Contents

    

Nanoscale Boundary Lubrication of Diamond-like Carbon Coatings with Fluorinated Compounds*

C. Kajdas

Institute of Chemistry and Central Petroleum Laboratory, Warsaw University of Technology, Poland

Abstract

Progress in the technology of magnetic media has brought about a remarkable increase in recording density. The most important factor determining the utility of magnetic disks is durability against head wear, and this durability is controlled by several factors. The present paper discusses the tribology of these media, particularly from the viewpoint of boundary lubrication. In that context there are two characteristic features of this lubrication regime: specific standard lubricants (fluoropolyethers such as Z-DOL and perfluoropolyetliers such as Z-15) and the newer application of these lubricants in the form of films only a few nanometers thick. Advanced phosphazene-type fluorinated compounds are of most interest at present, so these compounds are discussed in more detail. The emphasis is on X-1P lubricant used either alone or as an additive for fluoro- and perfluoropolyethers deposited on protective diamond-like carbon coatings.

Keywords

boundary lubrication, fluorinated lubricants, diamond-like carbon, computer disks

* — Published in : J. Synthetic Lubricants 18-1, April 2001. (18) 18

INTRODUCTION

New sophisticated techniques for measuring friction, wear, lubricant thickness, surface topography, and adhesion (all on a micro- to nanoscale) for imaging lubricant molecules, and the availability of superconductors for atomic scale simulations have led to the development of a new field of research called microtribology, nanotribology, molecular tribology, or atomic-scale tribology [1-3]. The converse of macrotribology, micro/nanotribology studies the wear behaviour of components at least one of which has extremely small mass. Since negligible wear occurs in these circumstances, tribological performance at this scale is mostly controlled by the lubrication of component surfaces. Under these specific nanoscale boundary lubrication conditions, lubricant films are very thin.
For solid lubricants, fluorinated or perfluorinated compounds are usually used. The perfluoroether (PFPE) class of lubricants is widely used on magnetic recording disks for computer data storage. In computers, commonly used lubricants include selected PFPEs as well as their polar derivatives containing various reactive end groups. A good lubricant should have a high degree of interaction between its molecules and the sliding surfaces. Head-disk interface interactions amistry of both head and disk surfaces as well as the lubricant chemistry. In a magnetic rigid-disk drive, recording/playback is accomplished by relative motion between a magnetic medium against a stationary read/write magnetic-head slider. Presently, diamond-like carbon (DLC) coatings lubricated with PFPE lubricants are widely used as the protective coating for thin-film magnetic storage disks. Recent developments in DLC films for magnetic media, and their chemical, mechanical, and tribological characteristics, are described elsewhere [4]. Another paper has reviewed the literature on interaction and degradation mechanisms of PFPE lubricants with protective carbon coatings used on magnetic media [5]. There is a boundary friction contact between the slider and the disk during starting and stopping (contact start-stop (CSS)). To reduce wear during the boundary friction operating conditions, hard coatings and liquid lubricant films are typically applied to the disk surface. To reduce static and dynamic friction on a thin-film magnetic recording disk used with a read/write head, the disk surface is typically lubricated with fluorinated or perfluorinated organic compounds. The computer head/disk interface is one of the most important factors controlling the performance and life of hard disk drives. Presently, the most widely used lubricants for disk drives are PFPEs with hydroxyl functional groups (Z-DOL and SA). The need to increase recording density has recently led to smoother disk surfaces and lower flying heights, but these changes have made the tribological environment at the head/disk interface much more severe. Accordingly, new lubricants and/or additives have been developed, which encompass cyclic phosphazene compounds that are environmentally stable and have better tribological performance than polar PFPEs.
The literature includes many reports about PFPE lubricant interaction with carbon coatings under both static and sliding conditions. Some of these papers and a number of other publications also relate to PFPE degradation mechanisms. The mechanisms include thermal decomposition, catalytic degradation, electron mediated degradation, and mechanical degradation processes. All these processes have recently been reviewed [5]. The goal of the present paper is to present new information on fluorinated compounds other than PFPE lubricants, namely phosphazene-type derivatives.

