar to what will actually be
117 experienced in low earth orbit However the MISSE 7 mission allowed for an opportunity for a number of samples to actually undergo tribologically testing in low earth orbit The tribological data is not presented in this current work, but the effects of exposure to low earth orbit on the surface chemistry of two sets of samples were studied. The experiments were performed on samples that were returned from orbit u nder extreme scrutiny and sample cleanliness procedures due to the rarity of the samples. Samples were mounted in passive experiment containers on either side of the International Space Station. The leading side or ram side of the space station involves ex posure to an aggressive flux of high energy oxygen ions, large temperature shifts, and exposure to micrometeoroid impacts. The trailing or wake side of the space station is a more pure vacuum environment. T wo sets of coatings were examined, a MoS 2 /Sb 2 O 3 /A u coating produced by the Hohman Plating Co. and a MoS 2 A ir Force Research Lab Reference coatings that had remained in terrestrial environments were investigated as well, with the Hohman coating only havi ng one sample that was mounted in the ram passive experiment container The chameleon coating had a sample mounted in both the ram and wake passive experiment containers The samples were investigated using x ray photoelectron spectroscopy to determine how exposure had changed the composition of the surface of the coatings. The results revealed a number of interesting conclusions I n both samples that had been mounted on the ram side of the station, all components of the coatings except for Au, had been ox idized to their highest oxidation state with no traces of the original compounds that had been deposited in the reference samples. The exposed samples
118 also had decreased amounts of adventitious carbon and the amounts of some elements detected in the refere nce samples were decreased. This was attributed to sputtering or etching by the high flux of atomic oxygen in the ram environments and the masking of the coating by contaminants in the singular wake sample. In addition, all of the samples that had been exp osed to low earth orbit had large amounts of SiO 2 on their surfaces. This was attributed to an undetermined source of contamination but it has been seen in other studies where samples were exposed to low earth orbit Also the chameleon sample that had been mounted in the wake passive experiment container showed a second contaminant ; F was detected by x ray photoelectron spectroscopy and its presence was also attributed to either the volatilization of a fluorinated grease or some other F containing sample in the more pure vacuum environment of the wake. All of these changes to the chemical nature of these coatings undoubtedly would undoubtedly have a negative effect on the tribological properties of the coatings. These coatings are specifically tailored for t he constituents to provide lubrication in different environments or under certain conditions The x ray photoelectron spectroscopy data showed that all of the MoS 2 in the surface region had been oxidized to MoO 3 which has poor frictional properties. This study therefore helps to understand how actual exposure to the environment where these coatings are expected to perform changes the coatings and will negatively affect the tribological properties of the coatings because in most cases oxides have higher fri ction coefficients and worse wear behavior. These negative impacts can lead to severe problems in cases such as space applications where the parts cannot be easily serviced.
119 Chapter 6 describes the initial study of the tribological properties of cadmium su lfide, an ionic solid with the hexagonal wurtzite crystal structure. C admium sulfide h as been investigated due to its similarities to other metal sulfide based solid lubricants. CdS is polymorphic with the more common form being the wurtzite structure, but it also has two cubic forms the first being the naturally occurring zincblende structure and a high pressure induced NaCl structure. This is opposite of the similar ZnS in which its more common form is the zincblende structure with the wurtzite being the less common MoS 2 and iron sulfides display low friction and are used extensively as solid lubricants however exhibit distinctly different structures. By studying metal sulfides of different structures, the fundamental nature of how crystal structure aff ects the frictional response of a material can be evaluated. S ingle crystal CdS was tested on a linear reciprocating tribometer, sliding against a 1.5mm sapphire ball with a speed of 400C 0.08N, 0.125N, and 0.25N leading to contact pressures of 330 MPa, 380M Pa, and 481MPa respectively The particular experimental set up due to the smaller dimensions of the sample prevented higher loads from being used. The frictional data revealed that CdS does not exhibit low friction under these sliding conditions All th ree tests indicated friction coefficient values between 0.2 0.3, yet steadily increasing with sliding distance. The wear rate of each test was calculated from atomic force microscopy scans to measure the amount of material removed in the wear tracks, and from knowledge of the normal force and the total sliding distance of the test. The wear rates of all three tests were calculated to be approximately 3x10 7 mm 3 /Nm. This is a relatively low value and is consistent with values seen in the literature for oth er similar ionic solids. While not a
