ISSN: 1402-1757 ISBN 978-91-7439-XXX-X Se i listan och fyll i siffror där kryssen är
LICENTIATE T H E SIS
Aldara Naveira Suarez Parameters Affecting the Functionality of Additives in Lubricated Contacts of ParametersFunctionality Naveira Additivesthe Lubricated Aldara Affecting in Suarez Parameters affecting the functionality of additives
Department of Applied Physics and Mechanical Engineering in lubricated contacts Division of Machine Elements ParametersEffect Affecting of base oil the polarity Functionality
ISSN: 1402-1757 ISBN 978-91-7439-077-3 of Additives in Lubricated Contacts Luleå University of Technology 2010 Effect of Base Oil Polarity
Aldara Naveira Suarez Effect of Base Oil Polarity Oil Base of Effect Aldara Naveira-Suarez
Parameters affecting the functionality of additives in lubricated contacts Effect of base oil polarity
Aldara Naveira-Suarez
SKF Engineering and Research Centre Kelvinbaan 16, Nieuwegein THE NETHERLANDS
Division of Machine Elements Luleå University of Technology SE-971 87 Luleå SWEDEN
February 2010 Printed by Universitetstryckeriet, Luleå 2010
ISSN: 1402-1757 ISBN 978-91-7439-077-3 Luleå www.ltu.se Chemistry ought to be not for chemists alone
Miguel de Unamuno, The tragic sense of life (1913)
ABSTRACT Traditionally rolling contact fatigue observed in bearing field applications was subsurface initiated. However, despite the improvement of steel properties, some factors such as downsizing in bearing design, extreme loading of the bearings as well as demanding application conditions (start up-stop cycles) have led to an increase on the cases of surface damage related to surface initiated fatigue, that comes basically from surface distress. Possible causes leading to surface initiated fatigue are: material and surface properties, marginal lubrication and lubricant chemical composition. Lubricants are formulated products composed of base oil, and an additive package designed for a specific application. Extreme-pressure (EP) and antiwear (AW) additives are chemically active additives, they react with the steel surfaces in contact to form a protective additive- derived layer, thus reducing friction and controlling wear. However, certain EP/AW additives that increase the performance of other machine elements, such as gears, can be detrimental for the bearings running in the same lubrication environment. In order to identify the plausible mechanisms that govern the detrimental effect of EP/AW additives on bearing performance, it is necessary to study closely the interactions occurring in the system form by the base oil, the additives present and the steel surface, as well as the influence of operating conditions. The focus of the present work is to identify the parameters affecting the additive-derived layer formation, as it is directly related to the additive reactivity towards the surface, and the tribological properties of the layer, that will determine the tribological performance. Zinc dialkyldithiophosphate (ZDDP), and two low viscosity model oils with different polarity were selected. The influence of base oil polarity on the additive performance was studied in the nanoscale using Atomic Force Microscopy and the tribological performance was evaluated using a ball-on-disc test rig under mixed rolling-sliding conditions in the boundary lubrication regime. An in-situ interferometry technique was used to monitor the additive derived reaction layer formation, and the chemical composition, morphology and nanomechanical properties were studies using X-ray Photoelectron Spectroscopy, Atomic Force Microscopy and Nanoindentation respectively.
i It was found that base oil polarity determines the transport of additives to the surface thereby controlling the maximum reaction layer thickness, friction and wear, as well as the morphology and nanomechanical properties of the additive-derived reaction layer. However the reaction layer chemical composition is not determined by the base oil polarity. Among the operating conditions, shear was identified as a fundamental parameter on the activation of additives on rubbing steel surfaces and the properties of the derived reaction layer.
ii LIST OF PAPERS
Paper A: The influence of base oil polarity on the tribological performance of zinc dialkyl dithiophosphate additives. A. Naveira-Suarez, M. Grahn, R. Pasaribu and R. Larsson (Submitted to Tribology International)
Paper B: Effect of base oil polarity on micro and nano friction behaviour of base oil- ZDDP solutions. A. Tomala, A. Naveira-Suarez, I.C. Gebeshuber and R. Pasaribu (Tribology: Materials, surfaces and interfaces, 2009, In press)
Paper C: The influence of slide-roll ratio on additive-derived reaction layer formation: Zinc dialkyl dithiophosphates A. Naveira-Suarez, M. Zaccheddu, A. Tomala, M. Grahn, R. Pasaribu and R. Larsson (To be submitted to Tribology Letters)
iii ACKNOWLEDGEMENTS
The work presented in this Licentiate thesis has been supported by the European Commission, through WEMESURF (Characterization of WEar MEchanisms and SURFace functionalities with regard to lifetime prediction and quality criteria – form micro to the nano range) research training network.
I would like to thank my supervisors Prof. Roland Larsson and Dr. Rihard Pasaribu, for their great support and guidance, for sharing their knowledge with me and for their constant enthusiasm, encouragement and commitment to our work.
My warmest thanks also to my colleagues at the Division of Machine Elements in Luleå University of Technology and at SKF Engineering and Research Centre, for creating an open and knowledgeable working environment. I would like to thank Mr Ralph Meeuwenoord for his great contribution to this work through his constant help in the tribological testing. Dr. Mattias Grahn and Fiona Elam have share many office hours with me and I must thank them for their patience.
I specially want to thank Dr. Maurizio Zaccheddu, from SKF ERC and Agnieszka Tomala, from Vienna University of Technology, that have contributed greatly to this work, revealing some of the mysteries of Physics to me. Working with them is a privilege and a pleasure.
I am especially grateful to Mr Hubert Koettrisch, for giving me the opportunity to join the team of the SKF Development Centre in Steyr, where I learned so much, and being an inspiring mentor to me.
I would like to thank my family and friends for being always so close to me, despite the time and the distance. I deeply thank my parents, for their love and support, for teaching me the importance of ideas and always challenging me to be critical and to wonder, what it is so important in my work and in my life.
iv CONTENTS
ABSTRACT ...... i LIST OF PAPERS...... iii ACKNOWLEDGEMENTS...... iv
1. INTRODUCTION ...... 1 1.1. Tribology...... 1 1.1.1. Friction...... 1 1.1.2. Wear...... 2 1.1.3. Lubrication...... 2 1.2. Lubricants...... 3 1.2.1. Lubricant additives: Extreme-pressure and antiwear additives ...... 4 1.2.2. Organo-metallic compound type: Zinc dialkyl dithiophosphates...... 7 1.3. Additive-derived layers...... 8 1.3.1. Additive-derived thermal layer...... 8 1.3.2. Additive-derived reaction layer ...... 9 1.4. Objectives ...... 11
2. MATERIALS AND METHODS ...... 15 2.1. Materials...... 15 2.1.1. Test samples ...... 15 2.1.2. Base oils and additives...... 16 2.2. Tribological testing ...... 17 2.2.1. Macro-tribological tests ...... 17 2.2.2. Nano-tribological tests...... 19 2.3. Additive-derived reaction layer formation ...... 21
2.4. Additive-derived reaction layer characterization ...... 22 2.4.1. Scanning Electron Microscopy and Energy Dispersive X-ray Spectroscopy ...... 22 2.4.2. Atomic Force Microscopy ...... 23 2.4.3. Nanoindentation ...... 23 2.4.4. X-ray Photoelectron Spectroscopy...... 23
3. RESULTS AND DISCUSSION...... 27 3.1. Low viscosity model base oils...... 27 3.1.1. Friction and wear performance...... 27 3.1.2. Reaction layer formation ...... 28 3.1.3. Morphology and topography of reaction layer ...... 29 3.1.4. Nanomechanical properties of reaction layer ...... 33 3.1.5. Chemical composition of reaction layer...... 34
v 3.2. Commercial base oils ...... 35 3.2.1. Friction and wear performance...... 35 3.2.2. Reaction layer formation ...... 38 3.2.3. Morphology and topography of reaction layer ...... 41 3.2.4. Chemical composition of reaction layer...... 42
4. CONCLUSIONS AND FUTURE WORK...... 47 4.1. Conclusions...... 47 4.2. Future work...... 48
REFERENCES……………………..…………………………………….49
vi Introduction CHAPTER 1
CHAPTER 1 1. Introduction
The first section of this chapter presents a short introduction to tribology and the related fields of friction, wear and lubrication. The second section describes the basic composition of commercial lubricants and summarizes the main categories of lubricant additives. The characteristics and functionalities of extreme pressure and antiwear additives are discussed, focusing on zinc dialkyl dithiophosphates. The third section describes the nature and properties of the additive-derived reaction layers formed on contacting steel surfaces, in terms of morphology, nanomechanical properties and chemical composition, and the possible mechanism leading to reaction layer formation. The last section summarizes the effect of extreme pressure and antiwear additives on bearing performance and presents the objectives of the thesis.
1.1. Tribology Tribology is defined as “the science and technology of interacting surfaces in relative motion”, and covers the fields of friction, wear and lubrication, including the interactions between solids, liquids and gases [1].
1.1.1. Friction The friction force is the resistance encountered, when one body is moved over other. Amontons’ (1699) four empirical laws of friction [1] are: • there is a proportionality between the maximum tangential force before sliding and the normal force when a static body is subjected to increasing tangential load, • the tangential friction force is proportional to the normal force in sliding, • the friction force is independent of the apparent contact area, • the friction force is independent of the sliding speed. These laws are valid for a large variety of tribosystems including metals under both dry and boundary lubricated conditions, but are limited, especially if dealing with polymers. The second law established the proportionality between normal load (P) and tangential force (F) [2] as follows: F = μ·P (1) being μ assigned later on the term “coefficient of friction”.
1 Introduction CHAPTER 1
1.1.2. Wear Wear is defined as the loss of material of a solid, caused by the mechanical stress of bodies in motion. Adhesive wear appears when two bodies or counterparts are in direct contact via asperities which adhere to each other, causing severe wear. Abrasive wear takes place when hard particles or counterparts remove material of a softer material, resulting in scratches and furrows. The process of abrasive wear involves both plastic flow and brittle fracture. Erosive wear, related to abrasive wear, occurs due to impacting particles, carried by a gas steam or a flowing liquid, and cavitation wear, is caused by fast flow liquids. Fatigue wear occurs by fatigue processes due to repetitive stresses under either sliding or rolling conditions, resulting in cracks, small cavities or delamination. Corrosive wear takes place, if the material suffers corrosion by the surrounding medium. When atmospheric oxygen is the corroding agent, oxidative wear occurs [1-3].
1.1.3. Lubrication To reduce friction and to prevent wear, the surfaces of the counterparts in motion are often separated by a lubricant film. This is in most cases an oil-based solution, but water- based lubricants are gaining in importance. For lubricated systems, the Stribeck curve describes the dependence of the friction coefficient on the dynamic viscosity η, the speed ν and the normal load L. Three regimes can be distinguished: Hydrodynamic lubrication regime. The surfaces are completely separated by a thick lubricant film (1-1000 μm) preventing contacts between the surfaces, and the load is carried by the lubricant film. The friction coefficient is low and arises from the shear forces in the viscous lubricant. The friction coefficient is calculated by Reynold’s equation where the friction coefficient increases as load increases or viscosity decreases or sliding speed increases. Mixed lubrication regime. The lubricant films are thin (0.01-1 μm) and some asperities from the two surfaces are in contact. The load is carried mainly by the lubricant film, but also by the asperities in contact. The two surfaces are separated from each other, partly by hydrodynamic forces and partly by thin layers of lubricant adhering to the surface
2 Introduction CHAPTER 1 contours. It is an intermediate state in which the strict laws of hydrodynamic lubrication are no longer fully applicable. Boundary lubrication regime. The lubricant film (1-100 nm) is squeezed out of the contact area and the whole load is carried by the asperities in contact. This typically results in high friction coefficients, influenced by the underlying surface as well as by the chemical constitution of the lubricant (bulk viscosity has little or no effect in the frictional behaviour), and high wear.
Boundary Lubrication Friction coefficient dependent on non-hydrodynamic characteristics
Mixed Lubrication Increasing coefficient of friction caused
μ by partial contact between the surfaces
Hydrodynamic Lubrication Friction coefficient determined by hydrodynamic theory Friction Coefficient,
Low speed viscosity • speed High speed High load normal load Low load Lubricant film thickness Figure 1 Stribeck curve, dependence of the friction coefficient on viscosity, speed and load for a lubricated sliding system. Adapted from [2]
1.2. Lubricants Lubricants are formulated products composed of a base oil (or base stock), which can be either mineral or synthetic, and various specialty additives designed for specific performance needs [4]. The base oil is, in most cases, a mineral oil with 20-30 carbon atoms. Either straight or branched chains are used, sometimes containing aromatic or aliphatic rings. Synthetic oils are more expensive and used for more demanding applications, where for example insulating, thermally or chemically resistant lubricants are required [5]. Lubricant additives are added to the base oil to optimize the performance of the lubricant in a certain application. Additive levels in lubricant range from 1 to 25% depending on
3 Introduction CHAPTER 1 the application. Lubricant additives can be categorized as chemically active or chemically inert. Chemically active additives, such as dispersants, detergents, antiwear and extreme-pressure, oxidation inhibitors, and rust and corrosion inhibitors, interact chemically with metals, forming protective layers, and the polar oxidation and degradation products, making them innocuous. Chemically inert additives, such as emulsifiers, demulsifiers, pour-point depressants, foam inhibitors, and viscosity improvers, improve the physical properties that are critical to the effective performance of the lubricant [4]. Most lubricant additives, except perhaps some viscosity improvers and pour-point depressants, consist of an oleophilic hydrocarbon group and a hetero atom (N,O,S, and P)-based polar functionality (Figure 2). The hydrocarbon group must have the sufficient carbon length to achieve the required solubility characteristics to the additive. The additives that require greater solubility in oil (dispersants, detergents, and viscosity improvers) usually contain large hydrocarbon groups. Those that require either lower solubility or greater surface activity (foam inhibitors and extreme-pressure) contain small hydrocarbon groups. The performance of the additive is strongly dependent on a proper balance of polar and non-polar characteristics [4].
Figure 2 Additive structure scheme [4]
1.2.1. Lubricant additives: Extreme-pressure and antiwear additives Extreme-pressure (EP) and antiwear additives (AW) control the lubricating performance of the oil in the mixed and boundary lubrication regimes. Performance enhancing properties of these additives are very important since, if oil lacks lubricating ability, excessive wear and friction can occur [4]. The main elements, which are responsible for the extreme-pressure and antiwear action, are sulfur and phosphorus respectively [6]. The solubility of additives is major issue in additive technology, because having these
4 Introduction CHAPTER 1 additives in a solution in the lubricant is essential to transport them close to the rubbing steel surfaces. As previously described, in general the structure of additives can be described as a functional core, or polar moiety, including the main elements, e.g. sulfur and/or phosphorus and a tail, or non-polar moiety, e.g. alkyl chain, which makes the additive soluble in the base oil. The functional core of the additives is generally composed of elements of different electronegativity. This leads to an unequal charge distribution leading to polarity. The polarity of an additive is responsible for the ability to recognize the surface and attach to it. The ability to attach and build layers on the surface depends therefore on the polarity, but also the required space for the tails plays an important role. If the tails are widely branched the near surface concentration and ordering of the molecules may be much lower than when unbranched additives are present [4]. EP/AW additives can be divided into four main groups by chemical composition: sulfur compound type, phosphorus compound type, chlorine compound type and organo- metallic compound (sulfur-phosphorus compound) type. Examples of widely used EP/AW additives [7] are summarized in Table 1.
5 Introduction CHAPTER 1
Table 1 Examples of commonly used extreme-pressure and antiwear additives [7]
Sulfur compound type
Sulfurized fatty oil
Sulfurized terpene
Sulfurized olefin CH3 CH3
CH2 C S S CH2 C
CH3 CH3
Sulfide
Phosphorus compound type
Phosphite Phosphate
Amine phosphate
Organo-metallic compound type
Dialkyldithiophosphate
Dialkyldithiocarbamate
Naphtanate
6 Introduction CHAPTER 1
1.2.2. Organo-metallic compound type: Zinc dialkyl dithiophosphates Dialkyl dithiophosphate compounds (MDTPs) have been used in lubricating oils due to their multifunctional performance as antiwear, extreme-pressure, friction modifying, antioxidant and corrosion resistant additives. Dialkyl dithiophosphates of different metals such as molybdenum [8], cadmium [9], copper [10], titanium, gadolinium [11], iron, antimony and other metals have been introduced in lubricants, being zinc dialkyl dithiophosphates (ZDDP) the most widely used [12-16].
ZDDPs can be classified in three groups based on their chemical structure, according to the tail chain of the phosphates: alkyl (primary and secondary) ZDDP, with single and branched carbon chains respectively, bound on the oxygen atoms, and aryl ZDDP, with an aromatic ring bound on the oxygen atom on which an aliphatic chain is bound (Table 2). The difference in tail composition determines to a large extent the solubility of the additive in the oil.
Table 2 Chemical formulae of primary, secondary and aryl zinc dialkyldithiophosphates (ZDDPs)
Primary alkyl ZDDP
R R O S S O R R P Zn P Secondary alkyl ZDDP R R O S S O R R
Aryl ZDDP
ZDDP can exist in neutral or basic form [17]. Basic ZDDP can convert to the neutral form and ZnO at elevated temperature [18] and their different antiwear properties have been studied [19]. The neutral ZDDP exists as equilibrium between monomer and dimer in solution [20], as illustrated in Figure 3. They can also exist as trimers or oligomers
7 Introduction CHAPTER 1 depending on the state of the ZDDP, crystalline or liquid, the concentration of ZDDP in solvent, and the presence of additional compounds [21].
Figure 3 Equilibrium between dimeric and monomeric forms. Adapted from [22]
1.3. Additive-derived layers EP/AW additives form a protective layer on the surface of the rubbing steel surfaces. Therefore in order to identify the plausible mechanisms that govern the interaction between the additives and the steel surface it is necessary to identify the parameters affecting the additive-derived layer formation and the tribological properties of this layer, as it is directly related to the additive reactivity towards the surface.
1.3.1. Additive-derived thermal layer ZDDP interacts weekly with steel surfaces at ambient temperature [23], however the iron surface is assumed to have a catalytic effect on the ZDDP decomposition, as a ZDDP- derived thermal layer forms at lower temperature than the temperature of thermal decomposition of the molecule [24]. The decomposition of ZDDP in oil solutions is induced by an isomerization of the original molecule (movement of the alkyl group from the oxygen to the sulfur) and the subsequent intramolecular (cis) elimination of an oleofin form the isomer), according to the hard and soft acids and basis (HSAB) theory. The isomer reaction product is assumed to be a precursor to the formation of long-chain polyphosphate layers on the steel surface [25]. The formation of thermal layers on steel surfaces by ZDDP is dependent on temperature, the concentration of the additive in solution [26], the decomposition time of the solution [25] and the type of ZDDP [27].
8 Introduction CHAPTER 1
The transparent, solid, thermal layers appear to grow as separate patches to form a mound-like structure which subsequently coalesce to form a smoother structure and can grow up to a thickness of 200 nm [25] . The layers have an indentation modulus of E≈35 GPa and hardness of H≈1.5 GPa [28].The layers consist of a mixture of short- and long- chain poly(thio)phosphates, determined by the temperature of the solution [24], and showing little evidence of iron, being zinc the main cation [26].
1.3.2. Additive-derived reaction layer ZDDP reaction layers form at lower bulk temperatures than thermal layers, the rate of layer formation increasing with temperature and only in the rubbing tracks, where actual sliding contact occur [29]. The reaction layers initially form as separate patches on steel surfaces and gradually develop to form a continuous pad-like structure [30]. A four-step process to describe the reaction mechanism of ZDDP in solution, under mild wear conditions, with rubbing steel surfaces [31]. The pads are solid-like with an indentation ≈ modulus of E ≈90 GPa and hardness of H≈3.5 GPa [32]. The reaction layer consist of glassy phosphate, with a thin outer layer of zinc polyphosphate (≈10 nm) grading to pyro- or ortho- phosphate in the bulk, being zinc/iron sulfides and /or oxides present as inclusions [16]. In the outer part of the layer, Zn is the main cation, but there is a increasingly higher presence of Fe towards the steel surface [33]. Zinc or iron sulfide may be present on the metal surface [34], however other studies have disputed it [33].
1000 nm Base oil / ZDDP / degradation products
Alkylphosphates (MPa hardness)
Long-chain glassy Zn/Fe poly(thio)phosphates (GPa hardness) Short-chain Zn/Fe phosphates 0nm STEEL FeO, FeS/ZnS
Figure 4 Scheme of ZDDP-derived reaction layer. Adapted from [35]
9 Introduction CHAPTER 1
A number of reaction mechanism have been proposed for the formation of the ZDDP- derived reaction layers, based mainly on thermal [27], catalytic [15](chemisorption on metal or hydrolytic) and oxidative [31] (by reaction with hydroperoxides or peroxy radicals) decomposition of ZDDP.
Heuberger [36, 37] summarized the reactions that may take place in the near-contact, off- contact and contact areas (Figure 5) under pure sliding conditions.
Off-contact area Contact area Near-contact area - Below 100°C: Adsorption of - Tribological energy leads to - Frictional heating leads to ZDDP and reaction products partial removal of reaction layer thermal decomposition and - Above 130°C: Thermal layer with wear phosphate layer formation formation - Frictional heating and shear - At 80°C: Alkylated sulfur reacts stress leads to the formation of with Zn and ZnS, which is reaction layer from the adsorbed deposited molecules - Reaction layer transfer to the ball
Figure 5 Scheme of reactions in the contact, near-contact and off-contact areas of a tribological contact lubricated with a ZDDP solution. Adapted from [38]
ZDDP reacts with the surface in contact to form protective reaction layers. However, several studies have proved ZDDP to have detrimental effects on wear under certain operation conditions [39], and to enhance friction when the system is operating in mixed and boundary lubrication regimes [40-42].
10 Introduction CHAPTER 1
1.4. Objectives Lubricating oils are designed and optimized for a specific application. For example an AW/EP additives that enhance the performance of other machine elements such as gears, can be detrimental for the bearings running in the same lubrication environment. Wan et al. [43] conducted a series of experiments with a commercial EP/AW additive package (S-P based) using a four-ball test rig and bearing life test machines. While the 4-ball tester showed a high weld load and low wear rate for the oil+additive blend, the rolling-bearing fatigue tests showed that under marginal film lubrication conditions (lambda =1.2) The reduction of bearing life is primarily attributed to the high chemical reactivity of the EP/AW additives in blended oil. Rolling-bearing fatigue tests showed that under marginal film lubrication conditions, lambda = 1.2 (ratio of lubricant film thickness and composite surface roughness), the EP/AW additive package blended lubricating oil reduced the bearing fatigue life by a factor of 4.5 compared with the base oil alone (Figure 6). Reduction of bearing life was attributed to high chemical reactivity of the EP additives in the base oil.
BASE OIL
EP-OIL
Figure 6 Effect of EP additive package on the life of 6309 DGBB IR [43]
In another study [44] tapered roller bearings were tested using lubricants containing S-P additives with modified sulfur extreme pressure (sulfur EP) components [44]. Three SAE 20 mineral oil lubricants containing S-P additives of different reactivity were selected for evaluation. The EP components were alkyl polysulfides with the general formula R-(S)n-
11 Introduction CHAPTER 1
R. The sulfur EP expected to be least active (LA) was an alkyl disulfide (n=2), the sulfur EP with standard activity (SA) had n>2, and the most active (MA) sulfur EP had an even larger sulfide chain length on average. Test lubricants were blended such that each contained approximately the same overall sulfur and phosphorus content. Surface layers and near-surface material (depth <500nm) were observed on the nanometre scale in cross-section using transmission electron spectroscopy (TEM). The antiwear layer appeared bright in the TEM images, suggesting it was less dense and/or containing more amorphous content than the sample near-surface material. The antiwear layer for the MA-S sample was ≈40-80 nm thick in the examined regions (Figure 7). Several shallow cracks (depth ≈50-80 nm) extended into the near-surface material, travelled in a uniform direction, and appeared to contain antiwear layer material.
