WO1998010115A1

WO1998010115A1 - Silicon-doped diamond-like carbon coatings for magnetic transducers and for magnetic recording media

Info

Publication number: WO1998010115A1
Application number: PCT/US1997/007222
Authority: WO
Inventors: David W. Brown; Bradley J. Knapp; Fred M. Kimock
Original assignee: Monsanto Company
Priority date: 1996-09-03
Filing date: 1997-04-30
Publication date: 1998-03-12

Abstract

A method is provided in which after completion of a finishing operation to define its shape, the surface of the magnetic transducer or magnetic recording media substrate is chemically cleaned to remove unwanted materials and other contaminants. In the second step, the substrate is inserted into a vacuum chamber (1), and the air in said chamber is evacuated. In the third step, the substrate surface is sputter-etched with energetic ions to assist in the removal of residual contaminants, i.e. hydrocarbons and surface oxides, and to activate the surface. Following the completion of the sputter-etch, a Si-DLC layer is deposited by ion beam deposition. Once the chosen thickness of the Si-DLC layer has been achieved, the deposition process on the substrates is terminated, the vacuum chamber (1) pressure is increased to atmospheric pressure, and the Si-DLC-coated substrates are removed from the vacuum chamber.

Description

SILICON-DOPED DIAMOND-LIKE CARBON COATINGS FOR MAGNETIC TRANSDUCERS AND FOR MAGNETIC RECORDING MEDIA

This application is a continuation -in -part of application Serial No. 08/607,657 (attorney docket 6051/53102) filed February 27, 1996, which is a continuation of application Serial No. 08/516,522 (attorney docket 6051/53012) filed August 17, 1995, which is a continuation of application Serial No. 08/205,701 (attorney docket 6051/52635) filed March 3, 1994.

Field of the Invention This invention relates to transducer assemblies and media utilized in magnetic recording systems. More particularly, the invention relates to magnetic recording media and transducers for use with magnetic recording media, e.g. thin film magnetic heads, magnetoresistive (MR) read heads, inductive heads, sliders, and tape heads which are coated with a thin, protective layer of silicon-doped diamond-like carbon (Si-DLC), and a process for deposition of the Si-DLC layer. Background of the Invention

The demand for increased storage capacity of magnetic disk drives continues to increase. Increased storage capacity demands increased areal density (number of bits per square inch) on the disk. Increased storage density can be achieved by various interdependent developments including: (i) higher sensitivity magnetic heads, such as magnetoresistive (MR) and giant magnetoresistive (GMR) heads, that can read smaller bits on the media; (ii) disks with higher magnetization and coercivity, so more magnetic flux from a smaller bit is available for sensing by the head; (iii) improved electronics for signal processing such as partial response, maximum likelihood (PRML); (iv) advances in bearings that reduce the nonrepeatabie runout of the spinning disk such that track width and track spacing can be reduced; and (v) slider and disk designs that allow operation at reduced magnetic spacing, i.e. reduced distance between the head and the magnetic layers of the disk.

Magnetic transducers which include thin film magnetic heads, magnetoresistive read heads, inductive heads, sliders and tape heads utilized in magnetic recording systems have been known in the art for many years. The susceptibility of all magnetic heads (especially sliders) to damage during operation is well known. The slider supports a thin film magnetic read/write head, which is formed by depositing layers of magnetic material, electrically conductive material, and electrically insulating material to form the magnetic poles, magnetic shields and magnetoresistive stripes which are the magnetic elements necessary for the transducing functions (reading and writing) with the magnetic coating on a magnetic recording medium. After lapping the magnetic head to a predetermined dimensions, a pattern of rails is produced on the lapped surface to form an air bearing surface which is used to "fly" the magnetic head over the magnetic recording medium. During operation, a magnetic head slider typically flies with its air bearing surface less than a few microinches (1 microinch = 254 A) above the magnetic recording medium. The air bearing surface of the slider contacts the magnetic disk during start-up and shutdown of the disk rotation, and sometimes during operation. In conventional sliders and recording media, this contact results in transfer of disk material (wear debris) to the slider, which degrades the aerodynamics of the slider and increases friction. The presence of the wear debris and the increased friction can result in catastrophic failure and loss of stored information.

In addition to mechanical wear, magnetic transducers such as thin film magnetic heads and magnetoresistive heads are fabricated from materials that are attacked by atmospheric constituents, such as moisture. Prolonged exposure of the transducer materials to atmosphere often results in degradation of performance due to oxidation and corrosion of the head materials.

Tape heads used with magnetic recording tape also suffer failure due to wear and corrosion. In this case, the magnetic elements are worn away by abrasive materials in magnetic tape. The magnetic materials on the tape head transducer are also degraded by environmental corrosion.

For many designs of magnetic transducers, it is critical for the protective coating to be electrically nonconducting. This is especially true for magnetoresistive designs of tape heads or sliders. In these magnetoresistive designs, there are magnetic shields in the heads which shield the magnetoresistive sensor from the other elements of the head. Those shields can be either grounded or at potential, but always at a different potential than the sensor. It is not allowable to provide a path for electrical conduction between the shields to the sensor. Additionally, it is not allowable to provide an electrically conductive path between the sensor to any other element in head which is maintained at a different electrical potential. Finally, there are situations in which it is deleterious to allow electrical charge to pass directly between the transducer and the recording medium when the transducer and recording medium come into direct contact. An electrically nonconducting protective coating layer can protect against this problem.

Currently, amorphous diamond-like carbon (DLC) coatings are used as protective coatings on magnetic data storage recording media (i.e. hard disks) and magnetic transducers (including sliders) to increase the wear resistance and durability of the head-disk interface. In this application, the utility of DLC coatings is due to their excellent mechanical properties such as high hardness and low coefficient of friction, and excellent resistance to abrasion, and resistance to corrosion by water, acids, bases, and solvents.

DLC films are so-named because their properties resemble, but do not duplicate, those of diamond. Some of these properties are high hardness of about 1,000 kg/mm² (10 GPa) to about 5,000 kg/mm² (50 GPa), low -friction coefficient (approximately 0.1), and transparency across the majority of the electromagnetic spectrum. At least some of the carbon atoms in DLC are bonded in chemical structures similar to that of diamond, but without long range crystal order. Although the term DLC was initially intended to define a pure carbon material, the term DLC is now used to include amorphous, hard carbon materials containing up to 50 atomic percent of hydrogen. Other names for these hydrogen-containing DLC materials are "amorphous hydrogenated carbon", hydrogenated diamond-like carbon, or diamond-like hydrocarbon. The structure of these hydrogen-containing DLC materials may be best described as a random covalent network of graphitic-type structures interconnected by sp³ linkages, although the definitive structure of the films has yet to be universally accepted. In keeping with the majority of the previous literature and art, the term DLC is used herein to refer to both the amorphous non-hydrogenated hard carbon materials, and the amorphous hydrogenated hard carbon materials. In some cases the DLC coatings may also contain nitrogen and are thus termed "nitrogenated DLC".

Many methods for preparation of DLC films are known in the prior art, including (i) direct ion beam deposition, dual ion beam deposition, glow discharge, radio frequency (RF) plasma, direct current (DC) plasma or microwave plasma deposition from a carbon-containing gas or vapor which can also be mixed with hydrogen and/or inert gas, (ii) electron beam evaporation, ion-assisted evaporation, magnetron sputtering, ion beam sputtering, or ion-assisted sputter deposition from a solid carbon target material, or (iii) combinations of (i) and (ii). It has been found that DLC coatings will impart improved corrosion resistance, wear resistance and durability to the head-disk interface only if the adherence of the DLC coating to the parent substrate is excellent. The most obvious and common approach to coating a substrate is to apply the DLC coating directly onto a clean surface which is free of residue. However, this approach often results in a DLC coating which displays inadequate adhesion, and therefore, poor wear resistance. Magnetic transducer assemblies are made up of a variety of materials, including the alumina-titanium carbide (Al₂O₃-TiC), alumina (Al₂O₃), nickel-iron (NiFe) alloys, and iron-silicon-aluminum (FeSiAl) alloys. Any protective coating should have sufficient adhesion to all of these materials. DLC coatings are typically under significant compressive stress, on the order of 5 x 10⁹ dynes/cm² (0.5 GPa) to about 5 x 10'° dynes/cm² (5 GPa). This stress greatly affects the ability of the coating to remain adherent to the substrate. Additionally, the surface of the substrate to be coated often contains alkali metals, oxides, and other contaminants which can inhibit bonding of the DLC coating. Even with an atomically clean surface the adhesion may be insufficient between DLC and particular substrate materials. Therefore, less obvious methods are required to produce a substrate with a highly adherent DLC coating which provides excellent abrasion resistance.

Protective DLC overcoats for thin film metal alloy disks are well known. A review of the field was presented by H. Tsai and D. Bogy in "Characterization of Diamondlike Carbon Films and Their Application as Overcoats on Thin film Media for Magnetic Recording", J. Vac. Sci. Tech. l (1987) 3287-3312. Illustrative from the prior art are the following references.

Aine, U.S. Pat. No. Re 32,464 teaches a magnetic recording medium coated with a sputter-deposited graphitic carbon protective layer having thickness between 1-5 microinches. Use of an adhesion-promoting carbide-forming layer between the disk and the carbon film is disclosed and claimed. Additionally, "the magnetic transducer head portion which occasionally sinks into contact with the recording medium is preferably formed of or coated with carbon, preferably in the form of graphite, to provide a low friction wear resistant contacting surface with the recording medium". The carbon layer may also be deposited by ion plating.

Michihide et al., EP 216 079 Al , teach a method for manufacturing a thin carbon film on the surface of a sliding member such as a hard disk of a magnetic recording device which contacts another member, in which the carbon thin film is formed by sputter deposition from a glassy carbon sputtering target. Howard, U.S. Pat. No. 4,778,582, describes a prior art sputter deposition technique for manufacture of a thin film magnetic disk coated with amorphous hydrogenated carbon. In this method, amorphous hydrogenated carbon is deposited by sputtering a carbon target in an atmosphere of argon and hydrogen. The deposition rate is approximately 6-7 A/minute. A 100 A thick titanium layer is used as an adhesion layer.

Meyerson et al. , U.S. Pat. No. 4,647,494, disclose an improved wear-resistant coating for metallic magnetic recording layers, where the improved coating is a non-graphitic hard carbon layer strongly bound to the underlying metallic magnetic recording layer by an intermediate layer of silicon, having a thickness less than 500 A. The silicon layer can be as thin as a few atomic layers. The hard carbon layer has a thickness in the range of about 25 A to 1 micron. The preferred method for depositing both the silicon intermediate layer and the hard carbon layer is claimed to be plasma deposition.

Japanese Laid Open Pat. Application (Kokai) No. 1-287819, Shinora, claims a magnetic recording medium (magnetic disk or magnetic tape) in which a diamond-form hard carbon thin film is located on a strongly magnetic metal thin film with a silicon or germanium layer, having thickness of 20 A to 50 A interposed between the two. The diamond-like carbon film has a thickness between 50 A and 100 A and can be formed by high frequency sputtering, ion beam deposition, or plasma acceleration. The DLC layer is then overcoated with a lubricant to form the final product.

Endo et al., U.S. Pat. No. 4,774,130, discloses a magnetic recording medium composed of a magnetic film formed on the surface of a disk-shaped substrate and a protective film further formed on the surface of the magnetic film. The protective film is composed of a first layer containing silicon, germanium, or chromium oxide, and a second (top) layer of amorphous carbon or graphite-containing amorphous carbon. The first and second layers can be formed by a variety of sputtering techniques including magnetron sputtering, diode sputtering, and ion beam sputtering. The thickness of a first layer of elemental silicon is in the range of 100 A to 300 A, and the second carbon layer has a thickness in the range of 200 A to 700 A.

Kurokawa et al. , U.S. Pat. No. 4,717,622, discloses a magnetic recording medium with a protective layer of high hardness carbon synthesized under a low temperature and low pressure gas plasma. The magnetic recording medium is useful in a system where the magnetic head contacts the magnetic recording medium. The diamond-like carbon film has a Vicker's hardness of more than 2,000 kg/mm² (20GPa), and a specific resistance of 10⁷ ohm-cm to 10¹³ ohm-cm.

Nakamura et al. , U.S. Pat. No. 4,804,590, teach an abrasion resistant magnetic recording member comprising a carbonaceous surface protective film on a surface of a magnetic film on the surface of a nonmagnetic substrate. The protective film has a lower layer of comparatively hard carbonaceous film and an upper layer of comparatively soft carbonaceous film. An intermediate layer of chromium, titanium, etc. may by used to improve the adhesion of the carbonaceous film to the magnetic film. The lower carbonaceous layer contains 5 atomic percent or less of hydrogen, fluorine, or a combination of hydrogen and fluorine, and the upper carbonaceous layer contains 6 atomic percent or more or hydrogen, fluorine, or a combination of hydrogen and fluorine. The lower carbonaceous layer may be a sputtered carbonaceous film, and the upper layer may be a plasma chemical vapor deposited (PCVD) carbonaceous film.

Additionally, protective DLC coatings on magnetic head sliders have been discussed in the prior art. Illustrative are the following references. IBM Technical Disclosure Bulletin, Vol. 25, No. 7A (1982) page 3173 discusses a magnetic slider made of silicon which is coated with an extremely hard surface layer of silicon carbide or DLC. The sufficient thickness of the silicon carbide or diamond-like carbon surface layer is in the range of 50 nm to 100 nm. The DLC layer can be produced by ion beam or plasma deposition. Bleich et al. , U.S. Pat. No. 5, 151 ,294, discloses a method for depositing a thin protective carbon film on the air bearing surface of a slider in a magnetic recording disk file by contacting the slider and a rotating magnetic disk for a time sufficient to cause transfer of the carbon from a carbon overcoated disk to the air bearing surface of the slider. The patentees teach that the carbon film formed on the slider is an essentially amorphous hydrogenated carbon film approximately 50 A in thickness.

Japanese Laid Open Pat. Application (Kokai) No. 3-25716, discloses a magnetic head slider, characterized by the fact the head slider body is made of a soft ceramic material, such as ferrite, and a hard ceramic membrane such as carbon, silicon, zirconium dioxide, aluminum oxide, or the like, which is bonded and molded onto at least the pressure-receiving surface of the slider which faces the magnetic disk. This disclosure is said to offer advantages in manufacturing of the slider, because softer ceramic substrate materials are used in processing. The hard ceramic membrane has a thickness in range of several hundred Angstroms, and may be made by a plasma chemical vapor deposition (PCVD) method.