120 success in terms of finding a new solid lubricant, the push for trial and discovery of materials for solid lubricity is important. These tests provided insight into the fact the presence of S in a structure does not lea d to lubricity but rather there must be a combination of both chemistry and crystallographic structure to achieve solid lubricity. Testing materials similar to those that exhibit solid lubricity is key in trying to find a way to establish what causes mater ials to have solid lubricity and identify materials that could provide interesting tribological characteristics. The field of solid lubrication has been growing since the discovery of solid lubrication itself and will only continue to grow as the number o f applications for solid lubrication increases. The growth and maturity of MoS 2 based lubricants serves as a good example of how the field of solid lubricity is moving more away from the discovery of new materials and into the refining and improvement of e xisting materials. However as mentioned numerous times previously in this document, the fundamental knowledge of man y solid lubricants remains relatively unknown and un explored due to the empirical and industrial drive behind solid lubricant innovation. Ho wever, fundamental knowledge such as that gained through the studies describes in this dissertation can help to provide information on how to improve the performance and properties of solid lubricants B y discovering how they function under sliding conditi ons their methods of failure, and the relevant wear mechanisms it will be possible to better determine how to pr ovide successful solid lubrication schemes required for future applications. Surface science and surface based techniques can be of particular usefulness. Lubrication is a phenomenon that occurs on the surface of materials and understanding more about the surface can yield important insight into solid lubrication. The use of x ray
121 photoelectron spectroscopy in particular is useful in that its on ly senses the top 1 10nm of a material. This gives extreme depth sensitivity and gives the exact knowledge of what chemical compounds are present on the surface of a material. Knowing this can help to see what really is interacting tribologically. Most mat erials will have some sort of passivation on the surface, whether it is a relaxation, reconstruction, oxidation, or adsorption of molecules to lower surface energy. These often result in the surface being chemically different from what the bulk of the mate rial even if only in very small layers on the surface. Similar studies to the one described in C hapter 4 and 5, where a solid lubricant was exposed to an extreme environment and then studied after by x ray photoelectron spectroscopy to determine how the i ntended environments changed the material are important and should be considered for a much larger variety of solid lubricants. As seen from the measured results the exposure to low earth orbit had a dramatic effect on the surface of the Hohman and Chamel eon coatings w h ere almost all of the deposited components of the advanced MoS 2 coatings were affected. As the applications and environments in which solid lubricants are used continue to grow it will be important to understand how the intended operational environments will affect the ir performance Too often components are designed to work in an application based on discussion in the literature on optimal performance, but little is known on the effects of the environment after exposure or the effects of ope ration on the surface chemistry of the component X ray photoelectron spectroscopy studies such as the one s performed on the MISSE 7 samples can reveal that the surface is completely different than the intended composition after exposure. This information can be used to try to better
122 develop coatings or add preventative measure s needed to increase performance even beyond what is originally intended. Vacuum tribology is a growing area of testing that can be easily integrated with other vacuum techniques to perform complete studies where samples do not have to leave the vacuum environment. In the Perry lab at the University of Florida the microtribometer used in the study of NiB is housed in a vacuum chamber attached to the main x ray photoelectron spectrosc opy vacuum system through a number of load locks U sing transfers arms the samples can be transferred directly from the tribometer to the x ray photoelectron spectroscopy analysis chamber. This opens up a large possibility of measurements that can be accom plished. Certain materials can undergo transformations under contact pressures, such as the transition of CdS to the NaCl structure, or can undergo tribochemical reactions where tribological interaction induces a chemical reaction. Using the vacuum tribome ter, samples such as these can be evaluated in the tribometer and then analyzed using x ray photoelectron spectroscopy to confirm such chemical changes without exposure to air where adventitious carbon could deposit or oxidation could occur. Also the chamb er where the tribometer is mounted includes a dosing system so that the sample can be tested in other environments besides vacuum or ambient air. The doser also allows for liquid lubricants to be specifically dosed onto the sample as it is running. As prev iously mentioned the chamber can be evacuated and then directly introduced to the x ray photoelectron spectroscopy analysis chamber removing the chance of contamination from ambient air.
123 Such a system can aid in the discovery of new lubricants. Many kno wn and potential solid lubricants are only tested in ambient conditions or under very specific operational conditions. Therefore, the number of solid lubricants known to perform well across many environments and the mechanisms by which they successfully pe rform is extremely small. The investigation of a broader class of lubricants under vacuum could potentially reveal new and interesting tribological properties that remain unknown simply due to the materials not being tested. This is an area in which vacuum tribologically can really prove its worth through the discover y of new materials for use in vacuum environments. As humanity continues to explore the universe around us the need s for additional and more technically advanced lubricants to use in space wil l continue to grow Vacuum tribology is well positioned to lead such innovation and through the discovery of new materials enabling the operation of moving mechanical assemblies of space applications.
