CA2186261A1 - Treatments to reduce frictional wear between components made of ultra-high molecular weight polyethylene and metal alloys - Google Patents

Abstract

The present invention provides methods for modifying surfaces of metal alloy and/or UHMWPE (ultra-high molecular weight polyethylene), preferably surfaces which are frictionally engaged, e.g., in an orthopaedic implant. The methods reduce the coefficient of friction of the metal alloy component, reduce the shearing of fibrils from the UHMWPE component. One method involves solvent immersion of the UHMWPE component to remove short chains of polyethylene at or near its surface, and to swell and toughen its subsurface. Another method also involves firmly coating the surface of the metal alloy component with an adherent layer of diamond-like carbon ("DLC") by creating a metal-silicide interface at the surface of the metal alloy to permit firmer adhesion of DLC. Although these methods are particularly useful in orthopaedic applications, they can also be used to treat similar components used in other applications.

Description

2 1 8626 1 Wo 95/26169 1 ~ . '7 TREATMENTS TO REDUCE FRICTIONAL WEAR BETWEEN COMPONENTS MADE OF
ULTRA-HIGH MOLECULAR WEIGHT POLYETHYLENE AND METAL ALLOYS

FIELD OF THE INVENTION
The present invention relates to ~iecreasing frictional wear and degradation in pL..JdU~;~S, typically u ~1.vyaedic implants, in which ~ made from ultra-high le~ Ar weight polyethylene 5 (UHMWPE) come into frictional contact with ~ , ~y, e.g., mads of a metal alloy, such as cobalt-c;1~L, ' RA~ J.J OF THE INVENTION
The replr- ~ ~ of destroyed or damaged human Joints i8 one of the great achi~ s of twentieth century OL I 1~ C surgery.
lO However, total Joint PLO~ eSeS~ - - d of various combinations of metal, ceramic, and polymeric , ~8, continue to suffer from distressingly limited service lives. For example, the current ut~ a~lon of high-load bearing yrvY-1-eses for the hip and knee have a typical l$fetlme on the order of 6-12 years. Generally, 15 falled implants can be replaced once or twice, which means that current tenhnn~ ogy provides a solution for--at most--about 25 years .
With human life e~ ncies steadlly increasing, there is a drlving need to increase significantly the effective lifetime of a 20 single implant. One of the problems encvu---e ed in ~e'3~n~n~ such p~vY~1.eses is the difficulty of finding materials which are both h~ -tible and also durable enough to replace a human Joint. In use, a human Joint is exposed to ~ubs~a,,-lal, repetitive loads and f rictional ,i ~L ~sses .