Head et al., U.S. Pat. No. 4, 130,847, describes a magnetic head slider having a protective coating preferably chromium over at least the magnetic head. The coating is produced in a recess within the slider body to a thickness as small as 10 microinches. Grill et al., U.S. Pat. No. 5, 159,508, teaches a magnetic head slider coated with an adhesion layer and a protective DLC layer. The coating is fabricated onto the substrate after a lapping operation, but before patterning of the rails onto the slider, which protects the magnetic head during the fabrication process. The slider has at least two rails on the air bearing surface, and the rails have a protective coating comprising an adhesion layer, typically about 10 A to 50 A in thickness, and a thin layer of amorphous hydrogenated carbon, less than about 250 A thick. Grill et al. teach that the two layers of the protective coating can be deposited by any suitable technique, e.g. PACVD, ion beam or laser techniques. The preferred technique is by the use of a DC biased substrate in an RF plasma deposition apparatus. Chang et al., U.S. Pat. No. 5, 175,658, claim a magnetic slider similar to that disclosed in Grill, et al , U.S. Pat. No. 5, 159,508, discussed above, but with an additional thin masking layer on top of the amorphous hydrogenated carbon layer. DC magnetron sputtering and RF magnetron sputtering are presented as preferred methods for depositing the amorphous hydrogenated carbon film. Japanese Laid Open Pat. Application JP 58-150, 122 describes a magnetic head which has a thin film of a material, e.g. carbon, having a lubricating effect on the surface of the head which faces the magnetic recording medium. The thickness of the film is disclosed to be in within the range of 200 A to 5000 A.

German Pat. Application No. DE 3,714,787 describes a storage system in which the surface of a magnetic disk is coated with friction-reducing carbon and the rails of the magnetic head slider are coated with a friction reducing carbon. The thickness of the carbon is 10 A to 1000 A.

Published Pat. Application PCT/US88/00438 discloses a magnetic head slider having a magnetic head which is built within one of the side rails. A wear layer is provided over the slider comprising a 50 A thick chromium layer and a 200 A thick carbon layer. Either component of the wear layer can be omitted.

The above prior art methods for application of DLC coatings on magnetic media transducers such as thin film heads and tape heads all suffer from one or more of the following deficiencies and shortcomings: (i) difficulty in precleaning of substrates prior to deposition; (ii) adhesion of the DLC coating; (iii) permeation of the amorphous carbon films by water vapor and oxygen; (iv) fabrication of coherent, dense coatings which perform well at thicknesses less than 200 A; (v) poor electrical resistivity of the coating due to the use of low resistivity Ti, Cr, or Si adhesion layers between the transducer and the DLC coating, or a "graphitic", low resistivity DLC layer; (vi) control of DLC coating properties during a deposition run and batch-to-batch variation of DLC coating characteristics; (vii) DLC coating thickness control and reproducibility of thickness; (viii) part-to-part and batch-to-batch control of DLC coating uniformity; (ix) production readiness and ability to scale-up the deposition process for mass production; and (x) difficulty in coating transducer assembly substrates of complex geometry or configuration. In copending application Serial No. 0/607,657, incorporated herein as a reference, Knapp et al. teach a direct ion beam process for deposition of DLC protective coatings on magnetic transducers such as sliders and tape heads, which provides remarkable performance compared to the prior art techniques discussed above. In their method, a silicon-containing adhesion layer consisting of amorphous silicon, silicon carbide, silicon nitride, silicon oxide, silicon oxy-nitride is deposited onto the magnetic transducer substrate prior to ion beam deposition of the DLC top coating layer.

As stated previously, increased storage density can be achieved by a reduction in the distance (i.e. magnetic spacing) between the head and the magnetic layers on the recording disk. One way to reduce the magnetic spacing between the head and the disk is to reduce the thickness of the protective coating on the disk and the transducer assembly containing the head. The ion beam DLC deposition process of copending application Serial No. 08/607,657 has been used to deposit protective DLC coatings at total thicknesses (thickness of interlayer plus thickness of DLC layer) of less than 65 A. However, to achieve the next generation of storage densities, it is estimated that the total thickness of the protective coating on heads must be reduced to less than 50 A, and possibly to as low as 20 A.

To achieve the required performance with coating thicknesses as low as 20 A, coatings and processes which are different from those of the prior art are required. In addition, it would be highly advantageous if a new coating was sufficiently adherent that it did not require an adhesion layer. Even if the wear performance of such a new material was equivalent to that of ion beam DLC, but it required no adhesion layer, the reduction in total thickness of the protective coating (by elimination of the adhesion layer), would be very beneficial in achieving substantial improvements in storage density. In addition to minimizing total coating thickness, there are other reasons to eliminate the adhesion layer. In certain applications it is acceptable for the DLC layer to wear away in some regions of the slider. However, upon complete wear removal of the DLC, the disk is exposed to direct contact with the adhesion layer material. Sliding contact of the disk with some typical adhesion layer materials, such as silicon, can lead to premature failure of the head-disk interface. This is to say that the exposed adhesion layer can generate more wear debris or have higher friction than even an uncoated slider. A new, low friction coating material that required no adhesion layer would solve this problem. Another reason to eliminate the adhesion layer is to reduce the number of manufacturing steps, thus reducing any yield loss and manufacturing complexity associated with deposition of the adhesion layer. A third reason involves quality control of the coated substrates. It is easier to analyze and assign cause to variation in a single layer coating process than in a two layer process, i.e. there would be no question about whether a manufacturing problem was associated with the adhesion layer or the DLC layer. From the above discussion it is clear that a protective coating for magnetic media and transducers is needed that exhibits friction and wear performance which is equivalent or superior to that of ion beam DLC but requires no adhesion-enhancing interlayer.

Summary of the Invention The invention provides a process for depositing a protective coating onto magnetic media and magnetic transducers to impart superior wear resistance, and improved lifetime. More particularly, this invention provides an ion beam deposited Si-doped DLC (Si-DLC) coating to the surface of magnetic transducers and magnetic media which is highly adherent without use of an adhesion enhancing interlayer, and exhibits greatly improved wear resistance and environmental durability. This invention also provides a low cost and efficient process for mass-producing the coated magnetic transducers and magnetic recording media with improved wear resistance and superior lifetime.

In the present invention, the wear surface of a magnetic recording media substrate or a magnetic transducer is provided with a single coating layer of Si-DLC that requires no adhesion-enhancing interlayer. The invention further provides a method for fabricating the protective Si-DLC coating on the surface of magnetic transducers and magnetic recording media.

In the method of the present invention, after completion of a finishing operation to define its shape, the surface of the magnetic transducer or recording media substrate is chemically cleaned to remove unwanted materials and other contaminants. In the second step, the substrate is inserted into a vacuum chamber, and the air in said chamber is evacuated. In the third step, the substrate surface is sputter-etched with energetic ions to assist in the removal of residual contaminants, i.e. hydrocarbons and surface oxides, and to activate the surface. Following completion of the of the sputter-etch, a Si-DLC layer is deposited by ion beam deposition. Once the chosen thickness of the Si-DLC layer has been achieved, the deposition process on the substrates is terminated, the vacuum chamber pressure is increased to atmospheric pressure, and the Si-DLC-coated substrates are removed from the vacuum chamber. Alternatively, the Si-DLC layer may be ion beam deposited onto the wear surface of a magnetic recording media substrate immediately upon completion of deposition of the magnetic material layer in vacuum.

The Si-DLC coatings of the present invention are characterized by having the following features: a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa, Raman spectral G-peak position in the range of about 1425 cm^"' to about 1530 cm^"1, and a silicon concentration in the range of about 1 atomic % to about 45 atomic % . Ion beam deposited Si-DLC materials of the present invention which contain hydrogen typically exhibit a hydrogen concentration in the range of about 25 atomic % to about 47 atomic %. Brief Description of the Drawings Further features and advantages will become apparent from the following and more particular description of the preferred embodiment of the invention, as illustrated by the accompanying drawings, in which like reference characters generally refer to the same parts or elements, and in which:

FIG. 1 is an illustration of the ion beam deposition apparatus used to manufacture Si-DLC coatings on magnetic transducers and magnetic recording media (e.g. hard disks) in accordance with the present invention;

FIG. 2A is a graph of frictional force versus the number of test revolutions in a constant speed drag test for an uncoated Al₃O₃-TiC slider;

FIG. 2B is a graph of touch down velocity change versus the number of test revolutions in a constant speed drag test for an uncoated Al₂O₃-TiC slider;

FIG. 2C is a graph of frictional force versus the number of test revolutions in a contact-start-stop test for an uncoated Al₂O₃-TiC slider; FIG. 2D is a graph of touch down velocity change versus the number of test revolutions in a contact-start-stop test for an uncoated Al₂O₃-TiC slider;

FIG. 3 A is a graph of frictional force versus the number of test revolutions in a constant speed drag test for a slider coated with a 25 A thick ion beam sputter-deposited Si adhesion layer, and a 100 A thick top layer of ion beam deposited DLC;

FIG. 3B is a graph of touch down velocity change versus the number of test revolutions in a constant speed drag test for a slider coated with a 25 A thick ion beam sputter-deposited Si adhesion layer, and a 100 A thick top layer of ion beam deposited DLC; FIG. 4A is a graph of frictional force versus the number of test revolutions in a constant speed drag test for a slider coated with a 25 A thick ion beam sputter-deposited Si adhesion layer, and a 50 A thick top layer of ion beam deposited DLC;

FIG. 4B is a graph of touch down velocity change versus the number of test revolutions in a constant speed drag test for a slider coated with a 25 A thick ion beam sputter-deposited Si adhesion layer, and a 50 A thick top layer of ion beam deposited DLC;

FIG. 5A is a graph of frictional force versus the number of test revolutions in a contact-start-stop test for a slider coated with a 25 A thick ion beam sputter-deposited Si adhesion layer, and a 100 A thick top layer of ion beam deposited DLC

FIG. 5B is a graph of touch down velocity change versus the number of test revolutions in a contact-start-stop test for a slider coated with a 25 A thick ion beam sputter-deposited Si adhesion layer, and a 100 A thick top layer of ion beam deposited DLC;

FIG. 6 is a representative laser Raman spectrum and curve fit analysis of an ion beam deposited DLC coating deposited using a feed gas of 12 seem methane and an ion beam energy of 150 Volts;

FIG. 7 is a laser Raman spectrum and curve fit analysis of a Si-DLC coating of the present invention deposited using a gas feed mixture of 3.6 seem of tetramethylsilane and 5 seem of argon, at a beam voltage of 350 Volts;

FIG. 8 is a graph of hardness and compressive stress versus ion beam energy ("beam voltage") for representative Si-DLC coatings of the present invention deposited using a gas feed mixture of 3.6 seem of tetramethylsilane and 10 seem of argon;

FIG. 9 is a graph of laser Raman G-peak position versus ion beam voltage for Si-DLC coatings of the present invention deposited using a feed gas mixture of 20 seem of tetramethylsilane and 10 seem of argon;

FIG. 10 is a graph of hardness and compressive stress versus the flow rate of tetramethylsilane added to 12 seem of methane in the precursor feed gas mixture for representative Si-DLC coatings of the present invention deposited at an ion beam energy of 200 Volts;

FIG. 11 is a graph of the atomic concentration of carbon, silicon, hydrogen and argon in Si-DLC coatings of the present invention versus tetramethylsilane flow rate added to 12 seem of methane in the precursor gas feed, for coatings deposited at an ion beam energy of 200 Volts; FIG. 12 is a graph of laser Raman G-peak position versus tetramethylsilane flow rate added to 12 seem of methane in the precursor gas feed, for Si-DLC coatings of the present invention deposited at an ion beam energy of 200 Volts;

FIG. 13 is a graph of hardness and compressive stress versus substrate angle for DLC coatings deposited at 150 Volts ion beam energy and a precursor gas flow rate of 12 seem methane;

FIG. 14 is a graph of hardness and compressive stress versus substrate angle for representative Si-DLC coatings of the present invention deposited at 350 Volts ion beam energy and a precursor gas mixture of 3.6 seem tetramethylsilane and 5 seem argon.

Detailed Description of the Invention The method of the present invention substantially reduces or eliminates the disadvantages and shortcomings associated with the prior art transducer assemblies for magnetic recording media (e.g. sliders, inductive heads, magnetoresistive head, tape head, thin film head, and similar devices, etc.) by providing:

(1) an improved transducer assembly for use with magnetic recording media, in which the transducer has an improved protective coating;

(2) for the manufacture of a transducer assembly for use with magnetic recording media, in which the transducer has an improved protective coating; (3) for the deposition of an amorphous Si-DLC coating onto the surface of a transducer assembly, which surface faces the magnetic recording media, in which the amorphous Si-DLC coating has the property of high adhesion to the substrate without the need for an adhesion layer;

(4) for the deposition of an amorphous Si-DLC coating onto the surface of a transducer assembly, which surface faces the magnetic recording media, in which the amorphous Si-DLC coating has the properties of impermeability to environmental elements such as water vapor and oxygen, high density, and extreme surface smoothness;

(5) for the deposition of a thin amorphous Si-doped DLC coating at layer thicknesses as small as 50 A or less onto the surface of a transducer assembly, which surface faces the magnetic recording media, in which the thin amoφhous Si-DLC coating provides a protective surface for the transducer;

(6) for the deposition of a protective amoφhous Si-DLC coating onto the surface of a transducer assembly, which surface faces the magnetic recording media, in which the layer thickness and uniformity of the amoφhous Si-DLC coating are reproducibly controlled to a high degree of accuracy; and

(7) for the deposition of an amoφhous Si-DLC coating onto the surface of a transducer assembly, which surface faces the magnetic recording media, in which the protective amoφhous Si-DLC coating is manufactured over large areas with high throughput.

In addition, the method of the present invention substantially reduces or eliminates the disadvantages and shortcomings associated with the prior art magnetic data recording substrates (e.g. hard disks) by providing:

(8) an improved magnetic recording substrate which has an improved protective coating;

(9) for the manufacture of a magnetic recording media substrate with an improved protective coating;

(10) for the deposition of an amoφhous Si-DLC coating onto the surface of a magnetic recording media substrate, which surface faces the magnetic transducer, in which the amoφhous Si-DLC coating has the property of high adhesion to the substrate without the need for an adhesion layer;

(11) for the deposition of an amoφhous Si-DLC coating onto the surface of a magnetic recording media substrate, which surface faces the magnetic transducer, in which the amoφhous Si-DLC coating has the properties of impermeability to environmental elements such as water vapor and oxygen, high density, and extreme surface smoothness;

(12) for the deposition of a thin amoφhous Si-doped DLC coating at layer thicknesses as small as 50 A or less onto the surface of a magnetic recording media substrate, which surface faces the magnetic transducer, in which the thin amoφhous Si-DLC coating provides a protective surface for the magnetic media; (13) for the deposition of a protective amoφhous Si-DLC coating onto the surface of a magnetic recording media substrate, which surface faces the magnetic transducer, in which the layer thickness and uniformity of the amoφhous Si-DLC coating are reproducibly controlled to a high degree of accuracy; and (14) for the deposition of an amoφhous Si-DLC coating onto the surface of a magnetic recording media substrate, which surface faces the magnetic transducer, in which the protective amoφhous Si-DLC coating is manufactured over large areas with high throughput.