Figure 7 TEM image from the MA-S cone specimen Surface layer (antiwear layer) and cone near–surface material showing shallow surface cracks [44]
Although nanometre scale surface damage was found in the MA-S sample (lubricant with the highest sulfur EP activity), the longest average test life was obtained using the MA lubricant. In order to identify the plausible mechanisms that govern the detrimental effect of those additives on bearing performance it is necessary to identify the parameters affecting the additive-derived layer formation and the tribological properties of this layer, as it is directly related to the additive reactivity towards the surface. The friction and wear behaviour of tribological systems lubricated with oils containing ZDDP have been mainly studied as a function of the additive structure, concentration, interaction with other additives or operation conditions, but the effect of the base oil in the ZDDP tribological performance has been hardly addressed [45-47]. When considering an additive molecule approaching the steel surface in a lubricating oil, it must be considered that the steel
12 Introduction CHAPTER 1 surface is already completely covered with oil molecules. Hence, it can be stated that the additive molecule has to compete with the oil for a place on the steel surface. The polarity of a molecule can be used to measure its affinity towards the steel surface. When the polarity of the molecule is low, its affinity towards the surface is low, while when the polarity of the molecule is high, its affinity for the steel surface is also high. Parameters, like polarity, that describe the base oil-additive interactions are needed in order to have a good running in, friction and wear model of lubricated contacts. Several operating conditions (in terms of lambda ratio, temperature and additive concentration) have been previously studied [29, 41, 42], showing that low lambda ratios, meaning high metal to metal contact, high temperature and high additive concentration lead to a high reactivity of the additives and therefore to thicker reaction layers.
This study focuses on the effect of slide-roll ratio and the interaction between additive- base oil molecules in reaction layer formation and tribological performance. The nature and properties of the derived reaction layers, as a function of operating conditions and base oil-additive interaction, in terms of thickness, morphology and nanomechanical properties, and chemical composition, were studied using a series of surface analysis techniques.
13 14 Materials and methods CHAPTER 2
CHAPTER 2 2. Materials and methods
The first section of this chapter presents the test samples and sample preparation method for the tribological testing. The physical and chemical properties of the lubricants, base oils and additive, used in the study are also summarized. The second section presents the set-up and experimental conditions for the tribological tests. A description of the different techniques used to characterize the additive-derived reaction layer, the measuring conditions and main analysis parameters are presented in the third section of the chapter.
2.1. Materials The test samples used for the tribological testing, described in Section 2.2, were selected to have material and surface properties matching bearing applications. The base oils tested were low viscosity reference oils [PaperA&PaperB]and commercial base oils [Paper C], were selected to have very similar chemical and physical properties, but different polarities what allows to study the influence of base oil polarity on the additive tribological performance and reaction layer formation. 2.1.1. Test samples The steel balls were of ASI 52100 steel with hardness 59-66 HRC and an average roughness (Ra) of 10 nm. The rings were washers (WS 81212) from SKF Cylindrical Thrust Roller Bearings (CRTB) of ASI 52100 steel with hardness 59-66 HRC and Ra=100, for reaction layer thickness tests. According to EN ISO 683-17, the steel contains 1.0% carbon, 1.5% chromium, 0.255% silicon and 0.35% manganese (all percentage by weight). Usual tolerances are ± 0.1% and both phosphorus and sulfur have to be lower than 0.025%. The rings are assembled in a holder to attach them to the rotating shaft of the WAM5 ball-on-disc test rig. The specimens were cleaned prior to testing by successive immersion first in an ultrasonic bath of petroleum ether for 10 min and then acetone for 10 min.
15 Materials and methods CHAPTER 2
2.1.2. Base oils and additives In order to study the influence of base oil polarity on tribological performance, initially two low viscosity model base oils were selected [PaperA&PaperB].Apolarbaseoil, diethylen glycol diethyl ether (DGDE) and a non-polar oil, n-hexadecane (HeD), both supplied by Acros Organics, NJ, USA. Iso-butyl-zinc dithiophosphate (A&S Chemie, Tubingen, Germany), with 99% purity, is employed in simple solution in both base oils without other additives present. The molecular structures of the base oils and additive molecules are presented in Figure 8.
(a)
(b)
(c) Figure 8 Molecular structures of (a) diethylen glycol diethyl ether (b) n-hexadecane and (c) iso- butyl-zinc dithiophosphate
Further two commercial fully synthetic oils, one polar, ester oil and one non-polar, poly- α-olefin (PAO), were used [Paper C]. PAO is chosen due to its purity instead of mineral oil because it has a relatively high concentration of sulfur which might interfere with the additive. The physical properties and sulfur and phosphor content of the base oils and the additive used, obtained by X-ray fluorescence (XRF) analysis before testing, are summarized in Table 3.
16 Materials and methods CHAPTER 2
Table 3 Base oil properties
BASE OIL TYPE Kinematic Kinematic Sulfur Phosphorus viscosity at viscosity at content content 40°C 100°C Code Description (wt%) (wt%) (mm2/s) (mm2/s)
Diethylen glycol DGDE 3.0 1.3 <0.00005 <0.00005 diethyl ether HeD n-hexadecane 1.9 1.7 <0.00005 <0.00005 Fully synthetic SKF Ester 26.8 5.2 0.00052 <0.00030 (Ester) Synthetic Fully synthetic 24.6 5.1 0.00055 <0.00030 Group IV (PAO) Iso-butyl-zinc ZDDP dithiophosphate ------23.400 11.300 NOTE1: The results for all elements are <0.0002 % It is assumed that the addition of the additives does not change significantly the viscosity of the bulk solution. An ultrasonic bath was used to dissolve the additives in the base oils. The temperature of the lubricant solution remained below 40°C during the dissolving procedure.
2.2. Tribological testing The friction and wear performance of the selected lubricants has been studied using a ball-on-disc test rig, under different operation conditions. The friction and wear depend directly on the chemical and physical properties of the atoms and molecules on the interface between the contacting surfaces in lubricated contacts. Therefore, a series of tests were performed using Atomic force microscopy to address the influence of atomic interactions in the friction response of the lubricants [Paper B].
2.2.1. Macro-tribological tests SKF-WAM5 ball-on-disc test rig (Wedeven Associates, Inc., Edgmont, PA, USA) enables to perform a variety of tests to evaluate the tribological performance of additive/base oil blends at controlled contact conditions. The ball and the disc are independently driven giving the possibility to simulate pure rolling and various slide-roll
17 Materials and methods CHAPTER 2 ratios (slip ratio). This rig is computer-controlled, which enables the ball and disc speeds, the load, and the test temperature to be controlled and varied in any desired programmed sequence (Figure 9).
UR
UB Figure 9 Schematic representation of WAM5 ball-on-disc test rig
Base oil polarity effect on the different additives behaviour is studied in mixed rolling/sliding contact to match closely the application conditions in bearings. In rolling/sliding conditions, wear is evenly-distributed around the tracks of both specimens which avoid the change in contact geometry that occurs during pure sliding tests, and thus only a slight change in contact pressure occurs [48]. The tribotests using low viscosity model base oils, were carried out at applied loads of 100 and 600N, resulted in a maximum Hertz contact pressure of 1.34 GPa (contact diameter 0.38 μm) and 2.43 GPa (contact diameter 0.69 μm), at a constant slide-roll ratio of 5%. [The slide-roll ratio (SRR), or slip ratio, is defined as sliding speed US=UB-UR divided by the entrainment speed, or rolling speed, U=(UB+UR)/2 where UB and UR are the ball and ring surface speed in contact respectively]. The entrainment speed is set a 0.50m/s and the temperature set constant at 90°C [Paper A]. The tribotests using commercial base oils were carried out at an applied load of 300N, for the 20 mm balls employed, resulted in a maximum Hertz contact pressure of 1.9 GPa (contact diameter 0.54 μm). The slide-roll ratios investigated ranged from -10 to +10%.. Temperature for all the tests was set constant at 90°C. Under these conditions the EHD film thickness is calculated at the centre of the contact to be ≈10 nm for the different systems. The specific film thickness or lambda ratio (the ratio of the central lubricant film thickness to the composite surface roughness of the two surfaces is contact) is set constant at 0.4 and the entrainment speed is accordingly set to 0.25 m/s. At such low lambda ratios it can be assumed that the system is operating in the boundary lubrication
18 Materials and methods CHAPTER 2 regime. The surface speed of the ball, the tested specimen in this work, is different for the SRR tested. In order to have the same number of rotations for all the specimens, the rubbing time is adjusted accordingly.
2.2.2. Nano-tribological tests Measurements were performed with an AFM MFP-3D the AFM MFP-3D atomic force microscope (by Asylum Research, Santa Barbara, CA, USA) in contact, constant force mode using non conductive silicon nitride cantilevers with a spring constant k = 0.2 N/m and a resonant frequency f0 = 38 kHz (Veeco Instruments, Santa Barbara, CA, USA). The main measurement parameters were: a scan size from 5x5 to 80x80 µm² (512 scan points and 512 scan lines), a scan rate from 0.5 to 2 Hz, scan angle of 90o, and a set point from 10 nN to 50nN in contact mode. The recorded data was both trace/retrace of height, deflection, and lateral force. Force curves (cantilever spring force as a function of z-piezo extension) were also determined at fixed positions on the specimens. Away from the specimen the lever maintains its free deflection, and first makes a surface contact at S (“snap-in”). In air, this initial contact is frequently with a surface layer of adsorbed water vapour or other contaminants, and is accompanied by formation of a meniscus around the tip-surface contact. Retraction of the lever results in an increasing cantilever spring force acting against the meniscus-related and other adhesive forces, until the “pull-off” force (P) is reached, when the cantilever jumps back to its free deflection position. Hysteresis during contact indicates that some plastic deformation has occurred whilst load was applied to the surface, due to the presence of a relatively soft surface film. In order to compare the conditions in the AFM tests with ball-on-disc experiments, the AFM single asperity contact pressure was estimated. The radius of the tip used for these measurements is 20 nm, at a maximum applied load of 50nN, corresponds to a contact pressure of approximately 100 MPa. Friction calculations Bhushan introduced two methods to measure friction [49]. For the present measurements the lateral force technique was used since it is described as more reliable and objective.
19 Materials and methods CHAPTER 2
The sample is scanned perpendicularly to the long axis of the cantilever beam and the lateral force signals in trace and retrace (LT, LRT) are recorded. In this arrangement, as the sample moves under the tip, the friction force will cause the cantilever to twist. Therefore the signal intensity between the left and the right detectors will be different, denoted as FFM signal [(L-R)/(L+R)]. This signal can be related to the degree of twisting, hence to the magnitude of friction force. By changing the set point parameter in the feedback loop, the normal force applied between probing tip and sample surface can be changed. The scan sizes were 5µm by 5µm, consisting of 512 scan lines with 512 scan points each. First, the average value of all of the 512 lines with 512 points for Lateral Trace (LTVavg) and Lateral Retrace (LRTVavg) from every scan were calculated. To obtain the friction force value (FFV), these two mean values have to be subtracted from each other, and divided by two. − | LTVavg LRTVavg | FFV = (2) 2 The measurements of the friction force values were repeated ten times in every environment to obtain representative and repeatable results. Assuming that the friction in the nanoscale follows Amonton's law, the friction force is given by: = μ ⋅ + FFV (SP F0 ) (3)
Where µ is the friction coefficient, the set point (SP) is the applied load and F0 is a force constant. Following the procedure suggested by Beake et al.[50], the force constant is nearly equal to the pull off force determined from the force distance curves.
μ = FFV + (4) (SP F0 ) Usually, the FFV and SP values are given in [V] Volts as acquired from lateral force measurements. However, the results can be easily compared with each other because findings in Volts are connected with the forces between tip and surface. In order to obtain commonly used units ([N] Newton), the lateral force needs to be calibrated by the determination of the slope of deflection vs. LVDT. The calibration delivers an accurate value of the inverse optical lever sensitivity (InvOLS) describing the sensitivity of the
20 Materials and methods CHAPTER 2 detector-cantilever combination. With the knowledge of the accurate value of InvOLS,it is possible to calculate FFV and SP in Newton: FFV [V] × InvOLS [nm/V] × spring constant [nN/nm] = FFV [nN] (5) SP [V] × InvOLS [nm/V] × spring constant [nN/nm] = SP [nN] (6)
2.3. Additive-derived reaction layer formation Thick ZDDP-derived reaction layers enhance friction mixed lubrication [48] and influence wear [51], therefore the study of the thickness and formation rate of the reaction layers is necessary to understand their tribological performance. In situ spacer layer interferometry was used to monitor ZDDP-derived reaction layer thickness on a steel ball surface. The tests to monitor the reaction layer formation are carried out by rolling/sliding a steel ball on a lubricated steel disc, to produce a wear track on both ball and disc [29]. Motion is then halted and a spacer layer and chromium-coated glass disc are loaded against the wear track on the steel ball. The lubricant is squeezed out from the contact, but any solid- like reaction layer remains. The test arrangement is shown schematically on Figure 10.
3CCD Camera 3CCD Camera
Microscope Microscope
Spacer Layer Spacer Layer
STEEL RING STEEL RING STEEL BALL STEEL BALL Thermocouple Thermocouple
Lubricant Lubricant supply Heater supply Heater system system Thermocouple Thermocouple
Figure 10 Schematic diagram of the spacer layer imaging method set-up on WAM5
When white light shines into the contact, a part of the beam is reflected from the semi- reflective chromium layer while some passes through the silicon dioxide layer and any transparent reaction layer present on the ball before being reflected back from the steel surface. Since the two beams of light have travelled different distances, upon recombination they undergo optical interference, so that some component wavelengths gain and some lose intensity, depending on the optical path difference. The result is a
21 Materials and methods CHAPTER 2 coloured interference image of the contact, where the colour at each point represents the composite effect of the path difference at that point on the intensities of all of the various wavelengths of the white light used. If this interference is frame-grabbed by a colour camera, the RGB (red-green-blue) colour of each pixel can be converted, via a suitable layer thickness/colour calibration [52-54], to a map of the reaction layer thickness present between the ball and flat [55].
Figure 11 Principle of spacer layer interferometry
The glass disc in then removed and the sliding/rolling of the steel ball on the steel disc continuous. This procedure is repeated over the total test time to obtain a series of interference images and thus maps of the variation of additive-derived reaction layer thickness on the ball over time.
2.4. Additive-derived reaction layer characterization
A series of surface analysis techniques have been used to characterize the additive derived layers in terms of thickness, morphology, nanomechanical properties and chemical composition.
2.4.1. Scanning Electron Microscopy and Energy Dispersive X-ray Spectroscopy Morphology and elemental chemical composition of the ZDDP derived reaction layers were characterised with Zeiss Supra 55 Scanning Electron Microscope (SEM) with a 15kV electron beam voltage in the secondary electron mode, equipped with Oxford Instruments Inca system for Energy Dispersive X-ray Spectroscopy (EDS). EDS spectra were collected at different locations inside and outside the wear track.
22 Materials and methods CHAPTER 2
2.4.2. Atomic Force Microscopy Atomic force microscopy (AFM) allows the study of surfaces at the nanoscale and has become a [56] widely-used tool in tribology to study dry lubricant films (i.e. coatings) and also layers formed by liquid lubricants [30, 32, 57] providing information of the topography of reactions layers formed by the lubricant. Topography and friction can be obtained by recording simultaneously lateral and height signals from the scanning tip.. By recording pull off forces-adhesion, elastic and viscoelastic properties of the sample to be investigated.
Measurements were performed with an AFM MFP-3D atomic force microscope (Asylum Research, Santa Barbara, CA, US) in contact, constant force mode using non conductive silicon nitride cantilevers with a spring constant k = 40 N/m and a resonant frequency f0 = 300 kHz (Veeco Instruments, Santa Barbara, CA, US). Images were captured continuously at a scan rate of 0.25 Hz and set point of 1V. The surface roughness parameters Ra and Rrms were obtained by processing the images.
2.4.3. Nanoindentation Nanoindentation measurements allow characterizing the nanomechanical properties of the reaction layers. Nanoindentation experiments were performed with a TI 900 TriboIndenterTM (Hysitron Inc., Minneapolis, MN, US). Indents were performed with a Berkovich diamond tip (Hysitron Inc., Minneapolis, MN, US) at a penetration depth of 20 nm and a loading rate of 10 μN/s. The nanomechanical properties of the layers were evaluated from their force- displacement curves. The elastic modulus (E) and hardness (H) were calculated with Hysitron analysis software, using an algorithm based on the Oliver and Pharr method [58, 59].
2.4.4. X-ray Photoelectron Spectroscopy X-ray photoelectron spectroscopy (XPS) provides not only elemental information (except H and He) but also allows the oxidation state of the elements to be assessed and provides quantitative information and thicknesses of multilayer films in the nanometre range.
23 Materials and methods CHAPTER 2
XPS has been widely used to investigate thermal and tribological reaction layers derived from the additive package present in lubricating oils [60], providing information about the chemical state of the main elements present in the reaction layer. Oxygen (1s) The O1s signal can be used to separate contributions arising from oxygen bound in an oxide or in a phosphate group. The O1s signal of an oxide can be found at approximately 530 eV, while oxygen bound in a phosphate is reported at 531.6-533 eV [37, 61-63]. The signal from a phosphate group [64] can be further separated into bridging oxygen (BO) , that links two phosphate groups together (P-O-P) at 532.9-534.0 eV and non-bridging oxygen (NBO), that terminates the phosphate group (P-O-, POH, P=O) at 531.5-532.0 eV [63, 65-67]. Polyphosphates of different chain lengths are thought to have different mechanical and rheological properties. Important parameters for the characterization of (poly)phosphate chains are P/O atomic ratio ,which is equal to n/(3n+1), and the ratio of bridging oxygen to non-bridging oxygen, which is equal to (n-1)/2(n+1), being n the phosphate chain length The estimation of the length chain therefore provides information on the structure of the formed layer [68]. XPS can give both P/O and BO/NBO ratios, however there are certain limitations. The decomposition of the O1s photopeak in the two corresponding contributions it can be questionable mainly because of the linewidth of the peaks compared with the chemical shift. Some difficulties may arise because sulfur may partially substitute for oxygen, O-P-S instead of O-P-O. XPS core level data on the layer clearly show that some sulfur in the sulfide form (S2-) but it cannot easily separate the metal sulfide from the thiophosphate due to the weak chemical shift [65].
Table 4 Phosphate glasses characterization parameters [64]
NAME CHAIN LENGTH (n) BO/NBO Charge/P
Orthophosphate 1 0 3 Pyrophosphate 2 1/6 2 Polyphosphate >3 1/6>x>1/2 2>x>1 Metaphosphate Infinite ½ 1 Ultraphosphate Crosslinked >1/2 <1
24 Materials and methods CHAPTER 2
Phosphorus (2p) Phosphorus is usually in an oxidized state [62]. The reported binding energy value of the
2p3/2 peak varies between 133.0 eV and 134.5 eV [65]. It was shown that this variation can be used to distinguish orthophosphates and metaphosphates from each other.. This width is due to the spin-orbit splitting of the 2p3/2 and the 2p1/2 electron level, which are separated by 0.85 eV and lead to a broadening of the signal. Sulfur (2p) The S2p signal is rather broad due to the spin-orbit splitting. Signals from tribostressed surfaces are generally found in the sulfide position at approximately 162.0 eV or when it is found at 169.0 eV can be then be assigned to sulfate [33, 60, 62, 65, 69]. It has been pointed that from the S2p position alone it is not possible to determine if the sulfide signal is due to zinc sulfides, iron sulfides or organic sulfide species. Zinc (2p, 3s, LMM)
The most intense signal from zinc is the 2p3/2 signal. Small differences in the chemical shift of the Zn2p peak can be observed between the chemical structures which are found in the ZDDP derived reaction layers (ZnO, ZnS, Zn (poly)phosphates) [69] and also the peak position of the Zn3s signal varies slightly [60]. Iron (2p)
Iron spectra shows two main peaks with maxima of Fe2p3/2 at 711 eV and of Fe2p1/2 at 724 eV (due to the spin-orbit splitting) [68]. The Fe2p spectra can be used to estimate the reaction layer thickness Ar+ ion sputtering [69]. The main three components identified in ZDDP derived reaction layers correspond to metallic Fe, and iron oxides, Fe(II) and Fe(III) [26, 38]. On the higher binding energy side there is a peak that can be assigned to iron phosphate [26]. However, iron is not always detected on reaction layers derived from ZDDP [33].
The X-ray photoelectron spectroscopy measurements were conducted in a PHI 5000 VersaProbe TM X-ray photoelectron spectrometer (Ulvac-PHI Inc, Chanhassen, MN, US) with a monochromatized Al Kα X-ray (1486.6eV) source. The diameter of the analysed area was 100 μm. The analyser was operated in the fixed-analyser-transmission mode.
25 Materials and methods CHAPTER 2
Two set of parameters (spot size, pass energy, time per step) were used for the XPS analysis. The first set was required for survey spectrum, while the second one was used for acquiring detailed spectra [Paper A & Paper C]. In the survey spectra of both the additive-derived reaction layers only the elements carbon, oxygen, iron, phosphorus, sulfur and zinc were found. Detailed spectra of C1s, O1s, Fe2p, P2p together with Zn3s, S2p and Zn2p were measured. Data was processed with CASA XPS software (Casasoftware Ltd., UK). The detailed spectra were fitted with Gaussian-Lorentzian curves and a Shirley background.
The signals of phosphorus, sulfur, iron and zinc of the 2p orbital exhibit two peaks: 2p3/2 and 2p1/2. This split arises from the coupling of the spin and orbital angular momentum.
The area of the 2p3/2 peak is always twice that of the 2p1/2 and binding energy of the 2p1/2 is the higher of the two peaks. In the phosphorus and sulfur signals these two peaks are close to each other, resulting in an asymmetric signal. The model curves used for fitting the experimental peaks were linked to each other: the ratio of 2p1/2/2p3/2 contributions was fixed at 0.5 and the difference in energy was maintained equal to 1.25 eV for sulfur and 0.85 eV for phosphorus. For quantification the signal arising from Zn3s is used in stead of the most intense zinc photoelectron line, Zn2p3/2 , since is in the same binding energy range as phosphorus, sulfur and oxygen and therefore exhibits a similar inelastic mean free path (IMFP) as the other elements present in the layer [38].
26 Results and discussion CHAPTER 3
CHAPTER 3 3. Results and discussion
This chapter presents the experimental results of tribological performance and the additive-derived reaction layer characterization. The first section is related to the results obtained using low viscosity model oils on the influence of base oil polarity on friction and wear, reaction layer formation and the characteristics of the additive derived reaction layers, in terms of topography, morphology, nanomechanical properties and chemical composition The second section extend the previous results using commercial base oils and introduce the slide-roll ratio as a fundamental parameter on additive activation and therefore in reaction layer formation. The properties of the additive- derived layers are also analysed using the previously described surface analysis techniques.