124 LIST OF REFERENCES 1. Hilton, M.R., Fleischauer, P.D.: Applications of Solid Lubricant Films in Spacecraft. Surf. Coatings Technol. 55, 435 441 (1992). 2. Allen, S.M., Thomas, E.L.: The Structure of Materials. John Wiley and Sons (1999). 3. Muratore, C., Voe vodin, A.A.: Chameleon Coatings: Adaptive Surfaces to Reduce Friction and Wear in Extreme Environments. Annu. Rev. Mater. Res. 39, 297 324 (2009). 4. Winer, W.O.: Molybdenum Disulfide as a Lubricant: A Review of the Fundamental Knowledge. Wear. 10, 422 45 2 (1967). 5. Miyoshi, K.: Solid Lubrication Fundamentals and Applications. Marcel Dekker, New York (2001). 6. Grattan, P.A., Lancaster, J.K.: Abrasion By Lamellar Solid Lubricants. Wear. 10, 453 468 (1967). 7. Deacon, R.F., Goodman, J.F.: Lubrication by Lamellar Solids. Proc. R. Soc. Lond. A. Math. Phys. Sci. 243, 464 482 (1958). 8. McCook, N.L., Burris, D.L., Dickrell, P.L., Sawyer, W.G.: Cryogenic Friction Behavior of PTFE based Solid Lubricant Composites. Tribol. Lett. 20, 109 113 (2005). 9. Riedel, A.: Electroless Nickel Plating. London: Finishing Publication, London (1989). 10. Brenner, B.A., Riddell, G.E.: Nickel Plating on Steel by Chemical Reduction. J. Res. Natl. Bur. Stand. (1934). 37, 31 34 (1946). 11. Sherwood, B.: Serendipity Produces a R adically New Plating Process: Electroless Nickel. Met. Finish. 106, 92 94 (2008). 12. Alloy Phase Diagrams. ASM Handbook Online. pp. p. 2.8 9.8. ASM International (1992). 13. Vitry, V.: Electroless Nickel Boron Deposits: Synthesis, Formation and Characte rization; Effect of Heat Treatments; Analytical Modeling of the Structural State., (2009). 14. Ohno, I.: Electrochemistry of Electroless Plating. Mater. Sci. Eng. A. 146, 33 49 (1991).
125 15. Ohno, I., Haruyama, S., Wakabayashi, O.: Anodic Oxidation of Redu ctants in Electroless Plating. J. Electrochem. Soc. 132, 2323 2330 (1985). 16. Delaunois, F.: Chemical Nickel Boron Deposits on Aluminum Alloys, (2002). 17. Delaunois, F., Petitjean, J.., Lienard, P., Jacob Duliere, M.: Autocatalytic electroless nickel b oron plating on light alloys. Surf. Coatings Technol. 124, 201 209 (2000). 18. Copeland, P.L., Glasoe, G.N., Johnson, R.P., Kennard, R.B.: No Title. Proceedings of the American Physical Society. pp. 913 945. Washington D.C. (1941). 19. Johnson, R.L., G odfrey, D., Bisson, E.: Friction of Solid Films on Steel at High Sliding Velocities. Cleveland, OH (1948). 20. Haltner, A.J., Oliver, C.S.: Effect of Water Vapor on the Friction of Molybdenum Disulfide. Ind. Eng. Chem. Fundam. 5, 348 355 (1966). 21. Ro ss, S., Sussman, A.: Surface Oxidation of Molybdenum Disulfide. J. Phys. Chem. 59, 889 892 (1955). 22. Spalvins, T.: Energetics in Vacuum Deposition Methods for Depositing Solid Film Lubricants. Lubr. Eng. 25, 436 441 (1969). 23. Holsinski, R., Gansheime r, J.: A Study of the Lubricating Mechanism of Molybdenum Disulfide. Wear. 19, 329 342 (1972). 24. Stupp, C.: Effects of Metals Co Sputtered with MoS2. Thin Solid Films. 84, 257 266 (1981). 25. Fleischauer, P.D., Bauer, R.: Chemical and Structural Effect s on the Lubrication Properties of Sputtered MoS2 Films. Trbiology Trans. 31, 239 250 (1988). 26. Spalvins, T.: Frictional and Morphological Properties of Au MoS2 Films Sputtered From A Compact Target. Thin Solid Films. 118, 375 384 (1984). 27. Simmonds, and Tribological Performance of MoS[sub x]/Au Co Sputtered Composites. J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 19, 609 (2001). 28. Lince, J.R., Hilton, M.R.: Metal Incorporation i n Sputter Deposited MoS2 Films Studied by Extended X ray Absorption Fine Structure. J. Mater. Res. 10, 2091 2105 (1995).
