DIAMOND-LIKE NANOCOMPOSITE THIN FILMS FOR AUTOMOTIVE POWERTRAIN COMPONENT COATINGS
FIELD OF THE INVENTION
The present invention relates generally to low friction diamond-like nanocomposite coatings comprising an amorphous matrix. More particularly, the present invention relates to low friction diamond-like nanocomposite compounds for use as friction reducing protective coatings on automotive powertrain assemblies.
BACKGROUND OF THE INVENTION
Internal combustion engines provide a harsh environment for the component parts, especially the parts which move. Components of the powertrain assembly in most internal combustion engines often move at high speeds and in close proximity to interacting surfaces. Such components must be machined and manufactured to specific tolerances that enable operating at high temperatures and low temperatures, without risking material stress, fatigue and/or failure. The components may or may not operate in the presence of lubricant.
However, the considerations of noise, vibration, and wear relative to frictional forces generated between interacting surfaces over prolonged periods at high operating temperatures remain.
Internal combustion engines were once made from heavier metals which were known to be durable and perform well under harsh temperature and frictional conditions. However, to achieve more cost-efficient and fuel- efficient vehicles, today's engines and powertrain assemblies require lighter components . Unfortunately, the lightweight metals of choice, such as aluminum and
aluminum alloys, are not as durable as their stronger, heavier predecessors. In addition, while aluminum pistons and bores (cylinders) are desirable from a cost and machining standpoint, the aluminum/aluminum interface of the bores, pistons, valves, etc. does not work if there is not sufficient lubrication. This is a problem for "cold starts", where the engine and parts must cycle briefly without adequate lubrication (i.e. before lubricant is pumped to the site) . As a result, cast iron liners, coatings or platings have been used in the aluminum bores to achieve the proper compatibility of the components during cold starts. In addition to the higher costs, the cast iron liners are heavy and must have a thermal expansion factor taken into account since the cast iron and aluminum have varying expansion coefficients over the operating temperature range of from about -100°C to about 350°C.
While cast iron plating may provide a workable solution, the plating process produces an undesirable amount of hazardous cyanide waste which is expensive to properly dispose of. It would be desirable to coat the aluminum and aluminum alloy components with a sufficiently strong and lubricious coating to facilitate the use of an all aluminum containing powertrain assembly and engine,
The application of hard and lubricious thin films, or coatings, formed from a combination of amorphous carbon-hydrogen-silicon to a substrate are known. However the film deposition conditions often require heating in a vacuum at 600°C. Heating at such temperature restricts the types of substrates which may be coated with such a coating, since the substrates could melt, lose original shape, or otherwise fatigue.
The use of deposition techniques such as radio frequency (RF) , ion beam and microwave plasma chemical vapor deposition (CVD) and physical vapor deposition (PVD) to deposit hard, lubricious coatings on substrates is known. However, delamination and non-adherence conditions pose problems for such coatings. Thicker deposited layers would seem to provide greater wear resistance and protection to the coated substrate. However, as thickness of the deposited coating layer increases, the likelihood of delamination of the coating also increases due to high internal stresses common in these coatings. A successful coating must have superior adherence, to a degree yet unknown to most hard coatings.
U.S. Patent No. 5,249,554 discloses a hard, wear-resistant coating system comprising a film and an interlayer which adheres to powertrain components resulting in reduced friction and wear. The films are deposited by physical or chemical vapor deposition of coating systems of various combinations of amorphous hydrogenated carbon, silicon-doped amorphous hydrogenated carbon, boron-, nitrogen-, boron nitride-, or metal- and doped amorphous hydrogenated carbon, silicon, silicon carbide, silicon nitride, boron nitride, and mixtures thereof. However, the coatings require the use of an interlayer. The disclosed interlayer mitigates the chemical incompatibility of certain coatings with certain substrates .
Coating systems having such additional layers, or interlayers, have many potential shortcomings. Depositing a required interlayer increases coating system complexity and cost. Further, interlayers themselves run the risk of delamination over time. Still further, such interlayers must be chemically compatible with both the substrate and the coating or outer film layer, thus increasing the complexity of factors taken into
consideration for a particular coating process, and limiting the number of possible substrate/interlayer coating combinations.