The ion beam deposited Si-DLC coating of the present invention protects the magnetic transducer assembly and magnetic recording media from wear and corrosion damage during normal operation and significantly extends the lifetime of the magnetic recording data storage system. The ion beam deposited Si-DLC coating eliminates the requirement for an adhesion layer, allowing for a substantial reduction of the total coating thickness of the protective coating, and for simplified manufacture. Additionally, the method for manufacture of the ion beam deposited Si-DLC coating substantially reduces or eliminates the disadvantages and shortcomings of prior art DLC coating processes. It is not intended by the discussion of a particular transducer assembly or magnetic recording medium to limit the method of the present invention to any particular type of transducer, e.g. sliders, or tape heads, or magnetic recording medium, e.g. hard disks.

It has been unexpectedly found that ion beam deposition of Si-DLC coatings on magnetic transducers provided remarkable performance improvement of magnetic transducers and magnetic transducer assemblies such as sliders. The remarkable performance is the result of the combination of the critical features and unexpected attributes listed below.

(1) The ion beam deposited Si-DLC coatings unexpectedly have tribological properties that are comparable to, or superior to standard DLC coatings as measured in typical mechanical tests used to evaluate protective coatings on magnetic heads.

(2) The ion beam deposited Si-DLC has unexpectedly good adhesion to the sliders without the need for an adhesion layer. As a result, the Si-DLC, without an adhesion layer enables: reduced total thickness of the protective coating, elimination of the manufacturing process steps and quality control associated with adhesion layer deposition, and elimination of adhesion layer materials that can cause premature failure when exposed during use of the slider.

(3) The ion beam process overcomes the difficulties in obtaining an atomically clean surface prior to deposition by sputter-etching the substrates using an ion beam of controlled shape, current, and energy. The ion beam "shape" is controlled by focusing the beam with electrostatic or magnetic fields. In this way, the beam can be efficiently delivered to the substrates to be processed, with maximum utilization. Control of ion beam current and beam energy to within 1 % is routinely achieved which results in a highly repeatable and predictable rate of removal of surface residual hydrocarbons and other contaminant layers. In addition, the ion beam sputter-etching process is conducted in high vacuum conditions, such that oxidation or contamination of the transducer surface with residual gases in the coating system is negligible. Finally, the apparatus geometry can be easily configured such that the sputtered contaminants deposit on the vacuum chamber walls, and do not redeposit onto the surface of the part as it is being sputter-etched.

(4) The ion beam deposition process produces a protective Si-DLC layer directly on the atomically clean ion etched substrate which adheres without use of an adhesion layer. The Si-DLC layer is preferably deposited immediately upon completion of the ion beam sputter-etching step to achieve maximum density and adhesion to the substrate. Deposition of the Si-DLC layer immediately upon completion of the ion beam sputter-etching step minimizes the possibility for recontamination of the sputter-etched surface with vacuum chamber residual gases or other contaminants.

(5) The ion beam deposition process produces highly dense Si-DLC coatings. This makes these Si-DLC coatings excellent barriers to water vapor and oxygen. The excellent barrier properties of the thin ion beam deposited Si-DLC coatings presumably result from the extremely high degree of ion bombardment during film growth. In addition, the Si-DLC coatings of the present invention are also exceedingly smooth, resulting in a surface with high resistance to wear. (6) Ion beam deposition produces coherent, dense ion beam Si-DLC coatings having thickness of 50 A or less and providing magnetic transducers and sliders with excellent wear protection. This result is also presumably due to the much higher degree of ion bombardment during film growth, which occurs in ion beam deposition as compared to other prior art deposition methods for DLC materials. Extremely thin protective layers are critically important for the newest technology of ultrahigh density magnetic recording media, in which the required distance between the magnetically active surface of the magnetic transducer material and the top surface of the recording medium must be as low as 100 A or less. The newest technology of direct contact tape heads is the most stringent example where this dimension is minimized.

(7) The ion beam deposition process produces highly electrically nonconducting Si-DLC coating layers. Use of these electrically nonconducting Si-DLC layers provides improved performance of magnetoresistive sliders, magnetoresistive tape heads, and other transducers, compared to prior art methods.

(8) The ion beam deposition process produces a coating with properties that do not change as a function of layer thickness, as is found in prior art RF plasma deposition processes. This attribute is achieved because the coating deposition step is preferably conducted with a charge neutralized ion beam. The use of charge neutralized ion beam deposition process also allows for coating of parts with complex geometry without interference to the process. Parts of varying geometry can be coated within a single coating run with no adverse effect on the deposition conditions. Complete slider assemblies, tape heads, other transducer assemblies, and magnetic recording disks can be easily coated. In addition, on substrates which contain electrically conducting and electrically insulating materials, all portions can be coated with the same high quality Si-DLC coating. In the case of the plasma deposition methods, Si-DLC coatings of different properties may be deposited on different locations of the same substrate, depending upon whether the area being coated is an electrical conductor or an electrical insulator, and the electrical connections between the substrates and the vacuum chamber. The lack of substrate geometry constraints of the present invention is in shaφ contrast to the plasma deposition methods of the prior art.

(9) The ion beam process allows easy fixturing of magnetic transducer assemblies and magnetic recording disks. Because of the ease of fixturing transducer substrates of nearly any shape or configuration, the ion beam process of the present invention can be used to apply a Si-DLC coating to a transducer during any part of the magnetic transducer fabrication or assembly process. For example, the Si-DLC coating can be applied (i) during fabrication of the transducer element, (ii) after fabrication of the transducer element, but before completion of the final transducer assembly, or (iii) after completion of the final transducer assembly.

(10) The ion beam process is capable of minimal batch-to-batch variation in the properties of the Si-DLC coatings. This is the case because process parameters such as ion energy, ion current density, gas flow rate, and deposition chamber pressure are largely decoupled in the ion beam deposition method of the present invention, and because each of these process parameters can be accurately controlled and reproduced to a high degree of certainty, often to within 1 % . In addition, the process endpoint for Si-DLC coating thickness is easily defined and reproduced.

(11) The ion beam deposition process is capable of tight part-to-part thickness uniformity, e.g. a variation of less than 2% can be easily achieved. This is the case because of the compatibility of the method of the present invention with commercially available substrate holders incoφorating motion, i.e. rotation and/or planetary motion.

(12) The ion beam process is readily scaled-up to accommodate mass production because large scale ion beam sources are commercially available. For example, commercially available 38 cm Kaufman-type gridded ion beam sources have been used to deposit DLC coatings simultaneously over four 18-inch diameter platens with a thickness variation across all parts of less than +/- 2% . Similar ion beam sources can be used to practice the Si-DLC process of the present invention. Plasma deposition systems for deposition of Si-DLC coatings on magnetic transducers and media are not presently commercially available on this scale. The inventors have discovered that the ion beam deposited Si-DLC coatings of the present invention provide magnetic transducers and recording media with an outstanding wear and corrosion resistant surface. This is a suφrising discovery, in view of the known problems encountered due to the presence of silicon at the head-disk interface. As previously stated, it was found that for DLC coatings deposited with a silicon adhesion layer, upon complete wear removal of the DLC, the disk is exposed to direct contact with the adhesion layer material. Sliding contact of the disk against the silicon often leads to a premature failure of the head-disk interface, such as a head crash. In this case, the exposed adhesion layer can generate more wear debris or have higher friction than even an uncoated slider. The Si-DLC coatings of the present invention form an excellent, highly wear resistant head-disk interface. Apparently, this is achieved by the bonding of the silicon in the DLC carbon matrix, which in essence encapsulates the silicon and eliminates detrimental effects of silicon at the head-disk interface. The inventors believe that the DLC matrix is capable of encapsulating the silicon, provided that the ratio of silicon-to-carbon atoms in the coating is less than 1.0, and preferably less than about 0.9. At a silicon-to-carbon ratio of 1.0, the coating material could have bonding characteristics of silicon carbide, or may begin to develop free, unbonded silicon, which would be detrimental to the head-disk interface. The apparatus for carrying out the preferred embodiment form of the invention is illustrated schematically in FIG.1. The coating process is carried out inside a high vacuum chamber 1 which is fabricated according to techniques known in the art. Vacuum chamber 1 is evacuated into the high vacuum region by first pumping with a rough vacuum pump (not shown) and then by a high vacuum pump 2. Pump 2 can be a diffusion pump, turbomolecular pump, cryogenic pump ("cryopump"), or other high vacuum pumps known in the art. The use of cryopumps with carbon adsorbents is somewhat less advantageous than other high vacuum pumps because those cryopumps have a low capacity for hydrogen which is generated by the ion beam sources used in the present invention for the deposition of Si-DLC. The low capacity for hydrogen results in the need to frequently regenerate the adsorbent in the cryopumps. It is understood that the process of the present invention can be carried out in a batch-type vacuum deposition system, in which the main vacuum chamber is evacuated and vented to atmosphere after processing each batch of parts; a load-locked deposition system, in which the main vacuum deposition chamber is maintained under vacuum at all times, but batches of parts to be coated are shuttled in and out of the deposition zone through vacuum-to-air load locks; or in-line processing vacuum deposition chambers in which parts are flowed constantly from atmosphere, through differential pumping zones, into the deposition chamber, back through differentia] pumping zones, and returned to atmospheric pressure. Magnetic transducer substrates or magnetic recording media substrates to be coated are mounted on substrate holder 3, which may incoφorate tilt, simple rotation, planetary motion, or combinations thereof. The substrate holder can be in the vertical or horizontal orientation, or at any angle in between. Vertical orientation is preferred to minimize paniculate contamination of the substrates, but if special precautions such as low turbulence vacuum pumping and careful chamber maintenance are practiced, the substrates can be mounted in the horizontal position and held in place by gravity. This horizontal mounting is advantageous from the point of view of easy fixturing of small substrates such as individual sliders. This horizontal geometry can be most easily visualized by rotating the illustration in FIG.1 by 90 degrees.

Prior to deposition, the transducer substrates are ion beam sputter-etched with an energetic ion beam generated in ion beam source 4. Ion beam source 4 can be any ion source known in the prior art, including Kaufman-type direct current discharge ion sources, radio frequency or microwave frequency plasma discharge ion sources, each having one, two, or three grids, or gridless ion sources such as the End Hall ion source of U.S. Pat. No. 4,862,032. The ion source beam is charge neutralized by introduction of electrons into the beam using a neutralizer (not shown), which may be a thermionic filament, plasma bridge, hollow cathode, or other types known in the prior art. In a preferred embodiment of the present invention, ion source 4 is provided with inlets for introduction of inert gases 5, such as argon, krypton, and xenon, for the sputter-etching, and for introduction of precursor gas mixtures 6, for deposition of Si-DLC layers. The precursor gas mixture is made up of silicon-containing compounds including, but not limited to silane compounds such as silane and disilane, and organosilane compounds such as methylsilane, dimethylsilane, trimethylsilane, tetramethylsilane (TMS) and diethylsilane (DES) which may be mixed with hydrocarbon compounds including, but not limited to methane, ethane, acetylene, cyclohexane, and mixtures thereof and inert gases such as argon, krypton, xenon and mixtures thereof. The precursor gas mixture may further contain nitrogen (N₂), or organosilane compounds or organic compounds which contain nitrogen. A critical feature is that a silicon-containing precursor gas is introduced into the ion beam source to provide the silicon doping level in the Si-DLC coatings which is required to obtain excellent adhesion and tribological performance of the Si-DLC coatings of the present invention. An additional ion source (not shown) can be used to co-bombard the substrates during Si-doped DLC deposition to alter the film properties.

An additional ion beam source 7, is also provided for ion beam sputter deposition of silicon and/or carbon atoms onto the substrate to alter the stoichiometry (silicon-to-carbon atomic concentration ratio) during deposition of the Si-DLC layer. Ion beam source 7 is provided with inlets for operation on inert gases 8, such as argon, krypton, and xenon, or for reactive gases 9, such as methane or nitrogen. The ion beam from source 7 is directed onto a sputtering target 10, which can be silicon, carbon, a mixture of silicon and carbon, or silicon carbide, depending on the desired stoichiometry of the Si-DLC layer, the operating conditions of ion source 4, and the precursor gases introduced in inlet 6. According to the method of the present invention, after preparation to define the final shape of the magnetic transducer assembly or magnetic recording media substrate, the surface of the substrate is first chemically cleaned to remove contaminants. Ultrasonic cleaning in solvents, or other detergents as known in the art is often effective; details of the cleaning depend upon the nature of the contamination and residue remaining on the part after the finishing operations. It has been found that it is critical for this step to be effective in removing surface contaminants and residues, or the resulting adhesion of the Si-DLC coating will be poor. This is because in many designs of magnetic transducers and recording media, only a very small amount, e.g. < 100 A, of surface material can be removed during the subsequent in-vacuum cleaning by ion beam sputter-etching step. For example, for the case of magnetic transducers, minimizing the sputter-etching step is critical in order to minimize pole-tip recession.

In the second step, the substrate is inserted into a vacuum chamber, and the air in said chamber is evacuated. Typically, the vacuum chamber is evacuated to a pressure of about 1 x 10^"5 Torr or less to ensure removal of water vapor and other contaminants from the vacuum system. However, the required level of vacuum which must be attained prior to initiating the next step must be determined by experimentation. The exact level of vacuum is dependent upon the nature of the substrate material, the sputter-etching rate, and the constituents present in the vacuum chamber residual gas. In the third step, the substrate surface is bombarded with energetic gas ions to assist in the removal of residual contaminants, e.g. any residual hydrocarbons, surface oxides, material smeared metal from lapping, and other contaminants, and to activate the surface. This sputter-etching of the magnetic transducer or magnetic recording media substrate surface is required to achieve high adhesion of the Si-DLC layer. The sputter-etching is typically carried out with inert gases such as argon, krypton, and xenon, but other gases (e.g. nitrogen) can be used if they due not adversely affect adhesion. Additionally, hydrogen may be added to the ion beam during sputter-etching to assist in activation of the surface. Typically, in order to achieve efficient and rapid ion sputter-etching, the ion beam energy is greater than 20 eV. Ion energies as high as 2000 eV can be used, but ion beam energies in the range of about 20 to about 500 eV result in the least amount of atomic scale damage to the magnetic transducer or magnetic recording media substrate.