3.1. Low viscosity model base oils Low viscosity model base oils, with similar physical and chemical properties, but different polarities were selected initially to address the influence of base oil polarity on tribological performance and reaction layer formation and properties. The molecular structure of the model based oils is known and allowed to complete the studies using Molecular Dynamics simulations, in order to gain a further understanding on the effect of base oil on the additive functionality towards the surface [70]. The presented results correspond to Paper A. Further analysis on the nano-scale origin of the effect of base oil polarity on additive performance, corresponding to Paper B, are also presented.
3.1.1. Friction and wear performance A series of ball-on-disc tests using low viscosity model base oils/ZDDP solutions were carried out to study the influence of base oil polarity on the friction performance of ZDDP additives. In the beginning of each test, the running-in phase, the coefficient of friction is decreasing as a function of time for a period of time, reaching a steady-state value afterwards. The average steady-state coefficients of friction for the different solutions at different normal loads are summarized in Table 5.
27 Results and discussion CHAPTER 3
Table 5 Steady state coefficient of friction after 1 h rubbing time for polar base oil diethylen glycol diethyl ether and non-polar oil n-hexadecane with different contents of ZDDP
Additive Diethylen glycol diethyl ether n-hexadecane concentration [wt %] P=1.3 GPa P=2.4 GPa P=1.3 GPa P=2.4 GPa
0% ZDDP 0.125 0.100 0.080 0.060 2% ZDDP 0.140 0.105 0.120 0.080 5% ZDDP 0.135 0.110 0.110 0.085
The coefficient of friction for both base oils increases when ZDDP is present in the lubricant which correlates with results presented in [10, 71, 72]. However, the degree of friction increase is higher for the case of n-hexadecane (non-polar) base oil. The antiwear properties of the reaction layers derived from the base oil + ZDDP solutions at different slide-roll ratios were also investigated. The wear track width (WTW) was measured from the balls using a calibrated microscope.
The WTW is measured to be ≈400 μm for the three DGDE solutions. For the HeD solutions, in the absence of ZDDP the wear track width is significantly lower, ≈250 μm, whereas when ZDDP is added the values of the WTW increase to ≈400 μm.
3.1.2. Reaction layer formation In situ spacer layer interferometry has been used to monitor the formation of reaction layers derived from ZDDP solutions on steel surfaces A series of interference images of the centre of the wear track derived from the base oil/ZDDP solution are shown in Figure 12 and Figure 13.
0 min 1 min 5 min 10 min 30 min 40 min 50 min 70 min 130 min
Figure 12 Series of interference images from wear track for DGDE + 5 wt% ZDDP
0 min 1 min 5 min 10 min 30 min 40 min 50 min 70 min 160 min
Figure 13 Series of interference images from wear track for HeD + 5 wt% ZDDP
28 Results and discussion CHAPTER 3
The development of a patchy ZDDP-derived reaction layer on the wear track during rubbing is indicated by the dark areas within the contact. The mean layer thickness, shown in Figure 14, is calculated from line profiles taken across the corresponding interferometry images.
70 DGDE + ZDDP HeD + ZDDP 60
50
40
30
20
10
Mean reaction layer thickness (nm) 0 0 1000 2000 3000 4000 5000 6000 Rubbing distance (m)
Figure 14 Influence of base oil polarity on the growth of ZDDP-derived reaction layer thickness with rubbing distance from images in Figure 12 and Figure 13
The reaction layer derived from the non polar base oil solution (HeD+5% wt ZDDP), developed more rapidly and is thicker than the layer derived from the polar base oil solution (DEG+5% wt ZDDP). The ZDDP, as a polar molecule, will have a higher probability to reach the steel surface when dissolved in non polar oil compared to the case when dissolved in polar base oil. This may be the reason that ZDDP in non polar base oil will form thicker reaction layer than that of non polar base oil if subjected to the same tribological conditions. The effect will be discussed further in section 3.2.2.
3.1.3. Morphology and topography of reaction layer Scanning Electron Microscopy and X-Ray Diffraction Spectroscopy analysis performed on the steel balls at the end of the tribological tests showed that the traces of elements present in the components of the additives (Zn, P and S) are only found inside the wear track (Figure 15).
29 Results and discussion CHAPTER 3
Figure 15 SEM micrograph inside the wear track and the corresponding EDS spectrum of the selected area inside the wear track of the steel ball lubricated with DGDE with 2 wt % ZDDP EDS spectra showed no traces of the additive components outside the wear track (Figure 16).
Figure 16 SEM micrograph outside the wear track and the corresponding EDS spectrum of the selected area outside the wear track of the steel ball lubricated with DGDE with 2 wt % ZDDP
A series of AFM images, 80x80 μm2 of the central region of the reaction layer formed on the wear track are displayed in Figure 17 with the corresponding line profiles taken transverse to the sliding direction of the wear track of the steel ball lubricated with DGDE with 0, 2 and 5 wt % ZDDP, after 1 h rubbing.
30 Results and discussion CHAPTER 3
80 80 80
60 60 60 60 60 60 40 40 40
20 20 20
40 40 40 µm µm µm 0 0 0 nm nm nm
-20 -20 -20
20 20 20 -40 -40 -40
-60 -60 -60
0 0 0 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 µm µm µm 80 80 80
40 40 40
0 0 0 Height (nm) Height (nm) -40 Height (nm) -40 -40 CENTRE CENTRE CENTRE -80 -80 -80 0 204060800 204060800 20406080 μm μm μm
Rrms = 13.8 nm Rrms = 49.9 nm Rrms = 29.2 nm
Figure 17 Topography of 80 x 80 μm2 scanned areas from the centre of the reaction layers and corresponding line profiles taken transverse of the rolling-sliding direction of the wear track of the steel ball lubricated with DGDE with 0, 2 and 5 wt % ZDDP, after 1 h rubbing at 1.3 GPa
• DGDE + 0 wt% ZDDP: A total coverage of the initial surface, with no visible surface finishing marks, is observed. No distinctive features can, however, be identified in the wear track. • DGDE + 2 wt% ZDDP: Elevated large pads (10-15um in length) elongated in the rolling-sliding direction can be observed in the wear track. Agglomeration of smaller pads (1-5 μm in length) and not elongated in the rolling-sliding direction are also present. The large pads are higher than the surrounding features and are flat and smooth, therefore believed to primarily support the load during the contact [57]. • DGDE + 5 wt% ZDDP: Elevated, smooth, larger pads (20-30 μm in length) can be identified in the wear track and form discontinuous strips. The smaller pads present in the DGDE + 2 wt% ZDDP solution are not present, being the rest of the surface homogeneously covered. This is visible as a reduction of the roughness of the film, which may explain the slight friction reduction with respect to the 2 wt% blend. Figure 18 present a series of AFM images, 80x80 μm2 of the central region of the reaction layer formed on the wear track with the corresponding line profiles taken transverse to the sliding direction of the wear track of the steel ball lubricated with HeD with 0, 2 and 5 wt % ZDDP, after 1 h rubbing.
31 Results and discussion CHAPTER 3
80 80 80
60 60 60 60 60 60 40 40 40
20 20 20
40 40 40 µm µm µm 0 0 0 nm nm nm
-20 -20 -20
20 20 20 -40 -40 -40
-60 -60 -60
0 0 0 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 µm µm 80 µm 80 80 40 40 40 0 0 0 Height (nm) Height (nm) -40 Height (nm) -40 -40 CENTRE CENTRE CENTRE -80 -80 -80 0 204060800204060800 20406080 μm μm μm
Rrms = 11.9 nm Rrms = 32.2 nm Rrms = 48.3 nm
Figure 18 Topography of 80 x 80 μm2 scanned areas from the centre of the reaction layers and corresponding line profiles taken transverse of the rolling-sliding direction of the wear track of the steel ball lubricated with HeD with 0, 2 and 5 wt % ZDDP, after 1h rubbing at 1.3 GPa
• HeD + 0 wt% ZDDP: Slight coverage of the initial surface, being the surface finishing marks still, what may indicate that little wear has occurred • HeD + 2 wt% ZDDP: It is possible to identify elevated areas composed of small streaks, being some bigger features elongated in the rolling-sliding direction, as well as the rest of the film. This structure suggests the load to be more evenly distributed through the reaction layer. • HeD + 5 wt% ZDDP: The morphology of the film is remarkably different. It appears more homogeneous. Attending to the detailed image, representative of the overall structure of this film, the surface is completely covered by pads that seem to have collided while growing.
The polarity of the base oil molecules determines their affinity towards the steel surface. When the polarity of the molecule is high, i.e. DGDE, its affinity towards the surface is low, resulting in a complete coverage of the wear track. While, when the polarity of the molecule is low, i.e. HeD, its affinity for the steel surface is also low, being the steel surface mostly uncovered. When ZDDP is added to the base oils, the ZDDP-derived reaction layers are non homogeneous and remarkably rougher than the layers derived from the base oils alone.
32 Results and discussion CHAPTER 3
3.1.4. Nanomechanical properties of reaction layer To evaluate the nanomechanical properties of the two different layers, the nanoindentation hardness, H, and the reduced elastic modulus, Er, were determined from the load-displacement curves. The hardness was calculated by dividing the maximum indentation load by the corresponding projected contact area, obtained from the area function evaluated at the corresponding maximum indentation depth. Using the Oliver and Pharr [58, 59] approach the reduced elastic modulus was estimated from the slope of the unloading portion of the indentation curve, evaluated at the maximum indentation load, and the projected contact area, which was also estimated for this maximum load. The elastic modulus is proportional to the force required to deform the sample elastically, and the hardness is a measure of the resistance of the material to plastic flow. As an indicator of the overall deformation behaviour of the layer, the reduced elastic modulus to hardness ratio, Er/H, was used to evaluate the resistance of the layers to plastic flow [73]. Imaging of the reaction layer derived from the HeD + 5 wt% ZDDP solution and the nanomechanical properties at selected points, including the Er/H ratio, are summarized in Figure 19 .
HeD + ZDDP
*1 Position Er [GPa] H [GPa] Er/H *2 Point 0 222.2 12.2 18.6 *0 Point 1 192.1 6.9 28.0 Point 2 168.0 6.4 26.1
Figure 19 Images of the selected indentation points inside and outside the HeD + 5wt% ZDDP derived reaction layer and the corresponding nanomechanical properties
The values obtained for Point 0, outside the wear track, correspond to previously reported values of reduced elastic modulus and hardness for the steel substrate [30, 32, 74]. The values for Point 1 and 2, taken inside the wear track, correspond with values reported for phosphate layers [73, 75].
33 Results and discussion CHAPTER 3
The reaction layer derived from the DGDE+5wt%ZDDP has also been analysed, by due to the thickness of the layer. A great influence from the substrate in the results was expected due to the small thickness of the layer. Atomic force microscopy experiments under oil were performed to establish the feasibility of imaging and friction force measurement, and to investigate the possibility of using the AFM tip to simulate a single asperity contact in a tribological situation, thereby providing a route to study additive derived layer formation processes and explain the differences observed in the macro-tribological tests. Atomic force microscopy distance curves indicate the existence of soft surface layer when ZDDP is added to the base oil for any additive concentration.
3.1.5. Chemical composition of reaction layer The influence of base oil polarity in the chemical composition of the ZDDP-derived reaction layers has been studied using X-ray photoelectron spectroscopy. Survey and detailed (high resolution) spectra were recorded on areas at the centre of the reaction layers. In the survey spectra, only signals from carbon, oxygen, sulfur, phosphorus and zinc were detected. Detailed spectra of carbon 1s, oxygen 1s, sulfur 2p, phosphorus 2p and zinc 3s and zinc 2p, were performed.
The survey and detailed spectra acquired on the reaction layer produced at SRR=5% in 2 wt-% solution of ZDDP on SKF Ester. All spectra are shifted to correct for sample charging by referring the main C 1s peak to 285 eV (aliphatic carbon). The C1speak at 284.7 eV was referred to 285 eV from aliphatic carbon (C-C, C -H), thus shifting all peak values. The second contribution can be assigned to C -S bond The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.70 eV. An initial very small peak was detected at 530.3 eV that can be assigned to iron or zinc oxide [37, 61]. Two main contributions were found: a peak at 532.3 eV, non-bridging oxygen (NBO), that can be assigned to phosphorus bound in an (poly)phosphate [26], and a peak at 533.9 eV, bridging oxygen (BO), that can be assigned to P-O-C or P-O-P bonding environment[26, 63]. The BO/NBO ratio indicates the chain length of the phosphates present in the reaction layer. The values calculated according to the recorded signal indicate the presence of polyphosphates.
34 Results and discussion CHAPTER 3
The S2ppeak was fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.7 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [37, 61, 68] and thiols or if sulfur substitutes oxygen in a phosphate [76]. The P2p peak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 134.3 eV, value corresponding to a phosphate [26]. The Zn 3s peak is found at 140.6 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV.
The results obtained for the reaction layer derived from the HeD+5wt% ZDDP, were very similar, showing that the base oil polarity has not a major influence in the chemical composition of the layer. However, further analysis will be performed using Angle- resolved XPS to study the chemical composition of the reaction layer in depth, with a special focus in the near-surface part of the layer, where the polarity of the base oil may affect the composition of the layer as it determines the additive access to the surface.
3.2. Commercial base oils Two commercial base oils of similar chemical and physical properties, but different polarity were selected to validate the previous results on the influence of base oil polarity on additive tribological performance. Also the influence of shear on additive activation and thus its influence on reaction layer formation and the properties of the derived layer are studied.
3.2.1. Friction and wear performance Initially, a friction force curve is capture for both base oil blends, as shown in Figure 20, in order to monitor the friction behaviour in relation to slide-roll ratio (SRR). The Synthetic Oil+ZDDP solution has a higher friction coefficient than the SKF Ester+ZDDP solution with increasing slide-roll ratio. It can also be observed how the absolute friction coefficient is slightly higher for negative SRR for both solutions, whereas the difference in the friction coefficient between both solutions is larger for positive SRR.
35 Results and discussion CHAPTER 3
0.12
0.08 μ 0.04
0 -15 -10 -5 0 5 10 15
-0.04 Friction coefficient,
-0.08
SKF Ester + ZDDP Synthetic Oil + ZDDP -0.12 Slide-roll ratio Figure 20 Friction curve for different base oil-ZDDP solutions
The friction coefficient increases very rapidly when the SRR ranges from 0 to 0.5%, being almost constant for higher SRR values. The slide-roll ratio values selected for further analysis, including reaction layer formation monitoring, cover pure rolling, the initial stage of high friction increase (±0.05 and ±0.5 %) and the steady-state stage ((±2, ±5 and ±10 %). In the initial region of high friction increase it was observed that the friction behavior of both oils was very similar, very close to pure rolling which might indicate that in the absence of shear the base oil molecules, that are initially attach to the steel surface [70], are primarily responsible for the lubrication of the surface as the additive molecules have not yet been activated
The antiwear properties of the reaction layers derived from the base oil + ZDDP solutions at different slide-roll ratios were also investigated. The wear track width (WTW) was measured from the balls using a calibrated microscope. Figure 21 shows the wear track width as a function of slide-roll ratio.
36 Results and discussion CHAPTER 3
620 SKF Ester+ZDDP Synthetic Oil+ZDDP 600
580
560
Hertzian point contact diameter 540540 Wear Scar Width (um)
520
500 -10% -5% -2% -0.50% 0% 0.50% 2% 5% 10% Slide-roll ratio
Figure 21 Wear track width as a function of slide-roll ratio
The WTW for Synthetic Oil+ZDDP solutions at different slide-roll ratios is always higher than those of SKF Ester+ZDDP. The wear scar increases with the slide-roll ratio for both solutions in either direction. The WTW for both solutions is very similar at SRR=-10%, however the values at SRR=10% are very different, being the WTW for SKF Ester Oil+ZDDP is approximately 50% lower than the corresponding Synthetic Oil+ZDDP. The higher wear exhibited by the layers formed by Synthetic Oil+ZDDP, despite the layer being thicker than the ones derived from SKF Ester+ZDDP solutions, may be due to the formation of thicker layer which cannot withstand the load and break away, as indicated by the morphology observed for those layers (see Section 4.4). The layers derived from the SKF Ester+ZDDP solutions at different slide-roll ratios, show WTW even lower than the contact diameter calculated from the Hertz theory for point contacts, at low SRR (ranging from -2 to 2%). At low SRR, the energy input to the system is still low, as is the activating rate of the additive; therefore the lubrication of the contacting surfaces is coming mainly from the base oil molecules. The polar base oil molecules have a higher affinity for the surface, attaching to it strongly, and thus providing a better protection for the contacting surfaces,
37 Results and discussion CHAPTER 3 explaining the lower WTW when compared to the wear track derived from the Synthetic Oil +ZDDP.
3.2.2. Reaction layer formation In situ spacer layer interferometry has been used to monitor the formation of reaction layers derived from ZDDP solutions on steel surfaces. Tests at different slide-roll ratio, in positive and negative direction, but with the same contact pressure, temperature and entrainment speed have enabled the characterization of the effect of base oil polarity, extended the previous studies with low viscosity model oils to commercial base oils, and shear stress on reaction layer formation. To illustrate the effect of base oil polarity on reaction layer formation, the results for a SRR=-5% are shown in Figure 23, as mean reaction layer thickness as a function of rubbing distance. The mean layer thickness is calculated from the line profiles taken across the corresponding interferometry images.
150 SKF Ester + ZDDP Synthetic Oil + ZDDP
100
50 Mean reaction layer thickness (nm)
0 0 500 1000 1500 2000 2500 Rubbing distance (m) Stage 1 ACTIVATION (Rate of formation > Rate of removal) - Reaction layer formation is very rapid Initial very rapid reaction - Process catalysed by chemical species generated or released SATURATION from the steel surface by rubbing - Amount of catalytic species proportional to rubbing distance Stage 2 WEARING-OUT (Rate of formation < Rate of removal)
Stage 3 EQUILIBRIUM (Rate of formation ~ Rate of removal) - Reach of “limiting thickness”
Figure 22 Stages on reaction layer formation
38 Results and discussion CHAPTER 3
It is possible to identify three different stages on the reaction layer formation process, as shown in Figure 22. Initially, an ACTIVATION stage can be identified, with a rate of layer formation higher that the removal rate: The additive molecules are activated by the tribological energy and approach the surface. The growth mechanism begins with distinct reaction events on micro-asperity contact at the steel surfaces. The reaction layer develops initially very rapidly as a consequence of the very rapid reaction, being the rate of formation higher that the rate of removal. This suggests that the layer formation may be strongly catalysed by chemical species generated or released during rubbing, such as soluble Fe2+ or Fe3+.By ligand exchange, the iron ion replace the zinc in ZDDP to form a less thermally stable metal dithiophosphate, that subsequently decomposes [77]. Thus it is possible that rubbing releases iron species, probably initially Fe2+ and which rapidly exchanges with ZDDP to form FeDDP. This iron dithiophosphate then decomposes at low temperatures to form iron phosphate glass in a reaction similar to that which occurs for ZDDP at much higher temperatures. This would also establish the autocatalytic reaction so that ZDDP can subsequently decompose directly.
The process can also be triggered by triboelectronic processes such as exoelectron emission and a subsequent negative ion reaction [78].
The reaction layer rapidly develops until reaching saturation: At this point the steel surface is completely covered by the reaction layer, which slows the rate of reaction layer formation. The next stage involves a WEARING-OUT of the layer, being the rate of formation lower than the rate of removal, until the layer reaches the final stage of EQUILIBRIUM,
39 Results and discussion CHAPTER 3 between the rate of formation and removal, when the layer reaches a constant thickness value, known as “limiting thickness” [79]. It can be observed how the polarity of the base oil influences the reaction of the ZDDP additive with the steel surface. A thicker layer is formed when the additive is blended in the non-polar oil, due to the higher affinity of the polar base oil molecules for the steel surface, that limit the access of the additive molecules to the surface and therefore their ability to attach and react with it to form a protective reaction layer. A series of interference images of the centre of the wear track derived from the Synthetic Oil group IV+ZDDP solution from the rubbing tests at different negative slide-roll ratios are shown in Figure 23.
Figure 23 Series of interference images from wear track for ZDDP2 (Synthetic Oil Group IV+2% iso-C4-ZDDP) at a entrainment speed of 0.25 m/s, 1.9 GPa, 90 °C at different negative slide-roll ratios
Figure 24 presents in detailed the results for the negative slide-roll ratios tested (for clarity (-) 0.05% is not plotted as the reaction layer formation was negligible). The reaction layer develops initially very rapidly, indicating a fast reaction, to subsequently slow down as the reaction layer grows to a stable thickness value (“limiting thickness”).
40 Results and discussion CHAPTER 3
It is also remarkable how, independently of the reaction layer thickness values reached in this initial stage, the layers reach a limiting thickness, being the values for all the SRR tested very similar (≈ 60-70 nm). The limiting thickness is determined by a balance between the rate of growth, depending on the additive concentration in base oil, contact temperature, and the rate of removal, determined by wear [51]. From these results we can conclude that SRR, once present, has a minor effect on the limiting thickness of the layer, but they also show how shear rate is a fundamental parameter on the initial activation of the additive towards the surface.
150 Synthetic Oil+ZDDP (-) 10%Slip Synthetic Oil+ZDDP (-) 5%Slip Synthetic Oil+ZDDP (-) 2%Slip Synthetic Oil+ZDDP (-) 0.5%Slip Synthetic Oil+ZDDP 0%Slip 100
50 Mean reaction layer thickness (nm) 0 0 500 1000 1500 2000 2500 3000 Rubbing distance (m) Figure 24 Influence of negative SRR on the growth of mean ZDDP-derived (Synthetic Group IV+ZDDP) reaction layer thickness with rubbing distance calculated from images in Figure 23
The overshoot in reaction layer thickness observed at the initial very rapid reaction layer formation, an effect also reported at low ratios and high temperatures [41, 77, 80], may indicate an initially very rapid reaction, which subsequently slows as a reaction layer develops, indicating a possible process catalysed by rubbing steel on steel.
3.2.3. Morphology and topography of reaction layer The morphology of the reaction layers has been studied using Atomic Force Microscopy. A series of topography images of the layers at the end of the total rubbing time are shown Figure 25.
41 Results and discussion CHAPTER 3
SRR =-0.5% SRR =-2% SRR =-5% SRR =-10%
80 80 80 80
100 100 100 100 60 60 60 60
50 50 50 50
40 40 40 µm 40 µm µm µm 0 0 0 0 nm nm nm nm
-50 -50 -50 -50 20 20 20 20
-100 -100 -100 -100
0 0 0 0 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 µm 150 µm 150 150 µm 150 µm 75 75 75 75 0 0 0 0 -75 -75 -75 -75 Height (nm) Height (nm) Height (nm) Height (nm) -150 -150 -150 -150 0 20406080 0 20406080 0 20406080 0 20406080 Contact (um) Contact (um) Contact (um) Contact (um)
Rrms = 16.1 nm Rrms = 15.8 nm Rrms = 34.4 nm Rrms = 57.3 nm
Figure 25 Topography of 80 x 80 μm2 scanned areas from the centre of the reaction layers and corresponding line profiles taken transverse of the rolling-sliding direction
At SRR=-0.5%, the surface is homogeneously covered, but no distinctive features can be identified. When the slide-roll ratio increases to SRR=-2%, a pad-like structure develops. The pads, 10-15μm in length, are higher and smoother, that the surroundings and are orientated in the rolling-sliding direction, and are consider as the load carrying features of the layer [32, 57], being therefore responsible for the antiwear protection. At SRR=-5%, the overall structure of the layer is very similar, but it start showing signs of a wearing-out, as part of the layer seem to be flaking off. At SRR=-10%, the flakes have developed further covering the whole reaction layer surface. The roughening of the layer that can be observed, it is in direct relation with the friction behaviour
It has been observed that a delamination or flaking process of the reaction layer derived from Synthetic oil + ZDDP solution occurs in this last stage when the SRR is >5%. The nanomechanical properties of the layer can be responsible for this and will be further investigated. Also the possible relation between this delamination process and crack formation in the surface will be address, by long run tests.