Table 1 Chemistry and selected properties of PFPE lubricants Z-15, Z-DOL, and SA, and the X-1P lubricant/additive [6]

LubricantFormulaEnd groupsMolecular weight
Z-15CF3—O—(CF2—CF2—O)m—(CF2—O)n—CF3—CF39100
Z-DOL*HO—CH2—CF2—O—(CF2—CF2—O)m—(CF2—O)n—CF2—CH2—OH               (m/n ~ 2/3)—OH2000
SA*F—(CF2—CF2—CF2—O)m—CF2—CF2—CH2—OH—OH3600
X-1P—CF31000
LubricantFormulaDensity
(x10³ kg/m³
at 20°C)		Kinematic viscosity (mm²/s)
Z-15CF3—O—(CF2—CF2—O)m—(CF2—O)n—CF31.84150 at 20°C
Z-DOL*HO—CH2—CF2—O—(CF2—CF2—O)m—(CF2—O)n—CF2—CH2—OH
(m/n ~ 2/3)		1.8180 at 20°C
34 at 40°C
SA*F—(CF2—CF2—CF2—O)m—CF2—CF2—CH2—OH1.8775 at 40°C
X-1P1.481826 at 20°C
205 at 40°C
* — Functional or polar lubricants

Emphasis is on X-1P lubricant used alone and as an additive for PFPEs deposited on protective DLC coatings. Table 1 shows some PFPE lubricants and the X-1P compound.

PHOSPHAZENE-TYPE COMPOUNDS

Chemistry and general tribological characteristics

Phosphazenes are ring or chain compounds consisting of alternating phosphorus and nitrogen atoms with two substituents attached to phosphorus. Owing to the presence of phosphorus and nitrogen, phosphazenes are inherently fire resistant. Cyclic phosphazenes are either liquids or low melting point crystalline solids. Their physical properties vary considerably with molecular weight and substituents. Phosphazenes are of interest in applications where fire resistance and thermal stability are important considerations. With the proper selection of substituents, thermally and hydrolytically stable ring and chain compounds, including fluids with low pour points and good thermal stability, have been synthesised [7]. Phosphazenes have properties necessary for advanced lubricants and/or additives for high performance applications [7-8]. Recently, a number of substituted aryloxycyclotriphosphazenes were investigated with a view to meeting the lubrication requirements of the integrated high performance turbine engine technology initiative [9]. Consideration of various properties and economic factors led to the identification of a derivative containing four meta-CF3, and two para-F substituents, as the leading fluid candidate.
Table 1 presents the chemical structure of this compound, usually denoted as X-1P. X-1P was found to be superior to 5P4E polyphenyl ether in pumpability, auto-ignition temperature, and vapour pressure at moderate temperatures [9]. It almost matches 5P4E in viscosity above room temperature and volatility, but is slightly inferior in vapour pressure at these temperatures, and in thermal conductivity and specific heat. Comparative four-ball tests on X-1P with different ball specimens have shown it to be superior to 5P4E polyphenyl ether on M-50 steel and SS 440C steel at 45 N and 135 N. However, with 52100 steel, X-1P and 5P4E showed almost identical characteristics at both loads [9].

X-1P boundary film formation

Recently, the dynamics of the formation and loss of boundary films formed during the lubricated sliding of steel surfaces were investigated over a range of temperatures and applied loads [10]: X-1P lubricant and mineral oil with and without zinc dialkyldithiophosphate (ZnDDP) additive were used in tests on a
Figure 1 Model for X-1P tribofilm [10]
cylinder-on-disc machine. The performance of these lubricants was found to be closely associated with boundary-film forming ability. Sliding is necessary to form the films and a certain amount of time is also required. It was concluded [10] that the films formed in X-1P lubricant develop more slowly than those in ZnDDP-containing mineral oil; however, they remain very thick even at high load and high temperatures.
The film was found to include compounds containing, Fe, O, C, F, and N. Below this layer was a mixture of iron oxide and oxyfluoride, and the base layer comprised iron fluoride and metallic iron. Of the compounds formed, iron fluoride is claimed to be the most important for determining the performance of the film. The iron fluoride only occurs in the sliding region, indicating that the films do not form without sliding. Based on the surface and subsurface composition of the X-1P lubricant surfaces, a model of the composition of the boundary film has been proposed [10]. Figure 1 shows the model of X-1P boundary tribofilm, in which the major constituents are indicated according to detailed X-ray photoelectron spectroscopy analytical data. The thickness and refractive index of the boundary films were monitored in situ with an ellipsometer.
Previous detailed studies of the tribological behaviour of thin organic films include PFPE [11], 5P4E [12], and hexa (3-trifluoromethylphenoxy) cyclotriphosphazene, and hexa(3-methylphenoxy)cyclotriphosphazene [13] lubricants. All these studies were performed under low-load and low-speed conditions. For high-temperature lubricants, surface stability and thermal stability are very important. In severe conditions it is not possible to maintain hydrodynamic lubrication and consequently the surfaces are only separated by a boundary film formed from a chemical reaction between the metal surfaces and the lubricant, as illustrated by the X-1P model. Figure 1. Note that surface organic-fluorine bonds occur because of partially reacted X-1P or residual lubricant not being removed during the hexane rinsing process. Surface fluorine-metal bonding was observed on the thinner films. Sputter depth profiling indicated subsurface fluorine-metal bonding. It was also found that the -CF3/C ratio increased with boundary-film thickness as measured by ellipsometry. This is consistent with a reaction of the X-1P lubricant with some of the -CF3, groups in the X-1P lubricant molecule.