Once the desired shape of the transducer assembly or magnetic recording media substrate is achieved by lapping or any other process used in fabrication, the shape and surface moφhology of the substrate should not be significantly affected by application of the protective coating. Due to the higher ion beam etch rates of magnetic materials compared to ceramics, a phenomenon known as "pole-tip recession" can occur in magnetic transducers if the ion beam sputter-etching process is operated at the incorrect condition, or for an extended period of time. Minimizing the sputter-etching time to remove < 100 A of material, operating with a beam of heavy inert gas (e.g. xenon) ions, and varying ion beam energy and angle can be used to minimize the effects of differential etching and pole recession.

Immediately following the sputter etch step, the Si-DLC layer is deposited by ion beam deposition. It is important to minimize the time between completion of the etch step and the start of the deposition of the Si-DLC layer. Deposition of the Si-DLC layer immediately after completion of the sputter-etching step minimizes the possibility for recontamination of the substrate surface with vacuum chamber residual gases or other contaminants. The thickness of the protective ion beam deposited Si-DLC coating is constrained to small dimensions since the coating thickness adds directly to the magnetic spacing between the magnetic transducer and the magnetic recording medium. Depending on the design and operation of the transducer, the Si-DLC coating thickness is typically in the range of about 20 A to about 2,000 A. Thicker Si-DLC layers are generally preferable in terms of providing increased protection against wear and corrosion, although outstanding wear and corrosion resistance is also obtained by ion beam deposited Si-DLC coatings at the low end of this thickness range. The actual thickness of the ion beam deposited DLC layer is chosen in practice based on either (i) the maximum allowable increase in magnetic spacing, or (ii) the minimum thickness that will perform without failure for a required period of operation. Several ion beam deposition methods may be used for the formation of the

Si-DLC coatings of the present invention, including: (1) direct ion beam deposition from silicon-containing precursor gas mixtures described above, (2) direct ion beam deposition from silicon containing gas mixtures in combination with a second ion "assist" beam directed at the substrate, i.e. "direct dual ion beam deposition", (3) ion beam sputter deposition from a silicon-doped carbon sputtering target, (4) ion beam sputter deposition with ion assist, i.e. "dual ion beam sputter deposition", (5) filtered cathodic arc ion beam deposition using a silicon-containing carbon cathode, (6) laser ablation ion beam deposition using a silicon-containing carbon ablation source, and (7) combinations of the above.

In direct dual ion beam deposition, the assist beam is typically composed of inert gas ions that are used to sputter unwanted material from the depositing film or to add energy to the surface region of the growing film to influence chemical bonding. The ion beam sputter deposition methods offer excellent control, uniformity, and flexibility of substrate geometry, but the deposition rate is slower than that of the direct ion beam deposition process. In dual ion beam sputter deposition, bombardment of the growing film by an additional ion beam is performed to maximize the film density and improve the electrical resistivity of the ion beam sputter-deposited Si-DLC coating. Ion beam sputter deposition from a carbon or silicon target can also be used in combination with direct ion beam deposition to add carbon or silicon atoms, respectively, to the growing Si-DLC film. In this way the elemental composition of the direct ion beam deposited Si-DLC film may be modified during deposition to achieve the desired stoichiometry. In this process the gas mixture fed to direct deposition ion source may not necessarily include a silicon-containing precursor gas (i.e. may be hydrocarbon gas only) if sufficient silicon is deposited into the growing film from the sputter deposition source. In the filtered cathodic arc and laser ablation ion sources, it is essential to perform some filtering of the beam to remove particles which degrade the coating. It may also be necessary to apply a bias voltage on the substrates to control the energy of the incoming ions.

For sake of process simplicity, rapid deposition, and ease of scale-up to mass production, the preferred deposition process for this invention is direct ion beam deposition from a silicon-containing precursor gas, which may be mixed with an inert gas. The most preferred silicon-containing precursor gas is TMS, but other gases such as silane and DES may be used. The inert gas may be chosen from any of the group VIII gases of the periodic table of the elements, but argon is most preferred due to its availability. Hydrogen and hydrocarbon gases, including but not limited to methane, ethane, and acetylene may also be introduced into the ion source plasma along with the silicon-containing precursor gas to modify the properties of the Si-DLC coating. The ion beam energy used in the Si-DLC deposition process may be in the range of approximately 20 eV to approximately 1000 eV. Ion energies in the range of about 20 eV to about 400 eV are most preferred to minimize heating of the substrate during deposition. In addition to the ion beam for direct deposition, an ion assist beam may be utilized, but is not required.

Once the chosen thickness of the Si-DLC layer has been achieved, the deposition process on the magnetic transducer or magnetic recording media substrates is terminated, the vacuum chamber pressure is increased to atmospheric pressure, and the coated substrates are removed from the vacuum chamber.

Alternatively, the Si-DLC layer may be ion beam deposited onto the wear surface of a magnetic recording media substrate immediately upon completion of deposition of the magnetic material layer in vacuum. The Si-DLC coatings of the present invention, which are direct ion beam deposited from precursor gases, are characterized by the following features: Nanoindentation hardness in the range of about 12 GPa to about 19 GPa, compressive stress in the range of about 0.4 GPa to about 1.8 GPa, Raman spectral G-peak position in the range of about 1463 cm^"' to about 1530 cm ', a silicon concentration in the range of about 1 atomic % to about 30 atomic % . Ion beam deposited Si-DLC materials of the present invention which contain hydrogen typically exhibit a hydrogen concentration in the range of about 25 atomic % to about 47 atomic % .

The Si-DLC coatings which are deposited by other methods, e.g. direct ion beam deposited using ions from solid sources, or by ion beam sputter deposition, are characterized by the following broader range of characteristics: a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa, Raman spectral G-peak position in the range of about 1425 cm^"1 to about 1530 cm^"1 and a silicon concentration in the range of about 1 atomic % to about 45 atomic % . Examples Examples 1-4 illustrate the characteristics of the ion beam deposited DLC coatings of Knapp et al. in copending application Serial No. 08/607,657. Examples 5-12 and the discussion which follows further illustrate the characteristics of the ion beam deposited Si-DLC coatings and superior performance of the ion beam Si-DLC-coated products of the method of this invention. In Examples 5-11 , Si-DLC coatings were deposited by direct ion beam deposition, using a precursor gas containing an organosilicon compound. In Examples 1-4, IBM 3380-type, 100% size magnetic head sliders were coated with ion beam deposited DLC. In Examples 5-9, Seagate Barracuda 4, 50% size thin film inductive nanosliders with NiFe pole pieces and an Al₂O₃-TiC air bearing body were coated with ion beam deposited Si-DLC and DLC. In Examples 5-9, the DLC and Si-DLC coatings were ion beam deposited onto the wear surface of the transducer which faces the magnetic recording medium. In Example 10, CoCrTa magnetic recording disks were coated with ion beam deposited Si-DLC and DLC. In Example 10, Si-DLC coatings were ion beam deposited onto the wear surface of a magnetic recording disk which faces the magnetic transducer. Example 12 demonstrates that adherent, hard ion beam Si-DLC coatings can be manufactured by an ion beam sputter deposition process. The examples are for illustrative puφoses and are not meant to limit the scope of the claims in any way.

Example 1 An Al₂O₃.TiC slider (IBM 3380-type) was coated with ion beam deposited DLC by the following method. The sliders already mounted onto their suspension system were first chemically cleaned with isopropanol and blown dry with nitrogen. The cleaned sliders were then attached to a 6-inch diameter graphite plate using adhesive tape. The graphite plate was then mounted onto a rotary stage, and the vacuum chamber was evacuated to a pressure of 4.8 x 10^'6 Torr. The slider was then sputter-etched using an argon ion beam from an 11 cm Kaufman-type ion source at an energy of 500 eV with a beam current of 137 mA for 2 minutes. The etch rate of a silicon witness coupon under these conditions was approximately 300 A/minute. After this sputter-etching step, a 1000 eV, 100 mA argon ion beam was used to sputter-deposit a 25 A thick layer of silicon by ion beam sputter deposition from a silicon target. After deposition of the silicon layer, a 100 A thick DLC layer was deposited by direct ion beam deposition using the 11 cm Kaufman-type ion source, operated at a beam energy of 75 eV and a beam current of 175 mA. Example 2

An Al₂O₃-TiC slider (IBM 3380-type) was coated with ion beam deposited DLC by the same procedure as in Example 1 , except the vacuum chamber was initially was evacuated to a pressure of 5.0 x 10^"6 Torr, and the thickness of the DLC coating was 50 A. Example 3

A magnetoresistive tape head was chemically cleaned, and mounted in an aluminum fixture. The fixture was installed in a stainless steel vacuum chamber and the chamber was evacuated to a pressure of 3.0 x 10^"6 Torr. The tape head was then sputter-etched using an argon ion beam from an 11 cm Kaufman-type ion source at an energy of 500 eV with a beam current of 137 mA for 15 seconds. The etch rate of a silicon witness coupon under these conditions was approximately 300 A/minute. After this sputter-etching step, a 1500 eV, 50 mA nitrogen ion beam was used to sputter-deposit a 25 A thick layer of silicon nitride by reactive ion beam sputter deposition from a silicon target. After deposition of the silicon nitride layer, a 225 A thick DLC layer was deposited by direct ion beam deposition using the 11 cm Kaufman-type ion source, operated at a beam energy of 75 eV and a beam current of 50 mA.

Example 4 Analog tape heads were cleaned with isopropanol and then blown dry with nitrogen gas. The samples were mounted in an aluminum fixture and the fixture was installed into a stainless steel vacuum chamber which was evacuated to a pressure of 4.6 x 10^"6 Torr. The tape heads were then sputter-etched for one minute using a 500 eV, 137 mA argon ion beam generated by an 11 cm Kaufman type ion source. The etch rate of a silicon witness coupon under these conditions was approximately 300 A/minute. After this sputter-etching step, a 1000 eV, 100 mA argon ion beam was used to sputter-deposit a 200 A thick layer of amoφhous silicon by ion beam sputter deposition from a silicon target. After deposition of the silicon layer, a 1200 A thick DLC layer was deposited onto the tape heads by direct ion beam deposition using the 11 cm Kaufman-type ion source, operated at a beam energy of 75 eV and a beam current of 100 mA. The durability of ion beam deposited DLC coatings of Knapp et al. copending application Serial No. 08/607,657 was demonstrated by applying the coating to Al₂O₃-TiC sliders (IBM 3380-type) used as transducers with magnetic recording disks. The sliders already mounted onto their suspension system (i.e. a head gimble assembly (HGA) in which the slider is mounted onto the gimble) were first chemically cleaned, then sputter-etched by an argon ion beam, followed by ion beam sputter deposition of a 25 A thick layer of silicon, and finally a layer of direct ion beam deposited DLC having thickness in the range of 50 A to 120 A.

The outstanding durability and performance enhancements due to these ion beam deposited DLC coatings were determined by performing two types of accelerated wear tests: a constant speed drag test, and a contact-start-stop (CSS) test. The CSS test is used by the magnetic disk recording industry as the ANSI standard method. A typical mode of failure in the tests is buildup of friction with revolutions of drag or cycles of CSS. For uncoated sliders, the friction starts at a coefficient level of approximately 0.2 and builds up to three or four times that initial value with several thousand cycles. The high levels of friction cause particle pullout with resulting abrasive wear and failure of the head-disk interface. Even if such wear does not occur, the high levels of friction are unacceptable, because the drive motor may not have sufficient torque to start the drive from the rest condition. Therefore, it is desirable to find a combination of protective overcoat and lubricant to achieve a high level of CSS or drag revolutions without unacceptable friction buildup.

Uncoated and ion beam DLC-coated sliders were measured on a standard tester developed by Hewlett Packard Laboratories, which consists of a computer-controlled spindle on which the disk is clamped and the slider-suspension assembly is attached to a strain gauge instrumented load beam to measure normal and frictional forces on the slider. Tests were conducted in a class 100 clean hood in room air at 28°C and 35% relative humidity. Standard commercially available 95 mm diameter thin film disks with a 200-300 A thick layer amoφhous carbon and a 20 A thick layer of lubricant were used in the tests.

The test procedure used was as follows. After the disk had been clamped to the spindle and the slider is loaded on the disk, a single revolution was made at low rpm during which the friction force of the dragging slider was recorded. Then a single CSS was made and the friction was measured during the first revolution to reveal the peak static friction, called the "stiction". An additional indicator of failure is the touch-down velocity (TDV), defined as the speed at which friction reaches a threshold value upon landing of the slider as the spindle revolution speed is reduced from the normal operating speed.

Next, the tester was programmed to perform either slow speed drag tests or CSS tests while taking one-revolution friction data periodically, at 100 cycle intervals. These data are averaged, and the result is plotted as a function of revolutions along with the TDV. An example of drag test data for an uncoated IBM 3380-type Al₂O₃-TiC slider is presented in FIG. 2A, and 2B, and an example of CSS test data for an uncoated IBM 3380-type Al₂O₃-TiC slider is presented in FIG. 2C and 2D.

Note in FIG. 2 A that the frictional force started at 2.07 gmf and increased with the number of test revolutions to a value near 5 gmf at 800 revolutions. Then, it remained nearly constant until increasing to a maximum of 6 gmf at approximately 2600 revolutions. The test was stopped at 3300 revolutions when the TDV showed two consecutive values 50% above the lowest value, a criterion chosen to maximize the accuracy of the determination of contamination on the sliders by wear products. An increase of TDV is an indication that material has transferred from the disk to the slider. Similarly, in FIG. 2C, the frictional force started at 1.93 gmf and increased to 5.2 gmf at approximately 300 start/stops. It reached a value near 7 gmf before the test was stopped after the TDV reached the termination criteria 1200 start/stops. Microscopic examination of the sliders after these tests revealed that there was a significant amount of wear debris adhering to the rails of the sliders. IBM 3380-type Al₂O₃-TiC sliders coated in Example 1 (100 A layer of DLC) and Example 2 (50 A layer of DLC) by the ion beam deposition process of the present invention were evaluated using the same test conditions used for the uncoated sliders. The tests on the DLC-coated sliders were set to terminate at 30,000 test revolutions if not interrupted by the increase of the TDV. FIG. 3A and FIG. 3B show the drag test results for a slider coated with 100 A of DLC in Example 1. The test result is substantially different and improved relative to the uncoated sliders. In FIG. 3A, the frictional force rapidly increased from 1.61 gmf to about 3.2 gmf and remained essentially constant for 30,000 revolutions. In FIG. 3B, the TDV decreased from 3.51 meters/second to 2.8 m/s and remained essentially constant for the duration of the test. After 30,000 cycles was completed, only a small amount of wear material was found on one rail of the slider. The rails were much cleaner than was the case for the tests of the uncoated sliders.