3.2.4. Chemical composition of reaction layer The influence of base oil polarity and SRR in the chemical composition of the ZDDP- derived reaction layers has been studied using X-ray photoelectron spectroscopy. Survey
42 Results and discussion CHAPTER 3 and detailed (high resolution) spectra were recorded on areas at the centre of the reaction layers. In the survey spectra, only signals from carbon, oxygen, sulfur, phosphorus and zinc were detected. Detailed spectra of carbon 1s, oxygen 1s, sulfur 2p, phosphorus 2p and zinc 3s and zinc 2p, were performed.
O1s
Fe 2p
S2p
C1s Zn 3s P2p
C1s O1s
NBO
BO oxide
P2p S2p
S2p3/2 P2p3/2
Zn 3s S2p3/2 P2p3/2
Figure 26 XPS survey and detailed spectra of C1s, O1s, P2p, Zn3s and S2p measured in the reaction layer derived from SKF Ester+ZDDP at SRR=5%
43 Results and discussion CHAPTER 3
Figure 26 shows the survey and detailed spectra acquired on the reaction layer produced at SRR=5% in 2 wt-% solution of ZDDP on SKF Ester. All spectra are shifted to correct for sample charging by referring the main C 1s peak to 285 eV (aliphatic carbon). The C1speak at 284.7 eV was referred to 285 eV from aliphatic carbon (C-C, C -H), thus shifting all peak values. The second contribution can be assigned to C -S bond The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.70 eV. An initial very small peak was detected at 530.3 eV that can be assigned to iron or zinc oxide [37, 61]. Two main contributions were found: a peak at 532.3 eV, non-bridging oxygen (NBO), that can be assigned to phosphorus bound in an (poly)phosphate [26], and a peak at 533.9 eV, bridging oxygen (BO), that can be assigned to P-O-C or P-O-P bonding environment[26, 63]. The BO/NBO ratio indicates the chain length of the phosphates present in the reaction layer. The values calculated according to the recorded signal indicate the presence of polyphosphates. The S2ppeak was fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.7 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [37, 61, 68] and thiols or if sulfur substitutes oxygen in a phosphate [76]. The P2p peak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 134.3 eV, value corresponding to a phosphate [26]. The Zn 3s peak is found at 140.6 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV.
44 Results and discussion CHAPTER 3
O 1s
Fe 2p
S2p C1s Zn 3s P2p
C1s O1s
NBO
BO oxide
P2p S 2p
P2p S 2p3/2 Zn 3s 3/2
S2p3/2 P2p3/2
Figure 27 XPS survey and detailed spectra of C1s, O1s, P2p, Zn3s and S2p measured in the reaction layer derived from Synthetic Oil+ZDDP at SRR=5%
Figure 27 shows the survey and detailed spectra acquired on the reaction layer produced at SRR=5% in 2 wt-% solution of ZDDP on SKF Ester. All spectra are shifted to correct for sample charging by referring the main C 1s peak to 285 eV (aliphatic carbon).
45 Results and discussion CHAPTER 3
Two peaks were used to fit the C1sspectra, the main peak being found at 285.0 eV assigned to aliphatic carbon (C-C, C -H). The second contribution can be assigned to C-S bonds. The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.65 eV. An initial very small peak was detected at 529.8 eV that can be assigned to iron or zinc oxide. Two main contributions were found: a peak at 531.8 eV, non-bridging oxygen (NBO) that can be assigned to phosphorus bound in an orthophosphate, and a peak at 533.4 eV, bridging oxygen (BO), that can be assigned to P-O-C or P- O -P bonding environment. The BO/NBO ratio indicates the chain length of the phosphates present in the reaction layer. The values calculated according to the recorded signal indicate the presence of polyphosphates. The S2phas two contributions, and both peaks were fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.9 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [37, 61, 68] and thiols or if sulfur substitutes oxygen in a phosphate [76]. The peak found at 168.9 eV corresponds with values reported for an oxidation state of -6, S(VI) that can be assigned to sulfate groups [37]. The P 2p peak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 133.7 eV. This value is typical for a phosphate. The Zn 3s peak is found at 140.5 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV.
Both reaction layers are very similar, being the main difference in the sulfur content and the oxidation state of the sulfur. In the layer derived from the polar base oil, the sulfur is present as a sulfide, whereas in the reaction layer derived form the non-polar base oil, sulfides and sulfates re present. The reaction layers for the negative slide-roll ratio for both base oils, have a similar composition to the one presented for the positive slide-roll ratio. The main significant difference is that the iron content for the negative SRR is very low, nearly not detectable.
46 Conclusions and future work CHAPTER 4
CHAPTER 4 4. Conclusions and future work
4.1. Conclusions The polarity of the base oil determines the tribological performance of the additives. The same type of additive behaves differently depending on the type of oil it is blended with. Base oil polarity influences the transport of additives to the surface thereby controlling the growth rate and maximum reaction layer thickness, friction and wear. It also influences structure and characteristics of the reaction layers. However, the chemical characterization of the reaction layers performed with XPS has shown little differences in the chemical composition depending on the polarity of the base in which the additive is blended. No traces of additives were found outside the wear tracks, proving that the reaction between the additive and the steel surface happens at asperity level, where the real contact occurs and that tribological energy is needed for activating the reaction between the additive and the surface. Shear was found to be a fundamental parameter on the activation of additives in the surface and influence the properties of the derived reaction layer Summarizing the results of this study we can conclude that: Reaction layer thickness Reaction layer kinetic
High lambda Low lambda Low temperature High temperature Low additive concentration High additive concentration High polarity base oil Low polarity base oil Low slide-roll ratio High slide-roll ratio
47 Conclusions and future work CHAPTER 4
4.2. Future work The relation between surface damage and reaction layer formation and characteristics, as a function of base oil polarity and operation conditions, will be further studied. Longer running tests will be carried out in order to monitor the wearing off of the layer under different operation conditions and address the influence on friction and wear. The mechanisms leading to the formation and removal of the additive-derived reaction layer and the characteristics of the reaction layer will be addressed. The effect of base oil polarity and operating conditions on ashless additives, phosphorus- based and sulfur-based, tribological performance will also be investigated The objective is to create a model for the formation and removal of the layer as a function of parameters, i.e. base oil polarity and slide-roll ratio, affecting the functionality of additives in lubricated contacts, and the effect on tribological performance.
48 REFERENCES
1. Stachowiak, G.W. and Batchelor,A. W., Engineering Tribology. 3rd ed. 2005, London: Elsevier Butterworth-Heinemann Ltd. 2. Hutchings, I.M., Tribology - Friction and wear of engineering materials, London: Elsevier Butterworth-Heinemann Ltd. 3. Bowden, F.P. and Tabor, D., The friction and lubrication of solids. 2001, New York: Oxford University Pres Inc. 4. Friction, lubrication and wear technology. ASM Handbook. 1992. 5. Bovington, C.H., Friction, wear and the role of additive in their control,in Chemistry and technology of lubricants, R.M. Mortier and S.T. Orszulik, Editors. 1997, Blackie and Son: London and Glasgow. 6. Sakamoto, T., et al., The reaction layer formed on steel by additives based on sulphur and phosphorus compounds under conditions of boundary lubrication. Wear, 1982. 77(2): p. 139-157. 7. Matsumoto, K., Surface chemical and tribological investigations of phosphorus- containing lubricant additives: EHT Zurich. 8. Sarin, R., et al., Additive-additive interactions: Search for synergistic FM-EP-AW composition. Wear, 1994. 174(1-2): p. 93-102. 9. Jianqiang, H., et al., Study on tribological properties and action mechanism of organic cadmium compound in lubricants. Wear. 259(1-6): p. 519-523. 10. Zhang, J., et al., Investigation of the friction and wear behaviors of Cu(I) and Cu(II) dioctyldithiophosphates as additives in liquid paraffin. Wear, 1998. 216(1): p. 35-40. 11. Boshui, C., et al., Tribochemistry of gadolinium dialkyldithiophosphate. Wear, 1996. 196(1-2): p. 16-20. 12. Barnes, A.M., et al., A review of zinc dialkyldithiophosphates (ZDDPs): characterisation and role in the lubricating oil. Tribology International, 2001. 34: p. 389-395. 13. Georges, J.M., et al., Mechanism of boundary lubrication with zinc dithiophosphate. Wear, 1979. 53(1): p. 9-34. 14. Watkins, R.C. and Spedding, H., Antiwear mechanism of ZDDP's - 2. Tribology international, 1982. 15(1): p. 13-15. 15. Spedding, H. and Watkins, R.C., Antiwear mechanism of ZDDP's - 1. Tribology international, 1982. 15(1): p. 9-12. 16. Gellman, A.J. and Spencer, N.D., Surface chemistry in tribology. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2002. 216(6): p. 443-461. 17. Wystrach, V.P., et al., Basic zinc double salts of O,O-dialkyl phosphorodithioic acids. Journal of organic chemistry, 1956. 21(6): p. 705-707. 18. Yamaguchi, E.S., et al., Dynamic light scattering studies of basic diisobutyi zinc dithiophosphate©. Tribology Transactions, 1998. 41(2): p. 233-240. 19. Yamaguchi, E.S., et al., Comparison of the relative wear performance of neutral and basic ZnDTP salts. ASLE transactions, 1996. 39(1): p. 220-224. 20. Kikabhai, T., et al., Proton and phosphorus 31 NMR study of ZDDP in solution. Journal of the Chemical Society. Dalton transactions, 1987. 807-813: p. 807-813.
49 21. Yamaguchi, E.S., et al., Dynamic light scattering studies of neutral diisobutyl zinc dithiophosphate©. Tribology Transactions, 1997. 40(2): p. 330-338. 22. Spikes, H., The history and mechanisms of ZDDP. Tribology Letters, 2004. 17(3): p. 469-89. 23. Bovington, C.H.,et al., Adsorption and reaction of decomposition products of zinc di-isopropyldithiophosphate on steel. ASLE transactions, 1984. 27(3): p. 252-258. 24. Fuller, M., eta l., Chemical characterization of tribochemical and thermal films generated from neutral and basic ZDDPs using X-ray absorption spectroscopy. Tribology international, 1997. 30(4): p. 305-315. 25. Suominen Fuller, M.L., et al., Solution decomposition of zinc dialkyl dithiophosphate and its effect on antiwear and thermal film formation studied by X-ray absorption spectroscopy. Tribology International, 1998. 31(10): p. 627-644. 26. Piras, F.M., et al., Combined in situ (ATR FT-IR) and ex situ (XPS) study of the ZnDTP-iron surface interaction. Tribology Letters, 2003. 15(3): p. 181-191. 27. Coy, R.C., eta l., Thermal degradation and EP performance of zinc dialkyldithiophosphate additives in white oil. ASLE transactions, 1981. 24(1): p. 77-90. 28. Aktary, M., et al., Morphological evolution of films formed from thermooxidative decomposition of ZDDP. Wear, 2001. 247(2): p. 172-179. 29. Fujita, H., et al.,, The formation of zinc dithiophosphate antiwear films. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2004. 218(4): p. 265-277. 30. Aktary, M., et al., Morphology and nanomechanical properties of ZDDP antiwear films as a function of tribological contact time. Tribology Letters, 2002. 12(3): p. 155-162. 31. Willermet, P.A., et al., Mechanism of formation of antiwear films from zinc dialkyldithiophosphates. Tribology International, 1995. 28(3): p. 177-187. 32. Graham, J.E., et al., Topography and nanomechanical properties of tribochemical films derived from zinc dialkyl and diaryl dithiophosphates. Tribology Letters, 1999. 6(3): p. 149-57. 33. Martin, J.M., et al., The two-layer structure of Zndtp tribofilms, Part I: AES, XPS and XANES analyses. Tribology International, 2001. 34(8): p. 523-530. 34. Bec, S., et al., Relationship between mechanical properties and structures of zinc dithiophosphate anti-wear films. Proceedings: Mathematical, Physical and Engineering Sciences, 1999. 455(1992): p. 4181-4203. 35. Gellman, A.J., et al., Surface chemistry in tribology. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2002. 216(6): p. 443-461. 36. Heuberger, R., et al., Pressure Dependence of ZnDTP Tribochemical Film Formation: A Combinatorial Approach. Tribology Letters, 2007. 28(2): p. 209- 222. 37. Heuberger, R., et al.,, XPS study of the influence of temperature on ZnDTP tribofilm composition. Tribology Letters, 2007. 25(3): p. 185-196. 38. Heuberger, R.C., Combinatorial study of the tribochemistry on anti-wear lubricant additives. 2007, EHT: Zurich.
50 39. Torrance, A., et al., An additive’s influence on the pitting and wear of ball bearing steel. Wear, 1996. 192: p. 66-73. 40. Tripaldi, G., et al., Friction behaviour of ZDDP films in the mixed, boundary/EHD regime. SP, 1996. 1209: p. 73-84. 41. Fujita, H., et al., Study of zinc dialkydithiophosphate antiwear film formation and removal processes, part I: Experimental. Tribology Transactions, 2005. 48(4): p. 558-566. 42. Fujita, H., et al., Study of zinc dialkyldithiophosphate antiwear film formation and removal processes, part II: Kinetic model. Tribology Transactions, 2005. 48(4): p. 567-575. 43. Wan, G.T.Y., et al., Effect of extreme-pressure additives on fatigue life of rolling bearings. Journal of physics. D, applied physics, 1992. 25(1): p. A147-A153. 44. Evans, R.D., et al., Effects of extreme pressure additive chemistry on rolling element bearing surface durability. Tribology international, 2007. 40(10): p. 1649-1654. 45. Minami, I., et al., Investigation of anti-wear additives for low viscous synthetic esters: Hydroxyalkyl phosphonates. Tribology international, 2007. 40(4): p. 626- 631. 46. Minami, I., et al, Anti-wear additives for ester oils. Technische Akademie Esslingen International Tribology Colloquium Proceedings, 2004. 14 III: p. 1593- 1598. 47. Kar, P., et al., Formation and characterization of tribofilms. Journal of tribology, 2008. 130(4). 48. Taylor, L., et al., Friction and wear behavior of zinc dialkyldithiophosphate additive. Tribology Transactions, 2000. 43(3): p. 469-479. 49. Bhushan, B., Nanotribology and nanomechanics - An introduction. 2005, Berlin: Springer. 5. 50. Beake, B.D., et al., Friction force microscopy study of diamond films modified by a glow discharge treatment. Diamond Realted Materials, 2000. 9: p. 8. 51. Palacios, J.M, et al., Films formed by antiwear additives and their incidence in wear and scuffing. Wear, 1987. 114(1): p. 41-49. 52. Gustafsson, L., E. Hoglund, and O. Marklund, Measuring lubricant film thickness with image analysis Proceedings - IMechE: J, Journal of Engineering Tribology. 1994. 208. 53. Lord, J., et al., A hybrid film thickness evaluation scheme based on multi-channel interferometry and contact mechanics. Journal of tribology, 2000. 122(1): p. 16- 22. 54. Cann, P.M., et al., The development of a spacer layer imaging method (SLIM) for mapping elastohydrodynamic Contacts©. Tribology Transactions, 1996. 39(4): p. 915-921. 55. Choo, J.W., et al., Influence of Surface Roughness Features on Mixed-Film Lubrication. Lubrication Science, 2003. 15(3): p. 219-232. 56. Topolovec-Miklozic, K., et al., Film thickness and roughness of ZDDP antiwear films. Tribology Letters, 2007. 26(2): p. 161-171. 57. Warren, O.L., et al., Nanomechanical properties of films derived from zinc dialkyldithiophosphate. Tribology Letters, 1998. 4(2): p. 189-198.
51 58. Oliver, W.C., et al., Improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of materials research, 1992. 7(6): p. 1564-1580. 59. Oliver, W.C., et al., Measurement of hardness and elastic modulus by instrumented indentation: Advances in understanding and refinements to methodology. Journal of materials research, 2004. 19(1): p. 3-20. 60. Bird, R.J., et al., The application of photoelectron spectroscopy to the study of e. p. films on lubricated surfaces. Wear, 1976. 37(1): p. 143-167. 61. Heuberger, R., et al., Pressure Dependence of ZnDTP Tribochemical Film Formation: A Combinatorial Approach. Tribology Letters, 2007. 28(2): p. 209- 222. 62. Bell, J.C., et al.,, Paper IX (ii) The Removal of Substrate Material through Thick Zinc Dithiophosphate Anti-Wear Films,inTribology Series,. D. Dowson, Editor. 1992, Elsevier. p. 387-396. 63. Eglin, M., et al.,, X-ray Photoelectron Spectroscopy Analysis of Tribostressed Samples in the Presence of ZnDTP: A Combinatorial Approach. Tribology Letters, 2003. 15(3): p. 199-209. 64. Brow, R.K., Review: the structure of simple phosphate glasses. Journal of Non- Crystalline Solids, 2000. 263-264: p. 1-28. 65. Grossiord, C., et al., UHV friction of tribofilms derived from metal dithiophosphates. Tribology Letters, 1999. 6(3): p. 171-179. 66. Brückner, R., et al., XPS measurements and structural aspects of silicate and phosphate glasses. Journal of Non-Crystalline Solids, 1980. 42(1-3): p. 49-60. 67. Onyiriuka, E.C., et al., Zinc phosphate glass surfaces studied by XPS. Journal of Non-Crystalline Solids, 1993. 163(3): p. 268-273. 68. Martin, J.M., et al., The two-layer structure of Zndtp tribofilms, Part I: AES, XPS and XANES analyses. Tribology International, 2001. 34(8): p. 523-530. 69. Choa, S., et al., Model for the boundary film formation and tribological behavior of a phosphazene lubricant on steel. ASLE transactions, 1995. 38(4): p. 757-768. 70. Naveira-Suarez, A., et al. Parameters affecting the functionality of additives in tribological contacts - An experimental and Molecular Dynamics simulation study. in IV World Tribology Congress. 2009. Kyoto, Japan. 71. Taylor, L.J., et al., Friction-enhancing properties of ZDDP antiwear additive: Part II - Influence of ZDDP reaction films on EHD lubrication. Tribology Transactions, 2003. 46(3): p. 310-314. 72. Taylor, L..J., et al., Friction-enhancing properties of ZDDP antiwear additive: Part I - Friction and morphology of ZDDP reaction films. Tribology Transactions, 2003. 46(3): p. 303-309. 73. Komvopoulos, K., et al., Nanomechanical and nanotribological properties of an antiwear tribofilm produced from phosphorus-containing additives on boundary- lubricated steel surfaces. Proceedings of the ASME/STLE International Joint Tribology Conference, IJTC 2004, 2004(PART B): p. 1631-1636. 74. Ye, J., et al., Evaluation of Local Mechanical Properties in Depth in MoDTC/ZDDP and ZDDP Tribochemical Reacted Films Using Nanoindentation. Tribology Letters, 2002. 13(1): p. 41-47.
52 75. Komvopoulos, K., et al., X-ray photoelectron spectroscopy analysis of antiwear tribofilms produced on boundary-lubricated steel surfaces from sulfur- and phosphorus-containing additives and metal deactivator additive. Tribology transactions, 2004. 47(3): p. 321-327. 76. Rossi, A., et al., Surface analytical studies of surface-additive interactions, by means of in situ and combinatorial approaches. Wear, 2004. 256(6): p. 578-584. 77. Fujita, H., et al., The formation of zinc dithiophosphate antiwear films. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2004. 218(4): p. 265-277. 78. Kajdas, C., et al., Importance of the triboemission process for tribochemical reaction. Tribology International, 2005. 38(3): p. 337-353. 79. Palacios, J.M., et al., Thickness and chemical composition of films formed by antimony dithiocarbamate and zinc dithiophosphate. Tribology International, 1986. 19(1): p. 35-39.
53
Paper A
THE INFLUENCE OF BASE OIL POLARITY ON THE TRIBOLOGICAL PERFORMANCE OF ZINC DIALKYL DITHIOPHOSPATE ADDITIVES
A. Naveira Suarez1,2*, Mattias Grahn1,3, Rihard Pasaribu1 and Roland Larsson2
1SKF Engineering and Research Centre, Kelvinbaan 16, 3439 MT, Nieuwegein, The Netherlands 2Division of Machine Elements, Luleå University of Technology, SE 97187, Luleå, Sweden 3Division of Chemical Engineering, Luleå University of Technology, SE 97187, Luleå, Sweden *Corresponding author (E-mail: [email protected])
ABSTRACT This study focuses on the effect of base oil polarity on the tribological performance of a primary ZDDP additives. Two types of model base oils, i.e. hexadecane (non-polar) and diethylen glygol dibuthyl ether (polar), pure and with the addition of 2 and 5 wt% ZDDP, were tested in the WAM5 ball-on-disc rig at different contact pressures (Maximum Hertzian contact pressure 1.3 and 2.4 GPa) and 5% slide-roll ratio. The results showed significant differences in terms of friction and wear between the two solutions, i.e. hexadecane + C4-ZDDP and diethylen glycol dibuthyl ether + C4-ZDDP. Under the studied conditions the addition of ZDDP to the base oils increases friction as also reported in the literature. However, the degree of friction increased for the hexadecane is higher than that of diethylen glygol dibuthyl ether. Different surface rubbing track roughness was observed for different oil polarity and tribological conditions. Further, significant difference in reaction layer thickness for the two different solutions (diethylenglycol + ZDDP and hexadecane + ZDDP) is observed. The difference on the reaction layer thickness is to be related to the observed differences in the measured roughness and friction increaseThese results display the importance of base oil polarity on the action of ZDDP additives in reaching the contacting surfaces. Scanning Electron Microscope (SEM) and Energy Dispersive X-ray (EDS) analysis showed an uneven distribution of the chemical elements on the rolling track. Elements contained in the molecular structure of the additive (Zn, P and S) were mainly found at the asperity tip, where the real contact actually occurs. The plausible mechanism on the effect of base oil polarity on the performance of ZDDP additives is discussed.