126 29. Stoyanov, P., Chromik, R.R., Goldbaum, D., Lince, J.R., Zhang, X.: Microtribological Performance of Au MoS2 and Ti MoS2 Coatings w ith Varying Contact Pressure. Tribol. Lett. 40, 199 211 (2010). 30. Lince, J.R., Kim, H.I., Adams, P.M., Dickrell, D.J., Dugger, M.T.: Nanostructural, electrical, and tribological properties of composite Au MoS2 coatings. Thin Solid Films. 517, 5516 5522 (2009). 31. Wahl, K.J., Seitzman, L.E., Bolster, R.N., Singer, I.L.: Low Friction, High Endurance, Ion Beam Deposited Pb Mo S Coatings. Surf. Coatings Technol. 73, 152 159 (1995). 32. Wahl, K.J., Dunn, D.N., Singer, I.L.: Wear Behavior of Pb Mo S Solid L ubricating Coatings. Wear. 230, 175 183 (1999). 33. Simmonds, M.C., Swygenhoven, H. Van, Pfluger, E., Savan, A., Hauert, R., Knoblauch, L., Mikhailov, S.: Magnetron Sputter Deposition and Characterisation of Ti/TiN, Au/TiN and MoSx/Pb Multilayers. Surf. C oatings Technol. 94 95, 490 494 (1997). 34. of Magnetron Sputter Deposited MoSx/Metal Multilayers. Thin Solid Films. 354, 59 65 (1999). 35. Zabinski, J.S., Donley, M.S., Dyhou se, V.J., McDevitt, N.T.: Chemical and Tribological Characterization of PbO MoS2 Films Grown by Pulsed Laser Peposition. Thin Solid Films. 214, 156 163 (1992). 36. Rigato, V., Maggioni, G., Patelli, A., Boscarino, D., Renevier, N.M.: Properties of Sputter Seposited MoS2 Metal Composite Coatings Deposited by Closed Field Unbalanced Magnetron Sputter Ion Plating. Surf. Coatings Technol. 131, 206 210 (2000). 37. Zabinski, J.S., Donley, M.S., Walck, S.D., Schneider, T.R., Mcdevitt, N.T.: The Effects of Dopant s on the Chemistry and Tribology of Sputter Deposited MoS2 Films. Tribol. Trans. 38, 894 904 (1995). 38. Teer, D.G., Hampshire, J., Fox, V., Bellido Gonzalez, V.: The Tribological Properties of MoS2/Metal Composite Coatings Deposited by Closed Field Magne tron Sputtering. Surf. Coatings Technol. 94 95, 572 577 (1997). 39. Fox, V.C., Renevier, N., Teer, D.G., Hampshire, J., Rigato, V.: The Structure of Tribologically Improved MoS2 Metal Composite Coatings and Their Industrial Applications. Surf. Coatings Te chnol. 116 119, 492 497 (1999).
127 40. Savan, A., Haefke, H.: MoS2 Based Alloys and Nanocomposites for Solid Lubrication. Lubr. Sci. 16, 229 238 (2004). 41. Deepthi, B., Barshilia, H.C., Rajam, K.S., Konchady, M.S., Pai, D.M., Sankar, J., Kvit, A. V.: Struc ture, Morphology and Chemical Composition of Sputter Deposited Nanostructured Cr WS2 Solid Lubricant Coatings. Surf. Coatings Technol. 205, 565 574 (2010). 42. Kao, W.H., Su, Y.L.: Optimum MoS2 Cr Coating for Sliding Against Copper, Steel and Ceramic Ball s. Mater. Sci. Eng. A. 368, 239 248 (2004). 43. Hilton, M.R., Jayaram, G., Marks, L.D.: Microstructure of Cosputter Deposited metal and oxide MoS2 Solid Lubricant Thin films. J. Mater. Res. 13, 1022 1032 (1998). 44. ., Swygenhoven, H. Van: Mechanical and Tribological Performance of MoS2 Co Sputtered Composites. Surf. Coatings Technol. 126, 15 24 (2000). 45. Dickinson, R.G., Pauling, L.: The Crystal Structure of Molybdenite. J. Am. Chem. Soc. 45, 1466 1471 (1923). 46. Levy, F. ed: Crystallography and Crystal Chemistry of Materials with Layered Structures. D. Reidel Publishing Company, Dordrecht, Holland (1976). 47. R.M.A, L. ed: Preparation and Crystal Growth of Materials with Layered Structures. D. Reidel Publishing Company, Dordrecht, Holland (1977). 48. Centers, P.W.: The Role of Oxide and Sulfide Additions in Solid Lubricant Compacts. Tribol. Trans. 31, 149 156 (1988). 49. Zabinski, J.S., Donley, M.S., Mcdevitt, N.T.: Mechanistic Study of the Synergism between S b203 and MoS2 Lubricant Systems Using Raman Spectroscopy. Wear. 165, 103 108 (1993). 50. Swygenhoven, H. Van: Morphology and Tribological Properties of metal (oxide) MoS2 Manostructured Multilayer Coatings. Surf. Coatings Technol. 105, 175 183 (1998). 51. H u, J.J., Bultman, J.E., Zabinski, J.S.: Microstructure and lubrication mechanism of multilayered MoS2/Sb2O3 thin films. Tribol. Lett. 21, 169 174 (2006). 52. Voevodin, A., Zabinski, J.: Nanocomposite and Nanostructured Tribological Materials for Space App lications. Compos. Sci. Technol. 65, 741 748 (2005).