SELECTED PHOSPHAZENE-TYPE COMPOUNDS AS ADVANCED TOPICAL LUBRICANT ADDITIVES AND/OR LUBRICANTS

General information

Mechanical interactions between the head and the medium are minimised by lubrication of the magnetic medium. Magnetic storage devices used for information (audio, video, and data processing) storage and retrieval are tapes, floppy disks, and hard disk drives. Magnetic media fall into two categories [14]: particulate media, where magnetic particles are dispersed in a polymeric matrix and coated on to a polymer substrate for flexible media (tapes and floppy disks), or on to a rigid substrate (typically aluminium, and more recently glass) for rigid disks; and thin-film media, where continuous films of magnetic materials are deposited on to the substrate by vacuum techniques.
During asperity contacts, disk debris can be generated by adhesive, abrasive, or impact wear. The primary function of the lubricant is to reduce the wear of the magnetic medium and to ensure that friction remains low throughout the operational life of the drive. An acceptable lubricant must exhibit properties, such as chemical inertness, high thermal, oxidative and hydrolytic stability, low volatility, shear stability, and good affinity for the magnetic medium surface. PFPE lubricants, which are the most stable lubricants, are used for topical lubrication of hard disks. The need for improved performance and more stable lubricants has led to research on new additives and lubricants. In recent years, research activities have been focused on selected phosphazene-type compounds which are considered as advanced additives and lubricants for head/disk interface lubrication [6, 15 -24].

X-1P additive and/or lubricant

General characteristics The X-1P product is a mixture of p-fluorophenoxy- and m-trifluoromethyl-phenoxy-substituted cyclic phosphazenes. The phosphazene ring in X-1P may exist in a puckered or planar form [25,26]. In the planar form, the phosphazene ring is sterically shielded by the six phenoxy groups, and it is unlikely that strong adhesive interaction occurs between partially fluorinated phenyl rings and the carbon coating. In the puckered form, a basal plane is defined by the nitrogen atoms and three phenoxy groups in the equatorial positions, and the three phenoxy groups in the axial positions are directed upward. A strong interaction is then possible between the exposed phosphazene ring and the carbon coating. Selected properties of X-1P and Z-DOL are presented in Table 1. Advanced analytical techniques have been used to investigate the degradation mechanism and oxidative stability of X-1P cyclophosphazene lubricant [27].

Trihological performance A cyclic phosphazene lubricant, X-1P, is considered as one of the most advanced lubricants for both rigid and flexible thinfilm magnetic media. It is a very special non-polymer compound, which can be used either as an additive for regular PFPE lubricant or just as a superb lubricant. Early information on the tribological properties of X-1P was published in 1992 [9]. The first papers considering the unique tribological performance relevant to the disk-drive industry were published in 1994 and 1995 [15-16]. These papers generated significant interest in the application of this special non-polymeric chemical as a lubricant for advanced recording media. It was shown that X-1P by itself performs well on hard disks during CSS and stiction tests with a lubricant thickness of about 0.5 nm [16]. Most of the early work testing X-1P involved its use as a pure lubricant coated by dipping or draining, resulting in a thickness of less than 1 nm. This work was published for both hot/wet environments [15] and ambient conditions [16]. Laptop computing had increased the importance of lubricant performance in adverse environments, especially high humidity (80% RH) and somewhat elevated temperature (30°C). Contact start-stop (CSS) test results with X-1P show it to have a tribological performance similar to or better than Z-DOL, as a topical lubricant on carbon coated thin-film rigid disks under ambient conditions; however, X-1P significantly outperforms Z-DOL in CSS testing under hot/wet (30°C, 80% RH) conditions [15]. Additionally, X-1P was found to be stable when heated in the presence of either water or a standard slider material Al2O3-TiC.
Figure 2 Durability, in revolutions, from the coefficients of kinetic friction presented elsewhere [22]
  • (disk speed = 0.75 m/s, normal load = 35 mN)
  • Z-DOL, h = 1.0 nm
  • X-1P, h= 1.0 nm
    1. 10wt.% X-1P in Z-DOL, h = 1 nm
    2. 5 wt.% X-1P in Z-DOL, h = 1.7 nm
    3. 5 wt.% X-1P in Z-DOL, h = 3.5 nm
    4. 5 wt.% X-1P in Z-DOL, h = 1.5 nm

    • Research emphasises that X-1P is advantageous for use in pseudo-contact recording due to protection of the head [19]. When working with very thin layers the lubricant serves as a boundary layer. It was also stressed that X-1P serves to protect the head from the catalytic decomposition of PFPE-type lubricants.