KEY WORDS: ZDDP, base oil polarity, friction, reaction layer thickness
1 1. INTRODUCTION Commercially available lubricants contain base oil and functional additive packages to achieve a desired performance for a specific application. Antiwear additives are used to react with the surfaces in contact to form a protective layer, which ensures that sporadic asperity contacts do not lead to severe wear [1]. Typically, these additives are phosphorus based, zinc dialkyl dithiophosphates (ZDDP) being the most frequently used in engine oils . ZDDP was initially used as an antioxidant, but their excellent antiwear properties were quickly recognized. It can also act as mild extreme pressure agent and corrosion inhibitor. ZDDP additives have been the object of a great deal of research due to that multifunction performance and the complexity of the mechanism that lead to it. The antiwear function of ZDDP is linked to the formation of sacrificial reaction layers on rubbing surfaces from ZDDP decomposition products rather than from intact ZDDP molecules. Although these additives may also reduce wear by preventing hydroperoxides from accumulating in the oil. A variety of mechanisms have been proposed [2], involving oxidative (by reaction with hydroperoxides or peroxy radicals) [3], catalytic (chemisorption on metal, hydrolytic [4]), and thermal [5] decomposition of the ZDDP. The resulting reaction layers have a heterogeneous composition with the chemical structure of the starting materials dictating their chemical composition [6, 7] and mechanical properties [8]. Examination of ZDDP tribofilms by X-ray absorption spectroscopy (XAS) has shown a phosphate structure coordinated by zinc and iron cations from ZDDP and the substrate [2, 6], Auger electron spectroscopy was used to show that the surface layer of phosphate is placed on top of a chemisorbed iron sulfide layer at the film/metal interface [9]. Using internal reflectance (IR), the antiwear film largely consisted of short-chain amorphous phosphates with metal cations [3, 10]. Tribofilms form from secondary alkyl ZDDP on steel surface are composed of a mixture of short and long chains of polyphosphate chemically inter-grown with the metal oxide surface [11] and have a high chemical and mechanical stability. Further studies using X-ray absorption near edge structure (XANES) found that the film has a layered structure with short-chain polyphosphates in the bulk, grading to pyro- or ortho-phosphates [12], and longer-chain polyphosphates on top [7]. A two layer structure for ZDDP derived tribofilms has been developed using a multi-technique approach (AES/XPS/XANES) [13]. The ZDDP solid-like tribofilm has a thin long-chain polyphosphate layer at superposed to a thicker short-chain mixed iron and zinc polyphosphate layer with gradient concentration, containing embedded nanocrystallites of metal sulphide precipitates (ZnO and ZnS). No oxide or sulfide layer was found in the interface between the phosphate and the steel surface. Scanning Force Microscopy (SFM) and the related technique interfacial force microscopy (IFM) have been used to examined ZDDP derived layers formed by thermal decomposition and by wear testing, showing a different morphology and nanomechanical properties [14]. Atomic force microscopy (AFM) investigations on the nature of the antiwear reaction layers have shown that they are laterally and vertically heterogeneous[15], but a common structure of ridge and valley regions can be identified. The ridge regions are composed by large pads, elevated flat regions elongated in the sliding direction, identified as load bearing areas, which possess higher indentation * modulus (ES ~85GPa) and hardness than the surrounding valley region, composed of * smaller pads (ES ~30GPa) [15]. The large pads are also laterally heterogeneous, with the
2 centre of the pads being stiffer than the edges. The chemical composition [16] and morphology of the reaction layers have been found to evolve with rubbing time, from a patchy initial structure to a densely packed series of pads that are concatenated in the sliding direction and form an almost complete layer over the steel surface, whereas the nanomechanical properties are mostly independent of the rubbing time [17]. Several studies have proved ZDDP to have detrimental effects on wear under certain operation conditions [18], and to enhance a friction increase when the system is operating in mixed and boundary lubrication regimes [19-21]. The friction and wear behaviour of tribological systems lubricated with oils containing ZDDP have been mainly studied as a function of the additive structure, concentration, interaction with other additives or operation conditions, but the effect of the base oil in the ZDDP tribological performance has been hardly addressed [22-24]. When considering an additive molecule approaching the steel surface in a lubricating oil, it must be considered that the steel surface is already completely covered with oil molecules. Hence, it can be stated that the additive molecule has to compete with the oil for a place on the steel surface. The polarity of a molecule can be used to measure its affinity towards the steel surface. When the polarity of the molecule is low, its affinity towards the surface is low, while when the polarity of the molecule is high, its affinity for the steel surface is also high. Parameters, like polarity, that describe the base oil-additive interactions are needed in order to have a good running in, friction and wear model of lubricated contacts. The work presented here shows an attempt to propose a base oil-additive parameter useful in modelling base oil-additive interaction. Our results shows that the tribological performance of ZDDP is different when blended in base oils with different polarity as presented further in the following sections.
2. MATERIALS AND METHODS 2.1. Lubricants and additives In order to study the influence of base oil polarity on tribological performance two low viscosity model base oils were selected. A polar base oil, diethylen glycol diethyl ether (DGDE) and a non-polar oil, n-hexadecane (HeD), both supplied by Acros Organics, NJ, USA. Iso-butyl-zinc dithiophosphate (A&S Chemie, Tubingen, Germany), with 99% purity, is employed in simple solution in both base oils without other additives present. The physical properties of the base oils and the additive are summarized in Table 1 and the molecular structures of the base oils and additive molecules are presented in Figure 1.
(a)
(b)
3 (c)
Figure 1 Molecular structure of (a) diethylen glycol diethyl ether, (b) n-hexadecane and (c) primary zinc dialkyldithiophosphate
Table 1 Properties of the tested base oils and additive
Viscosity at Dipole Density Code Description 40°C moment [kg/l] [mm/s2] [D] DGDE Diethylen glycol diethyl ether 1.936 0.883 0 HeD Hexadecane 2.982 0.773 1.12 ZDDP Zinc dialkyl dithiophosphate ------2.91
2.2. Test conditions WAM5 ball-on-disc test rig (Wedeven Associates, Inc., Edgmont, PA, USA) was chosen to evaluate using a single contact method, the tribological performance of additive/base oil blends at controlled conditions of temperature, load, speed and mixed rolling/sliding conditions. The ball and the disc are independently driven giving the possibility to simulate pure rolling and various slide-roll ratios (SRR). Rolling/sliding conditions were selected in stead of the most common pure sliding because wear is evenly-distributed around the tracks of both specimens which avoid the change in contact geometry and thus contact pressure and temperature that occurs during pure sliding tests [25]. After the ring was assembled, the temperature was raised to the test value, without applying load. Load was then applied; ball-ring relative position corrected to pure rolling and the desired slide-roll ratio was set afterwards. The ring is constantly lubricated, using a recirculation pumping system attached to the test rig.
Figure 2 Schematic representation of WAM5 ball-on-disc test rig
The steel balls (20 mm diameter) were of AISI 52100 steel with hardness 59-66 HRC and an average roughness (Ra) of 10 nm. The rings were washers (WS 81212) from SKF Cylindrical Thrust Roller Bearings (CRTB) of AISI 52100 steel with hardness 59-66 HRC and Ra=100 nm. The rings are assembled in a holder to attach them to the rotating shaft of the test rig. The specimens were cleaned prior to testing by successive immersion first in an ultrasonic bath of petroleum ether for 10 min and then acetone for 10 min.
4 Base oil polarity effect on ZDDP behaviour is studied in mixed rolling/sliding contact with a constant slide-roll ratio of 5% [The slide-roll ratio, or slip, is defined as entrainment speed, or sliding speed, US=UB-UR divided by the rolling speed, U=(UB+UR)/2 where UB and UR are the ball and ring surface speed in contact respectively]. The tribotests were carried out at applied loads of 100 and 600N, resulting in a maximum Hertz contact pressure of 1.34 GPa (contact diameter 0.38 μm) and 2.43 GPa (contact diameter 0.69 μm) respectively. The entrainment speed was set at 0.5 m/s. Temperature for all the tests was set constant at 90°C. The tests were conducted for 1 and 5 hours rubbing time. Pure base oil and two different concentrations, 2 and 5 wt% ZDDP, were tested. Under these conditions the elastohydrodynamic (EHD) film thickness is calculated at the centre of the contact to be ≈10 nm for the different systems which indicates that they are operating in the boundary lubrication regime.
2.3. Reaction layer formation Spacer layer interferometry imaging principle is as an in-situ (inside the tribometers, out of the contact) and post-mortem (after friction) method [26]. Tests are carried out by rolling/sliding a steel ball on a lubricated steel disc, to produce a wear track on both ball and disc. Motion is then halted and a spacer layer and chromium-coated glass disc are loaded against the wear track on the steel ball. The lubricant is squeezed out from the contact thus formed, but any solid-like reaction film remains. Because the glass disc conforms to the counterface, the interference images produced represent a map of the reaction film present in the contact. The glass disc in then removed and the sliding/rolling of the steel ball on the steel disc continuous.
2.4. Reaction layer characterization A series of surface analysis techniques have been used to characterize the additive derived layers in terms of thickness, morphology, nanomechanical properties and chemical composition. 2.4.1. Scanning Electron Microscopy and Energy Dispersive X-ray Spectroscopy Morphology and elemental chemical composition of the ZDDP derived reaction layers were characterised with Zeiss Supra 55 Scanning Electron Microscope (SEM) with a 15kV electron beam voltage in the secondary electron mode, equipped with Oxford Instruments Inca system for Energy Dispersive X-ray Spectroscopy (EDS). EDS spectra were collected at different locations inside and outside the wear track.
2.4.2. X-ray Photoelectron Spectroscopy The surface chemistry of the wear tracks was studied with X-ray photoelectron spectroscopy (XPS), analysing selected areas, both at different locations inside and outside the reaction layer. The X-ray source of the PHI 5000 (ULVAC-PHI, Chanhassen, MN, USA) is a focused and monochromatic AlKα beam with a diameter that can be chosen between 10 and 200 μm. The emitted electrons are collected and retarded with an Omega lens system at an emission angle of 45°. After passing a spherical capacitor energy analyzer, the electrons are collected by a 16-channel detector. The system is equipped with a high performance
5 floating-column ion gun and an electron neutralizer for charge compensation. The residual pressure was always below 5 x 10-7 Pa. The identification of the wear tracks was performed using scanning X-ray image (SXI), allowing the visualizing of the desired features in the sample. Small-area XPS spectra were collected with a beam diameter of 20 μm and a power of 4.25 W in the constant-analyzer-energy (CAE) mode, using a pass energy of 69 eV and a step size of 0.125 eV. Under these conditions, the full width half maximum height (fwhm) for Ag3d5/2 is 1.3 eV. Survey spectra were acquired with 280 eV pass energy and a step size of 1 eV. The whole set of spectra (detailed and survey spectra) was acquired within 60-80 min/area. XPS data was processed with CASA XPS software (Casasoftware Ltd., UK). The detailed spectra were fitted with Gaussian-Lorentzian curves and a Shirley background. Minor charging was observed and corrected by referencing the aliphatic carbon to 285.0 eV.
2.4.3. Atomic Force Microscopy Measurements on rubbed and non rubbed surfaces were performed with an AFM MFP- 3D atomic force microscope (by Asylum Research, Santa Barbara, CA) in contact, constant force mode using non conductive silicon nitride cantilevers with a spring constant k = 42 N/m and a resonant frequency f0 = 300 kHz (Olympus). The main measurement parameters were: a scan size of 80 x 80 μm² (512 scan points and 512 scan lines), a scan rate of 0.5 Hz, scan angle of 0o, and a set point of 1 nN in contact mode. The recorded data was both trace/retrace of height, deflection, and lateral force. Force curves (cantilever spring force as a function of z-piezo extension) were also determined at fixed positions on the specimens. Away from the specimen the lever maintains its free deflection, and first makes a surface contact at S (“snap-in”). In air, this initial contact is frequently with a surface layer of adsorbed water vapour or other contaminants, and is accompanied by formation of a meniscus around the tip-surface contact [27]. Retraction of the lever results in an increasing cantilever spring force acting against the meniscus-related and other adhesive forces, until the “pull-off” force (P) is reached, when the cantilever jumps back to its free deflection position. Hysteresis during contact indicates that some plastic deformation has occurred whilst load was applied to the surface, due to the presence of a relatively soft surface film.
2.4.4. Nanoindentation The mechanical properties of the substrates and AW films were investigated using a Hysitron TriboIndenter®. The three-sided diamond Berkovich indenter with an elastic modulus (Ei) between 1000 and 1140GPa, a Poisson ratio (νi) of 0.07 and a tip radius of curvature of 120nm was used. A certain preset force was applied and the resulting depth of penetration of the indenter was recorded. Both the loading of the preset force and the unloading were monitored. The applied load was calibrated to obtain a penetration depth of 20 nm. The elastic modulus was extracted using the Oliver–Pharr method from the unloading section of the force-distance (f-d) curve [28].
6 3. RESULTS 3.1. Friction A series of ball-on-disc tests using low viscosity model base oils/ZDDP solutions were carried out to study the influence of base oil polarity on the friction performance of ZDDP additives. In the beginning of teach tests, the coefficient of friction is decreasing as a function of time for a period of time, reaching a steady-state values afterwards. The average steady-state coefficients of friction for the different solutions at different normal loads are summarized in Table 2.
Table 2 Steady state coefficient of friction after 1 hour rubbing time at a contact pressure of 1.3 GPa and 2.4 GPa for non-polar oil, n-hexadecane, and polar base oil, diethylen glycol diethyl ether, with different concentrations of ZDDP
Additive Diethylen glycol diethyl ether n-hexadecane concentration [wt %] P=1.3 GPa P=2.4 GPa P=1.3 GPa P=2.4 GPa 0% ZDDP 0.125 0.100 0.080 0.060 2% ZDDP 0.140 0.105 0.120 0.080 5% ZDDP 0.135 0.110 0.110 0.085
As shown in Table 2, the coefficient of friction for both base oils increases when ZDDP is added to the base oils. Previous studies have shown how ZDDP and other phosphorus- based antiwear additives cause a significant increase in friction in thin film, high- pressure, lubricated contacts [25, 29, 30]. However, the degree of friction increase is higher for the case of n-hexadecane (non-polar) base oil. Also the concentration of ZDDP in the blend affects the overall behaviour. The antiwear properties of the reaction layers derived from the base oil + ZDDP solutions at different slide-roll ratios were also investigated. The wear track width (WTW) was measured from the balls using a calibrated microscope. The WTW is measured to be ≈400 μm for the three DGDE solutions. For the HeD solutions, in the absence of ZDDP the wear track width is significantly lower, ≈250 μm, whereas when ZDDP is added the values of the WTW increase to ≈400 μm.
3.2. Surface analysis: Morphology and chemical composition Scanning Electron Microscopy and X-Ray Diffraction Spectroscopy analysis performed on the steel balls at the end of the tribological tests showed that the traces of elements present in the components of the additives (Zn, P and S) are only found inside the wear track (Figure 3).
7 Figure 3 SEM micrograph inside the wear track and the corresponding EDS spectrum of the selected area inside the wear track of the steel ball lubricated with DGDE with 2 wt % ZDDP
EDS spectra showed no traces of the additive components outside the wear track (Figure 4).
Figure 4 SEM micrograph outside the wear track and the corresponding EDS spectrum of the selected area outside the wear track of the steel ball lubricated with DGDE with 2 wt % ZDDP
3.2.1. Chemical composition Survey and detailed high-resolution spectra were measured on the ball. Survey spectra were used for peak identification and to check the presence of contaminants. Detailed spectra of phosphorus 2p, together with sulphur 2p, carbon 1s, oxygen 1s, iron 2p and zinc 2p3/2 were recorded to identify the different chemical states of the species.
The survey and detailed spectra acquired on the reaction layer produced at SRR=5% in 2 wt-% solution of ZDDP on SKF Ester. All spectra are shifted to correct for sample charging by referring the main C 1s peak to 285 eV (aliphatic carbon).
The C1speak at 285.3 eV was referred to 285 eV from aliphatic carbon (C-C, C -H), thus shifting all peak values. The second contribution can be assigned to C -S bond The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.70 eV. An initial very small peak was detected at 530.3 eV that can be assigned to iron or zinc oxide. Two main contributions were found: a peak at 532.3 eV, non-bridging oxygen (NBO), that can be assigned to phosphorus bound in an (poly)phosphate [31], and a peak
8 at 533.9 eV, bridging oxygen (BO), that can be assigned to P-O-C or P-O-P bonding environment. The BO/NBO ratio calculated according to the recorded signal indicates the presence of polyphosphates. The S2ppeak was fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.6 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [13, 31, 32]. The P2ppeak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 134.3 eV, value corresponding to a phosphate [31]. The Zn 3s peak is found at 140.7 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV. The results obtained for the layer derived from HeD+ZDDP are very similar to the ones presented for the layer derived from DEG+ZDDP, indicating that the base polarity may not have a great influence on the reaction layer chemical composition C1speak was found at 285.2 eV was referred to 285 eV from aliphatic carbon (C-C, C -H), thus shifting all peak values. The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.67 eV. An initial very small peak was detected at 530.2 eV that can be assigned to iron or zinc oxide. Two main contributions were found: a peak at 532.2 eV, non-bridging oxygen (NBO), and a peak at 533.8 eV, bridging oxygen (BO). The values calculated according to the recorded signal indicate the presence of polyphosphates. The S2ppeak was detected at 162.5 eV, can be assigned to C-S bonded in sulfides. The P2p peak was found at 134.3 eV. The Zn 3s peak is found at 140.6 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV.
3.3. Reaction layer thickness To address the influence of oil polarity on the thickness of the ZDDP-derived reaction layer a series of experiments were performed using spacer layer interferometry imaging. The test with base oil + 5% wt ZDDP (steel-steel contact) were re-run at normal loads of 100N and 300N, resulted in a maximum Hertz contact pressure of 1.34 GPa (contact diameter 0.38 μm) and 1.93 GPa (contact diameter 0.51 μm). The tests were terminated in different time interval to allow reaction layer thickness measurement using space layer interferometry imaging as those done by Fujita and Spikes [20, 21]. However, due to the limitation of the current setup, the test can only be performed at 40°C. Under these conditions the EHD film thickness is calculated at the centre of the contact to be ≈15nm for the different systems. To keep the specific film thickness or lambda ratio (the ratio of the central film thickness to the composite surface roughness of the two surfaces is contact) constant the entrainment speed is varied accordingly (Table 3).
Table 3 Entrainment speeds
Lambda ratio Contact pressure Entrainment speed Description (λ) [GPa] [m/s] 0.4 DGDE + 5 wt% ZDDP 1.3 0.50 0.4 DGDE + 5 wt% ZDDP 1.9 0.55 0.4 HeD + 5 wt% ZDDP 1.3 0.32 0.4 HeD + 5 wt% ZDDP 1.9 0.36
9 The system is operating in the boundary lubrication regime. The interference images of the rubbing track at different rubbing time are shown in Figure 5, for the polar base oil solution, and Figure 6, for the non-polar base oil solution.
P = 1.3 GPa 0 min 1 min 5 min 10 min 30 min 40 min 50 min 70 min 160 min
P = 1.9 GPa 0 min 1 min 5 min 10 min 30 min 40 min 50 min 70 min 130 min
Figure 5 Series of interference images of the ball wear track for DGDE+5%ZDDP solution showing the evolution of the reaction layer thickness as a function of contact pressure and rubbing time
P = 1.3 GPa 0 min 1 min 5 min 10 min 30 min 40 min 50 min 70 min 160 min
P = 1.9 GPa 0 min 1 min 5 min 10 min 30 min 40 min 50 min 70 min 160 min
Figure 6 Series of interference images of the ball wear track for HeD+5%ZDDP solution showing the evolution of the reaction layer thickness as a function of contact pressure and rubbing time
The development of a patchy ZDDP antiwear reaction layer on the wear track during rubbing, indicated by the dark areas within the contact, can be clearly seen in Figure 6. It is also significant that for the non polar base oil solution (HeD+5% wt ZDDP), the reaction layers develop more rapidly and are thicker that the reaction layers derived from the polar base oil solution (DGDE+5% wt ZDDP). The line profiles obtained from the images showed that the initial layer thickness has a similar tendency for the two solutions and different loads. The initial reaction layer is typically 8-12 nm and the profile is very smooth. Only a significant difference can be seen for the non polar solution at high load, where already after only 1 minute of rubbing a rougher profile can be seen. It is also a general tendency that the profiles for intermediate times show a characteristic patchy shape of peaks and grooves, according to the uneven formation of the film on an early stage to stabilize at the end of the rubbing period at a thicker and slightly smoother layer. From line profiles across the centre of the interference images, reaction layer thicknesses values were calculated (Figure 7) in the central region of each image were determined.
10 60 DGDE+ZDDP(1.3GPa) 50 DGDE+ZDDP(1.9GPa) HeD+ZDDP(1.3GPa) HeD+ZDDP(1.9GPa) 40
30
20
10 Reaction layer thickness (nm) 0 0 1000 2000 3000 4000 5000 Rubbing distance (m)
Figure 7 Mean reaction layer thickness as a function of rubbing distance for hexadecane + ZDDP and diethylenglycol + ZDDP at different normal loads.
Considering mean film thickness, the two solutions show a similar behaviour, an unstable growth in the early stage of rubbing, where a maximum of film thickness is achieved. After this period, the behaviour of both solutions is remarkably different. The non polar solutions achieve stable thickness values, slightly lower that the maximum reached previously, while rubbing progresses, consistent with the findings of Topolovec-Miklozic et al. [33]. The film formed reaches a “limiting thickness”, consequent with equilibrium between film formation and removal. The polar solutions, in stead, reached the higher values at the end of the rubbing process, where the growth rate of the film is also significantly higher, than in the early stages. No stable value was reached at the end of the experiments, thus further experiments, with longer test time are required. Both oil types show the same behaviour when load is increased. The reaction film forms more rapidly and is thicker. The mean film thickness is approx. 10-25% higher
3.4. Topography and nanomechanical properties For further understanding on the origin of the friction increase the topography and nanomechanical properties of the reaction layers were studied using atomic force microscopy (AFM) and nanoindentation. A series of AFM images of 80x80 μm2 areas of wear track centre are shown, together with their corresponding line profiles taken transverse to the rolling-sliding direction. A reaction layer is formed only on the wear track and not on the non-contact area. The rolling-sliding direction is from bottom to top of the displayed images.
11 Figure 8 shows the wear track topography for the two types of base oil, DGDE and HeD, when tests were run with pure base oil, no additives present. The base oil DGDE presents a total coverage of the initial surface, no surface finishing marks are visible and a relatively homogeneous layer with no distinctive features can be observe in the wear track. Whereas, the reaction layer derived from HeD presents a slight coverage of the initial surface and the original surface finishing marks are still visible on the rubbed surface, also indicating that little wear has occurred. The polarity of the base oil molecules determines their affinity towards the steel surface. When the polarity of the molecule is low, i.e. HeD, its affinity towards the surface is low, while when the polarity of the molecule is high, its affinity for the steel surface is also high, i.e. DGDE. The line profiles show that both surfaces suffer a slight roughening process due to the running process, with signs of patchy film development for diethylen glycol diethyl ether, but the friction coefficient for DGDE is remarkably higher than for HeD. The difference is roughness is not enough to explain the difference observed in friction, therefore the explanation must be in the properties of the film itself.
DGDE + 0 wt% ZDDP HeD + 0 wt% ZDDP 80 80
60 60 60 60 40 40
20 20
40 40 µm µm 0 0 nm nm
-20 -20
20 20 -40 -40
-60 -60
0 0 0 20 40 60 80 0 20 40 60 80 µm µm 80 80
40 40
0 0 Height (nm) -40 Height (nm) -40 CENTRE CENTRE -80 -80 020406080 0 20406080 ?m μm
Rrms = 13.8 nm Rrms = 11.9 nm Figure 8 Topography of representative 80x80 μm2 scanning areas at the centre of the wear tracks of base oil (a) DGDE and (b) HeD
Figure 9 shows the topography in the centre of the wear tracks for the solutions of the two types of base oil when 2% wt ZDDP is added. Both reaction layers are non- homogeneous and remarkably rougher that the layer derived from the base oils alone. In the reaction layer derived from the solution DGDE+2% wt ZDDP it is possible to identify elevated large pads (10-15μm in length) elongated in the rolling-sliding direction. The trough regions exhibit an agglomeration of smaller pads (1-5 μm in length)
12 not elongated in the rolling-sliding direction. The large pads are higher that the surrounding features and are flat and relatively smooth, then primarily supporting the load during the contact [14]. In the reaction layer derived from HED +2%ZDDP it is possible to identify elevated areas composed of streaks, being the bigger features elongated in the rolling-sliding direction, as well as the rest of the film. The structure suggests that the load is more evenly distributed through the film.