128 53. Pearson, J.D., Zikry, M. a., Wahl, K.J.: Microstructural Modeling of Adaptive Nanocomposite Coatings for Durability and Wear. Wear. 266, 1003 1012 (2009). 54. Hu, J., Muratore, C., Voevodin, A.: Si lver Diffusion and High Temperature Lubrication Mechanisms of YSZ Ag Mo Based Nanocomposite Coatings. Compos. Sci. Technol. 67, 336 347 (2007). 55. Voevodin, A.A., Hu, J.J., Fitz, T.A., Zabinski, J.S.: Tribological Properties of Adaptive Nanocomposite Coa tings Made of Yttria Stabilized Zirconia and Gold. Surf. Coatings Technol. 146 147, 351 356 (2001). 56. Gardos, M.N.: The Synergistic Effects of Graphite on the Friction and Wear of MoS2 Films in Air. Tribol. Trans. 31, 214 227 (1988). 57. Tagawa, M., Yokota, K., Matsumoto, K., Suzuki, M., Teraoka, Y., Kitamura, A., Belin, M., Fontaine, J., Martin, J. M.: Space Environmental Effects on MoS2 and Diamond like Carbon Lubricating Films: Atomic Oxygen Induced Erosion and Its Effect on Tribol ogical Properties. Surf. Coatings Technol. 202, 1003 1010 (2007). 58. Zaidi, H., Mezin, A., Nivoit, M., Lepage, J.: The Influence of the Environment on the Friction and Wear of Graphitic Carbons: Action of Atomic Hydrogen. Appl. Surf. Sci. 40, 103 114 (19 89). 59. Lockemann, G., Oesper, R.E.: Friedrich Stromeyer and the History of Chemical Laboratory Instruction. J. Chem. Educ. 30, 202 204 (1873). 60. Huggins, M.L.: Evidence from Crystal Structures in Regard to Atomic Structures. Phys. Rev. 27, 286 297 (1 926). 61. Frerichs, R.: The Photo 594 601 (1947). 62. Apker, L., Taft, E.: Field Emission from Photoconductors. Phys. Rev. 88, 1037 1038 (1952). 63. Klick, C.C.: Luminescence and Photoconductivity in Cadmium Sulfide at the Absorption Edge. Phys. Rev. 89, 274 277 (1953). 64. Smith, R.W., Rose, A.: Space Charge Limited Currents in Single Crystals of Cadmium Sulfide. Phys. Rev. 97, 1531 1537 (1955). 65. Shulman, C.I.: Measurement of Shot Noise in CdS C rystals. Phys. Rev. 98, 384 386 (1955). 66. Guha, S.: High Field Distribution Function and Mobility in n Type Cadmium Sulphide. Phys. Rev. B. 2, 4971 4977 (1970).