    In 1996, the application of the X-1P compound as an additive to standard PFPE lubricants was disclosed [18]. According to that invention, in order to reduce the static and dynamic coefficients of a disk surface, the disk should be coated with a lubricant layer composed of the X-1P compound and a fluoropolyether oil such as Z-DOL. A lubricant composition containing a PFPE oil and X-1P compound, as well as a method for improving the lubricating properties of a PFPE oil, was also disclosed [18].
    A recent study compared the tribological performance of X-1P with Z-DOL using drag, contact start-stop (CSS), and ball-on-flat tests [22]. Drag tests were performed at 80% RH with Z-DOL, X-1P, and X-1P as an additive in Z-DOL.
    Figure 3 Ball-on-flat sliding results of Al2O3-TiC wafer against sapphire ball for Z-DOL and X-1P at 45 and 80% RH [22]

    It was found that an X-1P film of 1 nm exhibits a low coefficient of kinetic friction ranging from an initial value of 0.18 to 0.36 at 2 x 105 disk revolutions. The coefficient of friction for Z-DOL, at the same thickness, is initially similar to X-1P but then increases rapidly to reach 0.7 at 7 x 10³ disk revolutions. The coefficient of friction of Z-DOL with X-1P as an additive is initially at the same value; however, it increases rapidly after 104 revolutions to reach 0.55 after 2 x 104 revolutions. Figure 2 shows that X-1P has the highest durability, followed by Z-DOL with 10 wt.% X-1P additive.

    Ball-on-flat sliding results are presented in Figure 3. These test results and others [22] show that X-1P exhibits better performance than Z-DOL. The differences are particularly significant at high humidity for an Al2O3-TiC surface sliding against a lubricated disk. Z-DOL shows poor performance and may be degraded through interaction with water molecules, whereas X-1P is not.

    Action mechanism The experimental findings described above have been accounted for by a new mechanism approach with the concept of hydrogen bonding interaction and triboemission [6]. The new mechanism was specially developed for the Al2O3-TiC surface sliding against a disk with Z-DOL lubrication at high environmental humidity. The mechanism, a speculative one, also takes into account the role of X-lP lubricant in tribological performance.
      Figure 4 Diagram of water film interaction on
    1. Al2O3-TiC and
    2. Al2O3-TiC/DLC surfaces [6]
    • Dangling bond
    — Covalent bond
    « Hydrogent bond interaction

      This mechanism is as follows:
    1. at the interphase, there exist hydrogen atoms with partial positive charge;
    2. hydrogen bonding interactions at the sliding interphase results in high friction which depletes the lubricant at some sites;
    3. low-energy electrons are emitted from the sites with solid-solid asperity contact, including C-O bond scission through the interaction of low-energy electrons with lubricated molecules.

      The electron-induced degradation mechanism is based on the mechanism of lubricant component molecules by low-energy electrons [28]. The principal thesis of the model is that the lubricant components, for example, alcohols [29], form anions, which are then chemisorbed on the positively charged areas of rubbing surfaces. The general model of the negative ion-radical action mechanism of lubricating oil components assumes creation of two types of activated sites on the friction surfaces, that is, thermally activated sites and sites activated by the low-energy electron emission process (exoelectrons).
        Figure 5 Diagram of
      1. Al2O3-TiC and
      2. Al2O3-Ti-C/DLC and
        (a:C-Hx) disk/Z-DOL interaction via hydrogen bonding [6]

      It is assumed that the exoelectron energy is sufficient to cause the ionisation of lubricant molecules.