DGDE + 2 wt% ZDDP HeD + 2 wt% ZDDP 80 80
60 60 60 60 40 40
20 20
40 40 µm µm 0 0 nm nm
-20 -20
20 20 -40 -40
-60 -60
0 0 0 20 40 60 80 0 20 40 60 80 µm µm 80 80
40 40
0 0 egt(nm) Height -40 Height (nm) -40 CENTRE CENTRE -80 -80 0 20406080 0 20406080 μm μm
Rrms = 49.9 nm Rrms = 32.2 nm
Figure 9. Topography of representative 80x80 μm2 scanning areas at the centre of the wear tracks of base oil (a) DGDE+2% ZDDP and (b) HeD+2%ZDDP
In Figure 10, the reaction layer derived from DGDE+5% wt ZDDP presents elevated smooth large pads (20-30 μm in length). Those pads form discontinuous strips elongated in the rolling-sliding direction. The smaller pads observed in the previous blend are not present here, being the rest of the surface homogeneously covered. This is visible on a reduction of the roughness of the film, which can explain the friction reduction with respect to the 2wt% blend. For the reaction layer derived from HED+5% wt ZDDP, the morphology of the film is remarkably different. It appears more homogeneous, and attending to the detailed image, representative of the overall structure of this film, the surface is completely covered by pads that seem to have collided while growing.
13 DGDE + 5 wt% ZDDP HeD + 5 wt% ZDDP 80 80
60 60 60 60 40 40
20 20
40 40 µm µm 0 0 nm nm
-20 -20
20 20 -40 -40
-60 -60
0 0 0 20 40 60 80 0 20 40 60 80 µm 80 µm 80
40 40
0 0 egt(nm) Height Height (nm) -40 -40 CENTRE CENTRE -80 -80 0 20406080 0 20406080 μm μm
Rrms = 29.2 nm Rrms = 48.3 nm
Figure 10. Topography of representative 80x80 μm2 scanning areas at the centre of the wear tracks of base oil (a) DGDE+5% ZDDP and (b) HeD+5%ZDDP
It is also remarkable the presence of embedded wear particles in the films derived from HeD, absent for the films derived from DGDE. This may indicate a
To evaluate the nanomechanical properties of the two different layers, the nanoindentation hardness, H, and the reduced elastic modulus, Er, were determined from the load-displacement curves. The elastic modulus is proportional to the force required to deform the sample elastically, and the hardness is a measure of the resistance of the material to plastic flow. As an indicator of the overall deformation behaviour of the layer, the reduced elastic modulus to hardness ratio, Er/H, was used to evaluate the resistance of the layers to plastic flow. An example of imaging of the reaction layer derived from the DGDE + 5 wt% ZDDP and HeD + 5 wt% ZDDP solutions is presented in Figure 11.
DGDE + 5 wt% ZDDP HeD + 5 wt% ZDDP
*1
*1 *2 *0 *2 *0
Figure 11 Imaging of indentation points, inside and outside the additive-derived reaction layers
14 The nanomechanical properties at selected points (inside and outside the reaction layer), including the Er/H ratio, are summarized in Table 4.
Table 4 Nanomechanical properties of the additive-derived reaction layers
DBDE + ZDDP HeD + ZDDP
Position Er [GPa] H [GPa] Er/H Position Er [GPa] H [GPa] Er/H Point 0 229.8 10.8 21.2 Point 0 222.2 12.2 18.6 Point 1 207.3 11.5 17.9 Point 1 192.1 6.9 28.0 Point 2 215.8 12.5 17.2 Point 2 168.0 6.4 26.1
The values obtained for Point 0, outside the wear track, correspond to previously reported values of reduced elastic modulus and hardness for AISI 52100 steel. The values for Point 1 and 2, taking inside the wear track, correspond with values reported for phosphate layers. The reaction layer derived from the DGDE+5wt%ZDDP has also been analysed, by due to the thickness of the layer, a great influence form the substrate in the results is expected.
4. CONCLUSIONS • The polarity of the base oil determines the tribological performance of the additives. The same type of additive behaves differently depending on the type of oil it is blended with. • The reaction between additives and the surface happens at asperity level, where the real contact occurs. No traces of additives were found outside the running tracks, proving that tribological energy is needed for activating the reaction between the additive and surface. • Polarity of the oil influences the growth rate and film thickness of ZDDP antiwear films. The polarity of the molecules determines the way they cover and attach to the surface, influencing the final film thickness and the structure and characteristics of the reaction film. The reaction film for non polar base oil blends is thicker and formed more rapidly that for polar base oil blends.
Acknowledgement The authors wish to thank Prof. Stathis Ioannides (Product R&D Director of SKF at SKF Engineering & Research Centre in The Netherlands) for permission to publish this work. The work presented in this paper has been supported by the EC, Sixth Framework Programme, Marie Curie Action (WEMESURF research network entitled: ”Characterization of wear mechanisms and surface functionalities with regard to life time prediction and quality criteria-from micro to nano range” under contract MRTN CT 2006 035589).
15 5. REFERENCES
1. Hsu, S.M., Additives and tribochemistr. Proceedings of the World Tribology Congress III - 2005. 2005. 597-597. 2. Bancroft, G.M., et al., Mechanism of tribochemical film formation: stability of tribo- and thermally-generated ZDDP films. Tribology Letters, 1997. 3: p. 47-51. 3. Willermet, P..A.,et al., Mechanism of formation of antiwear films from zinc dialkyldithiophosphates. Tribology International, 1995. 28(3): p. 177-187. 4. Spedding, H., et al., Antiwear mechanism of ZDDP's - 1. Tribology international, 1982. 15(1): p. 9-12. 5. Coy, R.C., et al., Thermal degradation and EP performance of zinc dialkyldithiophosphate additives in white oil. ASLE transactions, 1981. 24(1): p. 77-90. 6. Fuller, M., et al., Chemical characterization of tribochemical and thermal films generated from neutral and basic ZDDPs using X-ray absorption spectroscopy. Tribology international, 1997. 30(4): p. 305-315. 7. Yin, Z., et al., Chemical characterization of antiwear films generated on steel by zinc dialkyl dithiophosphate using X-ray absorption spectroscopy. Tribology international, 1993. 26(6): p. 383-388. 8. Yamaguchi, E.S., etal., Comparison of the relative wear performance of neutral and basic ZnDTP salts. ASLE transactions, 1996. 39(1): p. 220-224. 9. Glaeser, W.A., et al., In situ wear experiments in the scanning auger spectrometer. Wear, 1993. 8(132): p. 162-164. 10. Willermet, P.A., et al., Formation, structure, and properties of lubricant-derived antiwear films. Lubrication Science, 1997. 9(4): p. 325-348. 11. Barcroft, F.T., etal., Mechanism of action of zinc thiophosphates as extreme pressure agents. Wear, 1982. 77(3): p. 355-384. 12. Yin, Z., et al., Application of soft x-ray absorption spectroscopy in chemical characterization of antiwear films generated by ZDDP Part II: the effect of detergents and dispersants. Wear, 1997. 202(2): p. 192-201. 13. Martin, J.M., et al., The two-layer structure of Zndtp tribofilms, Part I: AES, XPS and XANES analyses. Tribology International, 2001. 34(8): p. 523-530. 14. Warren, O.L., et al., Nanomechanical properties of films derived from zinc dialkyldithiophosphate. Tribology Letters, 1998. 4(2): p. 189-198. 15. Graham, J.F., et al., Topography and nanomechanical properties of tribochemical films derived from zinc dialkyl and diaryl dithiophosphates. Tribology Letters, 1999. 6(3): p. 149-157. 16. Yin, Z., et al., Application of soft x-ray absorption spectroscopy in chemical characterization of antiwear films generated by ZDDP Part I: the effects of physical parameters. Wear, 1997. 202(2): p. 172-191. 17. Aktary, M., et al., Morphology and Nanomechanical Properties of ZDDP Antiwear Films as a Function of Tribological Contact Time. Tribology Letters, 2002. 12(3): p. 155-162. 18. Torrance, A.A., et al., An additive’s influence on the pitting and wear of ball bearing steel. Wear, 1996. 192: p. 66-73.
16 19. Tripaldi, G., et al., Friction behaviour of ZDDP films in the mixed, boundary/EHD regime. SP, 1996. 1209: p. 73-84. 20. Fujita, H.H., at al.,Study of zinc dialkyldithiophosphate antiwear film formation and removal processes, part II: Kinetic model. Tribology Transactions, 2005. 48(4): p. 567-575. 21. Fujita, H.H., et al., Study of zinc dialkydithiophosphate antiwear film formation and removal processes, part I: Experimental. Tribology Transactions, 2005. 48(4): p. 558-566. 22. Minami, I., et al.,Investigation of anti-wear additives for low viscous synthetic esters: Hydroxyalkyl phosphonates. Tribology international, 2007. 40(4): p. 626- 631. 23. Minami, I., et al., Anti-wear additives for ester oils. Technische Akademie Esslingen International Tribology Colloquium Proceedings, 2004. 14 III: p. 1593- 1598. 24. Kar, P., et al., Formation and characterization of tribofilms. Journal of tribology, 2008. 130(4). 25. Taylor, L..J., et al., Friction-enhancing properties of ZDDP antiwear additive: Part I - Friction and morphology of ZDDP reaction films. Tribology Transactions, 2003. 46(3): p. 303-309. 26. Donnet, C., Handbook of Surface and interface Analysis - Methods for Problem- Solving. 1998. 27. Bhushan, B., Nanotribology and nanomechanics - An introduction. 2005, Berlin: Springer. 5. 28. Pharr, G.M., et al., Measurement of thin film mechanical properties using nanoindentation. MRS bulletin, 1992. 17(7): p. 28-28. 29. Zhang, J., et al., Investigation of the friction and wear behaviors of Cu(I) and Cu(II) dioctyldithiophosphates as additives in liquid paraffin. Wear, 1998. 216(1): p. 35-40. 30. Taylor, L.J., et al., Friction-enhancing properties of ZDDP antiwear additive: Part II - Influence of ZDDP reaction films on EHD lubrication. Tribology Transactions, 2003. 46(3): p. 310-314. 31. Piras, F.M., et al., Combined in situ (ATR FT-IR) and ex situ (XPS) study of the ZnDTP-iron surface interaction. Tribology Letters, 2003. 15(3): p. 181-191. 32. Bird, R.J.,et al., The application of photoelectron spectroscopy to the study of e. p. films on lubricated surfaces. Wear, 1976. 37(1): p. 143-167. 33. Topolovec-Miklozic, K., et al., Film thickness and roughness of ZDDP antiwear films. Tribology Letters, 2007. 26(2): p. 161-171.
17 19. Tripaldi, G., et al., Friction behaviour of ZDDP films in the mixed, boundary/EHD regime. SP, 1996. 1209: p. 73-84. 20. Fujita, H.H., at al.,Study of zinc dialkyldithiophosphate antiwear film formation and removal processes, part II: Kinetic model. Tribology Transactions, 2005. 48(4): p. 567-575. 21. Fujita, H.H., at al., Study of zinc dialkydithiophosphate antiwear film formation and removal processes, part I: Experimental. Tribology Transactions, 2005. 48(4): p. 558-566. 22. Minami, I., at al.,Investigation of anti-wear additives for low viscous synthetic esters: Hydroxyalkyl phosphonates. Tribology international, 2007. 40(4): p. 626- 631. 23. Minami, I., at al., Anti-wear additives for ester oils. Technische Akademie Esslingen International Tribology Colloquium Proceedings, 2004. 14 III: p. 1593- 1598. 24. Kar, P., at al., Formation and characterization of tribofilms. Journal of tribology, 2008. 130(4). 25. Taylor, L..J., at al., Friction-enhancing properties of ZDDP antiwear additive: Part I - Friction and morphology of ZDDP reaction films. Tribology Transactions, 2003. 46(3): p. 303-309. 26. Donnet, C., Handbook of Surface and interface Analysis - Methods for Problem- Solving. 1998. 27. Bhushan, B., Nanotribology and nanomechanics - An introduction. 2005, Berlin: Springer. 5. 28. Pharr, G.M., at al., Measurement of thin film mechanical properties using nanoindentation. MRS bulletin, 1992. 17(7): p. 28-28. 29. Zhang, J., et al., Investigation of the friction and wear behaviors of Cu(I) and Cu(II) dioctyldithiophosphates as additives in liquid paraffin. Wear, 1998. 216(1): p. 35-40. 30. Taylor, L.J., et al., Friction-enhancing properties of ZDDP antiwear additive: Part II - Influence of ZDDP reaction films on EHD lubrication. Tribology Transactions, 2003. 46(3): p. 310-314. 31. Piras, F.M. and A.A. Rossi, Combined in situ (ATR FT-IR) and ex situ (XPS) study of the ZnDTP-iron surface interaction. Tribology Letters, 2003. 15(3): p. 181-191. 32. Bird, R.J.,et al., The application of photoelectron spectroscopy to the study of e. p. films on lubricated surfaces. Wear, 1976. 37(1): p. 143-167. 33. Topolovec-Miklozic, K.K., at al., Film thickness and roughness of ZDDP antiwear films. Tribology Letters, 2007. 26(2): p. 161-171.
18 Paper B
# !$$"
9G '%%EI 9G 1% !%H(%!DICI4G8G5!$!$!EIBIA %" )G 0%%$D
H' $% &# # '$ ' % "%# %% eID@d ## #$ ''# # #""#% " &'# ' # "% &# ' " ## ## % '% &%' #$ &'# % $'$ ' # e0FF7d a 3$# '$! # &%' # # " # " % #&#$ '# " '$$ # #%# #$ ' "' " # ' " &%' "' c ' %' #!'$ # # " '$$ # '$ ' ##$#$a 4 % ''% ## # "#%# " &'# ' c #"##%# &'# ) #('$#%'# e 'd
'$ $# #!% e'd( ## &#$#$ $""## %%# ' " GUb0FF7c '$ # ## # #$a I '# " &'# `'$$ # ' $# #$ ' I5B TWXYY ## ' # '$ %'#$ ID@ % '% $# $# ' &&! # '$ '#$ '$ %$ a I ID@ #% # ' $###$ # ' # # %%% '# " "% %#""%# ! # ' % & # # $""##%# "% &# ' " # $# # '$$ " 0FF7a 1 $""## &'# )e #('$#%'# eb' &'# d '$ $# #!% e' d( # ' &##$ !"%' $""## "% &# ' e % %'# '$ ' %'#d $# # '$$ " 0FF7 %'#$ # &'# '#a 4 &#' ' &# ' & #$ # % & " &'# ' # 0FF7 '$$ # # "'%# % &# $%#$ # $# ' # '#a 4 ## # $' # '%# " &'# ' # "% &# ' " "' #$ &%' % '! '$$ #a
&CC3a E%! %a A#
4""#" $ % !""# %
! ()* () !""# +*, % $ $ - % (*. & / ) - $ + & !""# ' % $ 0 , V9, ,- F!0%"! 3a (!% )&! 0!#a ( '!"! @%,%! I VWXVTSa VWSW '!a F,% $ & U17A B!! %" 2!!%# D!!a 7!&$%% VQa TSTH 50 !""# 4!,0!!a 0! 4!!%" TC& 5%#! B!!a 6,!% 1 IPa 6,!% S 9,! 5#!!! %" 4%!!##a )&! 7!$%%% $ 23! 5%%%a STQWW )75 E%a 5%%% 6 RFDU0 F,% D!! D!!#! 0$a ( 7%%`1%?! !4 !5 Ua UPWW '!! 4!,%"a F,% 7 !""# D!" %,a !% $,G%Y,0!Y%#Y%
DFEF &G (G 2%!' P ( 3"" )!#! !" @ %##!"!" '$' WYXY 28A YYY 98 YYY X a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc !"# $% #
8 & 4 DE ?5& = 1 $ !""# //F &G5 . / % C 2 +) ; !""# $ 3 % !""# )) & &29 6)) D $ 71 0, C# )0. +,0 C# )6/ % 5:: &29 ))( H% 5:: I % &29 )( % + *+0 /) % : # + ( F !""# ? ; ) +, 7 1%"$#% "!" % +( +6 !""# 9 &29 92#0" % $ & : 5 &29 ; & ?5& !""# )+D )0. 71 J = 5 ; & ?5& &29 +8 % ' % &=5= (+ )) (/*66 7: !""# ) % 7.!!"% % B 1 2%#"$#% "!" ((.).) ( + ( + )(+71 /) <&9( < & = 5# ) () D > #& ?5& % 3 A 2 3 5:: % 11 $ $ 5:: & A 5 K L = % +) &=5= (+ )) (/*66 7: ) % <5 . + + 5@2 &=5= (!! $ (+ )) (/*66 7: )) 4 %# e9d YaS % 4 %# e69Cd YaTP YaSX % H '# &'% XT " G 4 '# &'% SY " I ) 4 #%"%' ) C## G' a a = A 4 #! W T V T 4 '$ e9d WY A#! c XUY B ">C $ 7"
W '$' WYXY 28A YYY 98 YYY a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc !"# $% #
) 2 ; +/ * : 0 K L ' 0 A 7 ) ? 22J 5# 7 = &29 J $ &29 = 2 +) A MJ"% $ () D % )) 9# = 4M5 * < = 4M5 ; +. 22J 5#D 2 22J J = 4M5 J ' % D 22J D , $
= 5# J = 4M5 J D 5# D ( % 229 H 3 I % )!" %" "# ; 6#" 5# & 3 % 1 (( ( + = ( + 2 ( + ( + % % + % 22J & % + !""# +. 7 $ ( %#! %%' + 5 5>9 % 22J - >"5 & D &L 2 + + & ) $ 5>9 - 5# 2
!""#
A!! 4
8@61%! 7@51%! 8@61%! 7@51%!