129 67. Vankateswaran, U., Chandrasekhar, M., Chandrasekhar, H.R.: Luminescence and Raman Spect ra of CdS under Hydrostatic Pressure. Phys. Rev. B. 30, 3316 3319 (1984). 68. Kroger, F.A., Diemer, G., Klasens, H.A.: Nature of an Ohmic Metal Semiconductor Contact. Phys. Rev. 103, 279 (1956). 69. Giaever, I.: Photosensitive Tunneling and Superconducti vity. Phys. Rev. Lett. 20, 1284 1286 (1968). 70. Giaever, I., Zeller, H.R.: Optical Phonons in Some Very Thin II VI Compound Films. Phys. Rev. Lett. 21, 1385 1388 (1968). 71. Mahapatra, A., Kornreich, P.G., Kowel, S.T.: Strain Induced Modulation of Photo conductivity in Thin Polycrystalline Films of Cadmium Sulfide. Phys. Rev. B. 18, 2766 2779 (1978). 72. Stephens, R.B.: Photoluminescence Determination of Minority Carrier Kinetics in Semiconductors. Phys. Rev. B. 29, 3283 3292 (1984). 73. Chuu, D.S., Dai C.M.: Quantum Size Effects in CdS Thin Films. Phys. Rev. B. 45, 11805 11810 (1992). 74. Sciacca, M.D., Mayur, A.J., Oh, E., Ramdas, A.K., Rodriguez, S., Furdyna, J.K., Melloch, M.R., Beetz, C.P., Yoo, W.S.: infared Observation of Transverse and Longitud inal Polar Optical Modes of Semiconductor Films: Normal and Oblique Incidence. Phys. Rev. B. 51, 7744 7752 (1995). 75. Marchman, K.R., Krick, B.A., Harris, K., Sawyer, W.G., Sinnott, S.B., Phillpot, S., Perry, S.S.: Examining and Predicting Wear in Ionic Solids with In Situ Scanning White Light Interferometry. Society of Tribologists and Lubrication Engineers Annual Meeting. Detroit, Michigan (2012). 76. Bangert Jr., C.E.: The State of the Global Coatings Industry. The Coatings Summit. pp. 15 17. Muni ch, Germany (2009). 77. Das, S.K., Sahoo, P.: A Parametric Investigation of the Friction Performance of Electroless Ni B Coatings. Lubr. Sci. 23, 81 97 (2011). 78. Das, S.K., Sahoo, P.: Roughness Optimization of Electroless Ni B Coatings Using Taguchi Me thod. Int. J. Manuf. Mater. Mech. Eng. 1, 53 71 (2011). 79. Das, S.K., Sahoo, P.: Optimisation of Tribological Performance of Electroless Ni B Coating Using Taguchi Method and Grey Relational Analysis. Tribol. Mater. Surfaces Interfaces. 5, 16 24 (2011)
130 80. Das, S.K., Sahoo, P.: Optimization of Electroless Ni B Coatings based on Multiple Roughness Characteristics. J. Tribol. Surf. Eng. 2, 85 108 (2011). 81. Das, S.K., Sahoo, P.: Wear Performance Optimization of Electroless Ni B Coating Using Taguchi D esign of Experiments. Tribol. Ind. 32, 17 27 (2010). 82. Contreras, A., Len, C., Jimenez, O., Sosa, E., Prez, R.: Electrochemical Behavior and Microstructural Characterization of 1026 Ni B Coated Steel. Appl. Surf. Sci. 253, 592 599 (2006). 83. Riddle, Y.W., McComas, C.E.: Advances in Electroless Nickel Boron Coatings: Improvement of Lubricity and Wear Resistance on Surface of Automotive Components. Proceeding of the 2005 SAE World Congress. pp. 11 14. Detroit, Michigan (2005). 84. Sahoo, P., Das, S. K.: Tribology of Electroless Nickel Coatings A Review. Mater. Des. 32, 1760 1775 (2011). 85. Das, S.K., Sahoo, P.: Tribological Characteristics of Electroless Ni B Coating and Optimization of Coating Parameters using Taguchi Based Grey Relational Analys is. Mater. Des. 32, 2228 2238 (2011). 86. Delaunois, F., Lienard, P.: Heat Treatments for Electroless Nickel Boron Plating on Aluminium Alloys. Surf. Coatings Technol. 160, 239 248 (2002). 87. Ziyuan, S., Deqing, W., Zhimin, D.: Surface Strengthening Pur e Copper by Ni B Coating. Appl. Surf. Sci. 221, 62 68 (2004). 88. Krishnaveni, K., Sankara Narayanan, T.S.N., Seshadri, S.K.: Electroless Ni B Coatings: Preparation and Evaluation of Hardness and Wear Resistance. Surf. Coatings Technol. 190, 115 121 (2005 ). 89. Li, P., David, M.: Curve Fitting Analysis of ESCA Ni 2p Spectra of Nickel Oxygen Compounds and Ni / AI203 Catalysts. 38, 880 886 (1984). 90. Nitridation Process o f Boric Acid. J. Alloys Compd. 224, 22 28 (1995). 91. Dervos, C.T., Novakovic, J., Vassiliou, P.: Vacuum Heat Treatment of Electroless Ni B Coatings. Mater. Lett. 58, 619 623 (2004). 92. Erdemir, A., Bindal, C., Zuiker, C., Savrun, E.: Tribology of natur ally occurring boric acid films on boron carbide. Surf. Coatings Technol. 86 87, 507 510 (1996).