Diamond-like coatings (DLCs) have been proposed for uses where wear-resistance is desired. However, most known DLCs will not adequately adhere to metals without additional compatibilizing interlayers, especially when the metal substrate to be coated must operate over a widely varying temperature range during its useful life. DLCs therefore continue to have thermal stability problems.
A hard, wear-resistant and lubricious coating which could be applied in thin layers directly to substrates, especially substrates used as components for powertrain assemblies, without the need for additional compatibilizing layers or interlayers would be extremely useful and desirable.
SUMMARY OF THE INVENTION
The present invention relates to coated powertrain components in internal combustion engines.
The powertrain component comprises a substrate made from a metal-containing or non-metal containing materials. The substrate is coated with a wear-resistant and friction reducing coating. The coating is made from a class of diamond-like solid state materials formed from interpenetrating networks comprising a first network of diamond-like carbon stabilized by hydrogen, a silicon network stabilized by oxygen, and optionally at least one network made from dopant elements or dopant compounds containing elements from Groups l-7b and 8 of the periodic table.
In a further embodiment of the present invention, an internal combustion engine is disclosed comprising a powertrain component. The powertrain component comprises a substrate made from a metal- containing or non-metal containing materials. The substrate is coated with a wear-resistant and friction reducing coating. The coating is made from a diamond¬ like atomic scale material formed from interpenetrating networks comprising a first network of diamond-like carbon stabilized by hydrogen, a silicon network stabilized by oxygen, and optionally at least one network made from dopant elements or dopant compounds containing elements from Groups l-7b and 8 of the periodic table.
In a still further embodiment of the present invention, a method for inhibiting wear of a component in a powertrain assembly is disclosed. A powertrain component is provided and coated with a coating made from a diamond-like material formed from interpenetrating networks comprising a first network of diamond-like carbon stabilized by hydrogen, a second silicon network stabilized by oxygen, and optionally at least one network made from dopant elements or dopant compounds containing elements from Groups l-7b and 8 of the periodic table.
In another embodiment of the present invention, a method for reducing friction of a powertrain component in a powertrain assembly is disclosed. At least one of two adjacent powertrain components are provided and coated with a coating with made from a diamond-like material formed from interpenetrating networks comprising a first network of diamond-like carbon stabilized by hydrogen, a second silicon network stabilized by oxygen, and optionally at least one network made from dopant elements or dopant compounds containing elements from Groups l-7b and 8 of the periodic table.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a schematic diagram showing the principle microstructure of two-network (A) , intermediate (B) , and three-network (C) nanocomposites.
FIG. 2 is a schematic diagram detailing the main method of fabrication of the DLN coatings.
FIG. 3 is a schematic diagram detailing the methods of fabrication of DLN coatings using reflected beam flow.
FIG. 4 is a schematic diagram detailing a preferred DLN fabrication and deposition chamber.
FIG. 5 is a graph of DLN and DLC hardness values plotted versus compressive stress values.
FIG. 6 is a schematic sectional view of an internal combustion engine showing a valve lifter assembly as illustrative of other power train components.
DETAILED DESCRIPTION OF THE INVENTION
The powertrain coatings of the present invention result in superior wear-resistance of the coated components, which would lead to significant improvements in the manufacture of internal combustion engines. The powertrain components comprise substrates made from metal-containing, ceramic-containing or plastic-containing compounds, or combinations thereof. The substrate is coated with a wear-resistant and friction reducing coating. The coating is made from a class of diamond-like solid state materials formed from interpenetrating networks comprising a first network of
diamond-like carbon stabilized by hydrogen, a second silicon network stabilized by oxygen, and optionally at least one network made from dopant elements or dopant compounds containing elements from Groups l-7b and 8 of the periodic table.
The diamond-like nanocomposite films and coatings (DLNs) used to coat the powertrain components of the present invention find their advantage in their superior hardness, adherence, temperature stability, and overall durability. The DLNs, especially the metal -doped DLNs, combine high microhardness with high elasticity. The microhardness values of the DLNs of the present invention range from about 6 to about 30 GPa, while displaying reduced intrinsic stress, especially as compared to DLCs. See Figure 5.