FIG. 4 A and FIG. 4B show the drag test results for a slider coated with 50 A of DLC in Example 2. The test results were nearly identical to those found in FIG. 3A and FIG. 3B for the slider coated with 100 A of DLC in Example 1. Again, there was only a very small amount of debris found on the rails of the slider for this test. The rails were much cleaner than was the case for the tests of the uncoated sliders. FIG. 2C and FIG. 2D show the results of a CSS test for an uncoated Al₂O₃-

TiC slider. It was found that the frictional force increased from an initial value of 1.93 gmf to 5.4 gmf at approximately 400 CSS cycles, then increased more gradually until about 800 cycles, after which it became erratic and the test was terminated by the TDV criterion at 1 ,200 CSS cycles. After termination of this CSS test, there was much more debris found on the rails of this slider than was found in the case of the previous drag tests on uncoated sliders.

FIG. 5A and FIG. 5B show the results of CSS tests on the same slider (100 A DLC coating) used to obtain the data in FIG. 3A and FIG. 3B. The data show that the frictional force increased rapidly from 1.77 gmf to approximately 3.3 gmf and remained essentially constant for 8,000 CSS cycles. The TDV showed only one instance of an increase, but otherwise showed a gradual decrease over the duration of the test.

These test results showed a remarkable difference in performance of the ion beam DLC-coated sliders compared to uncoated sliders in both drag tests and CSS tests. Other coated and uncoated sliders were also tested with similar results. In the drag tests, all of the uncoated sliders showed an increase of the friction to 6 gmf or more within the first 3,000 revolutions. This represented a 3-fold increase from the initial friction value. On the other hand, the tests with the DLC-coated sliders showed a rapid increase to slightly above 3 gmf, but the friction then remained constant for 30,000 revolutions. An increase to 6 gmf would be unacceptable even if the TDV failure criterion had not been met. However, an increase to 3 gmf is tolerable, and the results indicate that the tests could have progressed to significantly higher number of revolutions before a further increase in friction would occur. If doubling of initial friction is set as the failure criterion in the drag tests, the uncoated sliders all failed by 1 ,500 revolutions, while the ion beam DLC-coated sliders did not fail by 30,000 revolutions. Therefore, a factor of improvement of greater than 20 for the ion beam DLC-coated sliders was indicated by the drag tests.

Similarly, in the CSS tests, the uncoated sliders showed an increase of frictional force to 6 gmf by 1 ,000 cycles for three tests which were conducted. For the case of the ion beam DLC-coated slider, the frictional force did not increase above 3.4 gmf at any time up to the point the test was terminated at 8,000 cycles.

Again, if doubling of the initial frictional force from 2 gmf to 4 gmf is set as the failure criterion, all of the uncoated sliders failed by 200 CSS cycles, whereas the ion beam DLC-coated slider did not yet fail by 8,000 cycles. Therefore, a factor of improvement of greater than 40 for the ion beam DLC-coated sliders was indicated by the CSS tests.

The superior performance of the ion beam DLC coated sliders could be related to microscopic observations of debris on the rails of the air bearing surface. The ion beam DLC coated sliders showed significantly less debris as compared to the uncoated sliders. The reduced level of debris on the DLC coated sliders was most likely responsible for the superior performance.

In Examples 5-9, Si-DLC coatings of the present invention were direct ion beam deposited onto a variety of substrate materials utilized in the fabrication of magnetic transducers and magnetic recording media, and onto the wear surfaces of magnetic head sliders for evaluation of mechanical performance and comparison to ion beam deposited DLC coatings. In Example 10, Si-DLC coatings were ion beam deposited onto the wear surface of magnetic recording disks. In Example 11, Si-DLC coatings were deposited onto silicon wafers for characterization of the range of material properties of the ion beam deposited Si-DLC coating materials of the present invention. In Example 12, Si-DLC coatings were deposited by ion beam sputter deposition.

In Examples 5-9 and Example 12, all slider assemblies were chemically cleaned after mechanical lapping to the final geometry to remove lapping compound and other contamination prior to insertion into the vacuum coating chamber. The substrates to be coated were attached onto a 6-inch diameter aluminum mounting plate. The aluminum plate was then mounted onto a rotary stage, and the vacuum chamber was evacuated to a pressure in the range of about 1 x 10^'5Torr to about 1 x 10^"6 Torr, typically about 5 x 10^"6 Torr. Substrates were then sputter-etched using a 500 eV argon ion beam generated from an 11 cm Kaufman-type gridded ion source. Unless otherwise noted, the argon ion beam current density at the substrates was adjusted to give an etch rate of about 30 A/minute to about 70 A/minute on a silicon witness coupon, and the sputter-etching time was approximately 1 minute. The 11 cm ion beam source was also utilized for deposition of the Si-DLC and DLC materials. For the case of the Si-DLC coatings, immediately upon completion of the sputter-etching step, the ion beam energy was reduced, and a silicon-containing precursor gas, tetramethylsilane (TMS), was added to the argon feed gas to initiate the deposition of the Si-DLC layer. In some examples, methane gas was added to the TMS and argon feed gases for the deposition of Si-DLC. The same vacuum chamber conditions and 11 cm ion beam source were used for Examples 10 and 11. For the case of the DLC coatings in some of the examples, immediately upon completion of the sputter-etching step, the ion beam energy was reduced, and methane was added to the ion source plasma as a precursor gas, and the argon flow was shut off, to initiate the deposition of the DLC layer. In some of the DLC coating examples, after completion of the sputter-etching step, a 1000 eV argon ion beam was used to sputter-deposit an adhesion layer of silicon by ion beam sputter deposition from a silicon target prior to deposition of the DLC layer.

Example 5 For the purpose of comparative evaluation, a series of sliders mounted on HGA's were coated with Si-DLC, while some HGA's were coated with DLC using an ion beam sputter-deposited silicon adhesion layer. The undoped DLC coatings were made under deposition process conditions known to produce a DLC coating with a high hardness of about 20 GPa, in order to achieve a DLC coating with maximum resistance to wear. The HGA's coated with DLC and Si-DLC were subjected to nanowear tests and CSS tests. Four coatings were deposited onto the slider wear surface of HGA's for comparative nanowear testing: (1) 50 A Si-DLC in Run #6287; (2) 75 A Si-DLC in Run #6288; (3) 50 A DLC with a 15 A thick Si adhesion layer in Run #6291 ; (4) 75 A DLC with a 15 A thick Si adhesion layer in Run #6292. The process conditions for each of these coating runs are presented in Table 1 below:

lahifi .

Gas Feed Substrate Ion Beam Coating

Run Ar-TMS-CH₄ Angle Energy Thickness

NΩ (sςcjll) (degrees) (Volts) (A ) Coating Description

6287 5-3.6-0 30 350 50 Si-DLC on HGA

6288 5-3.6-0 30 350 75 Si-DLC on HGA 6291 0-0-12 20 150 50 Si interlayer* + DLC on

HGA

6292 0-0-12 20 150 75 Si interlayer* + DLC on

HGA

6331 0-0-12 20 150 50 Si interlayer* + DLC on HGA

6332 0-0-12 20 150 75 Si interlayer* + DLC on HGA

6338 5-3.6-0 30 350 50 Si-DLC on HGA

6339 5-3.6-0 30 350 75 Si-DLC on HGA

Note *: 15 A thick silicon interlayer deposited by ion beam sputter deposition using a 50 mA, 1000 V argon ion beam incident on a silicon sputtering target.

In the nanowear tests, the diamond tip of an atomic force microscope (AFM) contacted the coated surface of the HGA's with a predetermined force, and was then moved across the surface in a prescribed pattern to cause wear of the coating. This nanowear test produced two types of qualitative data which may be use for comparison of coated surfaces: wear resistance, and adhesion (whether or not the coating adhesion failed during the wear test).

The results of the nanowear tests on coated HGA's showed that both the 50 A thick (Run #6287) and 75 A thick Si-DLC (Run #6288) coatings performed as well as the 50 A thick DLC coating with a silicon adhesion layer (Run #6291). The HGA with 75 A of DLC with a silicon adhesion layer (Run #6292) failed prematurely due to loss of adhesion, probably as a result of surface contamination that was not properly removed prior to coating. Except for the sample from Run #6292, all the coatings remained adherent and the extent of the wear was small. Therefore, it was anticipated that all of these coatings would perform well in CSS tests. Under CSS conditions similar to those used to test the sliders coated in Examples 1 and 2, HGA's coated in Runs #6287, #6288, #6291 , and #6292 all passed the required 20,000 cycles without failure. Four additional runs were made to deposit Si-DLC and DLC coatings onto the slider wear surface of identical HGA's for comparative CSS testing: (1) 50 A Si-DLC in Run #6338; (2) 75 A Si-DLC in Run #6339; (3) 50 A DLC with a 15 A thick Si adhesion layer in Run #6331 ; (4) 75 A DLC with a 15 A thick Si adhesion layer in Run #6332. The process conditions for each of these coating runs are presented in Table 1.

Coated HGA's from Runs #6338, #6339, #6331, and #6332 were separated into three equivalent sets (Set 1 , Set 2, Set 3) containing four different HGA's in each set. Then, the coated HGA's were mounted in a CSS test setup similar to that used above. The CSS test results for each set of coated HGA's were as follows. In Set 1 , all four coated HGA's passed 20,000 CSS cycles without any failures. In Set 2, all four coated HGA's passed 40,000 CSS cycles without any failures. In Set 3, utilizing an accelerated CSS test condition, designed to produce failure in a reduced number of CSS cycles, all four coated HGA's passed 20,000 accelerated CSS cycles without any failures.

Therefore, the CSS and nanowear results for the ion beam deposited DLC and Si-DLC coatings on HGA's demonstrated that the Si-doped DLC coatings performed at least as well as undoped DLC protective coatings on magnetic head sliders.

Example 6 An HGA slider was coated with a ^"100 A thick coating of Si-DLC (Run #6186) by direct ion beam deposition using a feed gas mixture of 2.7 seem TMS and 12 seem methane, an ion beam voltage of 200 V, and an angle of incidence of the beam onto the substrates of about 10 degrees (see Table 2 in the discussion which follows). The coating was adherent, and free from delamination on all exposed surfaces of the HGA slider. Another HGA slider was coated with a ^"500 A thick coating of Si-DLC (Run #6179) by direct ion beam deposition using increased deposition time, but otherwise the same process conditions as in Run #6186. The adhesion of this 500 A thick Si-DLC coating was tested by a thermal shock test in which the coated part was alternately subjected to five cycles of boiling water and ice water exposure. Each exposure cycle consisted of immersion in a bath of boiling water for one minute, immediately followed by immersion in a bath of ice water for one minute. After five exposure cycles, the coated surfaces were examined by optical microscopy to determine if there was any evidence of coating delamination or damage to the coating. In previous studies, it was found that this boiling water-to-ice water thermal shock test is a rigorous test for the adhesion of DLC coatings. DLC coatings which did not exhibit excellent adhesion were typically found to delaminate from the substrate as a result of this test. The 500 A thick Si-DLC coating remained completely adherent after the thermal shock test, indicating that the adhesion of the Si-DLC coating to all materials on the HGA slider surfaces was excellent. Rxample 7 For the puφose of comparative adhesion testing, 300 A thick Si-DLC coatings and 300 A thick DLC (without an adhesion layer) coatings were deposited onto surfaces of a variety of materials which are used in the manufacture of magnetic transducer heads and magnetic recording media. The substrate materials investigated included a variety of metal alloys (NiFe, FeSiAl, CoCrTa) and ceramics (Al₂O₃, and Al₂O₃-TiC). Nickel and stainless steel substrates were also evaluated. NiFe is a material used in magnetic pole pieces of a slider; FeSiAl is another material commonly used in magnetoresistive heads. CoCrTa is a common magnetic recording disk material. Al₂O₃ is a common electrical insulator material used around the electrically conductive head material on a magnetic transducer. Al₂O₃-TiC is a common material used in the body and air bearing surfaces of magnetic head sliders.

Si-DLC and DLC coatings were ion beam deposited onto the wear surfaces of sliders on HGA's and onto specially prepared sample coupons. To facilitate evaluation of the adhesion strength of the Si-DLC and DLC coatings onto the magnetic materials, thin films of each magnetic material were deposited onto stainless steel coupons by argon ion beam sputter deposition from a metal alloy target. The following metal target materials were used: pure Ni; Ni with 19 wt. % Fe for the deposition of NiFe; Fe with 10 wt. % Si and 6 wt. % Al for the deposition of FeSiAl; Co with 8 atomic % Cr and 4 atomic % Ta for the deposition of CoCrTa. Highly polished Al₂O₃-TiC and stainless steel coupons were also utilized as substrates.

The adhesion of the Si-DLC and DLC coatings on the magnetic material substrates was evaluated by three methods: (i) optical microscopic inspection of the coating immediately after completion of the coating process, (ii) optical microscopic inspection after completion of a boiling water-to-ice water thermal shock test (conditions as discussed in Example 6), and (iii) a tensile pull adhesion test with a Sebastian II adhesion tester. In the tensile pull adhesion test, an aluminum pull stud of known contact area is first attached to the surface of a coated sample by epoxy. After curing the epoxy to final hardness, the pull stud attached to the specimen is mounted into the Sebastian II tester, and a force is applied to the stud in a direction peφendicular to the specimen surface in an effort to separate the stud from the specimen. The stud is pulled with increasing force until separation occurs. The force at which separation of the stud and specimen occurs is recorded. If the stud separates from the specimen due to disbonding at the interface between the substrate and the coating, then the recorded force is a measurement of the adhesion strength of the coating to the substrate. If separation occurs within the epoxy, within the substrate, within the bulk of the coating, or if the epoxy disbonds from the surface of the coating, then the adhesion strength between the substrate and the coating is greater than the recorded force. Optical microscopy examination of the test specimens after separation is performed to determine the origin of the separation. Previous work with DLC coatings has shown that coatings with excellent adhesion typically exhibit tensile bond adhesion strengths greater than about 5 kpsi (1 kpsi = 1 ,000 psi). Typically, for a well-bonded epoxy, it has been found that for coatings with the best adhesion to the substrate, failure occurs within the epoxy at values in the range of about 8 kpsi to about 12 kpsi. The epoxy bond may periodically fail at lower values.