131 93. Dvorak, S.D., Wahl, K.J., Singer, I.L.: Friction Behavior of Boric Acid and Annealed Boron Carbide Coatings Studied by In Situ Raman Tribometry. Tribol. T rans. 45, 354 362 (2002). 94. Hassan, A.K., Torell, L.M., Borjesson, L., Doweidar, H.: Structrual Changes of B2O3 Through the Liquid Glass Transition Range: A Raman Scattering Study. Phys. Rev. B. 45, 797 805 (1992). 95. Yano, T., Kunimine, N., Shibata, S., Yamane, M.: Structural Investigation of Sodium Borate Glasses and Melts by Raman Spectroscopy. III. Relation Between the Rearrangement of Super Structures and the Properties of Glass. J. Non. Cryst. Solids. 321, 157 168 (2003). 96. eill, J.P., Zabinski, J.S.: Nanocomposite Tribological Coatings for Aerospace Applications. Surf. Coatings Technol. 116 119, 36 45 (1999). 97. Tribology. Wear. 190, 139 144 (1995). 98. Deal, B.E., Grove, A.S.: General Relationship for the Thermal Oxidation of Silicon. J. Appl. Phys. 163, 3770 3778 (1965). 99. Erdemir, A., Erck, R.A., Robles, J.: Relationship of Hertzian Contact Pressure to Friction Behavior of Self Lubricating Bor ic Acid Films. Surf. Coatings Technol. 49, 435 438 (1991). 100. Cowley, J.M.: Structure Analysis of Single Crystals by Electron Diffraction. II. Disordered Boric Acid Structure. Acta Crystallogr. 6, 522 529 (1953). 101. Choy, T.: Materials International Space Station Experiment 7 (MISSE 7), http://www.nasa.gov/mission_pages/station/research/experiments/MISSE 7.html#description 102. Krick, B.A., Sawyer, W.G.: Space Tribometers: Design for Exposed Experiments on Orbit. Tribol. Lett. 41, 303 311 (2010). 1 03. Scharf, T.W.: Low Friction Coatings. Handbook of Lubrication and Tribology, Volume II Theory and Design. pp. 39 1 39 15. CRC Press (2010). 104. Scharf, T.W., Kotula, P.G., Prasad, S.V.: Friction and Wear Mechanisms in MoS2/Sb2O3/Au Nanocomposite Co atings. Acta Mater. 58, 4100 4109 (2010). 105. Dever, J.A.: Low Earth Orbital Atomic and Ultraviolet Radiation Effects on Polymers. Cleveland, OH (1991).
132 106. Goldstein, R.M., Goldstein, S.J.: Flux of Millimetric Space Debris. Astron. J. 110, 1392 1396 (1995). 107. Moulder, J.F., Stickle, W.F., Sobol, P.E., Bomben, K.D.: Handbook of X ray Photoelectron Spectroscopy. ULVAC PHI, Chigasaki (1995). 108. Morgan, W.E., Stec, W.J., Van Wazer, J.R.: Inner Orbital Binding Energy Shifts of Antimony and Bismuth Compounds. Inorg. Chem. 12, 953 955 (1973). 109. Dudder, G.J.: Surface Studies of Dry and Solid Lubricants Under Different Environmental Conditions, (2010). 110. Hamilton, M.A., Alvarez, L.A., Mauntler, N.A., Argibay, N., Colbert, R., Burris, D.L., Murat ore, C., Voevodin, A.A., Perry, S.S., Sawyer, W.G.: A Possible Link Between Macroscopic Wear and Temperature Dependent Friction Behaviors of MoS2 Coatings. Tribol. Lett. 32, 91 98 (2008). 111. Mcintyre, N.S., Spevack, P.A., Scrence, S., Briggs, D.: Effect s of Argon Ion Bombardment on Basal Plane and Polycrystalline MoS2. Surf. Sci. Lett. 237, L390 L397 (1990). 112. Tagawa, M., Yokota, K., Ochi, K., Akiyama, M., Matsumoto, K., Suzuki, M.: Comparison of Macro and Microtribological Property of Molybdenum Dis ulfide Film Exposed to LEO Space Environment. Tribol. Lett. 45, 349 356 (2011). 113. Dever, J.A., Miller, S.K., Sechkar, E.A., Wittberg, T.N.: Space Environment Exposure of Polymer Films on the Materials International Space Station Experiment: Results fro m MISSE 1 and MISSE 2. High Perform. Polym. 20, 371 387 (2008). 114. SiC/SiO[sub 2] Interface. J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 15, 1597 (1997). 115. Utigard, T.: Oxidati on Mechanism of Molybdenite Concentrate. Metall. Mater. Trans. B. 40, 490 496 (2009). 116. Baker, C.C., Chromik, R.R., Wahl, K.J., Hu, J.., Voevodin, A.A.: Preparation of Chameleon Coatings for Space and Ambient Environments. Wright Patterson AFB, OH (2 006). 117. Voevodin, A.A., Fitz, T.A., Hu, J.J., Zabinski, J.S.: Nanocomposite Tribological Surfaces, Film. 20, 1434 (2002).