      Figure 4(a) shows the interaction of water molecules with the Al2O3-TiC surface. Hydroxyl groups are chemisorbed from the environment on to the A12O3-TiC surface. In humid environmental conditions, an absorbed water film interacts with the chemisorbed hydroxyl groups through partially negatively charged oxygen atoms and partially positively charged hydrogen atoms. The higher the humidity, the more interacted water molecules are present.
      Figure 4(b) demonstrates that in the case of the A12O3-TiC surface with a DLC coating, groups other than hydroxyl groups (e.g., ester and carboxylic groups) are the most chemisorbed surface species. The water molecules interact by hydrogen bonding with these functional groups, forming a more strongly absorbed film. The water film is additionally strengthened by the interaction of water molecules with dangling bonds in the carbon coating.
      The hydroxyl and oxygen groups favour water adsorption so they belong to the hydrophilic category. Z-DOL has hydroxyl groups at both ends while DEMNUM-SA has only one hydroxyl end-group. Additionally, the backbone of Z-DOL has a higher frequency of oxygen linkage in the ether group (number of oxygen atoms per CF2 groups) than DEMNUM-SA. Accordingly, it is expected that Z-DOL will interact more with water molecules. The results of static friction, kinetic friction, and durability tests have confirmed that DEMNUM-SA is relatively insensitive to humid environments, compared with Z-DOL [30]. On the other hand, X- IP has no hydrophilic end groups, so there is no interaction between its end groups and water molecules.
      A diagram of head/disk interaction through hydrogen bonding is shown in Figure 5(a). The head is made of Al2O3-TiC, and the disk is coated with amorphous hydrogenated carbon and lubricated with Z-DOL. On the disk surface there are some functional groups (Figure 6) that also interact with the Z-DOL molecules. The strength of the head/disk interaction is controlled by the water film thickness on the Al2O3-TiC surface: the higher the humidity, the thicker is the water film. Some micro-water-droplets can penetrate the lubricant film, displacing it from the surface. The upward side of Z-DOL film, particularly at the -CF2—CH2—OH end, will interact with the Al2O3-TiC layer.
      Figure 5(b) shows that for an Al2O3-TiC head with carbon coating, the water molecules interact strongly with the functional groups of the carbon coating and dangling bonds (see Figure 6) of the head surface;
      Figure 6 Simplified diagram of the chemical structure of hydrogenated carbon overcoat surface [5]

        Figure 7 Schematic of
      1. Al2O3-TiC and
      2. Al2O3-TiC/DLC and
        (a: C—Hx) disk/X-lP interaction via hydrogen bonding [6]

      as a consequence the hydrogen bonding interaction between the head and disk is significantly reduced compared with Figure 5(a). Less hydrogen bonding prevents high friction at the interface. The lubricant film covers the surfaces and an increase in the electron emission during sliding through solid-solid asperity contact is alleviated. Lack of low-energy electrons postpones the degradation of Z-DOL lubricant molecules.

      Figure 7(a) shows that when X-1P is applied to the disk surface instead of Z-DOL, there is no hydrogen bonding between the two surfaces.
      Figure 8 Coefficient of static friction as a function of CSS cycles for disks without DLC coating, lubricated with Z-DOL, X-1P, and X-1P as an additive in Z-DOL in CSS tests, measured at 45 and 80% RH [22]

      Hydrophobic —CF3 and —F layers repel the water film. Similarly, in the case of an Al2O3-TiC/DLC head and DLC/X-1P disk system ( Figure 7(b)), water molecules are confined to the head slider because no hydrogen bonding occurs at the interface. The role of carbon in confining water molecules is not so critical to tribological performance as for Z-DOL. This accounts for similar tribological performance obtained from CSS tests between an Al2O3-TiC slider without and with the DLC coating at high humidity [22]. CSS test results obtained at 45% and 80% RH, using Z-DOL, X-1P, and X-1P as an additive in Z-DOL, and an Al2O3-TiC slider without and with DLC coating, are depicted in Figures 8 and 9. The mechanism is not substantiated yet, and should be considered speculative. However, as discussed in detail elsewhere [6], it explains the degradation process of Z-DOL, and the lesser degradation of X-1P.
      Figure 9 Coefficient of static friction as a function of CSS cycles for disks with DLC coating, lubricated with Z-DOL, X-1P, and X-1P as an additive in Z-DOL in CSS tests, measured at 45 and 80% RH [22]

      X-100 and XML-86 phosphazene type additives

      General information It has been demonstrated above that X-1P shows a lower stiction value and a smaller increase in the coefficient of stiction as a function of CSS cycles than the typical PFPE lubricant Z-DOL. It was also shown that X-1P used as an additive in Z-DOL reduces stiction and increases the stability of Z-DOL, particularly at higher environmental humidity. However, one possible concern with X-1P as an additive in Z-DOL is phase separation, because X-1P is immiscible with Z-DOL. Phase separation of phosphazene additives in PFPE lubrication was investigated as a function of additive concentration and lubricant layer thickness [24]. Two additives, X-1P and X-100, were tested. The amount of phase separation was found to be higher for X-1P than X-100. It was concluded, however, that phase separation can be reduced and possibly avoided if additives and disk properties are chosen
      Figure 10 Chemistry of X-100
      (a) and XML-86
      (b) additive compounds [23]

      judiciously. This is related to previous results [31] and attributed to the interaction between phosphazenes and the active carbon surface. It was shown that a complex is formed from this interaction, which is stable if the carbon is active when coated; near edge X-ray absorption fine structure analysis was used to investigate the presence of unsaturated carbon bonds at interphases lubricated with Z-DOL and Z-DOL/additives mixtures [24].