In addition to the protective and wear- resistant properties of the powertrain component coatings of the present invention, the coatings are strong and erosion resistant, such as to chemicals, erosion, abrasion, or ablation while also being highly thermally stable. The coatings would therefore be impervious to physical or chemical attack, and highly resistant to incident laser or other radiation. The resistance of the coatings of the present invention to erosion, reduces the possibility of, for example, physical chipping. This results in the surface of the substrate also being less likely to environmental corrosive forces. The coatings have excellent adherence to the substrate without applying an interlayer, and are resistant to thermal shock and elevated temperatures beyond those known to erode or graphitize known diamond-like coatings.
The term "powertrain components" encompasses the parts of a machine assembly responsible for generating power leading to work or motion. In an
automotive internal combustion engine, the powertrain assembly comprises the valves, pistons, rings, etc. of the engine block, the gears, linkages, rods and joints, etc . which extend from the engine block to the transmission, the components of the transmission itself, and all linkages extending from the transmission and connecting to the axel assemblies. The powertrain components may include more components as would be readily understood by those skilled in the field of engine and powertrain assemblies. hat is important is that at least two parts come into close proximity or contact in such a way, and at such temperatures, that frictional forces can lead to the wear and breakdown of the parts without sufficient lubrication and wear- resistant coatings.
Wear-resistance refers to the ability of a material to resist mass removal m an abrasion/wear environment. The lower the mass removal for a given set of conditions, the higher is the material's wear resistance.
The fundamental structure of the preferred wear-resistant atomic scale diamond-like nanocomposites (DLNs) used to coat the selected substrates is comprised of two or more self-stabilized random networks, each stabilized chemically by additional atomic species, while both networks also structurally stabilize each other. An example of a material with such a structure is the diamond-like nanocomposite (DLN) which is the subject of U.S. Patent No. 5,352,493 and U.S. Serial No. 08/249,167 filed May 24, 1994.
In the DLN, a random carbon network, mainly in the form of sp3 bonded carbon is chemically stabilized by hydrogen atoms, and a glass-like silicon network is chemically stabilized by oxygen atoms, resulting in a
purely amorphous structure. "Amorphous" as used herein refers to a random structure or arrangement of atoms in a solid state that results in no long range regular ordering, and no crystallinity or granularity. The DLN materials have an amorphous structure and do not contain clusters greater than 10 Angstroms. This absence of clusters at the atomic scale is a key characteristic of the DLN coatings of the present invention. Clusters can destroy amorphous nature of the structure, and can serve as active centers of degradation.
The DLNs of the present invention contain no clusters or ordering greater than that defined by one- third the radius of the coordination sphere. This structure has been confirmed via electron projection methods, scanning tunneling microscopy, atomic force microscopy, glancing x-ray and electron diffraction techniques and high resolution transmission electron microscopy (TEM) . Cluster formation is prevented in the sources, in the primary plasma, in the chamber space, and during film growth.
The atomic structure of the class of diamond¬ like nanocomposite (DLN) solid state materials of the present invention is shown in FIG. 1(A) . The materials may have one or more separate disordered networks of dopants, as shown in FIG. 1(B) and 1(C) . The dopants may be any one or a combination of the transition metals and non-metals. of the Groups l-7b and 8 of the periodic table, and all three types of networks (C-H; Si-0 and the dopant network, Me-Me) bonded to each other predominantly by weak chemical bonds. The network elements other than the C-H network may be referred to as alloying elements. Further, silicon and oxygen atoms may also be used in the dopant networks with other elements and compounds.
The DLN coatings of the present invention may comprise a two component network: the diamond-like carbon-hydrogen network interpenetrated with the glass- like silicon-oxygen network. A three component network may also be used comprising the Si-0 and C-H networks with one or more dopant networks, with the dopants being interspersed with the previously mentioned two interpenetrating networks. In this instance three or more interpenetrating networks will be present in the DLN to form a so-called Me-DLN (metal-diamond-like nanocomposite) network. It is understood that non-metal dopant networks, may also be incorporated as the optionally present dopant networks interpenetrating the C-H and Si-0 networks.
