CA2040039A1 - Floor covering with inorganic wear layer - Google Patents

Abstract

SURFACE COVERING WITH INORGANIC WEAR LAYER

Abstract of the Disclosure A surface covering is a laminate including a hard inorganic wear layer deposited on a support by a reduced pressure environment technique such as ion assisted physical vapor deposition. The support may be selected from metal foils, films or sheets and plastics, rubbers or mineral/binder systems. The preferred sup-port materials include organic materials. The wear layer is between 1 micron and 25 microns in thickness and has a CIE LAB value of total Delta E of less than 12. Preferably, the wear layer is deposited on the sup-port at a temperature of less than 175°C.

Description

003g SURFACE COVERING WIT~ INORANIC WEAR LAYER

Background of the Invention The invention relates to a surface covering.
~ore particularly, the invention relates to a surface covering having an inorganic wear layer which prefera~ly has been deposited on a support structure by a low pres-sure environment deposition technique. Further, the ~invention i6 direc~ed to a multilayered floor ~overing in which each l~yer contributes to the wear performance and installation characteristics and affects ~he per formance of the other layers.
Floor coverings having wear layers are well known in the ar~. Such wear layers protect the decoxa-tive layer of the floor coverings and lengthen the use-~ul life of the floor covering. With the exception ofceramic tile which are ri~id and must typically be installed on a mortar bed and metal floors such as steel plates, neither of which have a wear layer per se, inor-ganic material is not used as the wear ~urface of floor coverings. Inorganic materials are typically considerPd too brittle to be walked on; particularly if a "thin"
layer were to be placed over a flexible or con~ormable support layer. ~urther, low pressure environment depo-sition techniques have not been applied to ~he manufac-tuxe of floor covexings.
Reduced pressure environment techniques fordepositing films of hard inorganic materials include sputtering, plasma polymerization, physical vapor depo-- 2 - ~0~00~9 sition, chemical vapor ~lel~osi1,-",, i~,n pl;.;iJI~ and ion implantation. Hard inorganic materials which can be prepared using these techniques include metals, metal oxides, metal nitrides and mixtures thereof; such as aluminum oxide, silicon oxide, tin and/or indium oxide, titanium dioxide, zirconium dioxide, tantalum oxide, chromium oxide, alumi~um nitride, boron nitride, silicon nitride, titanium nitride, and ~irconium nitride.
Often the-partial pressures of key gases in the deposition environment are controlled to ef~ect the properties and compositions of the deposited material. Therefore, a film formed on a substrate by reactive sputtering or reactive deposition can be a com-pound derived from a metal and a controlling gas, i.e., aluminum oxide produced by sputtering aluminum in oxy-gen. Sometimes the controlling gases are used to sus-tai~ a plasma in the deposi~ion environment. Ion assisted deposition is a technique in which the con-trolled gas is ionized and is used to bombard the depo-sition surface to modify the morphology and physical properties of the resulting film.
A critical review of vapor deposition ~echnol-ogy related to haxd coatings was presented by ~. E.
Sundgren and H. T. C. ~entzell in J. Vac. Sci. Te_h.
A4(5), September/October 1987, 2259-2279. ~ more com-= . . _ plete review of techniques involved in formation of thin films in reduced pressure environments is the book edited by J. L. Vossen and W. Kern, Thin Film Processes, Academic, New York, 1978~
Recent articles on thin film preparation include Clevenger, Lo A. t Thompson, C. V. and Cammarata, R~ C., AP~l. Phys. ~etter, 52~10), 7 March 1988, 735-797 on usj~ng commercial photoresicts as supports; and Journal of Materials Science Letter~, (1986), 177-178.
Patents dealing with thin film deposition include: U.S. 4,604,1Bl and 4,702,963.
Reduced pre~sure environment techniques have - 3 - ~ ~4~0~

been used to coat plastics materials such as plastic bags to improve gas impermeability. However, such coat-ings have been limited to about 0.5 microns in thickness.
~hile reduced pressure environment techniques have been used to form hard coatings on surfaces such as automobile parts, there has been no suggestion tha~ such coatings could be ~uccessfully used as wear surfaces for floor co~erings. I~ fact, suoh coatings tend to be brittle when applied in a substantial thickness. Thus, one skilled i~ the flooring art would nct expect reduced pressure environment deposited ~aterials to function adequa~ely as ~ floor covering or on other support sur-faces which are flexible, par~icularly in the thickness deemed necessary to protect the decorative layer of a floor covering.
Allian~e Wall manufactures and sells wall cov-erings in which porcelain enamel is fused to a steel sheet. However, use of a material as a wall covering does not suggest tha~ it would be acceptable as a floor coveri~g. Again, one skilled in the ~looring art would no~ expect a thin she~-t of ceramlc to wi~hstand the long ter~ abu~e to which flooring is subjected, par~icularly when laid over a resilient support structure and walked on by a woman in high heels.
Further, while reduced pressure environment techni~ues have been used to prepare protective coatings on plastics, the thickness of ~he prior art protective coatings generally do not exceed 0.5 microns. Typically this is because the deposition of hard coatings at greater thicknesses causes the temperature to exceed the allowable use temperature of the support. In addition, it is widely believed that although a hard inorganic coating on a polymer would provide some protection func-tion, the brittleness associated with hard materials, usually is believed to be a severe limitation. In fact, we have found that the ~rittleness is not a limitation, - 4 ~ 0 and have prepared materials that function superbly as protective coatings on organic layer~ or substrates.
SummarY of the Invention An object of the invention is to provide a floor covering that has the appearance retention of ceramic tile (~ncluding stain resistance and gloss retention), and resi~ts cracking ~nd brittle failure.
A fur~her ~bject is to provide a floor cover-ing having an inorg~nic wear layer which is ~lexible enough to be rolled arsund a reason~bly sized mandrel and therefore can be ins~alled in a manner similar ~o present resilient floor coverings.
Another object i~ to provide a floor covering laminate having the above listed features and which is conformable to the subfloor on which it is laid.
A still further object is to provide a surface covering including a support and a weax layer deposited on the support by a reduced pressure environment tech-nique. The wear layer is a 1 micron to 25 microns thick hard inorganic material.
Such a ~loor ~overing has been made by depos-iting a wear layer of a hard inorganic ma~erial on a ~uppor~ by a reduced pressure environmen~ techni~ue.
The pref~rred reduced pxessure environment technique i5 ion assisted physical vapor deposition; and the pre ferred suppor~ is multilayered.
Hard inorganic materials include aluminum oxide, tungsten~ steel, silicon oxide, Zireonium oxide and titanium oxideO Soft inorganic materials include aluminum, gold and copper. The hard inorganic materials from which the thin films of the present invention ~re formed have a Mohs hardness in their ~ulk form of at least 5 Mohs, preferably at least 7 Mohs and most pref-erably at least 9 Mohs~
The preferred hard inorganic material is a metal oxide or metal nitride; most preferably oxides, nitrides and oxynitrides of aluminum and silicon. Also _ 5 ~ 2~

preferable are oxides, nitrideq, and oxynitride materi-als that contain mixtures of silicon and aluminum. The mixtures may be homoyeneous or layered, and may also contain additional elements for the purpose of making processing simpler or less costly, or to enhance the appearance of the final layer. Aluminum oxide, silicon oxide and silicon nitride form films which are colorless, ciear and of hardness similar to the dirt to which the surface covering is normally subjected. The oxides and nitrides of ~he present thin films are not necessarily ~toichiometric but arQ believed to be close stoichiometric.
In one preferred embodiment ~he supports include a metal component such as a foil, a film or a sheetO The metal ~upport may be from 0.001~ to 0.25"
thick, preferably 0.003" to U.l" thick. The preferred support is a stainless steel sheet of at least 0.007 inches in thickness. Although a low carb~n ~teel may b~
used its performance is poorer. Pxeferably, the support includes a decorative layer of fused glass or ceramic frit ovexlying the metal component.
Since the glass or ceramic îs a metal oxide which can be deposited by a r~duced pressure environment technique, the wear layer can be formed from a ~lass or ceramic material. That i8, the decorative layer can be the wear layer.
Depositing a hard inorganic material on sur-face of a plastic, rubber or mineral/binder system sup port substrate improves the wear resistance and falls within the scope of the present invention. The plastic may be either a thermoset or thermoplastic. The pre-ferred thermoplastic is polyethylene terephthalate. The preferred thermoset plas~ic is a crosslinked reinforced polyester such as polyester sheet molding compound sold by Premix, Inc. The thickness ~f the support should be between 0.0005" and 0025".
An additional preferred support is a colorless - SA -Z~
transparent polymer film. The film may be either ther-moplastic or thermoset. The film may contain a backprinted image, or it may be pigmented for a decora-tive purpose. An additional preferred support would include the ~bove film that has been laminated to ~nother composite layered structure that may contain a visual image for which the wear layer on the film would provide enhance appearance retention.
Reduced pressure environment techniques are used on a large scale for prepara~ion of thin film alu-minum coatings on plastic webs. These are used for dec-orative purposes, as microwave susceptors in food packaging, as antistatic coatings in electronics packaging, and as vapor barriers. On a lesser scale, reduced pressure environment techniques are used to pre-pare other inorganic thin films on polymer ~u~strates for applications such as electroluminescent screens, security systems, vapor barriers, antistatic coatings, and as protective layers.
However~ in most applications J the thickness of the inorganic layer is limit~d to less than one micronO Speci~ic applications in which reduced pxessure environment techni~ues have been used to deposit thicker layers include the application of protective zirconia layers to aircraft engine parts; the application of early tran~ition metal nitride films to metal substrates for decorative purposes; and the application of metal nitride and carbide coatings to tool steel to extend usable life. For these "thick" applications, the proc~-ess temperature of the reduced pressure technique exceeds the use temperature o~ most plastics~ In addi tion, each of the support systems is a rigid material that has the characteristic of good impact, scratch, abrasion, or fracture resistance, or a high yield strength.
It is generally accepted that resilient, soft materials such as most plastics would not be capable of providing support for a thin hard inorganic layer. The usual argument is that any amount of 1ex in the inor-ganic layer would be sufficient to allow the inorganic layer to fracture. Furthermore, it is generally accepted that the fracturing would ruin the protective function of the film.
An aspect of this invention is that formation of a layered composite ~tructure containing an inorganic protective layer is~possible. Furthermore, even though the protective layer may contain cracks or fractures~ it still functions adequately as a pro~ective layer. In addition, if during its lifetime, the protec~ive layer becomes fractured, its function as a protective layPr continues~ albei~ at a slightly lower level specifically regarding transmission of gases and fluids. ~owever, overall the performance of the protective layer exceeds ,that of other currently known protective layers such as organics and organic/inorganic composites, and it exceeds the performance of thin (less than one micron) layers of the same material.
Another property of this invention is the adhesion between the inorganic l~yer and the polymer film. It is generally accepted in the coatiny i~dustry that application of an inorganic layer, especially a nitride, oxide, or oxynitride, especially either eva-porative or sputtering technigues, will result in forma-tion of an inorganic layer that has an intrinsic stress. It is also generally accepked that as the thickness of a deposited film exceeds 0.5 micron, the stress builds to such a high level that spalling or flaking of the coatiny`occurs, Thus, this invention demonstrates that a well adhered coating of a metal oxide can be prepared at a thickness between 1 micron and 25 microns and still remain adhered to a plastic substrate. The general s~ruc,ture that has been invented is a thick (preferably 1.0 to 25 microns, more prefera-bly 2 microns to 25 microns, and most preferably 3 ~o~

microns to 15 microns), well-adhered, inorganic oxide, nitride or oxynitride, on a polymer support.
The specific ~pplication we have demonstrated good performance in is the use of this material as a floor covering. It could generally be used as any sur-f ace covering. Examples would be as counter or desk tops, windows~ autom~tive parts, textile protection.
The material could be used as an abrasive like sand-paper, or it could be ormed into a tread for use as a woven material or fcr use as an abrasive string ~like a weed wacker). ~he $abrics made from the material could be used in flame retardant applications or for use in other safety related applications like protective aprons t covers or gloves.
Brief Description of the Drawinqs Figure 1 is a perspective view of a first ~embodiment of the present invention~
Figure 2 is a perspective view of a second embodiment of the present invention.
Figure 3 is a perspective view of a khird embodiment of ~he present invention.
Figure 4 is a cross-sec~ional view ~aken along line ~-4 in Figure 3.
Figure 5 i9 a cross-sectlonal view of a fourth embodiment of the present invention.
Figure 6 is a schem~tic representation of test setup to measure rupturc strain.
Detailed ~escription of ~he Invention Broadly, the invention i5 a surface covering having a hard inorganic material wear layer and a sup-port including metal or plastic. While the preferred floor covering is a flexible laminate which has been deposited on a ~upport by a reduced pressure environment technique and which permits installation similar to con-ventional resilient flooring, including resilient tiles;
the invention is intended to include rigid floor cover-ings having a wear layer of reduced pressure environment - 5D - Z O ~

deposited hard inor~anic material, and conformable floor coverings having a glass or ceramic material applied to a metal support by means other than a reduced pressure environment technique.
The invention also includes surface coverings, in general, which have a thick, hard i~organic wear layer deposi~ed on a support by a reduced pressure envi-ronment technique. Preferably the wear layer is 1 micron to 25 micron~ in thickness, preferably at least 2 microns ~hick, and most preferably 3 microns to 15 microns thick.
~n ~ne preferred embodiment, the substrate is a therm~plastic su~h as a polyester or polyv~nyl, or a thermoset, ~uch as a crosslinked polyester or epoxy, which have relatively low melting points or temperatures at which the substrat~ material is destroyed or degraded~ Therefore, it is preferred that the wear layer be deposited at a temperature of no greater than 175C, more preferably no greater than 150C, and most preferably no greater than 100C.
At these tempera~ures, those skilled in the art believe the properties of the deposited layer have been degraded to an unacceptable level 7 ~owever, it has been found that the deposit~d layer retains sufficient properties to function as a wear layer and even at the thickness of greater than 1 micron the preferred layers remain substantially transparent and colorless.
Color is measured by CIE LAB tests. In the preferred embodiment total Delta ~ is less than 12; more preferably less than 6, and most preferably less than 3.