      Chemistry of X-100 and XML-86 phosphazenes The chemical structures of the X-100 and XML-86 compounds are shown in Figure 10. These are newly developed phosphazene-type lubricants with low vapour pressure and high thermal stability [23]. They have some structural similarities to Z-DOL and therefore better solubility in Z-DOL than X-1P. Comparing Table 1 and Figure 10, it can be seen that X-100 and XML-86 have a phosphazene ring structure that is similar to X-1P, the difference is in the chemistry of the phosphazene ring substituents. X-100 has fluoropolymer substituted side chains [H(CF2)4—CH2—O—] while XML-86 has a branch that is similar to the functional end group of Z-DOL [(HO—CH2—CH2—O— ) group in XML-86 and (HO—CH2—CF2—O—) group in Z–DOL]. These compounds are more soluble in Z-DOL than is the X-1P additive: about 2 wt.% solubility for X-100 and 0.05 wt.% for X-1P in Z-DOL.

        
    (a) Z-DOL / X-100
    (b) Z-DOL / XML-86
    (c) Z-DOL / X-1P
    (d) Z-DOL
    Figure 11 Coefficient of friction during start-up vs. time (stiction test with a 12 h period following a CSS test) [23]

         Tribological performance of X-IOO and XML-86 phosphazene additives Experimental investigations have been carried out on the tribological behaviour of these additives in Z-DOL, a commonly used PFPE-type hard-disk lubricant [23]. CSS tests, stiction tests, and constant-speed drag tests were performed to evaluate the stiction and friction behaviour of the two new additives compared with X-1P additive performance in Z-DOL: the lubricant/additive weight ratio investigated was 95:5. The disks were lubricated by dip coating and the thickness of the lubricant layer was about 2 nm for all disks, measured with FTIR spectroscopy. Test results showed that the Z-DOL/X-100 mixture had better tribological performance than Z-DOL. Disks lubricated with Z-DOL/X-100 have a lower coefficient of stiction and friction than disks lubricated with Z-DOL alone, the variations with Z-DOL/X-100 are generally small and the increase in stiction is only a weak function of CSS cycles. Both X-100 and X-1P are effective as additives, in enhancing the tribological performance of Z-DOL.
    The results from rest stiction tests with a 12 h dwell period after 10 000 CSS cycles are shown in Figure 11. It can be seen that disks lubricated with Z-DOL/X-100 have the lowest rest stiction (0.6), Figure 11(a), while disks lubricated with Z-DOL/XML-86, Figure 11(b), have the highest rest stiction (2.9). The major conclusion from this work is that the XML-86 compound is not a promising additive; on the other hand, both X-100 and X-1P phosphazene-type compounds are effective additives for enhancing stiction performance of the head/disk interface.

    Additive-lubricant interaction with the carbon coating The interaction mechanism of the X-1P additive in Z-DOL lubricant with disk and slider surfaces has been investigated [20]. It was suggested that the X-1P molecules would lie underneath the Z-DOL molecules and next to the carbon coating. To understand the interaction mechanism better, Figure 12 compares the contact angles of similar lubricants and water on different types of solid surfaces. It can be seen that lubricant X-1P and water exhibit higher contact angles than Z-DOL and X-1P as an additive in Z-DOL. These contact angle measurements
    Figure 12 Contact angles of lubricants and water on various solid surfaces [22]
    suggest that the thickness of lubricant X- IP on the disk surface may be lower than that of Z-DOL [22]. The thinner film minimises lubricant agglomeration due to surface tension.

    The presence of substituents in X-100 and XML-86 additives that are not in X-1P may influence the additive lubricant interaction with the carbon coating. This seems to be particularly important for XML-86, which causes a substantial increase in stiction over Z-DOL (see Figure 11). Although XML-86 is similar in chemistry to X-100 or X-1P, its viscosity is substantially higher; therefore it seems likely that the increased viscosity of XML-86 causes the overall increase in friction and stiction [23]. On the other hand, it is suggested that X-100/Z-DOL lubricant interaction with the carbon coating is similar to that for X-1P/Z-DOL.