The three networks (C-H matrix, Si-0 matrix and a dopant matrix) are bonded to one another mainly by weak chemical bonds. Carbide formation can be prevented even at metal concentrations as high as 50% (verified using Auger electron spectroscopy, electron spectroscopy for chemical analysis (ESCA) , extended x-ray absorption fine structure spectroscopy (EXAFS) and Fourier transform infrared spectroscopy (FTIR) ) . Again, the properties of these materials can be varied over wide ranges depending on the dopant and the concentration selected, as well as the deposition technique and parameters. As already mentioned, the structure of these composites can be tailored at the molecular level. Therefore, unique electrical, optical, and other desirable solid state properties with desired mechanical strength, hardness and chemical resistance can be imparted on the DLN coatings which can be used to coat metal-containing and non-metal containing substrates.
Preferred dopant elements to be used in the Me-
DLN network, and which are particularly effective for use as dopants in a corrosion-resistant Me-DLN coating are B,
Si, Ge, Te, O, Mo, W, Ta, Nb, Pd, Ir, Pt, V, Fe, Co, Mg, Mn, Ni, Ti, Zr, Cr, Re, Hf, Cu, Al, N, Ag and Au; with W, Cr, Zr, Ti and Hf being preferred. Preferred compounds which may be used as dopants include TiN, BN, AIN, ZrN and CrN; with TiN and ZrN being most preferred.
The carbon content in the diamond-like nanocomposite is greater than about 40 atomic % of the DLN preferably about 40 to about 98 atomic %, more preferably about 50 to about 98 atomic %. Although the DLN may theoretically be prepared without any hydrogen, the hydrogen content is preferably at least about 1 atomic % and up to about 40 atomic % of the carbon concentration. The sum of the silicon, oxygen and dopant elements and dopant containing compounds is greater than about 2 atomic % of the DLN. In one preferred embodiment, the ratio of carbon to silicon atoms is from about 2:1 to about 8:1, hydrogen to carbon atoms is about 0.01:1 to about 0.4:1, silicon to oxygen atoms is about 0.5:1 to about 3:1, and dopant to' carbon atoms is about
0:1 to about 1.5:1. Therefore, in the DLN network, for every 1 part carbon, there is from about 0.01 to about 0.4 parts hydrogen, from about 0.125 to about 0.5 parts silicon, and from about 0.0375 to about 1.0 parts oxygen. In such a scheme, if a third dopant network were present, for every 1 part carbon, there would be from about 0.01 to about 1.5 parts dopants depending upon the desired characteristics to be imparted to the Me-DLN network.
The low intrinsic stress found in the DLNs contributes to their wear-resistance and low coefficient of friction properties. By contrast, known DLC films typically possess high intrinsic stresses, and as a result usually suffer from pin holes and overall porosity. In addition, intrinsic stress limits the thickness of the DLC layer that can be achieved, as delamination or other adherence problems will often
occur. Due to the comparatively low intrinsic stress present in DLN films and coatings of the present invention, such coatings are pore-free, and therefore resist wear and permeation. As shown in Example 4 and Figure 5, the DLNs of the present invention display both low intrinsic stress and yet high hardness values.
The presence of the glass-like silicon network, stabilized by oxygen, serves to prevent the growth of graphitic carbon at high temperatures, to prevent metal cluster formation in metal-con aining three-network nanocomposites, and reduce the internal stress in the nanocomposite structure and thereby enhance the adhesion to substrates. This appears to lead to superior adherence of the DLNs of the present invention to the substrate material. The DLNs adhere well to both metal- containing and non-metal containing substrates including metals, metal alloys, plastics, ceramics and combinations thereof.
As already mentioned, to improve the adherence of DLC coatings, an intermediate layer, or interlayer, is required between the substrate and the DLC coating. Often, if the DLC coatings are too thick, delamination occurs due to high internal stress. Surprisingly, with the DLN coatings of the present invention, adherence is so good that an interlayer is not required. As a result, the DLN coating may be applied directly to the substrate, and more thickly without risking delamination from the substrate. The ability to apply a thicker layer of DLN coating results from the low intrinsic stress due to the Si-0 network, and is believed to contribute to the superior erosion resistance of the DLN-coated substrates. Of course, should an interlayer be desired for some reason, the DLNs may incorporate interlayers.