.. - 6 ~

Metals and hard inorganic ma~rial~ ~uch as eeramlcs have unique ~ropert~es~ Prop~rly ~elected ceramic~ are hard enough to resist bei~g ~cratched by the grit particles in dirt. Properly ~elected metals S should be hard enough to ~upport the ceramic and yet be-flexible. SUch a laminate ~an be made ln an atomistic deposition chamber by depositing on a th~n, properly tempered ~teel. This lamin~te could then be mounted on an organic-polymer ~upport ~ayer to form a flooring structure. The ~upport lay~r ~onfor~s to the subfloor irregularities and accommodates lateral movement of the subflooring str~cture. Although v~cuum te~hnigues rould be used in making ~uch a flooring structure, current te~hnology would enabl~ lt to be made on a continuous, air~to-air production line~
No organic sur~ace, ei~her currently ~n exis-tence or envisioned, posse ~ee suffi~ient resi3tance to loss of gloss and to other physical damage to fully meet desired pexformance. Thick ~1/4 ~nch), hard ceramic tiles (Mohs hardness of at least 7 and preferably ~5) resist loss of gloss and other phy~ical damage extremely well.
The Mohs hardnes~ of,grit particles in dirt probably range~ between 6 ~silicatea) and 7 ~silica).
rule of ~humb among tribologist~ ~ that if a ~urface is 1.5 Moh~ unit~ harder than A gr~t partlcle, the ~ur~ce w~ll not be ~cratched by the grit p~rticle. Thi~
applies when the grit particle is between two ~urfaces of equal hardness. In a flooring si~ua~ion, the grit particle is usually between the relatively ~oft bottom surface of a shoe and the floor surfaoe. There~ore, the maximum downward force on the grit particle is the resistance the bottom of the shoe offers to penetration by the grit particle. The softer the bottom of the shoe, the less downward force ex~rt~d on the particle.
Consequently, the difference in hardnes~ between the grit particle and the flooring ~urface may not need to - 7 ~

be quite as large ~8 1 . 5 Moh~ units . In any case, a Mohs hardnes~ of 8~5 is a reasonable goal for the cera-~ic film. However, wear layer of Mohs hardness of about S or 7 have b~en ~hown to retain gloss level despite larger scratches. Prior art organic wear layers have a Mohs hardness of less than 3. Therefore a Mohs hardness of 3 or greater will yield an improvement.
If formed by atomistic deposition, the ~eramic-film wearlayer envisioned for the laminate structure would be expected to be essentially stain proof and to retain ~ts gloss extremely well. The film would be expected to b~ essentially stsin proof because such films provide excellent corrosion resistance for the substrates on which they are deposited. The film retains its gloss and resists damage from grit particles because it can be made sufficiently hard, approaching the hardness of the grit particles in dirt, and may be 6upported on a support having proper stiffness.
Although ceramic film has both stain resis~
tance and 910S8 retention, its brittleness has prevented it from being used as a wear layer in a resilient floor covering. Brittlenes~ makes the ceramic film ~uscepti-ble to serious damage. However,.by combining the cera~
mic f~lm with a support ~uch a~ a sheet of metal or pla~tic wi~h the proper ch~racte~i~tics, a ceramic ~llm may be used. ~f the ~upport i~ su~iciently ~trong to give the floor coverlng ~he abil~ty to support a uniform 125 lb~/~q ft load with a defle~tion of not more than one-flve hundredths of the span, the floor covering may 130 be fre~ ~tand ~ g.~ ~ ~ile does noS have ~ abi-a~ to ~ r~ ~ Ch ~*~o~t~ e 3~ ~q ~
r ~ O lity~. L~mlnate must have the necessary physical proper-,4A ties as discu sed below.
In ord~r to understand why such a laminate should ~olve the problem of brittle damage, i~ is useful to divide the type~ of force~ causing damage int~ two categories: ~1) localized preYsure and (2) impac~.
Localized prassure occurs when a grit particle i~ pre~sed downward against the ceramic surface. If the particle can force the ceramic ilm down into the sup-port layer on which it has been deposi~ad, the ceram ~ ~ ~ ~ 3 9film i3 put into tenBion And ~ail3 . Cerami C~, ~lthough strong in compression, are weak in tens~on. To ~void such failure in ten~ion, the ~upport layer must resist 5 being indented when the grit particle i8 pressed against the ceramic film. Actuslly, all the ceramic film does in pro~ecting the support layer from indentation i~ to spread the force over a grea er rea before that force reaches the support layer. ~ardened steel appears to combine the desired hardnes~ ~up to a Moh of almost 7 and flexibility. Although lackinq the hardnes of steel, some organic polymer6, particularly engineering polye ters, ha~e provided adequate support.
A ~eramic/me~allic lamin~te also possess the properties needed to re~ist lmpact. I~pact occurs when a heavy ob~ect ~trikes the floor. Da~age is mo8t likely to occur when the pressure ttbat 18 ~orce per unit area) i3 large enouqh to cauQe an lndentat~on. ~ere the ten-sile strength of the ~teel should resi~t puttin~ the ceramic in tension.
An additional proper~y that the support layer should pos~es~ is the ability ~o produce a ~raduai con-tour rather than an abxup~ conkour, both when a grit particle exert~ a ~orce on it through the ~eramic f~lm 25.. and when it is ~ubjected ~o impac~. C~lculation~ nave shown that for a giv~n vertical displAcement, a gradual contour sub~ects the ceramic film to.les~ tension than do~ the abrupt contour. ~n order ~o produce a gradual contour, the support layer should be flexible but not 30 limp. Two materi31~ ~ha~ po~sess the desired properties are properly tempered epring ~teel and polyester based ~heet molding compound~.
The ceramic ~ilm 3hould have hardnes~ at least of about 6 Mohs and good.strength. To pos~ess the~e a~tributes, the ceramic must have the proper mlcros~ruc~
ture. In ~lms formed by ~tomis~ic ~eposition, d~sira~
ble micro~tructure can be attai~ed by ~ncreasing the temperature ~nd the bombardment energy. One of the advantages of using a ~teel ~upport ~ayer iB that a high enough temperature can be u~ed to get optimum microstxucture.
The ceramic film ~hould be applied ~o that it is under ~ompre~sion. Thi~ ~an be accomplished by depo-siting the metal-atom port$on vf the ceram~c ~ir8t and ~
~hen adding ~he other element later, either in the ~ame step or in a second step. Using a two-~tep proces~
allows better control for depo~ition of the nonmetallic atoms.
The ceramicimetallic lamina~e 1~ pre~erably adhered to a conf~rmable ~upport lay~r. Thi~ support layer must be hard enough to ~upport the ceramic/
metalli~ laminate but must ~180 be able to conform to any irregularities gn the ~ubfloor. To perform in a superior manner, the conformable ~upport layer should be capable of inelastic deflection, ~.e., capab}e of perma-ne~t deflectlon with or without residual fvrces ~uch as appli~d by adhesive~.
In addition, lf used in re3ilient sheet ~oods it must accommodate ~ome lateral ~ovement of the ~ubfloor. To be able to perform over all subfloor~
including particleboard, the floor covering ~hould h~ve a rupture s~rain in exce~ of 0.3~. Due to ~ea~onal change~ including temperature ~nd humidity~ particl0board subfloors expand ~nd contract ~bou~ 0.3~ during the year.
Plywood expand~ and contracts ~bout 0.15~. Therefore, to per~orm adequately ove~ a wooden ~ubfloor, the floor cov-er~ng including the wear layer should have a critical buckle strain of at l~a~t 0.1~ and pre~erably at l~ast 0.3%. Floor ~overings of the present ~nvent~on having plastic support structures meet this requirement.
~ he ~upport layer pre~erably i~ typlcally made ~rom an organic polymer. It i~ desirable to ~elect the polymer 50 that its v~coelastic character will allow it to conform to the floor and still ~nable it to resist indentation by a rapid impact.
Surface contour~ can readily be incorporated by embossing the metallic substrate layer before appli-cation of the ceramic film. Incorporation of a pa~tern could be done most readily by printing the pattern on the metalllc substrate before depoRltion of the cera~P~03~
film. Some of the ceram$c fllm~ thAt c~n be depo~ite~
atomis~ically nre colored, and they may be applied in patterns by use of ~tencils.
Although the focu~ of thi~ invention i~ on atomistieally deposited ceramics, the concept of a thin flexible metallic substrate layer could be used wit~
other type~ of ceramics. Colored ce~amic glazes vr inks used in conventional ceramic technology could be applied in a pattern on the metallic sub~trate layer to ~orm a wearlayer in place of ~he atomis ically deposited cera-mic film.
The ~asic concept i~ ~ombining thin, hard wear surfaces with decoratlve, ~upport fitructure~ o produce unique wear-resistant flexible floor~ng product~. The flooring products have the appearance reten~on appoxi-mating that of ceramic t$1e but ~re light weight ~nd easier to install.
A ~eries of ~norganic oxides and nitrid~s ~including aluminum oxide and ~ilicon oxide) h~s been used as the thin, hard inorganic we~r layer. ~he vari-ety of material3 uséd for the support layers include combinations of metals, plastic~, rubber and mineral/
binder systems. The mean~ of deccration include gla~
frits, holograms, sublimable dyes and pigmented inks.
The pla~tic~, rubber and mineral/binder sy~tems may be through color. Out~tArlding performance has been demon-~trated in an embodiment con6isting of three micron~
aluminum oxide over en micron~ gla~s decorative layer on ~even mils ~empered steel ~him ~tock bonded either to a ~illed vinyl tile or a layer of rubber and in ~ubli-mable ink decora~ed polyester sheet ~olding compound ~PS~C). ~luminum oxide coated ~SMC resi~ts ~cratche~
better than any organic or organic/inorganic co~ting tested.
Since each layer of the ~loor covering lami-nate affects performance, æ l~yer of rotogravure ink will change the appearance retention of a wear layer on ~, ; . . .

O~ J~
a plas~ic support. Therefore, inks, ~uch ~s ~ublimable Ink~, whl~h will diffu~e $nto the ~upport layer are pre-ferred.
The advantages of the flooring products of the present invention include an Rppearance retention ~n traf~c envir~nment~ in a pro~uc~ whlc~ can be llght in weight, which can be either rigid or conformable, which can be thinner than products currently ~n the ~arket place, which c~n be flexible, which can be more resil-ient than ceramic tile, and which can be in~talled withconventional resilient-~loor~ng tools.
One pre~erred embodiment of the floor covering 1 is shown in Figure 1. The support 2 is a metal, plastic, rubber or mineral~binder ~ystem. A wear layer 3 of hard inorganic material i~ depo~ited on the support by a reduced pres~ure environment technique. A decora-tive layer 4 i~ deposed between the ~upport layer and the wear layer. The preferred metal i~ ~tainless steel.
While such metals as ferroplate, bras~/ferroplate, steel/ferroplate, chromium-plated ~ra3s and 01 steel have been used, any flexible but stlff ~upport can be used.
The pre~erred thickness of the ~upport i8 ~rom about three to about nine mil, moa~ pr~fer~bly about fiYe to ~bout seven mll. Two and ~our micron alumina wear layer3 on ~hree, ~ive and seven mll tempered ~hlm steel did not crack even when the resulting laminate was supported by a deformable rubber of Shore hardne~ 70 and walked on by women in high heel~. The three-mil substrate could be pierced by high heels.
The preferred modulu3 is abou~ 3 x 107 lbs./inch2. A modulu~ of thi~ valu~ or less ensure~
~ that the laminate is sufficiently flexible to bend around a 2-inch ~andrel wi~hout the wear layer cracking9 even when ~he wear layer is on.the convex side~
Prefer~bly, the floor coverlng i3 ~ufficiently flexible to bend around a 2Q-inch mandrel without cracking.

~ he 3upport ~ub~trate may also be ~ decorated or undecorated plastic, rubber or mineral/binder system provided the support layer i~ ~ufficiently rlgid. The support layers tested include ~ polyester 6heet molding compound (PSMC), rigid polyvinylchloride IPVC~ OD a tilë
base, polyethersulfone on a glass base, gl~ss fiber reinforced polyester, fiber filled phenolic, polyetheretherketone with and without a glass base, polyimide ~n a glass base, polymethylmethacryla~e, a photographic polyester on a gla~s base, Teflon, snd PVC
on PSMC. A preferred polyester ~upport substrate ~at~-rial is PSMC or polyethy~ene terephthalate. A fiber filled polyester is more ~table and yields fewer cracks.
The thickness of the wear layer must be at least one micron. Preferably the thickness of the wear layer is at lea~t about three ~l~rons. Thickness of les~ than three microns tend to fail more ~requently.
Hardness o~ the wear layer equal to ~nd prefo erably greater than that of ~il$ca al80 i8 desirable.
Preferably the hardnes~ i5 at least 6 Mohs~ and more preferably 8.5 Mohs.
The invention includes wear layer~ of m~tal, met~l oxides ~nd metal nitrides. The preferred composi-tions include A1203, SiOx, AlN, S~3N~ and TiN. Flooring structures with ~ive to eight m~crons o~ A1203 and ~iOx ~upported on an undercoæted, r~inforced polye~ter ~ub strate had gloss retention superior to currently mar-keted wear layer material~. ~lthough individually visible ~cratches were apparent, the ~cratches did not affec~ gloss retention. The scratches can be eliminated or a~ least minimized by obtaining a good match between the mechanical properties of the ~ubstrate and the wear layer. Gloss retention and overall appearance rete~tion is increase~ by in~reasing wear layer hard~es~ an~ ~ub-strate hardness. Therefore, Si3N~ may be a ~uperiorwear layer to A123-~3 -~ ~ ~ O ~ 3 The decorative layer 4 is a gla~ or ceramic frit, or pigment. ~he use of printable inks enables the creation of intricate design~. However, -~ince ~he wear layer materials may be colored, the wear layer ~nd de~o-rative layer may be combined and a multi-colored we~r layer can be deposited with a low pre~ure env~ronmen~
technique with the use of stencils.
The s~ructure o~ the ~igure 1 embodiment is acceptable for a resil~ent floor~ng structure which i~
rolled for ~torage and transport to the installation sit~, provided ~he laminate i~ ~ufficiently flexible.
However, if the flooring structure is a 12 x 12 inch tile having a rigid support structure, the tile may not b~ capable of conforming to the irregularitles of a wood subfloor and therefore may require in3tallation proce dures similar to ceramic or ~arble~
To overcome this di~advantage the lam$nate may be bonded to a resilient or conformable layer S a3 shown in Figure 2. The conformable layer 5 has dimen~ionæ
slightly greater than the l~minate. ~hi~ allows for the difference in thermal expansion between the subfloor and the laminate. The conformable layer i~ capable o~ lne-lastic deflectlon under gravi~ational ~orces so ~ha~
over a reasonable leng~h of tlme, the lower ~ur~ace of the laminate c4n~act~ ~he ~ubiloor over ~ubst~nt~lly the entire sur~ace ~rea. The conforma~le layer i8 capa-ble o~ conformlng to ~he contour of ~he ~ubfloor, including a l/lS" ledge between two plywood ~heets form-~ng the ~ubfloor.
~he ~harp corner~ of the ~igure 2 ~mbodiment may cause problems since the til~ cannot be laid ~n a perfectly flat plane. Therefore, the corner~ tend to sn~g the soles of shoes. ~o avoid this proble~, the lile may be formed a~ ~hown an Figures 3 and 4. The laminate of support ~tructure 2, decorative layer 4 and wear layer 3 ~ formed~ ~hen the laminate is press molded into a cup-~hape and bonded to th~ resilient ., .~

support base 6. The side~ 7 of the lamlnate are sub-stantially perp~ndiculAr to the pl~ne o~ the conform~ble layer and are adjacent the sides of the ~onformable layer.
In another embodiment shown in Figure 5, ~he conformable layer 8 ha~ alignment mark~ 9 on the upper exposed surface. ~he tiles 1 are bonded to the confor-mable layer in alignment with the mark~ to give ~ pleas-ing decorative appearance and A dlscon~inuous wear surface. The disconti~uities i~prove flexibility o~ the floor covering and may extend down to a micron ~c~le.
The following examæles, while not ~ntended to be exhaustive, illustrate the practice of the ~nvention.
Procedure for the PreParation o~ Vapor De~osited Coatin~s Coating ~aterials. ~etal~ ~nd metal oxides were obtained in 99.94 nomin~l purity ~rom standard industrial sources. Water was removed from ga~es using molecular sieve traps. al2o3 ~99.99~) And SiO2 (99.99 were obtained from E. M. Indu~trie~; ZrO2 S99.7~) and Ta20s ~99.8%) was obtained from Ceraç, Inc.t TiO2 ~99.9~, was obtained ~ro~ Pure Tecb, Inc.
pParatus. The deposition 8y5tem ~ D~nton DV-SJ/26) included a 66 cm wide high ~acuum bell-jar assembly~ ~ h$gh ~peed pumping sy~em ~CTI Cryoyenic~
CT-10 cryopump and Alcatel ZT 2033 mechanical p~mp)J an ~lectron-beam vaporizat~on Rourc~ ~Tem~scal STIH-270-2MB
four-hearth ~Supersource", with an 8 kWatt Temescal CV-8 high-voltage controller and ~-beam po~er ~upply and Temescal XYS-8 sweep control~; a resistively heated vaporization 60urce ~Denton Vacuum, 4kWa t); a cold cath ode ionization sourc~- tDenton Vacuum model CC101 with - ~oth CClOl~PS and CClOlPS biased ~nd u~bia~ed power ~upplies); a residual g~s analy~er ~Inficon Qu~drex 200); a qu~rtz crystal ~ype depo3ition rate co~roller ~Inficon IC6000); four eight inch circular deposition targets af~ixea to a planet~ry rotation sub-3y~tem7 ~nd a 10" diameter ~ta~nless steel aperature for ~ocu~ing the e ~eam ~or thermally) evaporated material and the io~ pl~sma on the same deposition ~urface. ~he YariOUS