2 1 8626 t Wo 9~/26169 P~~

Although the ~ ~ Ic details may vary, a natural human hip, knee, or cho~ er ~oint generally ln~ D~ (a) a more-or-less spherlcal ball; (b) an al.t ' ~ t to a long bone; and, (c) ~1 hPm~RphPrlCal socket (the "acet~lbular cupn) ln a contlguous bony 5 ~Lu~;Lu~ ~ which retains the æpherical b811 so that the long bonu may pivot and articulate. In a healthy Joint, nature m1nlm~"S the friction between the Joint . , - 1.8 and p ~v~n~s bone-on-bone wear and des LL U~ ~lon by uslng matlng porous car~ ~ ag~ n~ c layers that provide "squeeze-film" synovial fluld lubrlcatlon, This 10 lubrlcatlon results ln a low coefflclent of frlctlon on the order f O . 02, When a human Jolnt has been destroyed or damaged by diseas~ or in~ury, surgical repl~ t (arthroplasty) normally 18 requlred.
A total ~oint rPrlr -t 1n~ d~s - ~ that simulate lS natural human Joint, typlcally: (~) a more-or-less spherlcal ceramic or metsl ball, often made of cobalt-chromlum alloy: (b) ~ttached to a "stem, " whlch gener~lly is implanted into the core of the adJacent long bone; and (c) a h~m~cph~Prlcal socket which takes the place of the acetabular cup and retains the spherlcal ball.
20 This hpm1cphprical socket typically is a metal cup affixed into the joint socket by mechanlcal a~t ' ~s and ~llned" with UHMWPE so that the ball can rotate wlthln the soCket, and so that the stem, via the ball, can plvot and articulate.
One of the dlff$cultles ln ~,.,nY~ u.;~lng any davlce for 25 lmplantatlon into the human body is the need to avold an adverse 2 1 ~62~ ~
WO 95/2616g immune ~ponse. The posslh~l1ty of an adverse lmmune ~ 0~8e i8 reduced when certain synthetlc materials are used. Cobalt-~l.r~
alloy, tltanium, and UHMWPE are ,e . ~e~ of such synthetic materials. U .~c, Lu..al ely, the use of UI~MWPE in the bearlng of a 5 total ~olnt r ~pl e 1. msy be the cause of ~t lea8t one typs of fallure of such devlces.
Three baslc problems may cause a total ~oint r~rl ~C ~ to fail or to have a limited servlce life. The first problem, which manifests itself at the bone-stem in~e~f~, ls not the focus of lO the present application. Bec~use the elastic modulus of the stem sreatly exceeds that of the bone, flexural ` loadlng caused by walklng creates local cyclic stress c~ e--L- c-tions due to the non-nce of the stem. These stresse~3 can be intense and evensevere enough to cause death of local bone cells. If this occurs, 15 pockets of non-support are created, and the stem may loosen or fail.
The two other basic problems, which are coupled, are thc subject of the present application. One of the8e problems, known a~ ball-cup frlction and wear, results from frlctlon~l wear between 20 the h~mtsE~h~r1cal bearing (which 18 "lined" wlth U}/MWPE~ and th~
pol t shed spherical ceramic or metal ball attached to the stem. The other problem, known as sub-surface fatigue, results from brittleness of the UHMWPE bearing and the re8ultlng ~ de~;y of the UHMWPE bearing to fail under reciprocating applled loads.

Wo 9~/26169 2 ~ 11 r ~ ~ A7A
For many years, the acetabular cup ln ~oint lmplants has been "lined" wlth UHMWPE, or other materials, in order to decrease th~
coefficient of frictlon of the soCket or bearlng. Unf~L~u,.a~ y~
~..lln1~ 1 experience hag shown that, at least when UHMWPE is used to S line the bcaring, elther thc surface of the UHMWPE ~bearing" and/or the surf~ce of the metal/ceramlc ball ultimately is deY~Luyed by frictlon-induced wear. Alternately, the acetabular cup loosens after a period of use, greatly increasing ball-cup friction and wear.
Some insight lnto the cause of fallure due to ball-cup frictlon and wear has been gleaned from hlstologlcal studies of th~
.u.Lu~..dlng tlssue. These hlstologlcal 8tudles show that the ~ -Lu~l--dln~ dlstressed tlssue typlcally contalns ~ 1 y small partlcles of UHMWPE whlch range `'rom sub-ml~;L, ~LY to a few 15 miuL~ L:~ in slze. Larger particles of UHMWPE appesr to b toler8ted by the body, as is the solid bulk of the UHMWPE bearlng.
However, the body ..~pa ~ntly does not tolerate smaller partlcles of UBMWPE. ~n fact, these small particles of UHMWPE cause powerful histlocytic reactions by whlch the body -a~s~rully a~ g to 20 ellmlnate the forelgn materlal. Agents released ln thls process attack the nelghboring bone to cause ~wenr debris-induced osteolysls" which, ln turn, leads to a loss of fixatlon and loos~nln~ of the p.~ l.esls due to "r~ '-lln!~ of the bone.
5Ahe f $rst step ln the 5~en~ 10n of the small partlcles of 25 UHMWPE appears to be the formatlon of a very thln layer of Wo95r26169 2 1 ~ 6~ P ~
s polyethylene between the spherical ball and the UHMWPE lining of the bearing. ~his thin film of polyethylene adheres to the "ball~
and serves as a soft, shearable, solid lubricant ~ , 9 i of ml 11 1 ~7ns: of 8ubm$.;-, ~èr particles . Adheslve wear between the 5 ball and the bearing p.odu~es strong, adheslve ~unctions on the ball. ~Yhen exposed to further friction, fibril8 of the polymer shear off of these adhesive ~unctions and are drawn into slender e~ Qec;Llng 1 ~, ~s, eventually producing 1 ~ t rupture.
This 1 ~, t rupture ~yy~Len(.ly yL~ ,e~ the lubricous, 10 e~L~ ly small particles of U~MWPE which eventually migrate to the bone-acetabular cup bond line. The reason for migration of these particles into the "crevice" between the ball and the cup are the mi,;Lo.:uLLe..~a that are yenêLated in the synovial fluid by joint motion. Once a sufficient number of small particles enter the 15 bone-cup crevice, th~ bone tlssue begins to degrade and the ~olnt rPpl e ~ êventually loosens and falls.
One way to reduce f riction between the metal and U8MWPE
C , ~a would be to oat one or both of the I a with diamond-like carbon (DLC), which is rh~m-r~lly inert, 20 b~ lble, and is known to have a l~w coefficient of friction.
Ullfo.~u--F~ely, the very properties of DLC that make it a desirable coating for parts that will be frictionally engaged make it difficult to achieve strong ~lhP~on of the DLC coating to thQ
auba~La~ê, particularly where deposition; - a~uLe8 must be low.
25 ~his limited ~lh~l on problem can be PY~r,-~h~ted by very high Wo 95/26169 rC l~ L~L