The environmental conditions faced by powertrain components are severe and extreme. Some components must face high speed revolutions or linear stroke cycles in excess of 5000 rpm. These moving components, as well as the stationary components contacted by the moving components, are subjected to extreme heat during operation, sometimes followed by periods of extreme cold. Operating temperatures may range from about -100°C during cold engine start up to about 350°C.
The adherence and wear-resistance of the DLN coatings of the present invention may be further assisted by the ability of the DLNs to have their characteristics, properties and performance tuned "tuned", or selectively altered according to the desired application. Such "tuning" is accomplished by incrementally altering the particular dopant as well as the dopant concentration. The DLNs may also have their properties altered when no dopants are included. Such changes in properties in the two-network system can be achieved by altering the deposition conditions in terms of temperature and pressure, etc.
The DLNs of the present invention have temperature stability far exceeding that of traditional diamond-like (DLC) materials. Crystalline diamond is stable to approximately 1100°C, upon which graphitization occurs. Quartz has long term thermal stability to 1470°C, and short term thermal stability up to 1700°C.
Traditional, non-alloyed diamond-like (DLC) films are stable only to about 250-600°C before graphitization occurs. By contrast, the DLN structures used to provide the corrosion and erosion resistant coatings of the present invention have long term stability to over 1000°C and short term stability to 2000°C. Therefore the thermal
stability of the DLNs exceeds that of DLCs while preserving the amorphous, diamond-like state.
Further, in the range of from about 600°C to about 1000°C, the chemical bonds of the carbon matrix of DLN materials partly change from sp3 to sp2. However, the general structure of the nanocomposite and their "diamond-like" properties are preserved. By contrast, under similar conditions, the usual "diamond-like" carbon (DLC) is graphitized and loses its diamond-like properties. In the range of from 400°C to 500°C (preferably 430°C) , a reverse transition is observed, whereby the ratio of sp3 to sp2 is increased.
The density of the C-H and Si-0 two network DLN varies from about 1.8 to about 2.1 g/cm3. The rest of the space is taken up by a random network of nanopores with diameters varying from about 0.28 to about 0.35 nm. The nanopore network does not form clusters or micropores . The properties of the two-network DLN may then be tailored by adding dopant. The dopants fill the nanopore network in a random fashion, eventually resulting, at a certain dopant concentration, in an additional network without clusters or microcrystalline grains, even at concentrations as high as 50 atomic %. At concentrations below about 10 atomic %, the dopants are distributed as separate atoms in the nanopores of the diamond-like matrix. The average distance between dopant atoms in this quasi-random structure can be controlled by the concentration of the dopant. When the relative concentration of the dopant element or compound reaches about 20-25 atomic %, the dopants form the third (Me-Me) network in the DLN structure as shown in FIG. 1(C) .
The electrical properties of the DLN structures of the present invention can be continuously varied over a wide magnitude (at least about 20 orders of magnitude)
from a highly dielectric state to a highly conductive state while preserving and improving the properties of the DLN state. A transition to a superconducting state, with the absence of electrical resistivity, is observed at low temperatures for certain three-network nanocomposite networks.
The DLNs may be synthesized via co-deposition by clusterless beams of ions, atoms or radicals of the relevant elements, where the mean free path of each particle species exceeds the distance between its source and the growing particle film surface, and each beam contains particles of well-defined energy. Carbon- containing particle beams can be produced by plasma discharge in a plasmatron and extracted as charged particles by a high-voltage field in a vacuum chamber and directed onto the substrate. At least 50% of the carbon- containing particles have kinetic energy above about 100 eV. The temperature of the substrate during growth should not exceed 500°C.