204~)a'39 power ~upplies, pres~ure and ga~ flow monit-ors were operated either automatic~lly uning Denton'~ cu~tomlzed process control ystem, or manually. Typically, a depo-sition run began with an automated pump-~own pro e85, was followed by a depo~ition proce~ controlled by ~he IC6000 and ended with ~n automated venting cycle.
~ e~tion ~rocess~ ~he following general procedure was ~ollowed for all ~eposition runs.
Following evacuation to < 1.0 x 10-5 Torr ~he tempe~a ture of the chamber, as mea~ured by a cen~ered ~hermo-couple at planet level, was ad~u~ted to the desired deposition . emperature and ~he planetary rotat~on was ~tarted. Next, Ar gaQ was admitted to increase the chamber pressure to abou~ 1 x 10-~ ~orr, and A plasma 300-600 mAmps/300 6Q0 Volts wa~ initiated at ~he cold cathode source tcurrent den~ity ~etween 95 and 500 u-~mps/cm2) which was used to sputter-clean the substra-tes, in situ, for ~ive minutes. The deposition process was thereafter controlled by an IC6000 process which typically included parameters ~uch s heating rateg and times, mat~rial dens~ties, desired deposition rates and thicknesses, and the number of l~yer~ de3ired. Prior to deposition, ~he ~ub~tra~es were ~hielded from the met~l, or metal oxide ~ource. Ion bombardment with an ion plasma began and the shi~ld3 wer0 removed ~lmultaneou~ly when the IC6000 si~nalled that the ~et~l or metal oxide had been he~ted to the temperature approprlate ~or vaporiza~ion. A ~uartz crystal microb~lance provided input for the IC6000 eedback loop ~y~ em which provided deposition rate control for the remainder o~ ~h~ pro-cess. After deposition of a ~peci~ied thickness, the ion source was turned off, the shields replaced, and the vapor sources allowed ~o cool.
Rupture Strain ~es~ for Thin Ceramic Coatings One surprising fe~ture of ~he presen~ inven tion is the rupture ~train of the thin hard inorganic coatings of the present invention. Obtaining the rup-~ ~6 ~

ture strain of a thln, hard inorganic film or coating ~uch ~s a ~eramic 1Q a di~flculk task as the coat~ng 1~
not thick enough to be self-supporting to be te3ted with conventional apparstus. Among the properties of yield stre~s, yield strain, modulu~ of ela~t~city, rupture or.
ultimate strain and Poisson's ratio, the yield Btrain i5 of most lmportance as the wear layer will undergo strain as determined by the underlying load support ~tructure.
~ro create a support ~tructur~, it i~ necessary to deter-mine how much strain can be tolerated by the wear l~yera~d ~hen make design adjustments of the ~upport parame-ters so that thi~ str~in will not be attained in ser-vice.
Ceramics are brit~le and charac~eristically, the yi~ld ~train is close to, ~nd in a practic~l sense, : ~8 equal to the ultima~e or rupture ~tr~n. A ductile region does not exist between yleld and rupture. This condition makes the test more deflnitlve as rupture is more xeadily detected than yield, l.e., a crac~ is observed at the ultimate strain or rupture~
An eYaluative test for measuring the ultimate strain to brittle fracture in A thin ceramic film was developed. The test 1~ para3itic in that it relies on a host to produce the elongation ~train in the cer~mic 2S coating. A ~hin, highly tempered ~tael strip iB coated with A very much thinner coating o~ the wear layer ~ratio of thicknesses of 250 to 1). ~he ~teel ~trip i8 bent in a cantilever fashion and being ~o thick compared to the coating, it bending performance i~ not affected by the pre~ence of the coating. By measuring the deflection of the cantilever, the surface ~train of the bQnt steel can be calculated by ~las~ic mechanics equations. The coating will experience th~ same elonga-tion ~train as the surface o~ the ~teel. The beam is progre~siYely deÇlected increaslng the surface ~tr~in of the steel. ~hen the rupture 8tr2in of the coatlng i~
attained 9 the coating rupture~ by cracking which i3 ViS-`, .

ually evident. ~easurement of the deflection of the beam and the po~it~on along the beam where the crack occurred are 9uf f ic~ ent data to calculate the ~train when the crack occurred.
The ~redibility of the te~t i5 dependent upon the following items: ~1) the coating ~ust be lOU~ and adhered to the cantilever surface, 52) the deflect~on of the beam mu~t be amall to insure a~cur~cy with use of elastic beam formulae and ~ the yield Qtrain of the cantilever beam mu~t be greater ~han the rupture ~train of the coating.
The detection of a crack and ~ts position mu~t b~ accurately determined. Detec~$on of a crack ln a three micron transparent film requires scrutiny.
lS Observance at 40x magnification ~nd ~llumination by collimated light appear.~ to be neces~ary to discover the existence of a typical tension crac~-~igure 6 depicts the instrument 3etup to detect and measure the position of the rupture.crack~ in the wear layer coat$ng. The clamp 10 holds the specimen 11 in a horizontal reference plane indicated by dashed line 12. Micrometer 13 both de~lects ~n~ mea~ure~ the distance of deflectio~ Ye~ The crac~s 14 in the wear layer 1~ are observed wi~h the aid of microscope 16 and collimated light sourca 17~
~he l¢ngth of tbe beam ~nd lts thickness ~re in~er-related and w~de vari~tion~ of ~he two are po~sible. A lenyth of two lnches and a thi~kness of 0.030 inche~ has been found suitable ~or creating ob~er-vable strain cracking of the wear layer. ~he test pro-cedure is ~l~o usable in evaluating compre~sive surfa~e 5 rains by ~mply mounting th~ beam 80 that the bending places the coating in compression, i.e., invertingO ~he uni~ then deflects up, not do~n. The percent ~urface strain at pos~ t$on X, ex, $8 calculated by the ~ollowing formula:
ex - 3t(1 - x)ye 100%

~ 18 ~ 9 Test evaluat$4n o~ the ~ethod and instrumenta-t1on wa~ done on one hal~ ~ch ~de pecimens wi~h a standard coating of 3 microns o~ A1203. Sp¢cimens 1 to 4 were coated by the procedure ~et forth ~bove. A g~as~
decal was also fused to a 0.031 inch ~hick 302 9tainless steel strip to form Sample 5 which had a coati~g thick-ness of 10 microns.
Specific values of these coating operations are as follows:
Cr~c~ Ob~erved Specimen 1 ~~ Strain @ Rupture lnchinch ~Calcula~ed) 1 1.750.030 0.60 2 1.750.030 0.71~
3 2.500.024 0~56%
4 2.250.030 0~334 2.500.031 >0.58~

~wo factors that contribute to the high strain valu~ are:
1. The coating i~ not a slngle crystal as it is deposited in a l~yer foxm which builds in some form of vo~dsO ~his is evidenced by repeated measurements of depo~it density of 160 lb~. per cubic ft. aq contrasted to 247 lbs. per cubic ft. for ~ingle crystal ~apphire.
~he coating structure conceivably has more extensibility before rupture.
2. Thi~ test detects elongation strai~-to-rupture on the as deposited-coating. The deposited coating may not and probably ~s Dot residua}-~train-free. Other sources of information and papers on depo-~ition cite condition~ creating high compression or tenqion deposition strains. I~ the coating 18 deposited with compressive strains, these 8trains must be dim~n-ished to zero by bending be~ore actual tension ~trains are created. Thu~ if ~he coating were under compression - 19 ~

from deposition, thi~ test would ~ea ure the ~um of the residual compressive strain plu~ the ~ctual tenston strain to failure.
Samples 1 to 4 present a r~ng~ of a~-deposited strain-to-rupture of 003 to 0.7~. ~he v~riation of strain of several ~amples from any one ~oating operation hss been experimentally found to be ~ 0.1~. ~hiæ BUg-gest3 ~ha~ there were either var~atlon~ ln the coating structure or reaidual trains in the ~amples tested.
Analysi~ of he crack~ng behavior and pattern~
discloses characterist~c~ of t~e coatin~. The observed cracking has been ~instantaneous~ which i~ typical of a brietle ceram~c 80 that oDe can conclude that cracks will propag~te once started. The cr~cks for these ~am-ples produced under progressive deflection were all per-pendicular to the g~nerated t~nsile ~tress, were ~11 parallel, ~nd were &urprisingly uniformly ~paced one from the other. The spacing wa~ small av~raging four tenths of a mil ~part. This,indicates a ~ightly bonded, uniform coating a~ no delamination occurred and the cracking progressed in repeti~ive ~ashion.
~he crack~ in the sample~ 1 and 2 were evident in the deflec~ed beam but could not be observed ~at 40x) when the beam wa~ removed from the in~trument and returned to the flat condition. ~aving cracked and being a cQramic, the cracks cannot heal to a once-ag~in continuous ~urface. A machi~i~t's dy~ on the eurface did not ~ake the cra~ks Vi8 ible. Thi~ suggest~ that the cracks were pushed together tightly when the ~pecimen wa3 returned ~o flat and that there was no debris thrown off from the edge~ o the crack. I~ could be ~urmised that the coating was under residual compre~sion atrains when deposited.
xamples 1-1 to 1-36 The followiny are example~ of hard inorganic materials which have been deposited on variou~ ~ubstra~
tes:

. .

Table 1 Z0~339 Example Pilm ~u~atrate Thlckne~3 No. o No. Mat' 1. M~teri~ u) Pilm Layer6 1-1 SiO2 S51 ~oil 19.4 11 1-2 SiO2 SS fc~il ' 7. 9 1-3 ZrO2 SS foil 2.5 1-4 A123 SS ~oil O.,S
1-5 A123 SS foil 1.5 2 1-6 Zr02 SS ~oil 4.8 1-7 A1203 SS foil 5.4 1-8 A1203 Ferroplate 11.3 32 1-9 A12~3 Ferroplate 3.2 26 1-10 A1203 ~erroplate 6.7 ~2 ~ 23 Ferroplate <1~0 1-12 A1203 Cera~ic Tile 1.0 1-13 A123 Cerami~ Tile 32 1-14 A123 ~r2ss Ferroplate 1.0 1-15 A1~03 Brass Ferroplate 20 1-16 A123 Bra~/Ferroplate 29 1~17 A123 Steel/Ferroplate 1-18 A123 Steel/Ferroplate 20 1-19 A1203 Steel/Ferropl~te 29 1-20 A123 l/BnThick 01 Ste~l 1 1-21 A1203 1/8~Thick 01 8teel 29 1-22 A1203 1/8U~hick 01 Ste~l 32 1 23 A1203 TEOS2/Ceramic T~le0.1 1-24 A1203 TEOS/Ceramic Tile 0.2 1-25 A123 TEOS/Ceramic Tile 0.5 10 1-26 Zr02 on A1203 ~erro~teel 0.1 1-27 A1203 Ceramic Tlle 1.2 3 1-2B ~123 Br~ss Ferroplate 1~2 3 1-29 A1203 Bra~s ~erroplate 1.0 1-30 TiNX Ferro Steel 0O3 1-31 TiWx Ferro Steel 1.0 1-32 TiNX Ferro ~teel 1.9 1-33 SiO2 Ferro Steel 1.1 1-34 ~123 Marble 3-0 1-35 ~Nx ~arble 2~4 1-36 TiNX on Marble 1/3 2 .. ,; "

- 21 - ~(34~1C3~

l Stainless Ste~l 2 Tetraethylorthos11~cate Samples approximately 8iX inches 3quare were tested in th2 Wal~ers Test ~n which ~ix female walkers ~-reached a total traf f ic count of 1200 D
On matte ~inish, hard tmanufacturer' 8 r~tings of Mohs 6.5 ~nd 8.5) ceramic t~les, ~lu~i~um oxide coat-ing did not ~cratch t~ ~ eignificant ~xtent. Increa~ed damage occurred in ~ample~ where the aluminum oxid~ was depositea onto ceramic sub~trates with ~ohs hardness le~s than 6.5.
On hardt ~hiny ceramic tile, aluminu~ oxide performed well. On ~ofter, unglazed tile, the coat$ng appeared to provide protection against large scratches during the first half of the te~t, ~nd ~t the end of the test there were fewer ~but noticeable) ~crat~hes on the coated ~han on ~he uncoated 88mple8. The ~luminum oxide coating prevents the formatlon of haze tmult~ple fine scratches) cn brass ferroplate. On ferropl~te, applica-tlon of aluminum oxide at 140C produced a coating thatperformed as well as one applied at 250C. Th0 best ferroplate samples were ones ~oated when other types of ~amples were nst 1n the cha~ber.
When aluminum oxide was appli~d to a ~hiny ceramic tile that was essentially not scratched ~n it~
uncoated state ~and on which scratche~ prasent, could be readily seen), the coating performed almost ag well a~ the uncoated tile. ~he coated tile had two fairly large, almost scuff-like scratches but otherwise was essen~ially as good as the uncsated tile.
Under the same test condi~io~s, the coated ferroplate ~amples--although exh~biting complete resis-tance to multiple fine scratches--had a number of large ~cratches on them. The ferroplate samples with the mo~t ~cratches were tho~e prepared at th@ same time as ~sm-ples other than ferroplateD These re6ults hlnt ~hat the . j . ., ;, - 22 - ~ ~4 coating may be ~dve~ely affected by contaminant~ from the other sample~.
On 50~ter9 unglaze~ t~le~ the co~ting ~ppeared ~o protect ~he tile from large ~crat~h~ during th~
first half of the test. At ~he end of the t¢~t, there were fewer but more noticeable 3cratche~ on the coated, with coat~ng removed ~long the ~cratches.
Exam~les 2-1 to 2-B
Performance of vapor-deposited aluminum oxide was evaluated us~ng the Walker ~e~t. Under the~e test conditions, ~he alumlnum-oxide-coat~a ferropl~te ~amples with the thicker coatings were ~he be~t performing flooxing prototypes. The only samples to retain their gloss in all areas were those with vapor-depo~i~ed alu-minum oxide coatings at least 2.5 microns thic~ onferroplate. The principal damage to these ~amples con-sisted of medium and large scratches.
Samples approximately six ~nches square were tested in the Walker~ Test ~n which 8iX emale walker~
reached a total traffic count of 1~36.
Because the ~ample~ were only ~ix ~nche~
square, the walkers either placed a single foot on each sample or had to make a special e~fort to place both feet on each s~mple. It was ob~erved that ~hen they placed both ~eet on a ~ample, they u~ually placed their feet on diagonally oppo~i~e quadrant~ of the 3ample.
This produced on mo~t ~a~ple~ two ~reas which were much more worn than other area ~ee re~ult~ ~n Table 2.