The deposition conditions for the Si-DLC (Runs #6393, #6422, #6385, #6186, #6179) and DLC (Runs #6454 and #6386) coatings, and the adhesion test results are presented in Table 2. For Run #6422, the argon ion beam sputter-etching process prior to Si-DLC deposition was carried out for at time sufficient to etch 180 A from the surface of a silicon witness coupon. For the coatings summarized in Table 2 below, no adhesion layer was deposited prior to deposition of either the Si-DLC layer or the DLC layer. Table 2

Gas Feed Substrate Ion Beam Coating Run Ar-TMS-CH₄ Angle Energy Thickness Substrate Adhesion Results £L (seem) {degrees) (Volts) (A) Material" Y ^h ιs^c Pull"

6393 5-3.6-0 30 350 300 Ni P P

10.8

6422 5-3.6-0 30 350 300 NiFe P

10.0 FeSiAl P

11.6

CoCrTa P

12.2

6454 0-0-12 20 150 300 NiFe F

FeSiAl P

CoCrTa F

6385 5-3.6-0 30 350 300 A P 6.7

B 5.0

C P P

D P P

E P P

6386 0-0-12 20 150 300 A P P

5.6

B F C P P D F E P P

6186 0-2.7-12 10 200 100 G P

6179 0-2.7-12 10 200 500 G P P Note a: A = polished Al₂O₃-TiC coupon, B = stainless steel, C = Al₂O₃ insulator material around conductive head material on slider of HGA, D = NiFe pole pieces on slider of HGA, E = Al₂O₃-TiC air bearing and body material of slider of HGA, , G = all materials on exposed slider surfaces of HGA.

Note b: Visually observed coating adhesion by optical microscopy immediately after completion of coating process. P = Pass (no coating delamination); F = Fail (coating delamination from the substrate).

Note c: Results of boiling water-to-ice water thermal shock adhesion test. P = Pass (no coating delamination); F = Fail (coating delamination from the substrate).

Note d: Tensile bond adhesion strength (kpsi) results from pull test. All values reported are values at which cohesive failure occurred within the epoxy bond.

The adhesion test results in Table 2 indicate that the Si-DLC coatings of the present invention exhibit outstanding adhesion to a variety of metallic materials (Ni,

NiFe, FeSiAl, CoCrTa) used in magnetic transducers and magnetic recording media. For the case of the tensile pull adhesion test, the bond strength between the Si-DLC coating and the substrate was greater than or equal to 10 kpsi (the cohesive strength of the epoxy due to the curing conditions utilized) for each of the magnetic metallic materials tested. As determined by the pull test, the adhesion of the Si-DLC coating (Run #6385) to stainless steel was also excellent, and exceeded the cohesive strength of the epoxy. In contrast, the DLC coating without an adhesion layer (Run #6386) on stainless steel spalled either during deposition or immediately upon removal from the coating chamber. The adhesion of the DLC coating (without an adhesion-enhancing interlayer) was poor on all of the metal substrates tested.

The adhesion of both the Si-DLC coating (Run #6385) and the DLC coating (Run #6386) to the Al₂O₃-TiC coupon was sufficient to pass the pull test; failure occurred within the epoxy adhesive at 6.7 kpsi and 5.6 kpsi, respectively.

Microscopic examination of the Si-DLC-coated HGA's (Runs #6385 and #6179) revealed good adhesion (no evidence of adhesion failure) before and after the thermal shock test. The pull test could not be performed on the HGA's because of their small size. Microscopic examination of the DLC-coated HGA (Run #6386) in the as-deposited condition revealed spalling (delamination) of the DLC coating on the NiFe pole pieces, and no indication of adhesion failure on the alumina surrounding the pole pieces or on the Al₂O₃-TiC body. The thermal shock test on this HGA caused the coating to be completely removed from the NiFe poles, but did not cause any coating adhesion failure on the Al₂O₃ or Al₂O₃-TiC slider materials. These results showed that the undoped DLC adhered well to the Al₂O₃-TiC body material and the insulating layers of Al₂O₃, but not to the NiFe pole material.

Adhesion test results on Si-DLC-coated HGA's in Runs #6186 and #6179 in Table 2 demonstrated that different deposition conditions (ion beam energy, flow rate of silicon-containing precursor gas, e.g. TMS, and flow rate of hydrocarbon precursor gas, e.g. methane) could be utilized to produce ion beam deposited Si-DLC coatings which have excellent adhesion to all the materials on the wear surface of a typical slider on an HGA. Overall, these results showed that the ion beam deposited Si-DLC coatings of the present invention have superior adhesion to a variety of metallic materials (e.g. Ni, NiFe, FeSiAl, CoCrTa) and ceramic materials (e.g. Al₂O₃ and Al₂O₃-TiC) used in magnetic head sliders and other transducers and magnetic recording media without the need for an adhesion layer. Because of their inherent chemical resistance, and demonstrated outstanding adhesion, the Si-DLC coatings of the present invention impart outstanding corrosion resistance to the magnetic materials used in magnetic transducers and recording media.

Example 8 A highly polished stainless steel coupon, and a slider mounted in an HGA were coated with an adherent DLC coating using a thin layer (about 30 A) of Si-DLC as an adhesion layer. After chemical cleaning and mounting in the vacuum chamber, the substrates were sputter-etched with an argon ion beam under conditions which produced an etch depth of about 130 A on a silicon witness coupon. Next, the deposition conditions summarized in Table 3 (Run # 6198) were utilized for the deposition of an Si-DLC layer, followed by a DLC layer. Table 3

Gas Feed Substrate Ion Beam Layer

Run Ar-TMS-CH₄ Angle Energy Thickness Substrate Adhesion Results Ω. ccm^ {degrees) (Volts) (A ) Material" Yis^h IS^C Eull^d

6198 6-3.6-0^c 30 75 30 B P P

12.7

0-0- 12^f 30 75 500 G P P

Note a: B = stainless steel, G = all materials on exposed slider surfaces of

HGA. Note b: Visually observed coating adhesion by optical microscopy immediately after completion of coating process. P = Pass (no coating delamination); F = Fail (coating delamination from the substrate).

Note c: Results of boiling water-to-ice water thermal shock adhesion test. P = Pass (no coating delamination); F ^■= Fail (coating delamination from the substrate). Note d: Tensile bond adhesion strength (kpsi) results from pull test. All values reported are values at which cohesive failure occurred within the epoxy bond. Note e: Ion beam source process conditions for deposition of Si-DLC layer. Note f: Ion beam source process conditions for deposition of DLC top coating layer.

The adhesion test results are also summarized in Table 3. The coatings on the stainless steel coupon and slider surfaces of the HGA passed the thermal shock adhesion test. The tensile bond strength of the coating to the stainless steel coupon, as determined by the pull test, was greater than 12.7 kpsi, the value at which cohesive failure occurred within the epoxy. Because of the small size of the slider on the HGA, a pull test could not be conducted. These results demonstrated that the ion beam deposited Si-DLC materials of the present invention can be used successfully as adhesion-enhancing interlayers for DLC coatings on magnetic head sliders.

Example 9 A series of coating runs was performed to investigate the effect of the silicon concentration in the ion beam deposited Si-DLC material on the adhesion of this coating to the wear surfaces of magnetic head sliders. The goal was to determine the minimum concentration of silicon in the ion beam deposited Si-DLC material required to achieve an adherent coating without an adhesion layer. The run conditions, and adhesion test results are presented in Table 4. The values of the approximate Si atomic concentration presented in Table 4 are based on the results of measurements of films deposited under similar conditions (see Table 6 in the discussion which follows).

Table 4 Gas Feed Ion Beam Layer Approx.

Run Ar-TMS-CH₄ Energy Thickness Si Cone. Adhesion Test Results Mo. (seem) (Volts) (A at. %) yjs" IS^C

6577 0-0.36-12 300 50 5.2 P P 6575 0-0.36-12 300 100 5.2 P P

6576 0-0.36-12 300 200 5.2 F F

6578 0-1.1-12 300 50 6.7 P P

6579 0-1.1-12 300 100 6.7 P P

Note b: Visually observed coating adhesion by optical microscopy immediately after completion of coating process. P = Pass (no coating delamination); F = Fail (coating delamination from the substrate). Note c: Results of boiling water-to-ice water thermal shock adhesion test. P = Pass (no coating delamination); F = Fail (coating delamination from the substrate).

The deposition conditions for Runs #6577, #6575, and #6576 were the same as those of Run #6153 in Table 6. The deposition conditions for Runs #6578 and

#6579 were the same as those of Run #6155 in Table 6. The 50 A and 100 A thick

Si-DLC films containing 5.2 atomic % Si and 6.7 atomic % Si showed no sign of delamination from the slider surfaces either before or after the thermal shock adhesion test. The 200 A thick Si-DLC coating made with a Si concentration of 5.2 atomic % (Run #6576) displayed partial delamination before the thermal shock test, and further delamination after the thermal shock test. Loss of adhesion for thicker coatings is the result of the compressive stress in the coatings. Therefore, for low Si atomic concentrations in the Si-DLC material, the coating adhesion is improved by reducing the coating thickness. Advantageously, this reduced coating thickness is one of the key objectives of the Si-DLC coating material of the present invention. For slider (and also magnetic media) applications, a typical coating thickness is less than about 100 A, and preferably less than about 50 A. Therefore, the results indicate that an atomic Si concentration of less than or equal to about 5 % results in sufficient adhesion of the ion beam deposited Si-DLC material to magnetic head sliders. Furthermore, as the coating thickness is reduced, even lower Si concentrations are required for sufficient adhesion. For example, for a 20 A thick Si-DLC coating applied to magnetic transducers and media, the inventors believe that a Si concentration as low as about 1 atomic % will result in an adherent ion beam deposited Si-DLC coating.

Example 10 Two deposition runs were conducted for the puφose of comparing the adhesion of the ion beam deposited Si-DLC coatings and DLC coatings to the surface of magnetic recording disks. Substrates for this evaluation were standard magnetic recording disks, with a CoCrTa magnetic alloy (typically cobalt with 8 atomic % chromium, and 4 atomic % tantalum), except the surface of the disks was not coated with the standard hard carbon coating. The disks were packaged in an air tight container immediately upon completion of the deposition of the magnetic material to minimize contamination, so that no chemical cleaning was required prior to ion beam deposition of DLC or Si-DLC.

After mounting one of the disks in a vacuum coating chamber, and evacuating the air to a base pressure of about 5 x 10^"6 Torr, a short ion beam sputter-etch was performed to remove oxide material that formed on the disk surface due to exposure of the metal surface to air. Then, one half of the disk was coated with a 300 A thick coating of Si-DLC by direct ion beam deposition (Run #6497, ion beam voltage of 350 Volts, 3.6 seem TMS + 5 seem Ar gas flow to the 11 cm gridded Kaufman-type ion beam source). The ion beam deposited Si-DLC coating on the disk was uniform, low friction, and smooth. This procedure was repeated on the other half of the disk, but in this case, a 300 A thick DLC coating was deposited (Run #6498, ion beam voltage of 50 Volts, 12 seem of methane gas flow to the ion source). Both coatings passed the thermal shock adhesion test, with no visible evidence of delamination. However, the DLC coating failed the tensile pull test- failure occurred at the interface between the DLC coating and the CoCrTa layer at a force of 3.6 kpsi. In shaφ contrast, the Si-DLC coating passed the tensile pull test- the epoxy bond broke at a force of 7.4 kpsi, indicating that the tensile bond strength of the Si-DLC to the CoCrTa layer was greater than 7.4 kpsi. These results demonstrate the superior adhesion of the Si-DLC coating compared to DLC when applied to the surface of magnetic recording media. In the optimum manufacturing configuration for ion beam deposited Si-DLC coatings on magnetic media, a chemical cleaning and ion beam sputter-etch step would not be required prior to the deposition of Si-DLC, because the Si-DLC coating would be deposited immediately upon completion of the magnetic layer, without breaking vacuum. In this way, the magnetic layer would not be exposed to air or contaminants prior to deposition of the Si-DLC coating. However, the results in this example clearly show that if the surface of the magnetic material becomes contaminated, the ion beam deposited Si-DLC coating can be made to adhere by use of an ion beam sputter-etch step prior to deposition of the Si-DLC layer.

Example 11 In Examples 5-10, Si-DLC was ion beam deposited using different process conditions such as ion beam voltage, precursor gas composition, and deposition angle (angle of incidence of the ion beam on the substrates). This example presents a summary of some of the representative properties of the ion beam deposited Si-DLC materials of the present invention as a function of various process conditions. Tables 5, 6, 7 and 8 and FIGS. 6-14 illustrate the dependence of Si-DLC material hardness, stress, elemental composition, and chemical bonding structure (as evidenced by Raman spectroscopy) on the ion beam voltage, precursor gas composition, and deposition angle using the ion beam deposition process of the present invention. The reported values for these material properties were obtained by analysis of Si-DLC films (typical coating thickness of about 2,000 A to about 7,000 A) deposited on silicon wafers. Nanoindentation hardness of the Si-DLC materials was measured with a Nano Instruments (Oak Ridge, TN) Nanoindentor II which dynamically measures the applied load and indentation depth during penetration of a small pyramidal-shaped diamond tip into the material. Because the indentation depths were small (in the range of 500 A to 1000 A) compared to the thickness of the

Si-DLC coatings the measured nanoindentation hardness values were independent of the substrate hardness. By analysis of the loading and unloading curves during the indentation process, the instrument allows for correction of the hardness value for elastic recovery during the indentation. Since the Si-DLC materials exhibit significant elasticity, this hardness measurement technique is required to obtain accurate hardness values. This method also provides a measurement of Young's modulus of the Si-DLC materials. The nanoindentation hardness values were scaled to a reference hardness value of 11.6 which was obtained for the Si(100) wafer substrates. The uncertainty in the nanoindentation hardness values is approximately - /-10%.

The compressive stress of the Si-DLC coating materials was measured by the wafer curvature method using an FSM, Incoφorated stress meter. The FSM instrument measures the bow imparted to a Si(100) wafer substrate by the coating material and calculates the coating stress using Poisson's equation (thin film assumption) based on the measured coating thickness (determined by surface profilometry), the wafer thickness and the known elastic properties of the wafer.