Yf 0FF7 YaYQY YaYSY YaXWT YaXYY Wf 0FF7 YaXWY YaYQY YaXUY YaXYT Tf 0FF7 YaXXY YaYQT YaXVT YaXXY
'$' WYXY 28A YYY 98 YYY V a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc !"# $% #
$ % $& ' "$( )!""#
5 $ U '$' WYXY 28A YYY 98 YYY a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc !"# $% # * # + # "$( ," % % & !""# ; = % 0 &29 )++ D 0) % 2 ( 22J ">C)F!""# ">C(F!""# ">C/)F!""# 6. 8), ">C 668 D = 5 7" !""# ">C % !""# % % !""# & 2 6 22J ">C 7" 2 , A & 2 6 A 22J * - 4319C "44% $ ! B $ ! B $! %!B $! %!B 4 0'#&)#$B 4%!B # ' %8 22#B # 22#B ' ' 8 #8 ' XXYaUYT XWaTW YaYYUW YaYUU YaWVT UUaXQ XSaQSV YaYYW XXYaUYT VRaTS YaYYUX YaYUW YaWVT UUaXQ XSaQSV YaYYW VVaaWXTXWaTW YaYXXY YaXXU YaWVT UUaXQ XSaQSV YaYYW VVXaWXT VRaTS YaYXQX YaXQQ YaWVT UUaXQ XSaQSV YaYYW TTWaYWT XWaTW YaYXTS YaXSV YaWVT UUaXQ XSaQSV YaYYV TTWaYWT VRaTS YaYWUT YaWTT YaWVT UUaXQ XSaQSV YaYYU FECTf 0FF7 XPaXYT XWaTW YaYYTQ YaYT YaWWU UYaSY RUaRRP YaYYX XPaXYT VRaTS YaYYRT YaYR YaWWU UYaSY RUaRRP YaYYX VWRaVXU XWaTW YaYXUT YaXV YaWWU UYaSY RUaRRP YaYYX VWRaVXU VRaTS YaYXUT YaXV YaWWU UYaSY RUaRRP YaYYX TUTaTWV XWaTW YaYWYQ YaXP YaWWU UYaSY RUaRRP YaYYW TUTaTWV VRaTS YaYXQT YaXR YaWWU UYaSY RUaRRP YaYYX FECPYf 0FF7 XPaTTS XWaTW YaYQYX YaRS YaWWP UXaSV SSaRX YaYXY XPaTTS VRaTS YaYQRS YaQV YaWWP UXaSV SSaRX YaYXW VWQaSRY XWaTW YaYSQQ YaSS YaWWP UXaSV SSaRX YaYYR VWQaSRY VRaTS YaXXSP XaXX YaWWP UXaSV SSaRX YaYXW TURaRQV XWaTW YaYQUY YaQY YaWWP UXaSV SSaRX YaYYR TURaRQV VRaTS YaYPXP YaQR YaWWP UXaSV SSaRX YaYYQ '$' WYXY 28A YYY 98 YYY T a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc !"# $% # . / # 2 / & $ .1 ">C 7" 2 8 22J 2 = 2 8 D !""# A !""# 22J % % = &29 &29 $ % 71 &29 )*() ()) D C# = <&9 )) 6)) D 71 0, C# )0. +,0 C# & )6/ = $ &29 !""# !""# 9 &29 % &29 $ 0) $ ? &29 !""# % .)* +) &29 +8 7 8# % A !""# D 7" !""# ">C % !""# % !""# !""# A !""# 9#2!"!!" % # 5 = # :G" " 5@2 5@2 > G : % D % > 5$ 2 # 9 & <>9>5?:2 ' K 1 L 9:%D % +))6 0 / 1 )0((./ S '$' WYXY 28A YYY 98 YYY a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc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a '%! !" bSSdWcVHUS TQHRHI ` 2!& PYRVX' d8% UW UWWTc Paper C THE INFLUENCE OF SLIDE-ROLL RATION ON ADDITIVE DERIVED REACTION LAYER FORMATION Aldara Naveira Suarez1,2*, Maurizio Zaccheddu1, Mattias Grahn1,3, Rihard Pasaribu1 and Roland Larsson2 1SKF Engineering and Research Centre, Kelvinbaan 16, 3439 MT, Nieuwegein, The Netherlands 2Division of Machine Elements, Luleå University of Technology, SE 97187, Luleå, Sweden 3Division of Chemical Engineering, Luleå University of Technology, SE 97187, Luleå, Sweden *Corresponding author (E-mail: [email protected]) ABSTRACT Functional additives, particularly extreme-pressure and antiwear additives, in formulated oil will compete to adsorb and function in tribological contacts. In this paper we study how the interactions between additives and base oil molecules and operation conditions influence tribological behaviour. One polar (ester oil) and one non-polar (poly-α-olefin) commercial base oil blended with zinc dialkyl dithiophosphates were studied. The tribological performance was evaluated using a ball-on-disc test rig under mixed rolling- sliding conditions in the boundary lubrication regime. An adapted in-situ interferometry technique was used to monitor the additive derived reaction layer formation. The properties of the additive-derived reaction layers were studies using different surface analysis techniques, i.e. X-ray Photoelectron Spectroscopy and Atomic Force Microscopy. It was found that base oil polarity determines the transport of additives to the surface thereby controlling the maximum reaction layer thickness, friction and wear, as well as the morphology of the additive-derived reaction layer. However the reaction layer chemical composition is not strongly influenced the base oil polarity. Among the operating conditions, shear was identified as a fundamental parameter on the activation of additives on rubbing steel surfaces and the properties of the derived reaction layer. KEYWORDS: base oil polarity, shear, reaction layer formation, 1 1. INTRODUCTION Extreme-pressure (EP) and antiwear additives (AW) control the lubricating performance of the oil in the mixed and boundary lubrication regimes. Performance enhancing properties of these additives are very important since, if oil lacks lubricating ability, excessive wear and friction can occur [1]. The main elements, which are responsible for the extreme-pressure and antiwear action, are sulfur and phosphorus respectively [2]. The solubility of additives is a major issue in additive technology, because having these additives in a solution in the lubricant is essential to transport them close to the rubbing steel surfaces. The structure of additives can be described as a functional core, or polar moiety, including the main elements, e.g. sulfur and/or phosphorus and a tail (non-polar moiety), e.g. alkyl chains, which make the additive soluble in the base oil. The functional core of the additives is generally composed of elements of different electronegativity. This leads to an unequal charge distribution that leads to polarity. The polarity of an additive is responsible for the ability to recognize the surface and attach to it. The ability to attach and build layers on the surface depends therefore on the polarity, but also the required space for the tails plays an important role. If the tails are widely branched the near surface concentration and ordering of the molecules may be much lower than when unbranched additives are present [1]. Dialkyl dithiophosphate compounds (MDTPs) have been used in lubricating oils due to their multifunctional performance as antiwear, extreme-pressure, friction modifying, antioxidant and corrosion resistant additives. Dialkyl dithiophosphates of different metals such as molybdenum [3], cadmium [4], copper [5], titanium, gadolinium [6], iron, antimony and other metals have been introduced in lubricants, being zinc dialkyl dithiophosphates (ZDDP) the most widely used [7-11]. ZDDP reacts with the surface in contact to form protective reaction layers. A variety of mechanisms have been proposed [12] for the formation of the ZDDP derived reaction layers, involving oxidative (by reaction with hydroperoxides or peroxy radicals) [13], catalytic (chemisorption on metal, hydrolytic [14]), and thermal [15] decomposition of the ZDDP. The resulting reaction layers have a heterogeneous composition with the chemical structure of the starting materials dictating their chemical composition [16, 17] and mechanical properties [18]. However, several studies have proved ZDDP to have detrimental effects on wear under certain operation conditions [19], and to enhance friction when the system is operating in mixed and boundary lubrication regimes [20-22]. Wan et al. [23] conducted a series of experiments using a four-ball test rig and bearing life test machine. The reduction of bearing life is primarily attributed to the high chemical reactivity of the EP/AW additives in blended oil. Rolling-bearing fatigue tests showed that under marginal film lubrication conditions, lambda = 1.2 (ratio of lubricant film thickness and composite surface roughness) the EP/AW additive package blended lubricating oil reduced the bearing fatigue life by a factor of 4.5 compared with the base oil alone.. Reduction of bearing life was attributed to high chemical reactivity of the EP additives in the base oil. In order to identify the plausible mechanisms that govern the detrimental effect of those additives on bearing performance it is necessary to identify the parameters affecting the additive-derived layer formation and the tribological properties of this layer, as it is 2 directly related to the additive reactivity towards the surface. The friction and wear behaviour of tribological systems lubricated with oils containing ZDDP have been mainly studied as a function of the additive structure, concentration, interaction with other additives or operation conditions, but the effect of the base oil in the ZDDP tribological performance has been hardly addressed [24-26]. When considering an additive molecule approaching the steel surface in a lubricating oil, it must be considered that the steel surface is already completely covered with oil molecules. Hence, it can be stated that the additive molecule has to compete with the oil for a place on the steel surface. The polarity of a molecule can be used to measure its affinity towards the steel surface. When the polarity of the molecule is low, its affinity towards the surface is low, while when the polarity of the molecule is high, its affinity for the steel surface is also high. Parameters, like polarity, that describe the base oil-additive interactions are needed in order to have a good running in, friction and wear model of lubricated contacts. Several operating conditions (in terms of lambda ratio, temperature and additive concentration) have been previously studied [27-29], showing that low lambda ratios, meaning high metal to metal contact, high temperature and high additive concentration lead to a high reactivity of the additives and therefore to thicker reaction layers. This study focuses on the effect of slide-roll ratio and the interaction between additive- base oil molecules in reaction layer formation and tribological performance. The nature and properties of the derived reaction layers, as a function of operating conditions and base oil-additive interaction, in terms of thickness, morphology and chemical composition, were studied using a series of surface analysis techniques. 2. MATERIALS AND METHODS 2.1. Materials 2.1.1. Test samples The steel balls were of SKF3 (ASI 52100) steel with hardness 59-66 HRC and an average roughness (Ra) of 10 nm. The rings were washers (WS 81212) from SKF Cylindrical Thrust Roller Bearings (CRTB) of ASI 52100 steel with hardness 59-66 HRC and Ra=100, for reaction layer thickness tests. According to EN ISO 683-17, the steel contains 1.0% carbon, 1.5% chromium, 0.255% silicon and 0.35% manganese (all percentage by weight). Usual tolerances are ± 0.1% and both phosphorus and sulfur have to be lower than 0.025%. The rings are assembled in a holder to attach them to the rotating shaft of the test rig. The specimens were cleaned prior to testing by successive immersion first in an ultrasonic bath of petroleum ether for 10 min and then acetone for 10 min. 2.1.2. Base oils and additives As base oils, two fully synthetic oils, one non-polar (poly-α-olefin, PAO) and one polar (ester oil), were used. PAO is chosen due to its purity instead of mineral oil because it has a relatively high concentration of sulfur which might interfere with the additives. Their physical properties and sulfur and phosphor content, obtained by X-ray fluorescence (XRF) analysis of an oil sample before testing, are summarized in Table 1. 3 Table 1 Base oil properties BASE OIL TYPE Kinematic Kinematic Sulfur Phosphorus viscosity at viscosity at content content 40°C 100°C Code Description [wt%] [wt%] [mm2/s] [mm2/s] Fully synthetic SKF Ester 26.8 5.2 0.00052 <0.00030 (Ester) Fully synthetic Synthetic Oil 24.6 5.1 0.00055 <0.00030 (PAO) NOTE1: The results for all elements are <0.0002 % It is assumed that the addition of the additives does not change significantly the viscosity of the bulk solution. A fully formulated primary zinc dialkyldithiophosphate (ZDDP), with 99% purity, is employed in simple solution in both base oils without other additives present. 2 wt% ZDDP solutions were prepared using an ultrasonic bath to dissolve the additives in the base oils. The temperature of the lubricant solution remained below 40°C during the dissolving procedure. 2.2. Test conditions 2.2.1. Thermal layer (Immersion tests) Thermal layer formation is studied by immersion of fresh, clean steel balls in heated base oil-additive solutions at different temperatures. At the end of the 24 hours immersion the balls are removed from the solution, assembled in the WAM5 without any rinsing and the spacer layer interferometry method, described in the following section, is used to obtain a map of the reaction layer thickness at several positions of the ball surface. 2.2.2. Reaction layer (Tribological tests) SKF-WAM5 ball-on-disc test rig (Wedeven Associates, Inc., Edgmont, PA, USA) enables to perform a variety of tests to evaluate the tribological performance of additive/base oil blends at controlled contact conditions. The ball and the disc are independently driven giving the possibility to simulate pure rolling and various slide-roll ratios (slip ratio). This rig is computer-controlled, which enables the ball and disk speeds, the load, and the test temperature to be controlled and varied in any desired preprogrammed sequence. Base oil polarity effect on the different additives behaviour is studied in mixed rolling/sliding contact to match closely the application conditions in bearings. In rolling/sliding conditions, wear is evenly-distributed around the tracks of both specimens which avoid the change in contact geometry that occurs during pure sliding tests, and thus only a slight change in contact pressure occurs. The tribotests were carried out at an applied load of 300N, for the 20 mm balls employed, resulted in a maximum Hertz contact pressure of 1.9 GPa (contact diameter 0.54 μm). The slide-roll ratios investigated ranged from -10 to +10%. [The slide-roll ratio (SRR), or slip ratio, is defined as sliding speed US=UB-UR divided by the entrainment speed, or rolling speed, U=(UB+UR)/2 where UB and UR are the ball and ring surface speed in contact respectively]. Temperature 4 for all the tests was set constant at 90°C. Under these conditions the EHD film thickness is calculated at the centre of the contact to be ≈10 nm for the different systems. The specific film thickness or lambda ratio (the ratio of the central film thickness to the composite surface roughness of the two surfaces is contact) is set constant at 0.4 and the entrainment speed is accordingly set to 0.25 m/s. The system is operating in the boundary lubrication regime. The surface speed of the ball, the tested specimen in this work, is different for the SRR tested. In order to have the same number of rotations for all the specimens, the rubbing time is adjusted accordingly. 2.3. Reaction layer characterization 2.3.1. Spacer layer interferometry Spacer layer interferometry imaging principle [30] is used as an in-situ (inside the tribometers, out of the contact) and post-mortem (after friction) method [31] to monitor reaction layer formation [32]. Tests are carried out by rolling/sliding a steel ball on a lubricated steel disc, to produce a wear track on both ball and disc. Motion is then halted and a spacer layer and chromium-coated glass disc are loaded against the wear track on the steel ball. The lubricant is squeezed out from the contact, but any solid-like reaction layer remains. The interference images produced are a map of the reaction layer present in the contact. The glass disc in then removed and the sliding/rolling of the steel ball on the steel disc continuous. The test arrangement is shown schematically on Figure 1. 3CCD Camera 3CCD Camera Microscope Microscope Spacer Layer Spacer Layer STEEL RING STEEL RING STEEL BALL STEEL BALL Thermocouple Thermocouple Lubricant Lubricant supply Heater supply Heater system system Thermocouple Thermocouple Figure 1 Schematic diagram of the spacer layer imaging method set-up on WAM5 2.3.2. Atomic force microscopy Measurements were performed with an AFM MFP-3D atomic force microscope (Asylum Research, Santa Barbara, CA, US) in contact, constant force mode using non conductive silicon nitride cantilevers with a spring constant k = 40 N/m and a resonant frequency f0 = 300 kHz (Veeco Instruments, Santa Barbara, CA, US). Images were captured continuously at a scan rate of 0.25 Hz and set point of 1V. The surface roughness parameter, root mean square roughness, (Rrms) was obtained by processing the images. 2.3.3. X-ray Photoelectron Spectroscopy The X-ray photoelectron spectroscopy measurements were conducted in a PHI 5000 VersaProbe TM X-ray photoelectron spectrometer (Ulvac-PHI Inc, Chanhassen, MN, US) with a monochromatized Al Kα X-ray (1486.6eV) source. The diameter of the analysed area was 100 μm. The analyser was operated in the fixed-analyser-transmission mode 5 with a pass energy of 46.95 eV (full width at half-maximum (fwhm) for Ag3d5/2 = 1.1eV). Samples were argon ion sputtered using 2KV ions over an area of 2x2 mm2. Two set of parameters (spot size, pass energy, time per step) were used for the XPS analysis. The first set was required for survey and detailed spectra, while the second one was used during depth profiling for the survey and Fe 2p detailed spectrum. In the survey spectra of both thermal layers and additive-derived reaction layers only the elements carbon, oxygen, iron, phosphorus, sulfur and zinc were found. Detailed spectra of C1s, O1s, Fe2p, P2p together with Zn3s, S2p and Zn2p were measured. Data was processed with CASA XPS software (Casasoftware Ltd., UK). The detailed spectra were fitted with Gaussian-Lorentzian curves and a Shirley background. The signals of phosphorus, sulfur, iron and zinc of the 2p orbital exhibit two peaks: 2p3/2 and 2p1/2. This split arises from the coupling of the spin and orbital angular momentum. The area of the 2p3/2 peak is always twice that of the 2p1/2 and binding energy of the 2p1/2 is the higher of the two peaks. In the phosphorus and sulfur signals these two peaks are close to each other, resulting in an asymmetric signal. The model curves used for fitting the experimental peaks were linked to each other: the ratio of 2p1/2/2p3/2 contributions was fixed at 0.5 and the difference in energy was maintained equal to 1.25 eV for sulfur and 0.85 eV for phosphorus. For quantification the signal arising from Zn3s is used in stead of the most intense zinc photoelectron line, Zn2p3/2 , since is in the same binding energy range as phosphorus, sulfur and oxygen and therefore exhibits a similar inelastic mean free path (IMFP) as the other elements present in the layer. 3. RESULTS AND DISCUSSION 3.1. Friction and wear Initially, a friction force curve is performed for both base oil blends, as shown in Figure 2, in order to monitor the friction behaviour in relation to slide-roll ratio. The Synthetic Oil+ZDDP solution has a higher friction coefficient than the SKF Ester+ZDDP solution with increasing slide-roll ratio. It can also be observed how the absolute friction coefficient is slightly higher for negative SRR for both solutions, whereas the difference in the friction coefficient between both solutions is larger for positive SRR. 0.12 0.08 μ 0.04 0 -15 -10 -5 0 5 10 15 -0.04 Friction coefficient, -0.08 SKF Ester + ZDDP Synthetic Oil + ZDDP -0.12 Slide-roll ratio 6 Figure 2 Friction curve for different base oil-ZDDP solutions The friction coefficient increases very rapidly when the SRR ranges from 0 to 0.5%, being almost constant for higher SRR values. The slide-roll ratio values selected for further analysis, including reaction layer formation monitoring, cover pure rolling, the initial stage of high friction increase (±0.05 and ±0.5 %) and the steady-state stage ((±2, ±5 and ±10 %). In the initial region of high friction increase it was observed that the friction behavior of both oils was very similar, very close to pure rolling which might indicate that in the absence of shear the base oil molecules, that are initially attach to the steel surface [33], are primarily responsible for the lubrication of the surface as the additive molecules have not yet been activated. The antiwear properties of the reaction layers derived from the base oil + ZDDP solutions at different slide-roll ratios were also investigated. The wear track width (WTW) was measured from the balls using a calibrated microscope. Figure 3 shows the wear track width as a function of slide-roll ratio. 620 SKF Ester+ZDDP Synthetic Oil+ZDDP 600 580 560 Hertzian point contact diameter 540540 Wear Scar Width (um) 520 500 -10% -5% -2% -0.50% 0% 0.50% 2% 5% 10% Slide-roll ratio Figure 3 Wear track width as a function of slide-roll ratio The WTW for Synthetic Oil+ZDDP solutions at different slide-roll ratios is always higher than those of SKF Ester+ZDDP. The wear track increases with the slide-roll ratio for both solutions in either direction. The WTW for both solutions is very similar at SRR=-10%, however the values at SRR=10% are very different, being the WTW for SKF Ester Oil+ZDDP approximately 50% lower than the corresponding SKF Ester+ZDDP. The higher wear exhibited by the layers formed by Synthetic Oil+ZDDP, despite the layer being thicker than the ones derived from SKF Ester+ZDDP solutions, may be due 7 to the formation of thicker layer which cannot withstand the load and break away, as indicated by the morphology observed for those layers (see Section 3.3.2). The layers derived from the SKF Ester+ZDDP solutions at different slide-roll ratios, show WTW even lower than the contact diameter calculated from the Hertz theory for point contacts, at low SRR (ranging from -2 to 2%). At SRR, the energy input to the system is still low, as the activating rate of the additive; therefore the lubrication of the contacting surfaces is coming mainly from the base oil molecules. The polar base oil molecules have a higher affinity for the surface, attaching to it strongly, and thus providing a better protection for the contacting surfaces, explaining the lower WTW when compared to the wear track derived from the Synthetic+ZDDP. 3.2. Thermal layer characterization 3.2.1. Thickness The thermal and thermo-oxidative decomposition of ZDDP in oil solution needs to be investigated to understand the processes that can contribute to the formation of a reaction layer on rubbing steel surfaces. It has been found that ZDDP decompose at temperatures between 150°C and 250°C to form insoluble polyphosphates, polythiophosphates, volatile mercaptans, hydrogen sulfide and olefins. When steel balls are immersed without rubbing in base oil-ZDDP solution heated at different temperatures, a transparent, solid reaction layer forms on the steel surface. Figure 4 and Figure 5 shows the thermal layer formation, with the correspondent interferometry images at the tested temperatures. The rate of thermal layer formation increases with temperature, being P-S, P=S and C-O bonds replaced by C-S, P-O and P=O bonds [15]. A significant layer only develops at temperatures of 120°C and above. It can also be observed how the thermal layer developed with increasing temperature when the ZDDP is dissolved in Synthetic oil Group IV is much thicker than the layer derived from the SKF Ester oil+ZDDP solution. 500 SKF Ester+ZDDP, after 24h 400 300 200 100 Reaction layer thickness (nm) 0 0 20 40 60 80 100 120 140 160 Temperature (oC) Figure 4 Influence of bulk temperature on thermal layer formation after 24 hours static immersion in a solution of 2% ZDDP in SKF Ester 8 500 Synthetic Oil+ZDDP, after 24h 400 300 200 100 Reaction layer thickness (nm) 0 0 20 40 60 80 100 120 140 160 Temperature (oC) Figure 5 Influence of bulk temperature on thermal layer formation after 24 hours static immersion in a solution 2% ZDDP in Synthetic oil Group IV 3.2.2. Topography Atomic force microscopy representative images, 80 x 80 μm of the central region of the thermal reaction layer formed on the fresh steel surface are displayed in Figure 6. ZDDP thermal layers appear to grow as separate islands to form a mound-like structure which subsequently coalesce to form, at least at high temperatures, a smoother structure. An indentation modulus of E*≈35 GPa and hardness of H≈1.5 GPa [34] has been reported for the thermal layers. 80 80 150 150 60 60 100 100 50 50 40 40 µm µm 0 0 nm nm -50 -50 20 20 -100 -100 -150 -150 0 0 0 20 40 60 80 0 20 40 60 80 150 µm 150 µm 75 75 0 0 -75 -75 Height (nm) Height (nm) -150 -150 020406080020406080 Contact (um) Contact (um) Figure 6 AFM images and line profiles of thermal layers from 24h immersion tests at 150°C for (a) SKF Ester+ ZDDP and (b) Synthetic Oil +ZDDP 9 A slight coverage can be observed in the surface when the steel ball is submerged in the SKF Ester + ZDDP, however the initial surface, shallow surface finishing marks, are visible. Whereas the thermal layer derived from the Synthetic Oil + ZDDP shows a completely different morphology. The surface coverage is inhomogeneous, while part of the initial surface can be still observed; irregular areas are covered by a thick layer (≈ 300 nm). Previous studies have show that thermal layers grow gradually and accumulate on the surface with increasing immersion time [35], even no obvious morphology changes were observed on thermal layers from 2 to 12 h. The surface roughness data are reported as root mean square (Rrms), being 16.4 nm for SKF Ester + ZDDP and 57.7 nm for Synthetic Oil + ZDDP. 3.2.3. Chemical composition The influence of base oil polarity in the chemical composition of the ZDDP-derived thermal layers has been studied using X-ray photoelectron spectroscopy. Survey and detailed (high resolution) spectra were recorded on the thermal layer that appears to cover the steel surface after the immersion tests. In the survey spectra, only signals from carbon, oxygen, sulfur, phosphorus and zinc were detected. Detailed spectra of carbon 1s, oxygen 1s, sulfur 2p, phosphorus 2p and zinc 3s and zinc 2p, were performed. The survey and detailed spectra of the thermal layer derived from the SKF Ester+ZDDP solution is shown in Figure 7. Two peaks were used to fit the C1sspectra, the main peak being found at 286.65 eV. The sample charging, probably caused by the insulating character of the thermal layer, was compensated by referring the C1s peak at 286.6 to 285 eV from aliphatic carbon (C- C, C-H), thus shifting all peak values. All the positions are corrected for sample charging. The second contribution can be assigned to C-S bond. In the O1speak three contributions were found. They were all fitted with a fixed FWHM of 1.97 eV. An initial very small peak was detected at 530.2 eV that can be assigned to iron or zinc oxide [36, 37]. Two main contributions were found: a peak at 532.2 eV, non-bridging oxygen (NBO), that can be assigned to phosphorus bound in an orthophosphate, and a peak at 533.9 eV, bridging oxygen (BO), that can be assigned to P-O-C or P-O-P bonding environment [38]. The BO/NBO ratio indicates the chain length of the phosphates present in the thermal layer. The values calculated according to the recorded signal indicate the presence of polyphosphates. The S2ppeak was fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.9 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [36, 37, 39]. The P2ppeak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 134.2 eV. This value is typical for a phosphate. The peak maximum of the Zn2p3/2 was found at 1022.9 eV. 10 O1s Fe 2p C1s S2p Zn 3s P2p C1s O 1s NBO BO oxide P2p S2p S2p3/2 P2p3/2 Zn 3s S2p3/2 P2p3/2 Figure 7 XPS survey and detailed spectra of C1s, O1s, P2p, Zn3s and S2p measured in the thermal layer derived from SKF Ester+ZDDP, 24 hours immersion test at150 °C 11 Figure 8 present the survey and detailed spectra of the thermal layer derived from the Synthetic Oil+ZDDP solution Two peaks were used to fit the C1sspectra, the main peak being found at 285.27 eV. The sample charging, probably caused by the insulating character of the thermal layer, was compensated by referring the C1s peak at 285.27 to 285 eV from aliphatic carbon (C-C, C-H), thus shifting all peak values. All the positions are corrected for sample charging. The second contribution can be assigned to C-S bonds. In the O1speak three contributions were found. They were all fitted with a fixed FWHM of 1.65 eV. An initial very small peak was detected at 530.3 eV that can be assigned to iron or zinc oxide [36, 37]. Two main contributions were found: a peak at 532.3 eV, non-bridging oxygen (NBO), that can be assigned to phosphorus bound in an orthophosphate, and a peak at 533.9 eV, bridging oxygen (BO), that can be assigned to P-O-C or P-O-P bonding environment [35]. The BO/NBO ratio indicates the chain length of the phosphates present in the thermal layer. The values calculated according to the recorded signal indicate the presence of polyphosphates. A very low signal was detected for the S2ppeak. It was fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 163.5 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [36, 37, 39]. The P2p peak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 134.5 eV. This value is typical for a phosphate. The peak maximum of the Zn2p3/2 was found at 1023.1 eV. 12 O 1s Fe 2p S2p C1s Zn 3s P2p C1s O 1s NBO BO oxide P2p S2p S2p3/2 P2p3/2 S2p3/2 Zn 3s P2p3/2 Figure 8 XPS survey and detailed spectra of C1s, O1s, P2p, Zn3s and S2p measured in the thermal layer derived from SKF Ester+ZDDP, 24 hours immersion test at150 °C 13 The P/S and P/Zn ratio of both ZDDP-derived thermal layers were calculated. The P/S and P/Zn ratio increased in comparison to the pure additive. This effect is related [40] to an enrichment of phosphorus in the contact area compared to other elements present from the additive. The P/S ratio obtained for SKF Ester+ZDDP solution has a value of 3.8, according to other reported in literature [41, 42], but the ratio for Synthetic Oil+ZDDP solution is much higher, indicating the little presence of S in the layer. 