. .

- ~3 ~ 0~9 ?able 2 Example Support Wear No. of A1203ThK.
~o. Substrate Layer Layers Total, u Commen~s 2-1 Brass A123 1 0.3 Purple-blue color;-many ~erroplate fine ~cxatche~ and very dull section~ thxoughout aample; A1203 ~ppeared ~o be removed by traffic ~ in 2 quadrants 2-2 Brass A1~03 1 0.5 Green to colorle~s; ~ome Ferroplate ~ine scratches, aome lar~er scratches, no dull areas A1203 appears to be intact __ 2-3 Brass A1203 1 1.0 Pink to colorles3; many Ferroplate fine scratches, A1~03 - partly removed ~ (uniformlY) 2-4 Brass A1203 5 2.0 Some fine ~cratches, Ferr~plate mo~t damage waS large~
sized scratche~; good _ _ ~loss re~en~ion ~
2-5 ~ra ~ A1203 2 2.0 More fine 8cratches than Ferroplate 2-4; some large ~i~e . _ ~cratch d~maqe 2-6 aras~ ~123 1 3~2 Almo~t no fine Ferrsplate ~cratche~, ~ome lArge ~ _. size ~cr~tches 2-7 Br~ss ~1233 ~ 2O5 Almost no fine ~erropl~te ~cra~che~, all damage 2 8 ~rass A1~93 1 3.2 medium to large ~ he performance of .alur~inum ox~d~-coa~ed f erroplate w~ th a coat~ng at lea~t 2 . ~ ~icron~ thiok wa~
35 superior to commercial wear layer~. The only ~ample~ o retain their gloss in ~11 the pivot areas were those - 24 ~

with aluminum oxide coatings at lea~t 2.5 microns thick on ferroplate. The principal damage to these ~amples consisted of a number of medium and large scr~tche~, each one of which is individually Yisible.
Indentations produced by spike heels on the :: aluminum-oxide-coated ferropla~e did not cause macro-cracking. Small parallel cracks were $ormed in the indentation but do not extend appreclably beyond the indentation.
Example 3 In this example, u~e of an ion gun duriDg A1203 deposition did not ~ignif$~antly ~ffec~ glo~
retention - for the~e flooring 8truc~ures. ~e/nonu~e o~ the ion gun dur~ng A1203 depositions on cera~ic 1~ decal/steel ~ubstrates generally has no s~gnif~cant effect on Walker Tes~ performance.
The performan~e level of the A120~ coated ceramic decal decorated steel ~tructure~ was llm~ted by the ~palling of the ~eramic de~al at its lnterf~c~ with the steel ~upport. Three-layer cer~mic decal samples on 7-m~l ~teel h~d fewex ~cratches than the coated single layer cer~ic decal/steel samples. ~he triple-decal ~amples were more ~everely ~arr~d due to thelr greater tendency toward spalling~
S~mples approximately ~1x lnche~ ~uare were ~ested in the Walkars Te3t. ~able 3 li~t~ aver~ge glo8s readings.
A1203 wear layer~ were ev~porated by the e-beam gun without the use of crucible l~ners. The cham-ber was baked out at 250C ~or 1 to ~ hours prior to each deposi~ion to minimize water vapor contamination.
The substrate temperature wa~ ~llowed to ~float~ start-ing at 30 - 90C durinq the depositio~ run~. For ~epos-iting done without the ion gun an 2 atmosphere of approximately 2.3x10-4 Torr wa~ m~intained.
The Decal used wa~ ~A2894 with ceramic overgl~ze color~, obtained from Philadelphia ~ecal. The teel was 7 mil ~tainle~s s~eel, obtained ~rom Lyon Industries.

- 25 - ~ 0 ~ ~ 0 Tab1e 3 Single ~ecal Single Decal Triple Dec~l Triple Decal No. of Ion Gun No Ion Gun Ion Gun No ~on Gun Passes Initial Final Initial Final In~t~al Final Init~al Final ~ 81.3 ~ 81.1 - -- 96.6 ~ 96.9 24 84.5 75.8 83.4 72.3 97.1 92.3 99.4 93.5 48 84.1 83.8 8146 78.6 96.4 96.0 96.4 g7.8 1~2 79.0 77.2 ~2.3 ~2.7 96.0 97.9 94.7 95.1 204 85.2 83.4 ... 83.1 83.2 97.9 97.2 95.9 95.4 ~02 75.1 76.8 75.9 73.9 9~.5 ~8.~ 95.8 95.0 8~4 80.5 79O9 8~.0 78.B 95.8 98.7 96.4 91.3 120~ 80.5 81.~ ~1.4 76,6 98.2 88.3 100~0 91.

Exam~ele~ ~ 1 to 4-23 Evalua~ions were made of (a~ alumina on a stiff but flexible sub~trate, (b) coatings prepared with and without the ~on gun, and ~c) l~yered ~oatings.
Alumina (2-4 ~icron3) on ~ flexible but ~tiff substrate (3-, 5-, or 7-mil tempered steel) did not crack in the Walkers Te~t when (1) the resulting l~mi-nate was supported by ~ deformable xubber ~8hore hard-ness 70) and ~2) even when high heel~ were included in the Walkers ~e3t. The laminate re~ist~d ine ~cratches, in ~ manner s~mllar ~o ferroplate te~te~ earlier/ but the ~everity of individually vi~lble scratches was 2S accentuated by failure 3f adhe~ion. ~he failure wa~
not, however, between the ~ubstrate ~nd coating but rather between the ubstrate 2nd a purplish layer ~hat was formed on the substrate~
The performance ~n the Walker~ Test o~ alumina ~0 an ferroplate was greatly improved by use of the ion gun during deposition.
The tandard, single-layer, alumina coating retained its appearance better than any of the layered coatings. The lB-layer chromium/alumina coating was a brilliant magenta.
In Table 4 ar~ ted the sub~trates and com-ments on the appearance of the ~amples after traf-ficking.

. ~:

~oo~

Table 4 Total Thickness Exa~ple Support Film No. of SSEM~
No. Substrate La~er_ Materi~l Layer~ rlicrons) _ Commeht~ __ Control 5-mil Shim Silicone Uncoat~d Matted. A number 4-1 Rubber o lndividu~l . 3cratche~. A few - __ _ heel dent~.
_ . _ _ . _ _ _ ~ , , . _ _ _ _ . _ _ _ . _ I Control 7-mil Shim Silicone Uncoated ~atted. A number : 4-2 Rubber of individu~l ~cratchesO No ~ ~ ~ heel dents.
Control 5-mil Shim Tile Uncoated Similar to above 4-3 _ 5-mil controlO
_ _ _ _ Control 7-mil Shim Tile Uncoa~ed Similar to above 4-4 7-mil control.
, 4-1 3-mil Shim Silicone ~1203 1 4.û Two-plece s~mple.
Stock Rubber No matting. A
number of indivi-dual ~cratches/
Som~ del~mination along center seam.
~;cratches ~ccentu-ated by adhe~lYe f~ilure. One heel 4 2 5-mil 8him Silicone A1203 1 4.0 N~ delamina~ion.
Stock Rubber No matting~ ~uch le~s scratching than ~xample 4-1~
Only a few barely di$c~rn~ble he~l dent~. Scratcbes accentuated by adhesive failure.

- 27 - ~O~L0~9 Table 4 t co~t ' d - ?
~otal ~hicknes~
ExampleSupport ~ilm No. of ~SEM, No. Substrate L~yer _ ~aterial LaYer~ microns) _ Comments 4-3 7-mil Shim Silicone A123 1 400 No delamin~tion, Stock Rubber No matting. Fewer ~cratches than .0 . Example 4~2. No discernibl~ heel I _ _ dent~. _ 4-4 Steel Tile A1203 1 4.0 No matting. A
Perro number of heel - dents. Number of scratches les~
thaD ~xample 4-2 bu~ more ~han ~-CLI
.
4-5 3-mil Shim Silicone A1203 1 4.1 Two-piece ~ample.
Stock Rubber No mattingA No delamination.
Slightly ~ewer scr~tches th~n ~xample 4-2.
Scratches accentu-~ted by adhe~ive failure. ~ number of heel dent~O
4-6 5-mil Shim Silicone ~1203 1 . ~.1 Similar to Ex mple Stock Rubber 4-2 4-7 7-mil Shim Silicone ~1203 1 4.1 No matting, no de-Stock Rubber lami~ation. ffany ~cra~ches whlch are accentuated by adhesive faîlure.
4-8 Steel Tile ~1203 1 4~1 Similar to Example Ferro ~4.

~, i ., To~ ~1 ~hickne33 ExampleSupport Film 2~o. of tSElt, No. _ Substrate Layer_ Materi~l L~yer~ mlcrons) Comment~
4-9 3-mil Shim ~ile ~12U3 1 2.1 No matting.
S~ock Slight delamina-tlon at n~ul ipl~
8craltches. Slgni-f$cantly more scra ches l~han Ex~mple 4-5.
S~r~'cche~ accentu-ated by a~he~ive ~ fAllureO
4-10 5-mil ;him Tile A1203 1 2.1 Similar lto Example Stock 4-6, but ~lightly f ewer scratches .
4-11 7-mil Shim Tile A1203 1 2.1 No matt~ng, no de-Stock , lan~in~tlon. Few-e~t scratches of any shim stock ~ample. cratche~
~ccentuated by __ Adhes ive f a i l,ure .
4-12 Steel Tile A1203 1 2.1 Similar to l:xan~ple 4-13 3-mil Shim Irile A1203 1 3.1 No matting, some Stock delamination.
Second most ~ratches .
4-14 5-mil Shim Tile ~1203 ~ 3.1 Mo~t sc:ratches of Stock any ~him stock r 4-15 7~mil Shim Tile A1203 1 3.1 More scratches Sto~k _ tharl E:xample 4-11.
4-16 Steel Tile Al;~03 1 3.1 S milar to ~erro _ __ Example 4-4.

'` 2~114~ 9 Table 4 ~Cont'd.) ~otal Thickness . ~ Example ~upport Film No. o~ ISEM, i o. Substrate ~aYer Material L~Yers icrons) ~ Comments 4-17 Steel ~ile ~12~3 1 3.0 8imilar to Ferro ~xam~le ~-4.
4-18 Steel ~ile A123 1 3.0 S1milar to Ferro Exam~le 4-4 - . _ - . _ - _ . _ r . ~ ~_ , '~ . __ _.
4-19 Steel ~ile A123 1 3.0 Large areas delam-Ferro inated ~before test). Delamina-tion along ~cr~tches.
4-20 Steel ~ile 8iO~A12035/5 2.0 8Ome matting, many 4-21 Steel Tile ~0tA1203 5/5 2.0 No ~at~ing but - Ferro many deep __ _ _scxatches.
4-22 Steel ~ile Cr/A1203 9/9 5 ~agenta. Worn Ferro thru on a pivot psint. Delamina-tion ~rou~d pivot Doint.
_ _ _ __... ~ ..... _ _ _ . .. ___ . _ _ , _ _ 4-23 St~l Tile ~ 1203 3/3 1.5 ~a~t~d area~.
Ferro Many scratches ln-cluding very fine cratches.

The A1203 metallic laminate wa~ suffi~iently flexibl~ that i~ ~ould be bent around a 2-inch mandrel without the A1203 ~racking, even when the ~1203 wa~ on the ~onvex side. The opti~um thicknes~ of the sub~rate layer appears to be 5 to 7 mils; the 3-mil ~ub3trate could b~ pierced by high heel~.
The alumina prevented the ~orm~tion of ~ine 3cratches on the shim steel. ~be severity of ~ndividu-- 30 - ~ ~4~3~

ally visible ~cra~ches was accentuated on the coa~ed ~ample~ by adhe~ive failure.
~ he use of the ion gun during depo~ition improved the performance of ~lumina on f~rroplate. The ~ample prepared without the ~on gun hAd many more scratches, significan~ ~dhe~ion failure along the scratche~, and an area a~out 1 x 2-1~2 $nches that delaminated before the t~st.
The 18-layer chromium/alumina coating was a brilliant magenta. ~he ~oating wa~ 5 microns th~ck, so thi~ situation was different tha~ one in which thin coating~ exhibit interference patterns.
The ~tandard ~oating reta~ned its appearance better than any of the ~ayered coatin~s (Example3 4-20 to 4-23).
Examples 5-1 to 5-15 ~utgassing during the deposition proces~ was demonstrated to ad~er3ely effect the scratch performance of A1203 thin f$1m The outga-~sing species was tenta-tively identified a~ water. This problem may be elimi-- nated by addition of a high temperature bake-out cycle to the depos~tion procedure. Outga~Ying wa~ ~hown to af~ect the ~cratch performance, and may greatly reduce scratch resi~tAnce~
For A1203 depo~lt~on, a 3-hour plate~u style bakeouk at 250~C suppressed the outga~slng ~ufficiently to prepare films which had reproducible scratch re~stance. rn the ab~ence of a bakeout, ~evere outga~sing occu~red wh~ch ~dver~ely affected ~cratch resistance in the A1203 films produced. The outgas~ing was probably due ~o thermal de~orption of w~er from A1203 on the wall~ of the depo~ltion chamber~ Direct id~ntification o the ou~gas~lng ~aterl~l ~u~t awa~
~n~tallation of ~ pres~ure adapter ~o~ the R~idual Ga~
Analyzer. If the ba~eout i~ not fea~ible due to thermal limitations of the ~ubstr~te mater~al, then the chamber ~hould ~e freshly cleaned a~d lined wlth new aluminum ~o~l immediately prior to deposition on that substrate.

,;

~o~

When the A1203 coated glas~ sub~trates rom deposition SERI~S A ~See Table 1~ were evaluated in the diamond stylus scratch test, two ~ajor observations were noted: both the Load to Incipient Damage tLID), and the type of damage at the LID changed from the first Memb~r of the series to tha last. The ~hange were not monoto-nic from the beginning of the seri~s to the end. For example, the first member of the serie~ tSpecimen 10) gave a LID of 50g due.to the ~ppearance of birefringeance along the ~crat~h track made by the dia-mond in th~ surface of ~he alumin3~ Scratching at load~
of up to 95g showed an in~rease ln the birefringea~ce, but at no point was any film delamination observed. In con~rast, Por the second ~2mber o ~he ~eries, birefringeance occurred ~t ~ LID of 40g; at 5Gg ilm delamination began and ~rack~ appeared normal to the scratch direction; and at 70 grams chipping was observ~d. For the third member of the ~eries, delamina-~ion and cracking were both ob3erved ~t ~n LID of only 25g, and film decohesion occurred at 40g. The rema$ning members of the serie~ were ~l~o charac~erized by low LID'~ due to delamination~ crac~ing and film decohe~ionO
~hese ~b ervations exempllfy a progre3~iv~ decrease ln adhesion between the vapor depo~lted A1203 and the gla~s ~5 substr~tes.