The elemental composition of the Si-DLC materials was measured by Rutherford Backscattering Spectrometry (RBS) and Hydrogen Forward Scattering (HFS) Analysis. In these methods, a beam of high energy helium ions bombards the coated samples. Helium ions are scattered from the beam by the sample, and hydrogen is forward scattered from the sample. From the known properties of the beam, the scattering geometry and the energy distribution of the detected helium and hydrogen, the composition of the coating can be determined. The uncertainty in the atomic concentrations as determined by these methods is approximately +1- 0.4 atomic % for argon, approximately +1-2 atomic % for silicon, approximately +1-1 atomic % for carbon, and approximately +/-10 atomic % for hydrogen. Information about the chemical bonding structure in the Si-DLC materials was obtained by Raman spectroscopy using a Renishaw micro-Raman spectrometer with a 514 nm wavelength laser. The Raman spectral signal is a measurement of the interaction of photons from an incident laser beam with the vibrational modes of a material lattice. These vibrational modes are determined by the chemical bonding in the material structure. The interaction causes the wavelength of the laser light to be shifted. Therefore, the Raman spectral signal is a signature of how the atoms in the Si-DLC material are chemically bonded. This information is complementary to the elemental analysis which determines only the concentration of elements which are present in the sample. DLC materials exhibit a characteristic peak in the Raman spectrum known as the G-peak. The position of the G-peak is commonly used to characterize and differentiate DLC materials. The discussion which follows clearly shows how the shape of the Raman peaks and the position of the Raman G-peak can differentiate the ion beam deposited Si-DLC materials from DLC, and can characterize the range of chemical bonding in the Si-DLC materials of the present invention.

The effect on the silicon dopant on the chemical bonding structure of ion beam deposited DLC materials was investigated with Raman spectroscopy. Raman spectra for representative DLC (Run #6304, 12 seem CH4 feed gas, ion beam voltage of 150 Volts) and Si-DLC (Run #6285, 3.6 seem TMS plus 5 seem Ar feed gas, ion beam voltage of 350 Volts) and materials are presented in FIGS. 6 and 7, respectively. These present the Raman spectral scans (raw data), the curve fit to the raw data, the background subtraction, and deconvoluted Gaussian peaks used to fit the raw data. For both materials, the deconvolution of the spectra revealed the existence of two peaks with the following peak position and peak width: D-peak at 1338 cm^"' (width 171 cm ¹) and G-peak at 1534 cm^"' (width of 107 cm ¹) for DLC; and D-peak at 1270 cm^"1 (width of 241 cm ¹) and G-peak at 1473 cm^"1 (width 111 cm^"1) for Si-DLC. For DLC materials, the peak at the larger Raman shift position is referred to as the G-peak (graphite peak) and the peak at the smaller Raman shift position is referred to as the D-peak (disorder peak). The inventors believe, but are not certain that the two peaks for the Si-doped material correspond to the D and G peaks. However, the difference in the peak positions found for the DLC and the

Si-DLC materials is certainly significant, and illustrate that the presence of silicon in the material has had a significant influence on the chemical bonding structure.

Table 5 and FIGS. 8 and 9 present the effect of ion beam voltage on the film properties. The precursor gas composition for all the deposition runs in Table 5 was 3.6 seem TMS and 10 seem argon, except for Run #6195, which utilized 3.6 seem TMS and 6 seem argon.

Table

Ion Beam Coating Depos'n. Hard- Composition Raman G- Run Energy Thickness Rate Stress ness C-Si-H-Ar Peak Posn.

NIL (Volts) (A) (A/min) (GPa) (GPa) ⁽atomic %) {cm^"1)

6195 75 4800 50 -0.71 12.0 NM NM

6255 150 4200 47 -0.44 13.4 38.1-20-41-0.9 1463

6253 200 3300 55 -0.77 15.5 39.8-20-39-1.2 1467 6262 250 3700 62 -0.94 NM NM 1470

6261 300 4100 68 -1.16 16.5 41.5-20-37-1.5 1470

6257 350 4200 70 -1.10 17.3 46.5-20-32-1.5 1469 6260 425 3900 65 -1.04 17.0 NM 1472

6258 500 4100 68 -1.06 17.3 43.2-20-35-1.8 1473 6259 650 3600 60 -0.99 16.5 45-20-33-2.0 1475

NM ^*= sample not measured.

In FIG. 8, hardness data is plotted as closed in squares and stress data is potted as open squares. Higher ion beam voltage causes the ions to impact the surface of the substrate with higher energy. Up to 300 V, the hardness and stress increase as a function of increasing ion beam voltage. Above 300 V, the hardness and stress are fairly constant, up to an ion beam voltage of 650 V, the maximum voltage studied. The very similar dependance of hardness and stress as a function of ion beam voltage was also found for ion beam deposited DLC. As in DLC, the lower hardness of Si-DLC at low ion beam voltage is attributed to lack of sufficient impact energy required to form the diamond-like chemical bonds which maximize the hardness of the material.

The hydrogen content of the Si-DLC material appears to decrease and the carbon content increases with increasing beam voltage. This trend is also similar to that found for ion beam deposited DLC and is attributed to removal of hydrogen by sputtering at higher impact energy. The argon content of the Si-DLC materials increases with increasing beam voltage. This is likely due to increased implantation of argon into the near-surface of the growing Si-DLC film as the impact energy

(proportional to the beam voltage) increases. In Table 5, the silicon concentration is constant at 20 atomic %, which suggests that the silicon concentration in the

Si-DLC material is controlled more by the precursor gas mixture than by the ion beam impact energy. The Raman G-peak shift shows a small, but systematic rise from 1463 cm^"1 to 1475 cm^'1 as the ion beam voltage is increased.

Table 6 shows the effects of varying the precursor gas composition by changing the argon flow between 5 and 20 seem while maintaining the TMS flow at

3.6 seem, for ion beam voltages of 350 V and 500 V.

Table 6

Gas Feed Beam Coating Hard- Composition Raman G-

Run Ar-TMS Energy Thick. Stress ness C-Si-H-Ar Peak Posn. No. (seem) (Volts.) (A) (GPa) (GPa) (at. %) (cm ¹)

6280 5-3.6 350 4400 -1.21 17.9 44-20-35-1 1473

6257 10-3.6 350 4200 -1.1 17.3 46.5-20-32-1.5 1469

6279 20-3.6 350 5400 -0.89 15.0 45.9-19-33-2.1 1474 6283 5-3.6 500 4300 -1.16 17.7 NM NM

6258 10-3.6 500 4100 -1.06 17.3 NM NM

6282 20-3.6 500 4100 -0.88 15.6 NM 1473 NM = sample not measured.

At both 350 V and 500 V ion beam voltage, the Si-DLC hardness and stress increased with decreasing argon flow rate. This suggests that decreasing the argon flow to zero and utilizing a precursor gas feed of only TMS would further increase the hardness. However, considering the small hardness change (0.6 GPa) obtained when the Ar flow rate was decreased from 10 seem to 5 seem, the increase in hardness expected for zero Ar flow is expected to be on the order of only 0.5 GPa or less. For a constant ion beam voltage of 350 V, increasing the argon flow rate had no significant effect on the Raman G-peak position or the elemental composition of the Si-DLC materials, except for an increase in the concentration of entrapped argon in the film.

The data in Table 7 and FIGS. 10, 11, and 12 show how the properties of DLC coating material which is ion beam deposited from 12 seem methane precursor gas is affected by addition of 0.36 seem to 2.7 seem of TMS to the precursor gas feed to the ion beam source.

Table 7

Gas Feed Coating Depos'n. HardComposition Raman

G-

Run Ar-TMS-CH₄ , Thickness Rate Stress ness C-Si-H-Ar Peak

Posn.

No. sccml {A) (A/min (GPa (GPa^ (atomic %)

{cm^'1)

6152 0-0-12 5200 43 -1.83 20.1 54.7-0-45-0.3 1538

6153 0-0.36-12 5000 42 -1.78 18.3 47.6-5.2-47-0.2 1529

6154 0-0.7-12 5700 48 -1.51 NM NM NM

6155 0-1.1-12 6400 53 -1.54 17.6 48.1-6.7-45-0.2 1513

6156 0-1.4-12 7000 58 -1.34 17.9 NM 1508

6157 0-1.8-12 7300 61 -1.24 17.9 50.3-9.5-40-0.2 1506

6159 0-2.7-12 6500 72 -1.34 18.9 49.9-12-38-0.1 1498

6191* 0-2.1-12 5100 53 -0.69 11.8 NM NM

*Note: Run utilized an ion beam energy of 75 Volts. All other coating runs utilized an ion beam energy of 200 Volts.

NM = sample not measured.

In FIG. 10, hardness data is plotted as filled in squares and stress data is potted as open squares. The hardness of the Si-DLC coatings made with TMS are about 2 to 3 GPa less than the hardness of DLC coatings (deposited without TMS). In addition, the hardness of the Si-DLC coatings is not strongly influenced by TMS flow rate over the range of flow rates presented in Table 7. All of the Si-DLC coatings are in compressive stress. The value of the compressive stress of the Si-DLC materials generally decreases as the TMS flow rate is increased.

In FIG. 11 , the atomic concentrations of carbon (closed squares), hydrogen (closed triangles), silicon (plus symbols), and argon (X symbols) in Si-DLC are plotted as a function of TMS flow rate added to 12 seem methane. The elemental composition data show that the silicon concentration in the coating material increases from 0 atomic % to 12 atomic % as the TMS flow is increased from 0 seem to 2.7 seem. The hydrogen concentration in the Si-DLC material decreases, while the atomic concentrations of carbon and argon content are fairly constant as a function of TMS flow rate. The Raman spectroscopy results show a large and continuous decrease in the G-peak position with increasing TMS flow. This large dependence of the G-peak position on the TMS flow rate correlates well to the large dependence of the silicon concentration in the Si-DLC materials on the TMS flow rate. Therefore, changes in the silicon concentration in the Si-DLC material can be determined by the position of the G-peak in the Raman spectrum. Comparison of the data obtained from Run #6191 with the other data in Table 7 shows that reducing the ion beam energy from 200 V to 75 V decreases both the compressive stress and hardness of the Si-DLC materials.

Table 8, Table 9, FIG. 13 and FIG. 14 show the effect of the deposition angle (i.e. angle of incidence between the impinging ion beam and the substrate surface) on material properties of ion beam deposited Si-DLC and DLC. The orientation of the ion beam source and the distance between the ion beam source and the substrates was the same for the deposition of the Si-DLC and DLC materials. In FIG. 13 (for DLC) and FIG. 14 (for Si-DLC), hardness data is plotted as closed squares, and stress data is plotted as open squares. For both materials, the hardness and stress are essentially constant as a function of angle at the lower angles but then decrease rapidly as a function of deposition angle above a critical angle. This critical angle for DLC is between about 40 degrees and about 50 degrees. The critical angle for Si-DLC is between about 60 degrees and about 67 degrees. For Si-DLC, the critical angle for the stress and hardness appear to be the same. Likewise, for DLC the critical angle for the stress and hardness appear to be the same. In general, for similar process conditions, the DLC hardness is linearly related (i.e. directly proportional) to the stress of DLC. Likewise, for similar process conditions, the Si-DLC hardness is linearly related (i.e. directly proportional) to the stress of Si-DLC.

Table 8

Deposition Coating Depos'n. Hard- Composition Raman G-

Run Angle Thickness Rate Stress ness C-Si-H-Ar Peak Posn.

NIL (degrees') (A) (A/min) (GPa) (GPa) (atomic %,) (cm^'1).

6382 0 5800 97 -1.1 1 17.2 40-19.5-39-1.5 1472 6381 15 5500 92 -1.13 17.1 NM 1472

6380 30 4400 73 -1.10 17.2 43.7-20-35-1.3 1471

6376 45 2500 42 -1.23 16.8 NM NM

6377 60 2400 27 -1. 18 17.0 41.9-19-38-1.1 1476 6379 67.5 3000 25 -0.65 13.4 NM 1469 6378 75 5000 28 -0.12 4.3 41.3-15-43-0.7 1470

Note: All coating runs utilized 3.6 seem TMS + 5 seem Ar gas feed to the ion beam source, and an ion beam energy of 350 Volts. NM = sample not measured.

Table 9

Deposition Coating Depos'n. HardComposition Raman

G-

Run Angle Thickness Rate Stress ness C-Si-H-Ar Peak

Posn. NIL (degrees) (A) (A/min) (GPa) (GPa) (atomic %) (cm^'')_

6318 0 6000 50 -1.52 17.3 52.8-0-47-0.2 1540

6319 10 6000 50 -1.64 17.2 NM NM

6320 20 5700 48 -1.67 18.0 NM NM 6314 30 5600 47 -1.29 18.3 52.8-0-47-0.2 1541

6315 40 4100 34 -1.76 18.3 NM NM

6316 50 3900 33 -1.04 15.3 NM NM

6317 60 4900 33 -0.15 6.4 61.9-0-38-0.1 1545

Note: All coating runs utilized 12 seem CH_d gas : feed to the ion beam source, and an ion beam energy of 150 Volts. NM = sample not measured. The elemental composition data in Table 8 for Si-DLC shows little if any change in composition as a function of angle below the critical angle. For angles less than 60 degrees, the elemental composition of the Si-DLC materials deposited from a precursor gas mixture of 3.6 seem TMS and 5 seem Ar, at an ion beam energy of 350 Volts is approximately 42 atomic % carbon, 19.5 atomic % silicon, 37 atomic % hydrogen, and 1.3 atomic % argon. Above the critical angle, a measurable change in composition is observed. At 75 degrees, the elemental composition of the Si-DLC materials is approximately 41.3 atomic % carbon, 15 atomic % silicon, 43 atomic % hydrogen, and 0.7 atomic % argon. The significant change is the decrease of the silicon concentration from 19.5 atomic % to 15 atomic % , and the increase in hydrogen concentration from 37 atomic % to 43 atomic % . Apparently, the Raman G-peak position was essentially unaffected by the change in deposition angle. Thus, the Raman G-peak position, although sensitive to changes in precursor gas mixtures (see Table 7), is not sensitive to small changes in the silicon concentration (or coating hardness and stress) which are associated with changes in deposition angle.

These data show that the range of angles over which hard Si-DLC can be deposited without significant change in material properties is larger than is the range for DLC. This is one example which illustrates that the ion beam deposition process for Si-DLC is more robust and capable for manufacturing than the process for deposition of DLC.