3.3. Reaction layer characterization 3.3.1. Thickness A series of interference images of the centre of the wear track derived from the SKF Ester Oil+2wt%ZDDP solution from the rubbing tests at different slide-roll ratios are shown in Figure 9. In all the images the sliding direction is from bottom to top. It can be seen that a reaction layer, as evidenced by the colour scale in the images, develops differently due to the variation on SRR. Figure 9 Series of interference images from wear track for SKF Ester Oil+2% iso-C4- ZDDP at a entrainment speed of 0.25 m/s, 1.9 GPa, 90 °C at different slide-roll ratios From the interference images, reaction layer thickness values were determined at each position across a vertical profile of the contact. From those values the average reaction layer thickness at the centre of the contact was calculated. The evolution of an average reaction layer derived from the Synthetic Oil group IV+ZDDP solution as a function of SRR is shown in Figure 10. It can be seen that reaction layer formation rate increases with increasing slide-roll ratio. 14 150 SKF Ester+ZDDP (-)5%Slip SKF Ester+ZDDP (-)2%Slip SKF Ester+ZDDP 0%Slip SKF Ester+ZDDP (+)2%Slip SKF Ester+ZDDP (+)5%Slip 100 50 Mean reaction layer thickness (nm) 0 0 500 1000 1500 2000 2500 3000 Rubbing distance (m) Figure 10 Influence of slide-roll ratio on the growth of mean ZDDP-derived (SKF Ester+ZDDP) reaction layer thickness with rubbing distance and slip calculated from images in Figure 9. A series of interference images of the centre of the wear track derived from the Synthetic Oil group IV+ZDDP solution from the rubbing tests at different negative slide-roll ratios are shown in Figure 11. 15 Figure 11 Series of interference images from wear track for Synthetic Oil Group IV+2wt% ZDDP at a entrainment speed of 0.25 m/s, 1.9 GPa, 90 °C at different negative slide-roll ratios Figure 12 presents in detailed the results for the negative slide-roll ratios tested (for clarity (+) 0.05% is not plotted as the SRR values were not stable during the test, ranging 0 to (+) 0.1% and the reaction layer formation was negligible). We can observed how the reaction layer develops initially very rapidly, indicating a fast reaction, to subsequently slow down as the reaction layer grows to a stable thickness value (“limiting thickness”). This may indicate that the process is catalyzed by chemical species generated or released from the surface due to the rubbing process, being more prominent during the running-in period, when surfaces are more reactive, and stabilizing afterwards, when the surface is protected by the reaction layer. It is also remarkable how, independently of the reaction layer thickness values reached in this initial stage, the layers reach a limiting thickness, being the values for all the SRR tested very similar (≈ 60-70 nm). The limiting thickness is determined by a balance between the rate of growth, depending on the additive concentration in base oil, contact temperature, and the rate of removal, determined by wear [43]. From these results we can conclude that SRR, once present, has a minor effect on the limiting thickness of the layer, but they also show how shear rate is a fundamental parameter on the initial activation of the additive towards the surface. 150 Synthetic Oil+ZDDP (-) 10%Slip Synthetic Oil+ZDDP (-) 5%Slip Synthetic Oil+ZDDP (-) 2%Slip ) Synthetic Oil+ZDDP (-) 0.5%Slip Synthetic Oil+ZDDP 0%Slip 100 50 Mean layer reaction thickness (nm 0 0 500 1000 1500 2000 2500 3000 Rubbing distance (m) Figure 12 Influence of negative slide-roll ratio on the growth of mean ZDDP-derived (Synthetic Group IV + ZDDP) reaction layer thickness with rubbing distance and slide- roll ratio calculated from images in Figure 11 16 The overshoot in reaction layer thickness observed at the initial very rapid reaction layer formation, effect also reported at low ratios and high temperatures [22, 32], may indicate an initially very rapid reaction, which subsequently slows as a tribofilm develops, indicating a possible process catalysed by rubbing steel on steel. Thus it is possible that rubbing releases iron species, probably initially Fe2+ andthatthese exchange rapidly with ZDDP to form FeDDP. This iron dithiophosphate then decomposes at low temperatures to dorm iron phosphate glass in a reaction similar to that which occurs for ZDDP at much higher temperatures. This would also establish the autocatalytic reaction so that ZDDP can subsequently decompose directly. Figure 13 show a series of interferometry images taken for the positive slide-roll ratio. Figure 13 Series of interference images from wear track for Synthetic Oil Group IV+2wt% ZDDP at a entrainment speed of 0.25 m/s, 1.9 GPa, 90 °C at different positive slide-roll ratios Figure 14 presents the results obtained for the positive slide-roll ratios. The positive SRR shows a similar tendency as presented for the negative SRR. The most significant difference is that the initial overshoot in layer thickness for SRR=5% is not present. The initial fast reaction may be too fast to be capture by the interferometry method, so when the first image is taken, the layer has already reached its limiting thickness. 17 150 Synthetic Oil+ZDDP 0%SRR Synthetic Oil+ZDDP (+) 2%SRR Synthetic Oil+ZDDP (+) 5%SRR 100 50 Mean reaction layer thickness (nm) 0 0 500 1000 1500 2000 2500 3000 Rubbing distance (m) Figure 14 Influence of negative slide-roll ratio on the growth of mean ZDDP-derived (Synthetic Oil+ZDDP) reaction layer thickness with rubbing distance and slip calculated from images in Figure 13 3.3.2. Topography A series of AFM images of 80x80 μm2 areas of wear track centre are shown in Figure 15, together with their corresponding line profiles taken transverse to the rolling-sliding direction. A reaction layer is formed only on the wear track and not on the non-contact area. The rolling-sliding direction is from bottom to top of the displayed images. SRR =-0.5% SRR =-2% SRR =-5% SRR =-10% 80 80 80 80 100 100 100 100 60 60 60 60 50 50 50 50 40 40 40 µm 40 µm µm µm 0 0 0 0 nm nm nm nm -50 -50 -50 -50 20 20 20 20 -100 -100 -100 -100 0 0 0 0 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 0 20 40 60 80 µm 150 µm 150 150 µm 150 µm 75 75 75 75 0 0 0 0 -75 -75 -75 -75 Height (nm) Height (nm) Height (nm) Height (nm) -150 -150 -150 -150 0 20406080 0 20406080 0 20406080 0 20406080 Contact (um) Contact (um) Contact (um) Contact (um) Rrms = 16.1 nm Rrms = 15.8 nm Rrms = 34.4 nm Rrms = 57.3 nm Figure 15 Topography of 80 x 80 μm2 scanned areas from the centre of the reaction layers and corresponding line profiles taken transverse of the rolling-sliding direction 18 At SRR=-0.5%, the surface is homogeneously covered, but no distinctive features can be identified. When the slide-roll ratio increases to SRR=-2%, a pad-like structure develops. The pads, 10-15μm in length, are higher and smoother, that the surroundings and are orientated in the rolling-sliding direction, and are consider as the load carrying features of the layer [44, 45], being therefore responsible for the antiwear protection. At SRR=-5%, the overall structure of the layer is very similar, but it start showing signs of a wearing-out, as part of the layer seem to be flaking off. At SRR=-10%, the flakes have developed further covering the whole reaction layer surface. The roughening of the layer that can be observed, it is in direct relation with the friction behaviour It has been observed that a delamination or flaking process of the reaction layer derived from Synthetic oil + ZDDP solution occurs in this last stage when the SRR is >5%. The nanomechanical properties of the layer can be responsible for this and will be further investigated. Also the possible relation between this delamination process and crack formation in the surface will be address, by long run tests. 3.3.3. Chemical composition The influence of base oil polarity and SRR in the chemical composition of the ZDDP- derived reaction layers has been studied using X-ray photoelectron spectroscopy. Survey and detailed (high resolution) spectra were recorded on areas at the centre of the reaction layers. In the survey spectra, only signals from carbon, oxygen, sulfur, phosphorus and zinc were detected. Detailed spectra of carbon 1s, oxygen 1s, sulfur 2p, phosphorus 2p and zinc 3s and zinc 2p, were performed. 19 O1s Fe 2p S2p C1s Zn 3s P2p C1s O1s NBO BO oxide P2p S2p S2p3/2 P2p3/2 Zn 3s S2p3/2 P2p3/2 Figure 16 XPS survey and detailed spectra of C1s, O1s, P2p, Zn3s and S2p measured in the reaction layer derived from SKF Ester+ZDDP at SRR=5% Figure 16 shows the survey and detailed spectra acquired on the reaction layer produced at SRR=5% in 2 wt-% solution of ZDDP on SKF Ester. All spectra are shifted to correct for sample charging by referring the main C 1s peak to 285 eV (aliphatic carbon). 20 The C1speak at 284.7 eV was referred to 285 eV from aliphatic carbon (C-C, C -H), thus shifting all peak values. The second contribution can be assigned to C -S bond The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.70 eV. An initial very small peak was detected at 530.3 eV that can be assigned to iron or zinc oxide [36, 37]. Two main contributions were found: a peak at 532.3 eV, non-bridging oxygen (NBO), that can be assigned to phosphorus bound in an (poly)phosphate, and a peak at 533.9 eV, bridging oxygen (BO), that can be assigned to P-O-C or P-O-P bonding environment [38, 46]. The BO/NBO ratio indicates the chain length of the phosphates present in the reaction layer. The values calculated according to the recorded signal indicate the presence of polyphosphates. The S2ppeak was fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.7 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [36, 37, 39] and thiols or if sulfur substitutes oxygen in a phosphate [47]. The P2ppeak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 134.3 eV, value corresponding to a phosphate. The Zn 3s peak is found at 140.6 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV. 21 O 1s Fe 2p S2p C1s Zn 3s P2p C1s O1s NBO BO oxide P2p S 2p P2p S 2p3/2 Zn 3s 3/2 S2p3/2 P2p3/2 Figure 17 XPS survey and detailed spectra of C1s, O1s, P2p, Zn3s and S2p measured in the reaction layer derived from Synthetic Oil+ZDDP at SRR=5% Figure 17 shows the survey and detailed spectra acquired on the reaction layer produced at SRR=5% in 2 wt-% solution of ZDDP on SKF Ester. All spectra are shifted to correct for sample charging by referring the main C 1s peak to 285 eV (aliphatic carbon). 22 Two peaks were used to fit the C1sspectra, the main peak being found at 285.0 eV assigned to aliphatic carbon (C-C, C -H). The second contribution can be assigned to C-S bonds. The O1sconsist of three peaks. They were all fitted with a fixed FWHM of 1.65 eV. An initial very small peak was detected at 529.8 eV that can be assigned to iron or zinc oxide. Two main contributions were found: a peak at 531.8 eV, non-bridging oxygen (NBO) that can be assigned to phosphorus bound in an orthophosphate, and a peak at 533.4 eV, bridging oxygen (BO), that can be assigned to P-O-C or P- O -P bonding environment. The BO/NBO ratio indicates the chain length of the phosphates present in the reaction layer. The values calculated according to the recorded signal indicate the presence of polyphosphates. The S2phas two contributions, and both peaks were fitted with a doublet with a fixed area ratio and a delta between the 2p3/2 and the 2p1/2 of 1.25 eV. The peak was detected at 162.9 eV, corresponding to oxidation state -2, S(II), can be assigned to C-S bonded in sulfides [36, 37, 39].The peak found at 168.9 eV corresponds with values reported for an oxidation state of -6, S(VI) that can be assigned to sulfate groups [36]. The P2ppeak was fitted with a doublet with fixed area ratio a delta of 0.85 eV. The peak position was found at 133.7 eV. This value is typical for a phosphate. The Zn 3s peak is found at 140.5 eV, while the peak maximum of the Zn2p3/2 was found at 1023.1 eV. Both reaction layers are very similar, being the main difference in the sulfur content and the oxidation state of the sulfur. In the layer derived from the polar base oil, the sulfur is present as a sulfide, whereas in the reaction layer derived form the non-polar base oil, sulfides and sulfates are present. The reaction layers for the negative slide-roll ratio for both base oils, have a similar composition to the one presented for the positive slide-roll ratio. The main significant difference is that the iron content for the negative SRR is very low, nearly not detectable. 4. CONCLUSIONS EP/AW additives for a protective layer on the surface of the rubbing steel surfaces. Therefore in order to identify the plausible mechanisms that govern the detrimental effect of those additives on bearing performance it is necessary to identify the parameters affecting the additive-derived layer formation and the tribological properties of this layer, as it is directly related to the additive reactivity towards the surface. In situ spacer layer interferometry has been used to monitor the formation of both thermal and reaction layer derived from ZDDP solutions on steel surfaces. It has been shown that for both ZDDP systems studied, negligible thermal layer formation takes place below a temperature of 120°C, but when that “activation” temperature is reached, the base oil polarity determines the thickness of the thermal layer. The layer derived from the ZDDP solution in non-polar base oil, PAO in this study, thicker than the one derived from polar base oil solution. ZDDP derived reaction layers on rubbed surfaces are generated at lower temperatures than the correspondent thermal layers. Tests at different slide-roll ratio, in positive and negative direction, but with the same contact pressure, temperature and entrainment speed have enabled the characterization of the effect of base oil polarity and shear stress on reaction layer formation. 23 To illustrate the effect of base oil polarity on reaction layer formation, the results for a SRR=-5% are shown in Figure 18, as mean reaction layer thickness as a function of rubbing distance. The mean layer thickness is calculated from the line profiles taken across the corresponding interferometry images. 150 SKF Ester + ZDDP Synthetic Oil + ZDDP 100 50 Mean reaction layer thickness (nm) 0 0 500 1000 1500 2000 2500 Rubbing distance (m) Figure 18 Stages of reaction layer formation It is possible to identify 3 different stages on the reaction layer formation process. Initially, an ACTIVATION stage can be identified, with a rate of layer formation higher that the removal rate: The additive molecules are activated by the tribological energy and approach the surface. The growth mechanism begins with distinct reaction events on micro-asperity contact at the steel surfaces. The additive molecules are activated by the tribological energy and approach the surface. The growth mechanism begins with distinct reaction events on micro-asperity contact at the steel surfaces, leading to distinct segregated pads. The reaction layer develops initially very rapidly consequence of a very rapid reaction, being the rate of formation higher that the rate of removal. This suggests that the layer formation may be strongly catalysed by chemical species generated or released during rubbing [27], such as soluble Fe2+ or Fe3+.By ligand exchange, the iron ion replace the zinc in ZDDP to form a less thermally stable metal dithiophosphate, that subsequently decomposes [48]. Thus it is possible that rubbing releases iron species, probably initially Fe2+ and that these exchange rapidly with ZDDP to form FeDDP. This iron dithiophosphate then decomposes at low temperatures to dorm iron phosphate glass in a reaction similar to that which occurs for ZDDP at much higher temperatures.The process can also be triggered by triboelectronic processes such as exoelectron emission and a subsequent negative ion reaction [49]. The reaction layer rapidly develops until reaching saturation: At this point the steel surface is completely covered by the reaction layer, which slows the rate of reaction layer formation. The reaction layer rapidly develops until reaching saturation: At this point the steel surface is completely covered by the reaction layer, which slows the rate of reaction layer formation. 24 The next stage involves a WEARING-OUT of the layer, being the rate of formation lower than the rate of removal, until the layer reaches the final stage of EQUILIBRIUM, between the rate of formation and removal, when the layer reaches a constant thickness value, known as “limiting thickness”. It can be observed how the polarity of the base oil influences the reaction of the ZDDP additive with the steel surface. A thicker layer is formed when the additive is blended in the non-polar oil, due to the higher affinity of the polar base oil molecules for the steel surface, that limit the access of the additive molecules to the surface and therefore their ability to attach and react with it to form a protective reaction layer. ACKNOWLEDGEMENT The authors wish to thank SKF for permission to publish this work. The work presented in this paper has been supported by the EC, Sixth Framework Programme, Marie Curie Action (WEMESURF research network entitled: ”Characterization of wear mechanisms and surface functionalities with regard to life time prediction and quality criteria-from micro to nano range” under contract MRTN CT 2006 035589). REFERENCES 1. Friction, lubrication and wear technology. ASM Handbook. 1992. 2. Sakamoto, T., at al., Reaction layer formation on bronze with an S-P extreme pressure additive in boundary lubrication under increasing load: I. Wear, 1985. 105(4): p. 307-321. 3. Sarin, R., et al., Additive-additive interactions: Search for synergistic FM-EP-AW composition. Wear, 1994. 174(1-2): p. 93-102. 4. Jianqiang, H., et al., Study on tribological properties and action mechanism of organic cadmium compound in lubricants. Wear. 259(1-6): p. 519-523. 5. Zhang, J., et al., Investigation of the friction and wear behaviors of Cu(I) and Cu(II) dioctyldithiophosphates as additives in liquid paraffin. Wear, 1998. 216(1): p. 35-40. 6. Boshui, C., at al., Tribochemistry of gadolinium dialkyldithiophosphate. Wear, 1996. 196(1-2): p. 16-20. 7. Barnes, A.M., et al., A review of zinc dialkyldithiophosphates (ZDDPs): characterisation and role in the lubricating oil. Tribology International, 2001. 34: p. 389-395. 8. Georges, J.M., et al., Mechanism of boundary lubrication with zinc dithiophosphate. Wear, 1979. 53(1): p. 9-34. 9. Watkins, R.C. at al., Antiwear mechanism of ZDDP's - 2. Tribology international, 1982. 15(1): p. 13-15. 10. Spedding, H., et al., Antiwear mechanism of ZDDP's - 1. Tribology international, 1982. 15(1): p. 9-12. 11. Gellman, A.J., at al., Surface chemistry in tribology. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2002. 216(6): p. 443-461. 12. Bancroft, G.M., et al., Mechanism of tribochemical film formation: stability of tribo- and thermally-generated ZDDP films. Tribology Letters, 1997. 3: p. 47-51. 25 13. Willermet, P.A., et al., Mechanism of formation of antiwear films from zinc dialkyldithiophosphates. Tribology International, 1995. 28(3): p. 177-187. 14. Spedding, H., Antiwear mechanism of ZDDP - 1. Tribology international, 1982. 15(1): p. 9-12. 15. Coy, R.C., at al., Thermal degradation and EP performance of zinc dialkyl dithiophosphate additives in white oil. ASLE transactions, 1981. 24(1): p. 77-90. 16. Fuller, M., at al., Chemical characterization of tribochemical and thermal films generated from neutral and basic ZDDPs using X-ray absorption spectroscopy. Tribology international, 1997. 30(4): p. 305-315. 17. Yin, Z., et al.,Chemical characterization of antiwear films generated on steel by zinc dialkyl dithiophosphate using X-ray absorption spectroscopy. Tribology international, 1993. 26(6): p. 383-388. 18. Yamaguchi, E.S., et al., Comparison of the relative wear performance of neutral and basic ZnDTP salts. ASLE transactions, 1996. 39(1): p. 220-224. 19. Torrance, A.A., et al., An additive’s influence on the pitting and wear of ball bearing steel. Wear, 1996. 192: p. 66-73. 20. Tripaldi, G., at al., Friction behaviour of ZDDP films in the mixed, boundary/EHD regime. SP, 1996. 1209: p. 73-84. 21. Fujita, H., et al., Study of zinc dialkyldithiophosphate antiwear film formation and removal processes, part II: Kinetic model. Tribology Transactions, 2005. 48(4): p. 567-575. 22. Fujita, H., et al., Study of zinc dialkydithiophosphate antiwear film formation and removal processes, part I: Experimental. Tribology Transactions, 2005. 48(4): p. 558-566. 23. Wan, G.T.Y., et al., Effect of extreme-pressure additives on fatigue life of rolling bearings. Journal of physics. D, applied physics, 1992. 25(1): p. A147-A153. 24. Minami, I., et al., Investigation of anti-wear additives for low viscous synthetic esters: Hydroxyalkyl phosphonates. Tribology international, 2007. 40(4): p. 626- 631. 25. Minami, I., et al., Anti-wear additives for ester oils. Technische Akademie Esslingen International Tribology Colloquium Proceedings, 2004. 14 III: p. 1593- 1598. 26. Kar, P., et al., Formation and characterization of tribofilms. Journal of tribology, 2008. 130(4). 27. Fujita, H., et al, Study of zinc dialkyldithiophosphate antiwear film formation and removal processes, part II: Kinetic model. Tribology Transactions, 2005. 48(4): p. 567-575. 28. Fujita, H. et al., Study of zinc dialkydithiophosphate antiwear film formation and removal processes, part I: Experimental. Tribology Transactions, 2005. 48(4): p. 558-566. 29. Fujita, H., et al., The formation of zinc dithiophosphate antiwear films. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2004. 218(4): p. 265-277. 30. Spikes, H., at al, The development and application of the spacer layer imaging method for measuring lubricant film thickness. Proceedings of the Institution of 26 Mechanical Engineers, Part J: Journal of Engineering Tribology, 2001. 215(3): p. 261-277. 31. Donnet, C., Handbook of Surface and interface Analysis - Methods for Problem- Solving. 1998. 32. Fujita, H., et al.,The formation of zinc dithiophosphate antiwear films. Proceedings of the Institution of Mechanical Engineers. Part J, Journal of engineering tribology, 2004. 218(4): p. 265-277. 33. Naveira-Suarez, A., et al. Parameters affecting the functionality of additives in tribological contacts - An experimental and Molecular Dynamics simulation study. in IV World Tribology Congress. 2009. Kyoto, Japan. 34. Aktary, M., et al.,Morphological evolution of films formed from thermooxidative decomposition of ZDDP. Wear, 2001. 247(2): p. 172-179. 35. Li, Y.-R., et al., Studies on ZDDP Thermal Film Formation by XANES Spectroscopy, Atomic Force Microscopy, FIB/SEM and 31P NMR. Tribology Letters, 2008. 29(1): p. 11-22. 36. Heuberger, R., et al., XPS study of the influence of temperature on ZnDTP tribofilm composition. Tribology Letters, 2007. 25(3): p. 185-196. 37. Heuberger, R., et al.,r, Pressure Dependence of ZnDTP Tribochemical Film Formation: A Combinatorial Approach. Tribology Letters, 2007. 28(2): p. 209- 222. 38. Piras, F.M., et al.,i, Combined in situ (ATR FT-IR) and ex situ (XPS) study of the ZnDTP-iron surface interaction. Tribology Letters, 2003. 15(3): p. 181-191. 39. Martin, J.M., et al., The two-layer structure of Zndtp tribofilms, Part I: AES, XPS and XANES analyses. Tribology International, 2001. 34(8): p. 523-530. 40. Yamaguchi, E..S., et al., Time-dependent film formation from ZnDTPs and nonphosphorus antiwear agents. Tribology Transactions, 2005. 48(1): p. 57-68. 41. Zhang, Z., et al., Tribofilms generated from ZDDP and DDP on steel surfaces: Part 1. Growth, wear and morphology Proceedings of the World Tribology Congress III - 2005. 2005. 617-618. 42. Zhang, Z., et al., Tribofilms generated from ZDDP and DDP on steel surfacUniversity Parkes: Part 2, chemistry Proceedings of the World Tribology Congress III - 2005. 2005. 619-620. 43. Palacios, J.M, at al.,., Films formed by antiwear additives and their incidence in wear and scuffing. Wear, 1987. 114(1): p. 41-49. 44. Warren, O.L., et al., Nanomechanical properties of films derived from zinc dialkyldithiophosphate. Tribology Letters, 1998. 4(2): p. 189-198. 45. Graham, J.E., et al., Topography and nanomechanical properties of tribochemical films derived from zinc dialkyl and diaryl dithiophosphates. Tribology Letters, 1999. 6(3): p. 149-57. 46. Eglin, M., et al.,, X-ray Photoelectron Spectroscopy Analysis of Tribostressed Samples in the Presence of ZnDTP: A Combinatorial Approach. Tribology Letters, 2003. 15(3): p. 199-209. 47. Rossi, A., et al., Surface analytical studies of surface-additive interactions, by means of in situ and combinatorial approaches. Wear, 2004. 256(6): p. 578-584. 27 48. Piras, F.M., et al., Growth of tribological films: In situ characterization based on attenuated total reflection infrared spectroscopy. Langmuir, 2002. 18(17): p. 6606-6613. 49. Kajdas, C.K., et al.,Importance of the triboemission process for tribochemical reaction. Tribology International, 2005. 38(3): p. 337-353. 28