Tabl~ 5: Fllm Dat~ Phy~ic~l ~nd Mechani~al Properties S~RI~ A:
Example Numb~r 5-1 5-2 5-3 5 4 5~5 5-6 5-7 5-8 LIDa ~grams)50 40 25 30 15 20 15 20 Thickne~sC (u) 3.17 3.74 3.554.033.394.03 4.22 3.70 Wt. Dep. ~mg)15.0 }7.3 16.11.85 1068 1.69 1.69 1.64 SERIES B:
Example Number 5-9 5-10 5-11 LIDa ~grams) 45 40 45 Thickne~C (u)3.31 2.98 3.31 SERIES C:
- Example Number 5-12 5-13 5-1~ 5-15 : ~IDa ~grams) 50 45 45 ~5 ThicknessC ~u)4.08 3~65 3~70 3.50 ..
5 a Load to Incipien~ Damage- Damage in excess of ~imple indentation b Calcula~edg base~ on SEM thickness and a coated ~rea of 16.75 cm2 c Obtained by Scanning Electron ~icro~copy ~ERI~S A:
Eight consecutive deposition ru~s were per-ormed. ~n e~ch ~88e, ubstrate~ ~n addi~ion to gl~
substrates were pr~ent in the chamber. The~e sub~r~-te~ included Perrosteel~ 5 m~l spring steel, chromed spring steelr thick ~01~ ~teel plate ~both chromed and untreated), ~ta~nle~s ste~ nd ~everal eDglneering plast~cq. In all but two of ~he runs ~n hls seri~, the samples were loaded into the depo~itlon chamber in la~e ~fternoon of the working day be~ore the run. For the ~xa~ple~ 5-5 en~ 5-7, however, the sample~ w~re loaded into the depo~ition chamber i~ the mornin~ and the ~ystem wa~ ~llowed to pump down over the lunch hour.
SERIES B:
Three consecutive depo3ition runs were per-formcd. ~hese runs conta~ned only gla~ ~ubstrate~.
The procedure was the sa~e ~ th~t for 8E~IES A ~xcept that Example 5-11 W8~ subjected to a three hour bakeout cycle at 250C while pu~ping overnight.
SERIES C:
Four consecu~ive deposition runs w~re per-formed~ Thes~ runs cont~ined additional sub~trate~
~apable of with~tanding a 250C heat treatment. ~or each of these runs the procedure includ~d 2n ov~rnight bakeout at 250C.

33 ~ 2 Diamond stylus scratch te~ result3 are reported here as Load to Incipient Dama~e (LI~) to the nearest five grams of ~tylus welght loading. Because the mechanism ~f cratching h~rd inorganic material~
does not include any macroscopically observable "recovery" mechanism, Lo~a to Incipient Damage ~
defined as that weight loading, in the LOM e~uipped with : a 45X cbjective, whers damage o~her ~h2n a simple inden-t tion is observed. For example: tbe LID ~ay be due to the observation of birefringennce ~t the edge of the scratch ~rack, by del~mination of the f1lm, chipping, or the development of crack Density measurements were obtained by d~viding the weight gain of ~ Perrost~el ~lide by the area exposed for deposition ~16.75 cm2) and the film th~ck-ness as determined by SEM. Control experiment~ showed that there was no detectable weight lo~ due to sputter-ing even after 20 minutes expo6ure to a 600mA/600V Ar+
ion plasma. In addition, a Ferrosteel slide su~jected ~o the entire deposition cycle but shielded from deposi-t~on experie~ced no detectable weight change.
A clue ~nto the cau~e of these ~dhesive dif-ferences was o~ered by a qualitative compari~on of Ion Gun voltage dur~ng the flrst ew mo~ent~ of eeveral o~
25 the SERI~S A deposition runs. A high bomb~rdment volk-~ge was atta~ed immediately at the ~tart of the depo~i-tion run and the voltag~ was sustained throughout the run. However, voltage dropped at the onset of deposi~
tion, and progressively longer ti~es were required to 30 reach snd sustain an io~ voltage of 600 volts. ~wo - lmportant facts are associated with this observation.
Yirst, the ion gun voltage ls Inver ely proportional to the chamber pressure" Thus a volta~e drop is ~ccompa-nied by a pressure ~urg~. 5e~o~d, A1203 films prepared 35 using a high voltage ion assi~t e~utperform those pre-pared w;th no ion assist. Therefore, a pressure surqe accompanied by a ~ol~cage drop will ~dver~ely effect the wear per ormance of ~uch a f i lm .

o~

The progressiYe nature o~ the deterioration in LID performance suggested ~n impurity buildup d~ A func-tion of chamber u~e. Thu~ lt was propo~ed that exce~
alumin~ deposited on the chamber walls gettered water vapor from the laboratory atmosphere whenever th~ cham-~ber wa~ opened to install ~r remove sub~trates.
Aluminum oxi~e ls a well known de~iccan~ which i~ acti-vated by heat ~reatment in a v~cuum. Radiat~on from the e-beam evaporation ~ource probably UactivatedY alumina 10 which had accumulated on the chamber wall3 during previ-ous runs and caused the ob~erved pressure ~urge~.
Direct verif ic~t~on of thi~ hypothesi~ using the Residual Gas Analyzer tRGA) wa3 not possible becau~e of its pressure l~itation.
The irst indirect confirmation that water vapor was being desorbed was obtai~ed using the RGA
under predepo~tion conditions. The RGA, UpOD evacua-t~on of the chamber to a pre~sure of lO-6 Torr.showed a constan~ tuncalibr~ed) water vapor pressure of 5xlO-5 ~orr. When the guartz heaters ln the chamber were energized, however, an i~mediate pressure surge due to an increa~e in water vapor pres~ure w~s observed~
Un~ortunately, the cutoff pre~ure for the RGA i~ 10-4 Torr, which i~ the vapor pressure in the chamber during most depo~ition runs. Therefore, the RGA cannot be used during the runs to directly conflrm the water vapor hypothesis.
A ~econd i~dixect confirmakion of the role of water vapor during the deposition process was obtained 30 by examination of the scratch ~est result~ obtained from . dep~ition SERIES B t~ee Table l). ~he first tw~ depo-sitions in this ~erie~ were run on consecutive day~, under the ~ame cond~tlons ~g the fir~t two member~ of ~E~I~S ~. For the firs~ two depo~it~on run~ in both ~RIES A and SERIES B, trends ~howing ~ decrea~e i~
~cratch LID, and an increase ln voltage a~ab~ ation time was observed tthe mag~itude of the pres~ure ~urge was mitigated by the ~hamber operator ~ecxeasing the flow rate through the ion gun~. Addition of a bakeout cycle to the deposition procedure for the third deposi-tion run in 8ERIES B resul'ced in recovery of the ~cratch 5 behavior obssrved in the f lr~ ~embers of ~o~h ~ERIES R
and B, nd de~reased the time required ~o obtai~ a ~ta~
ble ion gun voltage.
SERIES C was run in order to test the repro-ducibility of ~cratch tests obtained from run~ which included the bakeout cycle. In contrast to 5ERIES A, no significant change in scratch perfGrmance from the beginning of S~RIES C to he end was observed.
~ he early moments of the depositlon runs in SERIES C were not accompanied by the voltage drop~ and pressure surges that were observed in SERIES A. Al o no change ~as shown in the type of scratch damage o~ser~ed at the LID.
Examples 6-1 to 6-14 Increa~in~ the thickness of the decorative layer improved the performance of the glass a~d ceramic decal~, both coated and uncoated except that o~ ~he 20-micron ~hLck gla8~ decorative layers. Coating the decora~ive layer wi~h ~luminum oxide improved the over-all appea~ance retention in all ca~es. The a$1ure mode ~or the glass and ceramic decal~ Appear~ to be dif-ferent. D$amond 8tylu8 scratch test~ show that the glas3 decoratiYe layer cru~ble3 under relatively high ~tylus load wher~ the ~ramic decorative layer chips.
~n previous Walkers Te~t~, the de~orative l~yer which ~on~i~ted of 5-mi~ron-thlck glas dec~
~howed large individually di~ern~ble ~cratches that broke through to ~he metal substrate. 8ince the decora-tive layer al~o ~upport~ the aluminum oxide layer a ~hi~ker glass l~yer ~hould provide be~ter suppor~.
Samples made by l~yering glas~ decals were run ~D ~he W~lkers Te~t to test thi~ ~dea.
The glass-ink decal has 3 nano-hardness of 6 o~

Gpa. The ~est aluminum oxide has a nano-hardne~s of 10 Gpa. Tnere exist cesamic inks which form harder decora-tive layers than the glass ink~. These ceramic inks have a nano-hardness value of 11 ~paO Decals made from these ceramic inks not only should provide better support for the aluminum oxide layer but conceivably could act as a wear layer itselfO Single and mult~ple layer samples were prepared to evaluat~ the effe~t of thicknes~ on performance.
Samples approximately 5iX ~nche~ ~quare were tested in the Walkers Test. ~he ~amples ~ere supported by a vinyl ba~e kile to which they were attached by adhesive transfer t~p~. Six walkers reached a total traf f ic count of 1200.
Before and after trafficking, sixty-degr~e glo~s measurements were made with the Mallinckrodt Glossm~ter. A measurement was made at the center and at the center of each of f our guadrants of the sample for a total of fi~e meaqurements.
Sample descrip~ions and glo83 valu~s are listed in Table 6. The glass decals were 5 microns thick. Th~ cerami~ decal~ w~re 10 microns thick.
Table 6 Sample Glo~ Values No. ~ L!~ Ele~L8~ Inltial Final 6-1 1 layer glass dec~l/7 mil 302 steel 70 60 6-2 Alumlnum oxide co~ted 1 layer gla~s decal~ 53 54 7 mil 302 6-3 2 layer glass decal/7 mil 302 steel 96 70 6-4 Aluminum oxide coated 2 lay~r glass decal/ 63 59 7 mil 302 6-5 3 layer gla~s decal/7 mil 302 steel 110 8 6-6 Aluminum oxide coated 3 layer gla~s decal/ 86 85 7 mil 302 6-7 4 layer glass decal/7 mil 302 ~teel B3 37 6-e Aluminum oxide coated 4 layer glass decal/ 73 58 7 mil 302 2~)0~3 Table 6 ( Continued ) Sample Glo_~ Value~
- No. SamPle Descriptions Ini tial Final 6-9 1 layerceramic d~cal/7 mil 302 Q~ceel 78 67 6-10 Aluminum oxide coated 1. l~yer ceramicdecal/ 7~ _ 72 7 mil 302 6-11 2 layer ceramic: decal/7 mil 3û2 steel 87 ~4 6-12 Aluminum oxide co~ted 2 l~yer s:eramic decal/ 90 90 7 ~il 302 6-13 3 layer ceramic decal/7 mil 302 3~eel 89 80 6-14 Aluminum oxlde coa~d 3 l~yer ceramic decal/ 92 91 7 ~il 302 Increa ing ~he th~s:knes~ of the decora'cive layer improved the performanc~ o~ the gl~ nd ceramic de~als, both coated and uncoated. The qample with 'che best appearance and gIos~ re~ention wa~ thç! aluminum oxide-coated triple-layer ( 30 miLcron) ceramic-decal sample. In general, ~he multilayer ceram1c decal3 resisted large scratches b~ter than the ~ultilayer glass decals.
Coa~ing the dec4rat~ve laye~r w1th aluminum oxide i~proved the ov6~rall ~ppearance and glo53 reten~
tion in all s:ases exc~pt that o the 5- ~nd 10-mic~ron thick glaa~ decor~'cive layer~ . The ~luminum ox~ de coat ing lmprov~d the 91O3s rete~tion of both 8y9'cem5 ~ with the coated ceramic dec~l haviny the be~t glos~ r~ en-t~on. On the ceramic decal~, the aluminuol oxide r@d~ced the number of large scra~ches. On the glass decals, the aluminum oxide reduced th~ number of small scratchesO
With the ceramic decals, som~ of the ~cratches appeared conf ined to th~ alumînum oxide ~:oating .
The s~eramic dec~ls appeared to adhere less well to he steel than ~ the gla~s decals.. At 10 microns, the ceramic decals resi~ted f irle scratch~s be~-35 ter than ~he glass decals but had more ~cratch~3~ to the metalO At 20 microns, the ceramic decal~ r0si~ted ~oth ~ 38 - ~ ~ 4 fine and large 6cratches better han the gl~8 d~cal~
but still had more ~cratche~ to the met~l. The ~h1pping around the area o~ the ~cr~tcheq in the ceramic decals, seems to $ndicate an adhesion failure, po~sibly due to d;ff~rences in th~ coefficlent of thermal expan~lon.
The failure mode for the gl~s and ceramic decals ~ppeared to be different. Diamond stylus ~cratch tests on the ~ame 3ample~ that ~ade up thi~ Walker~ ~e~t ~howed that ~t relatively bigh load~ (60-95 g~ams), the glass decal~ tended to ~rumble where the ceramlc ~ecal~
did not. The crumblinq decal left granule~ of material on either side o~ the ~cratch. In tbe multi-layer cera-mic decal samples any fa~lure noted could be de~cribed a~ a ~hipping failure. The seratc~ ~rom the s~ylus loo~ed similar to aluminum oxide scratche~ but h~d intermittent ~rea~.where the ceramic ~nk chips away from the re~t of the coat~ng. It appeared that the ceramic ink ~n the decorative layer had a ~reater inherent ~trength than the gla~ lnk. However~ when ~tres~ed to the point of failure, the cera~c $nk exhibited a brit-tle failure where the gla~s crumbled.

Addition of a ceramic primer to ~he co~po~ite s~ruc~ure ellminated ~he spalling of ~h~ decor~tlve layer ~een in previou~ walker testing. Damage wa~ lim-it~d to large, indivldually visible scratches and can be yroup~d intv three types~ ~a) damage to the A1203 layer only; (b) damage to ~he decorative layer7 and ~c) damage to the metal substra~e. There wa~ no deglo 3ing due to ~ine s~ratches. The two ceramic primers performed e~ually well.
Sample~ were te~ted in the Walker~ Test.
T~ble 7 l~t~ the ~ample dat~ and tbe gloss values as mea~ured. Two ceramic/~etal composite ~ategor~e~ were tested, ~h~y werer (1) A1203-coated ~eramie decal on ~34001 pximer o~ 7 mil 302 ~teel~ ~nd ~2~ A1203-coate~
ceramic decal on J-~600001 primer on 7 mil 302 steel.