Figure 2 shows one preferred embodiment of the coating chamber used for the DLN coating deposition process. A vacuum deposition chamber 1 is provided to coat a substrate sample. A precursor inlet system 13, comprises a metal tube and a porous ceramic material 3 through which a liquid precursor, preferably a polysiloxane, is injected. The precursor inlet system 13 is shown incorporated into the chamber through the chamber base plate 11. The thermocathode 2 comprises a resistively heated thoriated tungsten filament 4. Substrates 5 to be coated with DLN film are attached to the substrate holder 6. The power supply 8 is used for biasing the substrates (DC or RF) . In practice the system is "pumped down" using normal vacuum pump down procedures. A gate valve (not shown) located on port 7 is closed and the system is backfilled with dry air,
nitrogen or argon until the chamber reaches atmospheric pressure. The door of the chamber, 9, is then opened and substrate to be coated 5 are attached to the substrate holder 6 using any of many possible methods (spring clip, screw, clamp, etc.) . Special fixtures may be required for substrates of special shapes. The substrate holder is designed in a way that it will also hold a cylinder sample (not shown) , which, in operation, rotates both about the axis of the central drive shaft 10, and its own axis which is perpendicular to 10. In this way, the axis of the cylinder would be perpendicular to the axis of 10.
When the substrates are loaded, the door of the chamber is closed, the chamber evacuated, and the gate valve opened to bring system pressure down to at least 10" 5 to IO"6 Torr, which is the desired range of system base pressure. When the above base pressure is achieved, argon gas is introduced into the chamber via a needle valve or mass flow controller, until the chamber pressure reaches approximately 5xl0"5 to lxlO"3 Torr, preferably about l-3xl0"4 Torr. The filament current, the filament bias and the electromagnet power supply are then switched on. The filament current is the current that passes through the thermocathode (also called the filament or the cathode) . The filament bias is the constant floating voltage applied to the filament (approximately -150V in relation to ground) . Plasma current is measured as the current between the filament and the base plate or ground. This voltage provides the field that moves electrons emitted by the filament to the base plate 11. The electromagnet power supply provides current to the electromagnet, which creates a magnetic field that results in the electron path becoming a spiral, increasing the electron path length and improving the probability of collisions between the electrons and the vapor molecules created due to precursor evaporation.
The substrate bias power supply is concurrently switched on.
Switching on these power supplies results in creation of an argon plasma, which is used to clean the substrates prior to deposition. After the required duration of cleaning, the precursor supply is opened. Precursor flow is controlled via a needle valve and occurs due to the difference in pressure between the chamber and the outside atmosphere. When precursor flow and vaporization in the chamber has stabilized, the argon gas flow is turned off. The ionized precursor vapors form a stable plasma, ions from which are accelerated towards the substrate holder due to the substrate bias . Thus, deposition of DLN film onto the substrate occurs.
Co-deposition of a dopant material is carried out as follows. Argon flow to the magnetron is commenced and the magnetron 8 is switched on after the base pressure has been reached. A shutter 12 is used to prevent deposition while the substrate is cleaned via sputtering'. When cleaning has been accomplished, the shutter is opened and sputtering is carried out at the desired power level. This may occur prior to commencement of DLN film deposition, during DLN film deposition, after DLN film deposition, or intermittently during DLN film deposition, depending on what kind of film structure and composition to be deposited are desired. Using DC or RF sputtering, materials of all kinds (metals, ceramics, alloys, etc.) can be used for co-deposition.
The growth conditions for nanocomposite films are as follows, with reference to FIG. 2. The pressure in the deposition chamber 1 should not exceed IO"3 torr, with the pressure in the active zone of the plasma generation 2, in the range from about 1.0 x IO"3 to about 5.0 x IO"2 torr. The temperature of the substrate should
not exceed about 200°C with the temperature of the cathode filaments being in the range from about 2100 to about 2950°C. The current in the cathode filament is from about 70 to about 130 angstroms, with the voltage across the filament being from about 20 to about 30 V. The voltage with respect to the ground is from about 70 to about 130 V with the plasma current being from about 0.5 to about 20.0 angstroms. The voltage of the substrate holder is from about 0.1 to about 5.0 Kv, with all the carbon- containing and Si-containing species having kinetic energy in the range of from about 100 to about 1200 eV and from about 25 to about 300 eV respectively. The metal beams consist of free atoms or monatomic ions. The kinetic energy of the metal atoms/ions does not exceed from about 25eV. With a precursor flow rate from about 0.5 to about 5.0 cc/hour, the growth rate of the DLN is from about 0.1 to about 2.0 micrometers (microns) /hour.