    Degradation mechanism of X-1P lubricant

    In accordance with the PFPE lubricant catalytic degradation mechanism [21], X-1P efficacy has usually been attributed to the passivating action of the catalytically active slider material (Al2O3-TiC). This idea might be important if the catalytic degradation mechanism were predominant; but this is not the case. The most recent review of the PFPE lubricant degradation mechanism [5] suggests that catalytic degradation is not relevant because the kinetics are very slow at asperity temperatures. Based on a wide variety of reviewed experimental data it is reasonable to emphasise that anionic intermediates produced by low-energy electrons play an important part in both the electron-mediated degradation process of PFPE lubricants and the chemical bonding of PFPE lubricant films with DLC surfaces under sliding conditions. Therefore, the X-1P mechanism question seems to be open, and more detailed research is needed to understand this issue better.
    Much less information exists on the degradation mechanism of X-1P. It has been reported [27] that degraded X-1P revealed oligomerisation to be the dominant mode of degradation. Analytical results confirmed the existence of a cyclotetramer structure. Although X-1P has much longer durability than Z-DOL, it has also been shown [12] that after sliding in a high vacuum environment for 55.8 km the lubricant degrades. Fragments such as CFO, CF3, C6H4FO and C6H5CF2 are produced, corresponding to a sharp increase in the friction coefficient.

    CONCLUSIONS

    This paper has reviewed and discussed selected literature concerning micro- and nanotribology of magnetic recording media, and particularly the boundary lubrication process. Specific standard PFPE lubricants along with advanced phosphazene-type fluorinated additives and/or extremely effective lubricants are considered in detail. Emphasis is put on the X-1P phosphazene compound used alone, and as an additive for Z-DOL, deposited on protective DLC coatings. Speculative action and degradation mechanisms related to the X-1P additive lubricant are discussed, mostly focusing on a new suggested mechanism of PFPE lubricant degradation in the presence of the X-1P additive. This mechanism has been specially developed for Al2O3-TiC surface sliding against the disk at high environmental humidity, with Z-DOL containing the X-1P phosphazene compounds. More research is required to understand the complexity of this X-1P lubricant/additive action/degradation mechanism better. The same is true for the newest phosphazene-type additive, X-100.

    Acknowledgements

    The author wishes to express his sincere thanks to Professor Bharat Bhushan and Dr. Zheming Zhao from the Computer Microtribology and Contamination Laboratory, Ohio State University, Ohio, USA, for their help, fruitful discussions, and suggestions while writing this paper.