- 3g - ~04 Tablc 7 203 50b Gloss Sample ~hk. 3 Walker Description No. (micron) CYcles Initial Final A1203/Ceramic Decal~
H34001 Primerl 7-1 4.8400 89.4 81.0 7-2 4.8B00 88.8 79.3 7-3 4.81200 B6.1 ~4.4 7-4 3.831200 8~.6 80O8 A1203/Ceramic Decal/
J-M600001 Primer2 7-54.34 400 98.2 91.9 7-6 3.83 800 94.2 ~0.1 7~7 4.34 120~ g7.5 90.0 7-~ 4.34 1200 95,6 89.7 1 Manufactured by ~eraeus, Inc.
2 Manufactured by Johnson ~atthey 1~ 3 Light optical microscope th~cknes~ de~ermlnation.
4 Average of four SEM me~surements.

Examples 8-1 to ~-19 Uncoated and A1~03-co~ked, 35 micron thick decora~ive layer ~mples had the ve~y good appesrance ~0 retentlon. A1~03~coated white H34002 primer t30 ~iCron) sa~ple~ were ~arginally bet~er ~ha~ ~he 30-mioron, uncoated whlte H34002 primer ~ples. Al203-co3ted 10-micron c~r~mic decal on 20 m~cro~ o~ white H34002 primer ~onta~n~d no scratche~ to ~he me~al subs~r~te.
Si~-inch ~quare samples were te~ted ~n the Walker~ T~st. Table 13 shows ~he sample de~cription~
and gives the raw data~ -~hree categorie~ of wear layer3 ~er~ preparedr They were:
1. 30 micron-H34002 prlmer on 7-mll 302-steel.
2. A1203-coated, 30 micron-H34002 primer on 7 302-steel.
3. A1203-coated, 13 micron-ceramic decal on 20 m~cron-H34002 primer o~ 7-mil 302~steel~

~O~L~O~
~ 11 samples tested h~d fewer scr~tches to the metal. AQ previously ~een, the pre~encQ o~ the primer coat had eliminated spalling o~ the ~eramic layer from the damage area. The damage of the A1203-coated decal 5 samples w~Y limited to khe aluminum oxide lay~r and th~
decal only.
~able 8 Example Thk. Walker 60 Glos~
Description No. tMicron? Cycles Initial Pinal 30 M;cron H34002 Primer 8~1 - 200 89.9 95~5 8-2 - 400 90.8 93.~
~-3 - 800 ~.7 91.9 8-4 - 1200 94.1 92.3 ~-5 - 1200 g2.2 91.6 8-6 - 1200 94~4 g3.7 8-~ - 1200 93.3 90.8 A1203/30 Micron ~34002 Primer 8-8 5.1 200 133.6 105.7 ~ 9 5.1 400 102.0 lOB.l 8-10 5.1 800 100.1 103.3 ~-11 5.1 120~ 99.1 ~0~.4 8-12 4.8 1200 103.3 106.9 ~5 8-13 4.8 1~00 1~2~ 105~6 8-14 4.8 1200 101.2 10505 8-15 4.8 12~0 1~1.5 102.7 A1203/Ceramic Decal/
20 Micron ~34002 Pximer 8-16 4.~ 200 87~6 96.2 8~17~.~ 400 85.8 94.0 8-18 ~.~ 8~0 9~.9 92.7 8-1~ 4.6 1200 89.6 94.5 xample~ 9-1 to 9-41 Structures wer~ ~abricated uslng Ion ~ssi~ted Physical Vapor Deposition tIAPVD) to deposit A1203 or - 41 - ~040~

si~ ~ceramic" wear layers onto undecor~ted pl~stic subs~rates. These s~ructure~ h~d the same ~verage gloss retention profile as ceramic tile. Scratch ~nd Walk~rs ~ests demonstrated the ~ynergist~c relationsh~p between the co~ting and sub3trate prsperties in these com- ~-posites. Nanoindentation showed r~latlonsh$ps between hardnes~, chemistry and the proce3~s used to prepare the wear layers.
Flooring ~ructures with 5-8 ~crons of A1~03 or SiOx ~upported by an undercoated, reinfor~ed ps~lyes-ter substr~te have gloss retention ~uperior to currently mar}ce~ed wear layer materi~ls. ~owever, indlvidually vi3ible scratches were apparent in these structure~0 Althou~h these scratcheQ did not af~ect glos~ retent~on, the post-traf~icklng appearance of the co~ted tructure~
would be ~mproved ~f all ~cratcbe~ were prevented. The key to that prev~ntion lies in obtaining a good match between the mechanical properties of the plastic sub-strate and the hard wear layer. In hese examples, the aluminum oxide coating provided only llmited improvement to the performance o any organic-containlng ~ubstr~te where ad~esion failure taluminum oxide removal) was a major factor.
Diamond s~ylu~ scra~ch te~ting ~nd nonoi~den-t~tion wer~ the two main char3c~erization tests tv moni-tor mecha~ical prop~rty re~ponse for the ~i~le structure3. ~he A1203 and ~ix supported by polye~ter ~heet molding compound ~PSMC) sh~w th~ h~ghest ~tylus ~SP ~oad to Substrate Pene~ration), ~hich 1~ con~is~nt w~th th~ ~uperior glos~ performance of ~uch ~tructuresO
Nonoindentation results show that A1~03 ~s the hardest wear layer material tested in ~n ~ctual flooring proto type in which a plastic ~upport was employed. ~i3N~ is suggeste~ as an alterna~ive material.
Ion Assi~ted Physical Vapor D~position ~IAPVD~
was used to produce films f~r wear layer on plastic substrates. Metal or metal oxide vapor was evaporated - q2 ~ o~

by heating with ~n Q~ectron beam until it vaporized.
When the vapor depo~ited on a sub~tr~te, simultaneo~6 bombardment by an ion beam helped to form a dense, defect free film. Materials deposited onto plaQtle S substrates include A1203, SiOX, AL203-SiOx, and SnOx-SiOx. Test 3tructures prepared by this teehnique are listed in Tables 9A; 9B and 9C.

IAPVD A1203 Wear layers on Non-decorated ?lastic ~ubstrates Sample Thickne~ LDP
No. Wear Layer tmicrons 3 SUDPOrt Sub~trate 9~23 1.5 PES2 2x tape/glass 15 9-2A123 4-9 None P5MC3 --9-3A123 6~0 None PSMC ~35 9 4~1203 ~~ PVC4 W~5 ---9-5A123 0.480 PVC WT 10 9-6A1203 0.52~ PYC WT ~~~
9_7Al2~3/siox 0.432 PVC WT ~-~
9-8A1203/SiOx 0.336 PVC WT ---9-9A1203 4.03 None G~RP6 ~35 9-10A123 3.89 None FFP7 ~30 9-11A1203 4.03 None PSM(:: ~35, <50 9-12~123 1.79 None PSMC ~40 9 13 A123 1.78 None P~EK8 30 9-14A1203 4-5 None Formica 23 9-15A1203 2.0 None Formica --~
9-16~123 2.0 None Formica ~-1 Load to Substrata Penetration 3~ ~ Polyethersulfone 3 Polyester Shee~ Molding Compound 4 Polyvinyl~hloride 5 Nonlasbestos v~nyl white ile base 6 Glass ~iber R~inforced Polye~ter 7 Fibar Filled Phenolic 8 Polyetheretherketone - 43 ~ ~ ~4~0~

Table 9B

IAPVD SiOx Wear L~yers on Nondecorated Plaatic Substrates Sample Thickness ~spl 5No. ~microns) ~ Sub~trate _ Sgrams) 9-17 2.3Kapton22x tape/Glas~>15 9-18 1.2 None P~MA 15 9-19 2.3 None PMMA >15 9-20 1.1 None PMMA >15 9-21 1.8 None PMMA 10 9~22 1~2Xapton2x tape/Glass25 9-23 1.2 PES42x t~pe/Glass 15 9-24 1.2 PEEK52x tap~/Glass15 9-25 1.2Cr~nar62x tape~Gla~25 9-26 1.2 None ~eflon7 C5 9-27 0.4 PVC8 WT9 5-10 9-28 1.4 PVC WT 10 - 9-29 2.3 PVC WT 15 9-30 3.8 PVC WT 15g-1~
9-31 3.7 PVC WT 15g-lB
9-32 2,8 PVC PSMC10 23 9-33 5.3 None PSMC -28 1 Load to Substrate Penetra~ion 2 DuPont Polyimide 3 Polymethylmethacrylate 4 Polyether~ulfone Polyetheretherketone 6 DuPont Pho~ographic Polyester 7 DuPont Polytetrafluroethylene 8 Polyvinylchloride 9 Non~asbestos Vinyl White Tile Base 1~ Polyester Sheet Moldin~ Compound 4~- 2~

Table 9C
Miscellaneous IAPVD Coatin~ on Plastfc 8ubatrat~s Sample Thlckne~s No. ~microns) Structure _.

9-34 4.37 A1203/PVCtCWT
9-35 4.~1 A1203/PYC/CwT
9-36 6.~0 Si~X/PVC/CW~
9-37 ~10 A10x/PYC/CWT
9-38 ~3 ~1~03~P~c/cw~
9-39 0.48 SiOx/~ologram ~-40 ---- SnU~/Rigid PVC
9-41 -~-- SnG~/SiOx/PVC

~or floor~ng structures w~th A1203 or S~Ox th~n hard coating~ on sele~ted pla~c ~u~tratess ~1) a~ increase ln wear l~yer hardne~s re~ulted i~ an lncrease in glos~ retention and overall ~ppe~ranc~
retention; and S2) ~n incre~se in substrate hardnes~
resulted in an ~ncrease in ~1099 retention ~nd over~ll appearance xe~ention.
Glo~s retention for flooriag ~tructures with .hin hard wear l~yer~ occur~ b~au~e the h~rd co~ting re~lst~ penetration and ~ubsequent removal~ The hard coating 3erve~ ~8 a bsrrier that protect~ ~h0 les scratch re~i~tant plastlc material. Therefore, the ~cratch te~t results reported here us~ an alternatiYe term, ~Load to ~ubstrate Penet~ation~ ~SP) rather t~an ~Load to Incipient D~mage~ ~LID). ~he L~P refer~ to the weight loading at which the diamond ~tylu~ p~netrate~
the hard protec~ive layer and enters the substrate below. ~or example, irreversi~le damage i~ caus~d by a ~tylus load of 15 grams ~or ~ A1203 coat~ng Qn P~MC, ~nd this low LID impl~es that poor glo~ retent~on will be observed. However, the oppo~ite ~ rue. Glos~ reten-tion by th~n, hard coatings depend~ upon both coating - 45 ~ 0 and subRtrate properties, and ~he LSP xeflect~ th~t syn ergistic relat~onRhip be~ter ~han does the LID.
Tables 9A, 9B and 9C contain ~he LSPs for mo~t o the coatingY that h~ve been prepared.
Results from the ~cratch tests are clearly in agreement with the Walker~ Test data regarding the supe-riority of A1203 as a wear layer~ The LSP for A1203 on PSMC is higher than that of SiOx.
The r~sults f or Si x on bo~h PSMC ~nd PVC/CWT
~how that scratch re~is~ance $mproves with the thickness of the hardcoat. Thi~ iB consi~ent with the observa-tion of improve~ gloss and appearance retention for th;ck vs thin coatings on soft pla~tic ~upports.
The high LSP3 for both SiOx and A1203 on P8MC
predict, in agreement with Walkers Te~t data, that th~
PSMC should be the best support.
Examples 10-1 to 10-4 This Walkers Te~t demon~trated that good gloss retention i~ obtained from a flooring structure con~i~t-in~ of a A1203 wear layer supported by a rigid plastic, like polyester sheet ~olding ~ompound ~P~MC). The per-formance rating of the A1203 coated ~etal substrates was complicated by ghe fact that water vapor contamination wa~ present during ~ome of the runs. The ~iN~, blue-2~ black in color, had gloss performance simil~r to A1203.
A structure con~i~ting of 4-microns o A1203 on ~ thick pla~e of polyester sheet molding compound tPSMC) remained essentially free of ~mall scratchesO
showing no hazing and retaining 86~ of it~ measured gloss a~ter 1200 walker cycle~. It had, however, a num-ber of individually visible ~cratches.
A1~03 wear layers on (a) abric filled pheno-- lic (FFP)~ (b) ~lack PSMC, and Ic) glass fibero reinforced polyester retained a lesser but still substantial portion of their origin~l gloss. The uncoated controls, in contrast~ were completely deglossed and covered with fine s~ratches that resulted in a final ha~y appearance.

2~0 ~6 --The sample~ wer0 te~ted ~n the Walkers Test.
Gloss mea~urement~ wer~ obtA~ned or the sample3 ~nd listed in Table 10.
Table 10 Gloss values fsr 4 micron thick A1203 coated and uncoated rigid polymer ~ubstrates before and after Walkers ~est traficking Sample No. Descr1ption Initial Final ~ 4 Loss 1010-1 A1203/Whi~e PSMCl 46.539~,7 -6.8 -14.6 C10-1 ~hi'ce PS~C ~7.5 2.5 -55.0 -95.7 10-2 A1203/Blaclc PSMC 46 . 837 . 9 -8 . 6 -18 . 4 C10-2 Black PSMC 64.22.6-61~.6 -96.0 10-3 A1203/GFP2 25 ,.116 . 7 --8 . 4 -33 . 5 15C10-3 GPP 2~.2 11.. 6 -8.6 -J.2.6 10-4 A1203~FFP3 69 .145 O 6-23 . 5 -34 . 0 C10-4 FFP 52.19.7 -42.4 -81.4 1 1/4" Thick Polyester Shee'c Molding S::ompound 2 1/4~ ~hick Glass Filled Pslyester 20 3 1/4" Thick Fabri~: Fil~.ed Phenoll~

l~xample 11 ~ 1203 and Slx we~r lay~rB on PStqC showed no ~iql~ificant glo~ reduct$on a~ter 1200 walker Qycl~s, however there were ~ome viE~ible scra~che~. Te~t ~looring 25 3tlructur~s u~lng commercial wear layer material~ 211 were completely deglossed ~nd vi~lbly ~cratched ~o a matte finlsh ~fter the same tes~ period. Al~o be~ween tho~e extreme~ was a 3econd new ~loo~ing structure with a 5 -8u wear layer of SiOx on a relatively non-rigid sub~trate ~n~n-decora~ed rigid ~VC fllm la~inated to non-asbe3tos vinyl white tile base). A1~03 cl~arly out-performed SiOx or structures having a common substrate, and comparable wear layer tbi~kness. ~he observation~
~rom this and other Walkers ~e3ts clearly demonstrate - ~7 ~

that important aspects in the performance of hard inor-ganic wear layers on plastic ~ub~trates include wear layer thickne~s, wear layer hardness, and 8upport rigi-dity.
S ~uperior gloRs retention and ~cratch resi~tance have been observed with new ~tructures consist~ng of a reinforced plastic support and an inorgan~c wear layer.
The ~upport ~aterial was polyester 3heet ~olding com-pound, and the wear layer con~i~ted of a five to eight mi~ron "thi~k~ film of either A12G3 or SiO~, prepared by IAPVD.
~he SiOx ~nd A1203 coatings on PSMC were abov0 the critical thickne~s required ~or wear resistance applicatio~s. Above that thicknes~ limit, further increases in coa~ing thicknes~ bave no apparent effect on either gloss retention or scratch re3i3tance. 8cratch tests ~uggest that the crossover point is in the one to three micron range.
Hardne~s of the coating mat~rial i~ a factor in determining glos~ reten~ion and ~cratch re~i3tance. Por example, despi~e being 2 ~icrons thinner than lt~ SiOx coated ~nalog, the 5-6 micron ~thick~ A1~03 coated PS~C
~amples ~tar~ed and flnished the Walkers ~est at higher glo~s, and with fewer visible ~cratches. Thi~ i~ coasis-tent with the pr~viou~ observatio~ that IAPVD A1203 l~ a har~er material than I~PV~ SiOx.
Sample~ approxlm~tely 8iX inche~ s~uare w~re tested in the Walker~ ~est u~ing the serpentine ~ample srrangement. Table~ llA and ll~ t average gloss read-ings from the Walker~ Test.
PSMC was obta~ned as 12~ x 12~ x 0.125~ p~nels of ~15402 Premi-Glass 1100-05, Cameo Colored, from Premix, Incorporated.
A1~03 and SiOx wear layer~ were evaporated from the e-beam gun without the use ~f crucible liners, and ~he chamber was oleaned and re~oiled immedia~ely prior ~o each deposition to avoid water vapor con-tamination~

~ 48 --Tllble llA

Gloss Yalue~ for Polye~ter Sheet Mol~ing C:c~mpound ~PSMC) and Ceramic Wear Layer~ on PSMC ~fter Walkers Test Tra~f ickin~

B ~i~ron ~hick 5-6 micron ~hick Control PSMC SiOx on PSMC A1203 on PSMr Pas~e~ Initial ~inal Initlal ~inal Ini~lal Final 0 5 0 t 7 4 7 . 9 5 4 . 1 24 62.0511,0 46.6 47.9 53.3 53.7 10~8 61.534.1 48.4 ~9.1 60.3 69,~
1~2 66.31507 38.5 42~2 4~.6 46.9 204 58.77.8 48.6 5~.B 55.~1 55.9 402 50.53.1 53.2 52.6 61.0 61.7 ~04 ~5.93.1 . 55.2 50.8 43.2 44,.B
}~120U 60.02.~ 53.1 49.û 58.7 59.4 Table 11B
. .