For the powertrain component coatings of the present invention, the useful range of thickness is thought to be from about 0.1 to about 50 micrometers, more preferably from about 1 to about 10 micrometers, and is most preferably from about 2 to about 5 micrometers.
The preferred range of operation for most applications is a pressure of about l-3xl0"4 Torr, a plasma current of about 1 amp., a filament current of from about 60 to about 75 amp., a substrate voltage of from about 600 to about 1000 V DC, or forward power of about 100 W in RF mode. The preferred frequency for RF mode is from about 90 to about 300 KHz. The preferred magnetron power depends on the type of material, composition and structure desired for the DLN coating.
In a further preferred embodiment, a preferred method of deposition uses a plasma discharge in a triode plasmatron, as shown schematically in FIG. 3, with the
plasma energy density above about 5 Kwh/gram-atom of carbon. The charged particles are extracted by a high voltage field in the vacuum chamber and directed onto the substrate 5. It is preferable that the potential of the substrate holder is from about -0.3 to about +5.0 Kv, and most preferably 1.0 +/- 0.2 Kv for DC and RF. In the RF mode, the frequency is in the range of from about 0 to about 25 Mhz for DC, and more preferably from about 90 to about 300 kHz for RF. The ratio of the electron emission to the carbon precursor flow in the plasmatron is from about 0.5 to about 1.5 electrons per particle.
Organosilicon compounds, such as siloxane, are preferred precursors for C, H, Si and O. One preferred organosilicon compound is polyphenylmethylsiloxane, containing-1 to 10 Si atoms. The high boiling point siloxanes may be introduced directly into the active plasma region through a porous ceramic or metallo-ceramic (3 in FIGS. 3 and 4) which is heated via radiation thermocathodes 4. The photon and electron emission of the thermocathodes affect the evaporation, fragmentation and ionization of the precursor molecules on the surface of the ceramic, which thereby functions as an ion source for the plasma generator. An alternative method for injection of the siloxane precursors is to use direct injection from a diffusion pump.
The formation of dopant-containing beams may be realized by any one of, or combination of, the following methods: 1) thermal evaporation; 2) ion-sputtering; 3) ion beams. - The dopant-containing beams are directed onto the growing film surface through the vacuum chamber to exclude interparticle collisions in the deposition chamber itself. Substrates are placed in an adjacent chamber on a rotating substrate holder, (for example a drum) which ensures double rotary motion, said adjacent chamber being connected to the plasma generation chamber
by an opening for the emission of the atomic or ionic beams, as shown schematically in FIG. 2. Alternatively, the plasma generation may be carried out within the chamber containing the substrates (FIG. 2) . A DC or a radio frequency potential is generally applied to the substrates during the deposition process. No external substrate heating is required. The substrate holder may be designed specifically to hold parts of different shapes such as cylinders, as would be apparent to one skilled in the field.
Useful variation of the above described methods for deposition of DLN films include the use of sputtered silicon and oxygen gas as precursors for the Si and 02, the use of sputtered carbon and hydrogen or hydrocarbon gas used as carbon and hydrogen precursors, or any combination thereof.
For deposition on non-conducting substrates, such as plastic, a method whereby a flow of neutral radicals is reflected from a high voltage target and directed to the substrate as shown schematically in FIG. 4. The process employs depositions similarly to those shown in FIG. 3, except that a reflecting electrode is used to generate a neutral beam. This process eliminates surface damage of the substrate resulting from charged and/or fast particles impinging on the substrate during growth.
A preferred method for depositing thin undoped
DLN films comprises ion bombardment (e.g. Ar+ or K+ with energy in the range of from about 30 to about 150 eV) through a vacuum chamber which has been backfilled by siloxane vapor (about 3 x IO"4 torr) . This results in a self-stabilized growth of a nanocomposite film, with the maximum thickness controlled by the maximum tunneling
distance for the relaxation of the charge of the absorbed radicals.