Example 12 In Example 12, Si-DLC coatings were deposited onto polished stainless steel coupons, polished Al₂O₃-TiC coupons, silicon wafers, and the magnetic transducers on HGA's by argon ion beam sputter deposition in Runs #6387 and #6389. In these runs, a 1000 eV, 50 mA argon ion beam generated by a 5 cm gridded Kaufman- type ion beam source was used to sputter-deposit the Si-DLC layer by ion beam sputter deposition from a silicon carbide target. The deposition rate of the Si-DLC coating was approximately 28 A/minute, and the compressive stress was 1.2 x 10¹⁰ GPa. Excellent adhesion of the ion beam sputter-deposited Si-DLC coating was verified on all substrates. A 300 A thick Si-DLC coating deposited at these conditions on all substrates passed the boiling water-to-ice water thermal shock adhesion test. Using the tensile bond adhesion pull test, the adhesion of the 300 A thick ion beam sputter-deposited coating onto the polished Al₂O₃-TiC coupon was greater than 11.8 kpsi, the force at which cohesive failure occurred within the epoxy bond.

In Run #4344, another 5900 A thick ion beam sputter-deposited Si-DLC coating was made for the puφose of material characterization. In this case, the same silicon carbide sputtering target was used, but the deposition rate was increased to 98 A/minute by operating at an ion beam energy of 1500 eV, and an the ion beam current of 150 mA. The compressive stress of this Si-DLC coating was 1.24 GPa, the hardness was 26 GPa, the Raman G-peak position was 1427 cm^'1, and the atomic composition of the material was 51.8 atomic % carbon, 46.0 atomic % silicon, 1.8 atomic % hydrogen, and 0.3 atomic % argon. The atomic concentrations were measured by RBS and HFS as described in earlier examples. As evidenced by the existence of the same compressive stress in the Si-DLC material deposited in Run #4344 and in Runs #6389, the Si-DLC material deposited in Run #4344 is believed to be approximately the same as the materials deposited in Runs #6387 and #6389.

The prior examples illustrate ion beam deposited Si-DLC coatings with silicon concentrations between about 5 atomic percent and about 46 atomic percent. The lower limit of the useable silicon concentration is the silicon concentration which is required to promote good adhesion of the Si-DLC coating. For any particular set of Si-DLC coating process parameters (i.e. gas mixture, ion beam voltage, and coating thickness), the adhesion of the Si-DLC coating is dependent on the silicon concentration, and substrate material. Because the Si-DLC coatings are under internal compressive stress, the adhesion strength of the Si-DLC coatings improves as the coating thickness is reduced. Therefore, as the magnetic spacing, i.e. Head-to-disk flying height, is reduced, the necessary silicon concentration in the Si-DLC coating may be reduced. The inventors believe that for the thinnest Si-DLC coatings, on the order of 20 A or less, a silicon concentration as low as 1 atomic percent is sufficient to produce a coating with adequate adhesion. In most circumstances, the preferable atomic concentration of silicon is about 2 atomic percent or greater.

The upper limit of the silicon concentration in the ion beam deposited Si-DLC materials of the present invention is determined by the ability of the DLC matrix to isolate silicon atoms, as previously discussed. When the silicon-to-carbon ratio becomes one, the deposited material takes on characteristics of silicon carbide. When the silicon-to-carbon ratio exceeds one, free silicon remains in the material. Both of these cases are detrimental to the head-disk interface. Therefore, the upper limit of the silicon-to-carbon ratio is about 0.95 for Si-DLC materials with low hydrogen content (such as those deposited by ion beam sputter deposition, deposition from a laser ablation ion source or a cathodic arc ion source), and preferably less than or equal to about 0.9 as described in Example 12.

The CSS and nanowear results of Example 5 for ion beam deposited Si-DLC and DLC coatings on HGA's, demonstrated that the Si-DLC coatings of the present invention performed equal to or better than DLC as protective coatings for magnetic head sliders.

The thermal shock test, tensile bond pull test, nanowear, and CSS test results discussed in Examples 5, 6, 7, 8, 9 and 12 showed that ion beam deposited Si-DLC material, made under a variety of conditions, adheres well to the materials that make up magnetic head sliders (Al₂O₃, Al₂O₃-TiC, NiFe, and FeSiAl).

Furthermore, Examples 7 and 8 showed that DLC does not adhere well to NiFe and FeSiAl without an adhesion layer, thus demonstrating that the Si-DLC materials of the present invention have superior adhesion without the need for an adhesion layer. Example 10 illustrated the Si-DLC coatings of the present invention applied to magnetic recording media.

Representative material characteristics of the ion beam deposited Si-DLC coatings of the present invention, which are direct ion beam deposited from silicon- containing precursor gases, (as described in Example 11) include the following: Nanoindentation hardness in the range of about 12 GPa to about 19 GPa, compressive stress in the range of about 0.4 GPa to about 1.8 GPa, Raman spectral G-peak position in the range of about 1463 cm^"' to about 1529 cm^"1, a silicon concentration in the range of about 1 atomic % to about 30 atomic %. Ion beam deposited Si-DLC materials of the present invention which contain hydrogen typically exhibit a hydrogen concentration in the range of about 25 atomic % to about 47 atomic %. Example 12 demonstrated that adherent, hard ion beam Si-DLC coatings can be manufactured by an ion beam sputter deposition process.

Because of their inherent chemical resistance, and demonstrated outstanding adhesion, the Si-DLC coatings of the present invention impart outstanding wear resistance, and corrosion resistance to the magnetic materials used in magnetic transducers and recording media. Since an adhesion-enhancing interlayer is not required, the ion beam deposited Si-DLC coatings of the present invention can be made extremely thin (about 20 A thick), thus allowing for higher data storage densities, while maintaining outstanding wear resistance and corrosion resistance. From the foregoing description, one of ordinary skill in the art can easily ascertain that the present invention provides an improved method for producing highly protective and wear resistant Si-DLC coatings on magnetic transducers and magnetic recording media. Highly important technical advantages of the method of the present invention include outstanding adhesion of the ion beam deposited Si-DLC coating without the need for an adhesion-enhancing interlayer, and ease and flexibility of scale-up to mass production of Si-DLC-coated substrates.

Without departing from the spirit and scope of this invention, one of ordinary skill in the art can make various changes and modifications to the invention to adapt it to various usages and conditions. As such, these changes and modifications are properly, equitably, and intended to be, within the full range of equivalents of the following claims.

Claims

WHAT IS CLAIMED IS:

1. A method for producing a protective, wear resistant silicon-doped diamond-like carbon coating on the wear surface of a magnetic transducer comprising the steps of: (a) chemically cleaning the surface of said magnetic transducer to remove contaminants; (b) sputter-etching the surface of said magnetic transducer in an evacuated deposition vacuum chamber with a beam of energetic gas ions to further remove residual contaminants; (c) ion beam depositing a silicon-doped diamond-like carbon coating to a predetermined thickness; (d) increasing the vacuum chamber pressure to atmospheric pressure; and (e) recovering a silicon-doped diamond-like carbon coated magnetic transducer having improved resistance to wear, abrasion and corrosion.

2. The method of Claim 1 wherein said contaminants are selected from the group consisting of residual hydrocarbons, surface oxides and other unwanted materials.

3. The method of Claim 1 wherein said silicon-doped diamond-like carbon coating is deposited by direct ion beam deposition.

4. The method of Claim 1 wherein said silicon-doped diamond-like carbon coating is deposited by direct ion beam deposition from a silicon-containing, carbon-containing, and hydrogen-containing precursor gas and has a Nanoindentation hardness in the range of about 12 GPa to about 19 GPa, compressive stress in the range of about 0.4 GPa to about 1.8 GPa, and a silicon concentration in the range of about 1 atomic % to about 30 atomic %.

5. The method of Claim 1 wherein said silicon-doped diamond-like carbon coating is deposited by direct ion beam deposition using carbon ions and silicon ions generated by a filtered cathodic arc ion source and has a

Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa and, a silicon concentration in the range of about 1 atomic % to about 45 atomic % .

6. The method of Claim 1 wherein said silicon-doped diamond-like carbon coating is deposited by direct ion beam deposition using carbon ions and silicon ions generated by a laser ablation ion source and has a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa and, a silicon concentration in the range of about 1 atomic % to about 45 atomic % .

7. The method of Claim 1 wherein said silicon-doped diamond-like carbon coating is deposited by ion beam sputter deposition from a sputtering target comprised of silicon and carbon and has a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa and, a silicon concentration in the range of about 1 atomic % to about 45 atomic % .

8. The product by the method of Claim 1.

9. The product by the method of Claim 4.

10. The product by the method of Claim 9 wherein said coating has a Raman spectral G-peak position in the range of about 1463 cm^"1 to about 1530 cm^"1.

11. The transducer of Claim 9 wherein the thickness of said silicon-doped diamond-like carbon coating is in the range of about 20 A to about

2000 A.

12. The product of Claim 8, wherein said magnetic transducer consists of a slider.

13. The product of Claim 8 wherein said magnetic transducer consists of a tape head.

14. The product of Claim 8 wherein said magnetic transducer consists of a magnetoresistive slider.

15. The product of Claim 8 wherein said magnetic transducer consists of a magnetoresistive tape head.

16. A transducer for use with magnetic recording media composed of a transducer assembly substrate, and silicon-doped diamond-like carbon coating deposited on and bonded to said transducer, whereby said transducer has improved wear and corrosion resistance.

17. The transducer of Claim 16 wherein the thickness of said silicon-doped diamond-like carbon coating is in the range of about 20 A to about 2000 A.

18. The transducer of Claim 16 wherein said silicon-doped diamond-like carbon coating has the properties of a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 5 GPa, a silicon concentration in the range of about 1 atomic % to about 45 atomic % , and a hydrogen concentration in the range of 0 to about 47 atomic % .

19. The transducer of Claim 16 wherein said silicon-doped diamond-like carbon coating has a Raman spectral G-peak position in the range of about 1463 cm^"' to about 1530 cm^"'.

20. The transducer of Claim 16 wherein said coating is deposited by direct ion beam deposition from a silicon-containing, carbon-containing, and hydrogen-containing precursor gas.

21. The transducer of Claim 16 wherein said coating is deposited by direct ion beam deposition using carbon ions and silicon ions generated by a cathodic arc ion source.

22. The transducer of Claim 16 wherein said coating is deposited by direct ion beam deposition using carbon ions and silicon ions generated by a laser ablation ion source.

23. The transducer of Claim 16 wherein said coating is deposited by ion beam sputter deposition from a sputtering target comprised of silicon and carbon.

24. The transducer of Claim 16 consisting of a slider.

25. The transducer of Claim 16 consisting of a tape head.

26. The transducer of Claim 16 consisting of a magnetoresistive slider.

27. The transducer of Claim 16 consisting of a magnetoresistive tape head.

28. A method for producing a protective, wear resistant silicon-doped diamond-like carbon coating on the wear surface of a magnetic recording medium comprising the steps of: (a) ion beam depositing a silicon-doped diamond-like carbon coating to a predetermined thickness onto the surface of the magnetic recording material in an evacuated deposition chamber; (b) increasing the vacuum chamber pressure to atmospheric pressure; and (c) recovering a silicon-doped diamond-like carbon coated magnetic recording medium having improved resistance to wear, abrasion and corrosion.

29. The method of Claim 28 wherein said silicon-doped diamond-like carbon coating is deposited by direct ion beam deposition.

30. The method of Claim 29 wherein said silicon-doped diamond-like carbon coating is deposited from a silicon-containing, carbon -containing and hydrogen-containing precursor gas and has a Nanoindentation hardness in the range of about 12 GPa to about 19 GPa, compressive stress in the range of about 0.4 GPa to about 1.8 GPa, and a silicon concentration in the range of about 1 atomic % to about 30 atomic % .

31. The method of Claim 29 wherein said silicon-doped diamond-like carbon coating is deposited using carbon ions and silicon ions generated by a filtered cathodic ion source and has a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa and, a silicon concentration in the range of about 1 atomic % to about 45 atomic % .

32. The method of Claim 28 wherein said silicon-doped diamond-like carbon coating is deposited by direct ion beam deposition using carbon ions and silicon ions generated by a laser ablation ion source and has a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa and, a silicon concentration in the range of about 1 atomic % to about 45 atomic % .

33. The method of Claim 28 wherein said silicon-doped diamond-like carbon coating is deposited by ion beam sputter deposition from a sputtering target comprised of silicon and carbon and has a Nanoindentation hardness in the range of about 12 GPa to about 70 GPa, a compressive stress in the range of about 0.4 GPa to about 10 GPa and, a silicon concentration in the range of about 1 atomic % to about 45 atomic % .

34. The product manufactured by the method of Claim 28.

35. The product by the method of Claim 28 wherein said coating has the properties of a Nanoindentation hardness in the range of about 12 GPa to about 19 GPa, a compressive stress in the range of about 0.4 GPa to about 1.8 GPa, and a Raman spectral G-peak position in the range of about 1463 cm^"' to about 1530 cm^"1.

36. The product of Claim 28, wherein said magnetic recording medium consists of a magnetic hard disk.

37. The product of Claim 28, wherein said magnetic recording medium consists of a magnetic floppy disk.

38. The product of Claim 28, wherein said magnetic recording medium consists of a magnetic recording tape.

39. A magnetic recording medium, composed of a magnetic recording medium substrate, and a silicon-doped diamond-like carbon coating deposited on said substrate, whereby said recording medium has improved wear and corrosion resistance.

40. The recording medium of Claim 39 wherein said coating thickness is in the range of about 20 A to 2000 A.

41. The recording medium of Claim 39 wherein said coating has the properties of a Nanoindentation hardness in the range of about 12 GPa to about

70 GPa, and a compressive stress in the range of about 0.4 GPa to about 10 GPa.

42. The recording medium of Claim 39 wherein said coating has a Raman spectral G-peak position in the range of about 1425 cm^"1 to about 1530 cm '.

43. The recording medium of Claim 39 wherein said coating is deposited by direct ion beam deposition from a silicon-containing, carbon-containing, and hydrogen containing precursor gas.

44. The recording medium of Claim 39 wherein said coating is deposited by direct ion beam deposition using ions carbon ions and silicon ions generated by a cathodic arc ion source.

45. The recording medium of Claim 39 wherein said coating is deposited by direct ion beam deposition using carbon ions and silicon ions generated from a laser ablation ion source.

46. The recording medium of Claim 39 wherein said coating is deposited by ion beam sputter deposition from a sputtering target comprised of silicon and carbon.

47. The recording medium of Claim 39 consisting of a magnetic hard disk.

48. The recording medium of Claim 39 consisting of a magnetic floppy disk.

49. The recording medium of Claim 39 consisting of a magnetic recording tape.