Gloss values for S10X on PVC, ~nd Polyether ul~one af ter Walker3 q~e~t ~rae~lg 4-6 micron ~rhick 3 mil ~hick PES on SiOx on PVC Citation Tile ~ase Passe~ Initl~l FinalIniti al Final 0 52.3 ~ 98.7 ~
24 52.5 52.89B.B 22.9 ~8 59.4 60.3102~,0 ~0.~
25... 102 56.8 42.2 }00.1 05.6 204 56 . 8 42 . ~ 109 . 0 04 . 5 402 `53.5 34.991.2 Ul.
804 48 . 1 21 ~ 9 95 . 5 û~ . 3 1200 ~!16.1 13.294.~ ~10O4 ~3=L~

Samples approximately ~ix inche~ Square were tested in the Walkers Test. Initial ~nd ~inal glo~s xeadings were made using a.Mallinckrodt 60 Pocket Gloss S Meter and B. A. Newman's template. Table 12A 118t~
average gloss readings for the samples. Descriptions of the samples ~re given in Table 12B.
Alumina wear layer~ were deposi~ced onto the samples by evaporating A1203 from the ~-beam gun without the use of crucible liners. The procedure included a bakeout at 250C for 1 hour prior ~o each deposition ~o minimize water vapor contamination. For most run~, the substrste temperature was al1Owed to ~float~ starting at 30~40C during the deposit~on run~. ~or depositions done without the ion gun, ~n 2 atmosphere o -2.3 x 10-4 Torr was maint~ined. Plasma cleaning, when employed, was for five ~inute~ at a pres~ure of about 3 - 6 x 10 4 Torr.
The substrate3 consisted of about 30 mils of Heraeus H34000 serie.~ White Overqlaze Frit~ on a 7 mil ~tainl~ss steel base. Ink fus~on ~as done using either oven~ or moving belt furnace.
Thickne~s me~surement~ were done u~ing the Amray Scanning Electron Mic~osc~pe ~SE~) or the Nikon Polarized Light Micro~cope ~PLM).

2~:)40~3~

Example 60 ~;loss at W~llcer Count No 0 200 800 1200 12--1 91.4 ~ 84.5 12-2 93.2 ----~ 91.3 12-3 98.0 ~ -- 95.6 12--4 g8 . 7 ~ 94 . 1 12-5 to -7108 . 2a 106 . 389. 3105 . 3 12-8 to -10103.6a 100.8 99.9107.B
12 11 ~o -13 99.9~ 105.7 98.3 105.1 12-14 t~ -16 1û1.6~ 7(~.7 102.~ 107.8 12-17 to -19 95 . 2a 100 . 5 97 . ~) 103 ~ 4 12-20 to-2~ 86.5a ~4.9 84.1 94O3 12-23 to -25 88.7a 100.2 8~.6 86.0 12-26 ~o-28 90.1a . 81.5 _99 5 ~_90.6 a average o~ three samples ~able l2 WearLayer Deposition Deposition I~n 2~
Example ~hickn~ss R~te ~emperature Cleaning Ion A123 No.~microns) ~A/S)C(C) Ga~ ~ A~ t Purity 12-12.2 7.259-196 Ar Yes 99.99 12-24.9 3190-123 A_ Yes 9g.99 12-33~6 12ll3-129 Ar Yes 99,99 12_45.3 33114-134 Ar Yes 99.99 12-53.4 1570-129 ~r Yes 99~8%

l2-63.4 1570-129 ~r Yes 99.8 12-73.4 1570-129 Ar Ye-~ 99.8~
12-83.4 1561-134 Ar Ye~ 99.5%
12-93.4 1561~134 Ar Ye~ 99.5%
12-103.4 1561-134 Ar Yes 99.5%

lS 12-ll3.4 15138-170 Ar Yes 99.9g 12-123.4 15138-170 ~r Yes 99.99 12-133.4 15l38-170 Ar Yes 99.99 12-147.7 1771-170 Ar Yes 99.99 12-157.7 1771-170 Ar Yes 99.99 12-167.7 1771-170 Ar Yes 99.99 12-172.9 40130-170 Ar Ye~ 99.5%
12-182.9 40130-17Q Ar Yes 99.5 12-192.9 40130-170 Ar Yes 99.5 12-2011.6 40100~-206 Ar Yes 99.5~

12-21lla6 40100-2n6 Ar Yes g9.5%
12-2211~6 40100-206 Ar Yes 99.5%
12-233.3 3095-160 Ar No 99.5 12~243.3 3095~160 Ar ~o 99.5%
12-253.3 3095-160 Ar No 99.5%

12-26B.6 S0 250 Ar Yes 99.5~
12-278.6 60 250 Ar Yes 99.5%
12-288.6 60 - 2$0 Ar Yes 99.5 . :

- 52 - ~ ~4 ~he results showed that ~10~8 retentlon p0r-formance is relatively in~ensltive to A12Q3 deposition parameters. ~hickne~s between 3u and 12u5 deposition rates between 7 A/S and 60 A~S; ~nd A1203 purity between 99.5 and 99.99% ~for i~ost~tically pr~s~ed powder~ or crystal~) did not effect Walkers ~est performance.
Exam~les 13-1 and 13-2 Samples were tested in the Walkers Te~t. One sample each was pulled at 200 and 800 counts while two samples were trafficked to 1200 counts.
PB~C was decor~ted by ~ublimation 1mprinting.
A1203 wear layers were evaporated by electron beam. No bake-out was used prior to evaporation.
Table 13A liRt~ the data and 9108~ valu~s for the samples tested. ~tain re~istanc~ tests were done by applying ea~h reagent for a per~od of four hours. The samples were cleaned witb ~icro ~nd water followed by a~etone. Delta E valu2~ were cal~ulated from L, a, b reading~ on a Hunter Laboratory, Model D25 op ical ~en-sor. Table 13B lists the samples tested or stain resistance and their Delta E values.

Table 13A
~earLayer Example Thickne~s 6D- ~lo~ w-l~er Count No. Description tmicrons~ 0 200 B00 1200 A1203/5ub.
13-1 Imprint/PBMC 4.7 ~ 0.354.6 45.9 4~.8 49.5 A1203/Marble 13-2 PBMC 4.0 + 0.3 61.15607 56.5 51.7 Tab1e 13B
- Sanford Shoe Hair Example Ink Iodine Poli~h Dye Ball Point A~phal~ Total NoO Del~a-E Delta-E Delta-E Delta-E Ink Delta E Delta_E Delta E
13-1 9.34 2.17 3.91 2.39 8.~7 1.87 2~.53 13-2 6,75 1.35 2.g4 1.~0 17.39 2.64 32~37 0~ 3 No difference was observed ~n wear per~ormance or adhesion of A1203 applied over decorated 5~ublimation imprin~) and non-decorated P~MC. Overall wear perform-ance was good. W~ar performance of marbled P~MC with A1203 wac similar to that of ~ublimation impri~ted PB~C~
with A123 The ~ample maintained a fairly level gloss curve~ The ~ample~ had very few fine scra~che~. The large ~ratch@3 were not nu~erous. ~he ~cr~tches become-readily visible wh~ they penetrated the A1203 anddestroyed the pri~t. ~he white color of the ~cratches was apparently caused by ~tre~s whitening of the PBMC.
Examples 14-1 to 1~-14 Example 14-1 to 14-7 were ~ormed by deposit-ing two to three micron~ of SiOx by g-beam ev~poration using a Web coater onto the 24-inch wide, 7-mil filled mylar sheets. Coating ~peed wa~ about 30 ft/min.
Examples 14-8 to 14-14 were ormed by depo~it-ing 3 micro~s of SiOx by E-beam evaporation onto back si~e of samples formed by the procedure of Examples 14-1 to 14-7. The procedure utilized a freshly cleaned and refoiled depo~ition chamber, ~n 2 ion assi~t, but not a bakeout. The samples were ~ot preclea~ed prior to load ing into the depo~tio~ chamber, And were not 0ubjected to a pla~ma ~leaning ~tep after loading.
The SiOx on mylar w~ adhered to ~ bulk ~old~
ing compound u5ing adhe3ive release tape. ~able 14 giv~s average ylos~ v~lue~ at each traffic lntervalO

~able 14 Exam~le No. _ 60~ Glo~s at Walker Count _ 0 20~ ~00 ~00 1200 1~-1 to 1~-3 19.1 31.3 ~ 17.2 1~,3 14-4 to 14-7 23.7 16.7 14.9 ~.9 ~.7 140~ to 1~-10 35.4 40.9 ---- 27.0 2~.5 14-11 to 14-14 29.2 3307 2~o3 21~2 9~5 _ 5q _ ~ 0 8ix inch ~guare ~amples were te~ted in the Walkers Te~t. ~able 15 li~ts the ~ample descrlptions and the respective gloss valuesO
Examples 15-1 t~ 15-12 were prepared with 7-mil, 302 stainles~ ~teel 3ub~trate~. ~xamples 15-13 to 15-15 were prepared with 14-mil cold rolled st~el supplied by Chicago Vitreou~ with their cera~ic ground coat. The su~strates were coated as follow~:
: 10 Examples 15-1 to 15-4: 30 micro~-~340Q2 primer and 10 micron-~34002 textured pattern w~th 20~ mattin~ agent ~7903.
Examples 15-S to 15 8: 30 ~icron-H34002 primer and 10 micron-~34~02 textured pattern with 20% mat~ing agent H7003 and a 5.95 micron thick clear~
Heraeus ~30011, protective ceramic slaze.
Examples 15-3 to 15-12: 30 ~icron-~34002 primer and 10 ~icron-H34002 textured pattern with 20~ matting agent H7003 and a 2.40 micron thick alumi-num oxide layer.
Example~ 15-13 o 15-15: 56~6 micron of grou~ ~oa~
29.1 micron-~34002 pri~er as the wear layerO
The primer and matting agent were m nufactur~d by Beraeu~.

Z~

Table 1 5 Total A123 Nomi nal Enamel Wear Layer ExampleThickness Walker = 60 Glos~_~hickness No.__ ~Micron) Cycle~ Initial inal~Micron)_ 15-1 40 200 ~2.8 ~7.5 15-2 40 800 ~6 . 5 64 . 4 15-3 40 1200 64.8 67.7 15-4 40 1200 61 . 0 59 . 7 ~ù 15~S 40 200 57.4 ~5.6S.gS
15-6 ~0 ~0~ 55.2 53.35.95 15-7 40 1200 65.3 63.55.~5 15-8 40 -1200 61 . 7 64 ~ 6 5 . 95 }5 9 4~ 200 71.9 75012.40 lS-10 40 ~00 72.5 71.02.40 15-11 40 1200 67.8 73.02.~10 15-12 40 1~0~ ~6.3 ~6.62.40 15-13 86 200 89 . B 95 . 5 15 14 800 90 D 6 96 . 0 15-15 12~ 88.6 87.1 A~ ~hown in Table 15, ~he ylo~s value~ of the thr~P sta~ nless~steel-~ubstral:e ca~egorie~ are e-~sen-~ially uncharlged aft~r a total trafic coulat of 1200.
Appearanc~ retention differenc~ between the cakegories 25 were noted howevex~ ~rhe tainle~s steel structure with-ou~ a wear layerD ~xhibi~ed ~ore visually objectionable scratches than the glaze-coated and aluminum nxide-coated structuresO These later two categorie~ bad hard protective wear l~yer~ which appear to fford i~creased 30 scral:ch resis~ance.