Therefore, the above-described flexible coatings of the present invention may be deposited on the selected substrate in thicknesses ranging from a few nanometers to a few microns, preferably from about 20nm to about 12 micrometers, depending only on the desired application of the coated substrate. The deposition may be tailored or "tuned" to meet the properties required for a particular application. The random interpenetrating of the two- or three-network DLNs guarantee uniform strength of the structures in all directions. The structures are free of micropores even at thicknesses of 80 Angstroms (8 nm) . The DLNs are therefore extremely stable and possess unique combinations of chemical, mechanical, electronic, and superconducting properties.
The following examples serve only to further illustrate aspects of the present invention and should not be construed as limiting the invention.
EXAMPLE 1
Coating of Samples with DLN
Square polished coupon samples of aluminum 6061 measuring 2"x2"xl/4" were coated with a three micron thick Zr-DLN coating. Several coupons were cleaned with alcohol and mounted in the coating chamber on the substrate holder plate using metal clips. The chamber was closed and pumped down to 3xl0"5 Torr. Argon gas flow was started and increased until the chamber pressure reached 1.5xl0"4 Torr. The magnetron shutter was closed and the magnetron power supply was switched on. Ion cleaning of the coupon began at a power level of 200 W.
After 20 minutes of cleaning, the filament current was turned on and increased to 65 amps while the filament bias voltage was -158 V and electromagnet current was 250 milliamps. Substrate RF bias was switched on to 100 W forward power at a frequency of 210 kHz. Substrate holder rotation was started and maintained at 7 rpm. After 10 minutes, of plasma cleaning of the substrate (coupon) , the precursor needle valve was opened to a setting of 3 cc/hour. After about 5 minutes, the argon gas flow was stopped. A plasma current of about 1 amp was obtained. Concurrently, the magnetron shutter was opened to a setting of 450 W. Substrate load power was about 80 W. After about 4 hours, the precursor valve was closed, and the power supplies switched off. The chamber and substrates within were allowed to cool . The chamber was then backfilled with nitrogen and the DLN coated aluminum coupons removed from the substrate holder.
EXAMPLE 2 Wear-Resistance
Coated aluminum samples are prepared as described in Example 1, except that only a 2 micron thick layer of Zr-DLN is deposited. The coated samples are subjected to a pin on disk wear tester. The rounded tip of a steel pin having a 10mm diameter is placed against the DLN-coated sample surface. No lubricant is used and a force of 1 to 5 N is applied to the pin. The pin moves at a sliding velocity of 1 to 400 cm/s. The pin motion on the sample is either circular or reciprocal. Friction is measured continuously by means of transducers attached to the drive mechanism. Up to 200,000 cycles, the samples would show no wear.
EXAMPLE 3
Thermal Stability of DLN Films
DLN films were tested for stability under high annealing temperatures, both in air and in vacuum. Thermal stability and maintained adhesion is required for the protection of weapon systems which undergo severe thermal shock and cycling under operating conditions. Annealing at 400°C in air for 9 hours followed by 600°C for 1 hour in vacuum did not change the film structure or degrade its properties. No film delamination was observed following the thermal testing that was carried out on DLN films. Analyses were carried out via visual observation, mechanical property determinations on annealed films and FTIR spectroscopy to determine changes in film structure. The values were found to deteriorate only approximately 10% after exposure to elevated temperatures of from about 400-500°C.
EXAMPLE 4
Mechanical Properties of DLN Films - Hardness
High hardness and mechanical modulus measurements were obtained on 9 different compositions of DLN and doped-DLN films. Measurements were carried out using a nanoindenter (Nanoinstruments, Knoxville, TN) Hardness ranged from about 5 to about 21 GPa. Elastic modulus of approximately 60-220 GPa was achieved, which matches approximately the modulus of steel . Hardness/modulus degradation in the films was minimal after exposure to 500°C. The hardness of the doped and undoped DLNs were plotted as a function of compressive stress . Both the doped and undoped DLNs showed a material with higher hardness values for a given stress
than the sampled amorphous hard carbon (a DLC) . See Figure 5.
Many other modifications and variations of the present invention are possible to the skilled practitioner in the field in light of the teachings herein. It is therefore understood that, within the scope of the claims, the present invention can be practiced other than as herein specifically described.