133 118. Hamilton, M.A.: Evaluation of Tribological Res ponse of Molybdenum Sulphide Based Coatings to Varying Environments, (2007). 119. Pomfret, M.B., Stoltz, C., Varughese, B., Walker, R.A.: Structural and Compositional Characterization of Yttria Stabilized Zirconia: Evidence of Surface Stabilized, Low Vale nce Metal Species. Anal. Chem. 77, 1791 5 (2005). 120. Galtayries, A., Cousi, C., Zanna, S., Marcus, P.: SO2 Adsorption at Room Temperature on Ni(111) Surface Studied by XPS. Surf. Interface Anal. 36, 997 1000 (2004). 121. Yan, X.B., Xu, T., Yang, S.R., Liu, H.W., Xue, Q.J.: Characterization of Hydrogenated Diamond like Carbon Films Electrochemically Deposited on a Silicon Substrate. J. Phys. D. Appl. Phys. 37, 2416 2424 (2004). 122. Wells, A.F.: Structural Inorganic Chemistry. Oxford University Press (1 984). 123. Li, Y., Zhang, X., Li, H., Lin, C., Xiao, W., Liu, J.: High Pressure Induced Phase Transitions in CdS up to 1 Mbar. J. Appl. Phys. 113, 083509 (2013). 124. Oladeji, I.O., Chow, L.: Optimization of Chemical bath Deposited Cadmium Sulfide Thin F ilms. J. Electrochem. Soc. 144, 2342 2346 (1997). 125. Reisfeld, R.: Nanosized Semiconductor Particles in Glasses Prepared by the Sol Gel Method: Their Optical Properties and Potential Uses. J. Alloys Compd. 341, 56 61 (2002). 126. Uda, H., Yonezawa, H., Ohtsubo, Y., Kosaka, M., Sonomura, H.: Thin CdS Films Prepared by Metalorganic Chemial Vapor Deposition. Sol. Energy Mater. Sol. Cells. 75, 219 226 (2003). 127. Moon, B. S., Lee, J. H., Jung, H.: Comparative Studies on the Properties of CdS Films Deposit ed on Different Substrates by R.F. Sputtering. Thin Solid Films. 511 512, 299 303 (2006). 128. Goto, F., Shirai, K., Ichimura, M.: Defect Reduction in Electrochemically Deposited CdS Thin Films by Annealing in O2. Sol. Energy Mater. Sol. Cells. 50, 147 15 3 (1998). 129. Jordan, J.F., Lampkin, C.M.: Photovoltaic Cells, (1978). 130. Sihvonen, Y.T.: High Performance Photoresistor, (1965).
134 131. Ma, X., Xiao, D.: Nanocomposite Coating of High Lubricity and Low Friction Fabricated by Plasma Spray Processes. P roceedings of the 19th International Conference on Surface Modification Technologies. pp. 150 155. St. Paul, Minnesota (2005).
135 BIOGRAPHICAL SKETCH Kevin Lamar Gilley was born in Louisville, Kentucky on September 19 th 1986 duation from Southern Baptist Theological Seminary in 1987, his family moved to Gainesville, Florida. In December of 1995 his family would move again to Plant City, Florida and remain there until the summer of 2002, when they would return to Gainesville, F lorida. After graduating from Santa Fe High School, in Alachua, Florida, and being a lifelong Gator fan, Kevin knew there was only one University he would apply to. After gaining acceptance to the University of Florida, Kevin graduated with his B.S. in Mat erials Science and Engineering on May 1, 2009. One day later on May 2, 2009 Kevin married his high school sweetheart Kailey. During his senior year at UF, Kevin decided that continuing his education and pursuing a doctorate would be the course of his futur e. After applying to multiple schools, Kevin chose to continue his education at the University of Florida and to work with Dr. Scott Perry. Near the beginning of his graduate career, Kevin began working on tribology and continued working in this facet of r esearch for the remainder of his graduate career. He received his Ph.D. in materials science and engineering in the fall of 2013 and plans to find a career that is both interesting and challenging. Go Gators!