20~L~0;3~

Although the low-carbon-steel structure exhib-ited excellent gloss retention, scratch resistance was poor compared to the other structures. Most of the scratch damage was limited to the upper~most ceramic layer which was the Hera~us H34002 system. The type of damage present indicated a poor level of adhesion bet-ween the ground coat and the Heraeus cexamic.
Those of ordinary skill in the art can readily deposit hard inorgan^ic films onto metallic substrates.
However, little work has been done in the area of depos-iting hard inorganic materials on organic ~ubstrates.
This is most likely due to the fact that those of ordi-nary skill in the ar~ believe that the properties of the hard inorganic material would be degraded to a point at ~hich the material would not be us~ful if it were depos-ited at a temperature low enough ~o allow deposition on the organic substrate without destroying the substrate.
The present inventors have determined that good wear layer properties, particularly gloss retention, scratch resistance and stain resistance, are achievable even if ~he inorganic layer is deposited at a temperature of less than 175C, preferably less than lSO~C and most preerably less than 100C.
It is also generally accepted that as khe thickness of a depo~ited film exceeds 0.5 microns, the stress builds to such a high level that spalling or flaking of the coating occurs. ~owever, the present inventors have shown that 1 micron to 25 microns thick inorganic materials can be deposited on organic materi-als with sufficient adherence to perform as surfacecoverings.
Two problems have been associated with the vapor deposi~ion of inorganic materials onto or~anic substrates. Films 1 micron to 25 microns thick depos-ited on organic substrates tend to discolor and crack.While the cracks and fractures degrade the ability of the deposited layer to prevent gas and liquid transmission, the overall performance of the protective layer, including appearance retention, exceeds the per-formance of thinner (less than one micron) low pressure environment deposited layers on both inorganic and orga-nic substrates.
These problems have been minimized by keeping the organic substrate relatively cool~ The substrates are radiation cooled by proximity to a cooled surface ~e.g., water and glycol, or liquid nitrogen coolant) during the time the substrate is not in the deposition zone. The substrate typically spends 3 out of every 12 seconds in the deposition zone. The use of radiation cooling makes the fabrication of aluminum oxide/PVC com-posites possible.
Inorganic materials deposited on organic substrates by low pressure environment techniques tend ~to discolor more than when deposited on inorganic substrates. All of the polymeric substrates deposited on to date have discolored somewhat during the deposi-tion process~ The absolute amount of discoloration has been quite small - typically around 3 to 6 total Delta E~ Some of the samples have been less than 3 total Delta E and some have been less than 1 ~otal Delta E. For a discussion of Delta E see Richard S.
Hun~er, The Measurement o~ Ap~earance, a Wiley-Inerscience Publication, John Wiley & 50ns, 1975.
Discoloration levels for alumi~um oxide coat-ings on free-standing films, uch as polyether sulfone (PES) and polyvinyl chloride lPVC), were done with a sheet of white bond paper behind the sample. Controls for the measurements were (as appropriate) the backside of the coated sample/ the virgin suxface of an uncoated sample/ or a piece of uncoated film on top of white bond paper.
Although discoloration has not been affected by chamber tempera~ures between 50 and ~00C during the aluminum oxide deposition on polyester sheet molding compound, it i5 believed that the discoloration is caused by trapping low molecular weight polymer frag-ments tha~ are outgassing from the polymer support in the growing inorganic film. If the ~emperature were hgh enough to degrade the organic ~ubstrate, additional fragments would likely be trapped and the discoloration increased. There is strong indication that the dscolor~
ation is in the inorganic coating. ~lakes removed from the organic substra~e are discolored and the discol-oration disappears from the substrate when the c~atingis dissolved.
Examples 16-1 to 16-26 No correlation between chamber te~perature and discoloration has been evident. Aluminum oxide was lS coated on a variety of PSMC formulations and surface treatments. The ~amples were prepared under a wide variety of deposi~ion conditions. See Table 16A. The range of final chamber temperatures reported for the PSMC was from 50C ~o 155C. Thus the substrates in these runs were exposed to differing thermal histories.
All of the ~amples showed similar, rela~i~ely low Delta E values.
~able 16A
~emp Thickness ExamPle_No.Deg. C Microns _ Ion Gun Delta E
16-1 50 4.10 Yes 9.09 16-~ 126 4.~0 Yes 5.16 16-3 135 4.00 No 5.24 16-4 135 4.00 No 5.50 16-5 145 4.70 Yes 2.50 16-6 153 4.32 No 3.60 16-7 154 5.80 Yes 5.60 16-8 155 4.50 Yes 4.14 The results are similar for other substrates.
See Table 16B. The polyether sulfone (PES~ sample had the highest level of discoloration which was probably caused by the inability to remove all adhesive from the ~o~

sample prior to the Delta E measurement. The discol~ra-tion of the polybutylene terephthalate (RBTP) was simi-lar to the PSMC, Table 16B
Temp. Thickness Example No~ Sub~trate Deq~ C MicronsDelta E
16-9 PVC/TILE 50 4.1 8.13 16-10 PVC/TILE 54 0.95 5.4 16-11 PVC/TI~E 54 0.95 4.3 16-12 PVC/TILE 54 1.42 3.54 16-13 PVC/TILE 54 1.42 3.19 16-14 PBTP 106 4.7 6.06 16-15 PBTP 117 4.5 2.14 16 16 PBTP 147 4.32 4.71 16-17 PES 104 4.7 4.81 16-18 PES 162 4.22 15.22 16-19 P5MC 50 4.1 9.09 16-20 PSMC 126 4.8 5.16 16 21 PSMC 135 4.1 5.24 16-22 PES/PSMC 135 4.1 5.5 16-23 PSMC 145 4~7 2~5 16-24 PSMC 153 4.32 3.6 16-25 P5MC 154 5.8 5.6 16-26 PSMC 155 4.5 4.14 xamples 17-1 to 17-4 The level of discoloration appears to be dependent on the thickness of the inorgan;c layer depo-sited. Aluminum oxide was deposited sequentially on PSMC substrates with a 600mA/600V 2 ion assist. The incxeased thickness also corr~sponds to increased expo-sure time.

- 60 ~ 4~0~

Table 17 Nomi na 1 Depos i t i on Time ~h i cknes s Ave Example No.Minutes _ Micron_ Delta E
517-la 110 10 12 . 82 17-}b 110 10 9 . 67 17-2a 56 5 8 . 57 17-2b 56 5 6.69 17-3a 11 1 2.14 1017-3b 11 1 3.79 17-4a 1.1 0 .1 0 . 53 17-4b 1.1 0.1 1.77 Examples lB-l to 18-5 The level of discoloration appears to be more dependent on thickness than on the coater u~ed to depo-sit the inorganic layer. Examples 18-1 to 18-5 were coated with aluminum oxide wear layers using three dif-ferent batch coater~.
~able 18 20Example Thickness No. Substrate Coater Microns Delta E
18-l Smooth Whi~e PBTP ~3) 1.73 0.41 18-2 Smooth White PBTP (l) ~.32 3.59 18-3 Cameo PSMC ~3) 1.73 1.06 18~4 Cameo PSMC ~2) 3~4 3.77 18-5 Cameo PSMC ~l~ 4.32 4.47 Examples l9-1 to 1~-20 The following examples show the effects of chamber temperature, ion source, and extraction grid on cracking. Example l9-1 was a typical deposition of alu-minum oxide on~o PSMC. Four microns of aluminum oxide was deposited onto PSMC, under typical conditions of about 20 Angstroms per s~cond in oxygen at a pressure of 2.5 x 10-4 Torr. The maximum temperature measured in the chamber was 153C. The ion source was not used.
The coating was cracked and sliyhtly discolored ~Delta E
of 5.16).

.

o~

As evident ~rom Example 19~2, the reduction of average chamber temperature reduces the degree of cracking, but does not seem to affect the discoloration.
About 2.9 microns of aluminum oxide was deposited onto PSMC which was radiation cooled by proximity to a cooled surface ~water and glycol coolant at -23C) during the time the substrate was not in the deposition zone. The ~ubstrate typically spent 3 out of every 12 seconds in the depo~ition zo~e ~(at 5 rpm). The cracking of the aluminum oxide film was reduced, (the typical area of uncracked film was larger). The Delta E was 5.06.
In Example 19-3, the aluminum oxide coating on PSMC in a si~ilar deposition system utilizing a Kaufman ion source yielded continuous coatings that were disco-lored. The aluminum oxide was deposited onto the PSMC
~ubstrate at an average rate of about 20 A/s and with a .Kaufman type ion source operating at 500 eY per ion and a tot~l beam current of about 40 mA. The thickness of the aluminum oxide was 3.2 microns. The coating was continuous and discolored. The Delta E was 3.77.
~here is some evidence that increasing the ion energy flux from a ~en~on cold ca~hvde lon source redu-ces ~racking. This experim~nt compared a flat ion source extraction yrid which concen~ra~ed the ion beam more than the convex extraction grid normally used for aluminum oxide coatings previously discussed. Four thicknesses of aluminum oxide were deposited se~uen-tially onto four sets of various composition stationary substrates all during the same pumpdown of the vacuum ~ystem. For the flat grid, the thicknesses obtained were 0.1, 0.4, 3 and 6 microns (Examples 19~4 to 19-11).
For the convex grid~ the thicknesses obtained were es~i-mated to be 0.1, 0.4 and 0.8 microns ~Examples 19-12 to 19-17).
A primary observa~ion for this experiment was the existence of a re~ion on the PSMC subs~rate that was crack free with the concentrated ion beam even at a Z1~)4~

thickness of about 6 microns (Examples 19-}0 and 19-11).
There was no such region on ~he thickest ~amples pre-pared with the convex grid (Examples 19-16 and 19-17), even though the thickness was csnsiderably less. A sim~
ilar trend was observed when alumiaum oxide was depos-ited onto a stationary 12" x 12" P5MC sample ~Example 19-18)~ with the flat extraction grid.
A combination of radiation cooling and chang-ing the ion energy ~lux produced aluminum oxide c~atings on PSMC that were continuous a~d coatings on PVC lami-nated to tile that were only slightly cracked.
~Examples 19-19 and 19-20.) Both coatings were disco-lored, About f our microns of aluminum oxide was depos-ited onto P8MC and PVCjtile substrates which were raaiation cooled by proximity to a liquid nitrogen cooled surface. The ion source was equipped with a flat grid as in the prior paragxaph~ The aluminum oxide on the PSMC was not cracked. The aluminum oxide on the PVC/tile was cracked but significantly less than previ-~ 20 ous attempts.

~ 3 Table 19 Nominal Chamber Degree Example Thickness Tempe~ature Total of No. Micron De~. Ç. _elta E Crackinq 5 19-1 4.32 153 5.16 normal 19-2 2.9 101 5.06 light 19-3 2.6 290 3.77 none 4 0.1 57 .53 19~5 0.1 ~ 57 1.77 1019-6 0.4 90 2.14 19-7 0.4 90 3.79 19-8 3 98 8.57 19-9 3 98 6.69 19-10 6 95 12.82 none to severe 1519-11 6 95 9.67 19-12 0~1 34 none 19-13 0.1 34 none 19-14 0.4 67 none 19-15 0.4 67 none 2019-16 0.8 95 < normal 19-17 0.8 95 < normal l9-lg 6 85 lg-l9 4.1 none 19-20 4.1 51 light ExamPleq 20-1 to 20 4 Cracking can occur because of dimensional changes Qf the substrate ~thermal expansion, stress relaxation or phase change), or a build up of stresses in the coatiDg due to growth mechanisms (intrinsic film stress). The relative importance of each contributing factor depends on the structure and composition of the composite, the conditions governing the growth of the film and the thermal history of the evolving composite.
These examples provide evidence that the conditions (i.e., eV/Atom ratio) governing the growth of the film, in the particular case of aluminum oxide on PSMC, are significant for the fabrication of continuous thick coatings.

2~

The continuous aluminum oxide coating on PSMC
was fabrica~ed without at~empting to limit the tempera-ture of the polymer samples during the deposition. The temperature probe was located close to the primary source of heat, the electron beam evaporator, therefore the ~emperature in the region of the samples should be lower than what was observed at the thermocouple probe.
The pxobe temperature was typically 300C. Based on preliminary measurements w~th a second probe, the corre-sponding temperature ~n the region of the samples was estimated to be about 150C to 200C. Only the probe temperature is reported in Table 20.
The crystal sensor ~or the deposition rate controller failed during the thir~ crucible of each sample. The depositions were continued in "time power mode", a mode of the controller where the power is held constant at a level determined by averaging over a short period prior to crystal failuxe, for a time period cal-culated to give the desired ~hickness assuming the desired deposition rate.
Four sets of polymer samples were coated. The first set of samples were ~mooth plain white PSMC, unsa-turated polyester resin filled with 70% ~eldspar and glass fibers. The second ~et included two types of tex-tured PSMC and the last kwo sets were textured PBTP, polybutylene ~erephthal~te filled with 55~ mineral fillers. Table 20 contains deposition and evaluation information.
The deposition conditions were chosen from the conditions used on the decorated ceramic steel samples which yielded a compressive stress on Kapton coupons and with the mildest ion bombardment conditions. These con-ditions included a beam voltage of 500 volts and beam current of 40 milliamperes.
The combination of deposition rate and ion bombardment yielded 3 micron continuous films of alumi-num oxide on the smooth PSMC. The texture of the other - 6~ -polymer substrates prevented evaluation of ontinuity by optical microscopy. Discoloration was noted after depo-sition on all of the sarnples. The initial gloss of the continuous aluminum oxide coatings on the smooth PSMC
was significantly higher than on di3continuou~ coatings~

u c l l ~
o l l ~ ~
_, .~ , , a) O u~ X
E~ O E~ O
I I tn o ~ l l ~ o o ~c ~ c t o ~ ~
~ lco l ~ ~ o h C) I ~ I O O O O ~
,~e ,, I _I I ç~ o o o o Lt ~1 tq U~
rl E Ll I ~ ~Q

t Ch U Q

G~ I I 1` ~ (1 o -~ X I ~ S ~D E3 J~ a~ I I o o n o ~) u~ ~ ~1 ,1`
u ~ I I r~
U~ _I 5.~ H I I ~ XO C: ~) O
4~ O
E~ ~ I I O ~ ~ U
~ ~rl U

~ c m I I ~ q o ~
I ~ r~ ~ O U ~
dP I I ~ ~ C ~ ~ U C
~ ~ 1 o u l l ~ x ~
I I ~ O C ~ rl rl un o ~ ~ Dl o~
~ ~ I e I ~ o Io o c:, o ~ . o r~ 0 o a~ O . c~ u o c:~ o o ~ ~ E
C: QJ X
~ O .C
U~ ~ C I ~ ~ ~ ~ C O C .Y V~

~ .a s o o ~ c :~
U I U I ~ V ~ ~ ~ ~ .r~
E~ U) I I~ ~ ~ ~ ~ 3 o S -- ~ lJ U~
o O ~ C
c ~ a~ o I ~ I ~ cn ~ C
o ~J Q~ r f~ O U ~ ,.C O X O ~ ~
H C ~ I ~J I ~ O U a) .C
a~ ~ 3 In al ~n a) rn n ~ a) u~ ~ In o I
~ Z; I I o o o X I I

Claims (16)

1. A surface covering comprising a support and a wear layer deposited on said support by a reduced pressure environment technigue, said wear layer being from 1 micron to 25 microns in thickness, said wear layer comprising a hard inorganic material.

2. The surface covering of claim 1, wherein the wear layer is 3 microns to 15 microns in thickness.

3. The surface covering of claim 1 or 2, wherein the hard inorganic material is selected from the group consisting of inorganic oxides, inorganic nitrides and inorganic oxynitrides.

4. The surface covering of claim 3, wherein the hard inorganic material is aluminum oxide.

5. The surface covering of any of claims 1 to 4, wherein the wear layer is deposited on a polymeric support.

6. The surface covering of claim 5, wherein the polymeric support comprises a thermoplastic selected from the group consisting of thermoplastic polyester, thermoplastic polyurethane, thermoplastic polyacrylate, polycarbonate and polyvinyl.

7. The surface covering of claim 5, wherein the polymeric support is a thermoset selected from the group consisting of thermoset polyester, thermoset polyurethane, thermoset polyacrylate, polyether and epoxy.

8. The surface covering of any of claims 1 to 7, wherein the wear layer is deposited on the support at a temperature of less than 175°C.

9. The surface covering comprising of any of claims 1 to 8, wherein the surface covering is a floor covering.

10. The floor covering of claim 9, wherein the support comprises a metal component selected from the group consisting of a foil, a film and a sheet.

11. The floor covering of claim 10, wherein in the metal component has thickness of between 0.15mm and 12.5 microns.

12. The floor covering of any of claims 9 to 11, wherein the support further comprises a conformable layer capable of inelastic deflection.

13. The floor covering of any of claims 9 to 12, wherein the wear layer is discontinuous.

14. A floor covering comprising a metal sup-port layer and a wear layer consisting essentially of hard inorganic material.

15. The floor covering of claim 14, wherein the support layer is capable of inelastic deflection.

16. The floor covering of claim 14, wherein the hard inorganic material is a fused ceramic.