    References

    1. Bhushan, B., 'Micro/nanotribology and its applications to magnetic storage devices and MEMS', Tribol. Int., 28 (1995) 85-96.
    2. Bhushan, B., Micro/Nanotribology and Its Applications, Kluwer, Dordrecht, 1997.
    3. Bhushan, B., Micro/nanotribology using atomic force/friction force microscopy. State of the art', Proc. IMechE, Part J: J. Eng. Tribol., 212 (1998) 1-18.
    4. Bhushan, B., 'Chemical, mechanical and tribological characterization of ultra-thin and hard amorphous carbon coatings as thin as 3.5 nm: Recent developments', Diamond Rel. Mater., 8 (1999) 1985-2015.
    5. Kajdas, C., and Bhushan, B., 'Mechanism of interaction and degradation of perfluoropoly-ethers with DLC coating in thin film magnetic rigid disks - A critical review', J. Info. Storage Syst., 1 (1999) 303-20.
    6. Zhao, Z., Bhushan, B., and Kajdas, C., 'Tribological performance of PFPE and X-1P lubricants at head-disk interface. Part II. Mechanisms', Trib. Lett., 6 (1999) 141-8.
    7. Singlcr, R.E., and Bieberich, M.J., 'Phosphazenes', Synthetic Lubricants and Hig/i-Performance Functional Fluids, ed. R. Shubkin, Marcel Dekker, New York, 1993, pp. 215-28.
    8. Singler, R.E., Deome, A.J., and Dunn, D.A., 'Preparation and properties of phosphazene fire-resistant fluids', Ind. Eng. Chem. Prod. Res. Dev., 25 (1986) 46-57.
    9. Nader, B.S., Kar, K.K., Morgan, T.A., Pawloski, C.E., and Dilling, W.I., 'Development and tribological properties of new cyclotriphosphazene high temperature lubricants for aircraft gas lurbinc engines', Trib. Trans., 35 (1992) 37-44.
    10. Choa, S.H., Ludema, K., Potter, G.E., DeKoven, B.M., Morgan, T.A., and Kar, K.K., 'A model for the boundary film formation and tribological behavior of a phosphazene lubricant on steel', Trib. Trans., 38 (1995) 757-68.
    11. DeKoven, B.M., and Meyers, G.F., 'Friction studies in ultrahigh vacuum of Fe surfaces with thin films from exposure to perfluorodiethylether', J. Vac. Sci. Technol., A9 (1991) 2570-7.
    12. DeKoven, B.M., and Milchell, G.E., 'HREELS, XPS and in-situ friction studies of thin polyphenyl ether films on steel surfaces', Appl. Surf. Sci., 52 (1991) 215-26.
    13. DeKoven, B.M., and Mitchell, G.E., 'Chemical and tribological behavior of thin film organic lubricant films on surfaces'. Application of Surface Science and Advances in Tribology, ACS Symp. Ser., eds. Y.W. Chung, A.M. Homola, and G.B. Street, 47 (1992) 15-42.
    14. Bhushan, B., 'Magnetic recording surfaces', Characterisation of Tribological Materials, Butterworth-Heineman, Boston, 1993, ch. 7.
    15. Yang, M., Talke, F.E., Perettie, D.J., Morgan, T.A., and Kar, K.K., 'Environmental effects on phosphazene lubricated hard disks', IEEE Trans. Magn., 30 (1994) 4143-5.
    16. Yang, M., Talke, F.E., Perettie, D.J., Morgan, T.A., Kar, K.K., DeKoven, B.M., and Potter, G.E., 'Cyclophosphazenes as potential lubricants for thin-film hard disks', Trib. Trans., 38 (1995) 636-44.
    17. Perettie, D.J., Morgan, T.A., and Kar, K.K., 'X-1P as a dual purpose lubricant for pseudocontact recording', IDEMA Insight, 9, 3 (1996) 1-3.
    18. Chao, J.L., Lee, Y.C., and Chao, C., Lubricant composition and method, US patent 5 587 217 (1996).
    19. Perettie, D.J., Johnson, W.D., Morgan, T.A., Kar, K.K., Potter, G.E., DeKoven, B.M., Putzig, C.L., Chao, J., Lee, Y.C., Gao, C., and Russak, M., 'Cyclic phosphazenes as advanced lubricants for thin-film magnetic media'. Adv. Info. Stor. Syst., 7 (1996) 157-63.
    20. Mate, C.M., Kasai, P.M., Tyndall, G.W., Lee, C.H., Raman, V., Pocker, D.J., and Waltman, R.J., 'Investigation of phosphazene additive for magnetic recording lubrication', IEEE Trans. Magn., 34 (1998) 1744-6.
    21. Kasai, P.H., 'Degradation of perfluoropolyethers and role of X-1P additives in disk files', ./. Info. Stor. Syst., 1 (1998) 23-31.
    22. Zhao, Z., and Bhushan, B., 'Tribological performance of PFPE and X-1P lubricants at head-disk interface. Part I. Experimental', Trib. Lett., 6 (1999) 129-39.
    23. Zhao, Z., Kang, H.J., Fu, L., Talke, F.E., Perettie, D.J., and Morgan, T.A., 'Tribological study of phosphazene-type additives in perfluoropolyether lubricant for hard disk applications', Lubr. Eng., 55, 3 (1999) 16-21.
    24. Kang, M.J., Zhao, Q., Talke, F.E., Perettie, D.J., DeKoven, B.M., and Morgan, T.A., 'The use of cyclic phosphazene additives to enhance the performance of the head-disk interface', Lubr. Eng., 55, 3 (1999) 22-7.
    25. Allcock, H.R., Phosphorus-Nitrogen Compounds, Academic Press, New York, 1972.
    26. Allinger, N.L., 'Calculation of molecular structure and energy by force-field methods', Adv. Phys. Org. Chem., 13 (1976) 1-82.
    27. Keller, M.A., and Saba, C.S., 'Oxidative stability and degradation mechanism of a cyclotriphosphazene lubricant', Anal. Chem., 68 (1996) 3489-92.
    28. Kajdas, C., 'Importance of ionic reactive intermediates for lubricant component reactions with friction surfaces', Lubrication Science, 6, 3 (1994) 203-28.
    29. Kajdas, C., 'About an anionic-radical concept of the lubrication mechanism of alcohols', Wear, 116 (1987) 167-80.
    30. Bhushan, B., and Zhao, Z., 'Friction/stiction and wear studies of magnetic thin-film disks with two polar perfluoropolyether lubricants', IEEE Trans. Magn., 33 (1997) 918-25.
    31. Perettie, D.J., Morgan, T.A., Kar, K.K., DeKoven, B.M., Putzig, C.L., and Kastl, P.E., 'Phosphazene lubricants as a stabiliser for the head-disk interface', MIPE, MR-34 (1997).
    32. Zhao, X., and Bhushan, B., 'Lubrication studies of head-disk interfaces in a controlled environment: Part 1 - Effects of disk texture, lubricant thermal treatment and lubricant additive on durability of head-disk interface', Proc. IMechE, Part J (in press).

    This paper was first presented at the 12th International Colloquium on Tribology, Technische Akademie Esslingen, Germany.