es~lve stress, such as that found ln a plasma-deposlted DLC
( up to 8 GP~ ) . Therefore, some have concluded that DLC--or at least plasma-deposited DLC--cannot be used ln oL ~ ae~llc appllcatlons .
Energetlc ion b~A ass~;lated DLC has a far lower residunl stress than rl r leposited DLC, and is a better Csndldste for a high lnte~rlty DLC. The su},~Late materl81 to whlch all forms of carbon adhere most ~ ,rully 18 slllcon. Thls 18 because strong covalent Sl-C bonds are. easlly formed between the costln; and thQ
sillcon :,ub~LLaLe. Some have a~l . Led to lmprove the A~hPQ~oA, of DLC to other materlals, such ss metal alloys, by formlng an inL~ry~,~ed sillcon bond-coat to whlch the DLC wlll adhere morQ
strongly .
Unfortunately, thls slmple approach does not result ln ~rlhPe:iAn that survives in appllcatlons, 8uch as oL~-)yae~c appllcatlons, where the DLC coating ls sub~ected to :,ul,~La.,Llal frlctlon and stress. The slmple formatlon of a 811icon bond-coat on ~ met~l alloy appears to create another relztlvely weak interface between the slllcon and the metal or alloy.
~lleLC:0L~: a method 18 needed by which 8 DL~ coating can be strongly adhered to A metal surfaCe, and by whlch the shearlng of polymer flbrlls from an UHMWPE . AAt can be pL~:v~ L~d. The method would be most efflclent if lt rendered the UHMWPE ~ ~u -less brlttle 80 thst sub-surf~Ce fatl~ue fallure was reduced.

21 8~
Wo g~/Z6169 li ~~

FIGURES
Flg. l is a diagrammatlc L~plese~ L~tion of a mixed pLo:-~)e-,~lve and cros~ sectional side vlew of the wea~. test machlne used ln the followlng experiments.
Fig. 2 ls a chart of the number o f wear cycles against the volume (mm~) logt durlng those cycles.
SUMMARY OF T~IE INVEWTION
The present inventlon provldes a method for modlfylng the ~UL ~ S lnvolved, typlcally a metal alloy spherical ball and an UHWPE bearing, to reduce: (1) irictlonal wear between such ac~s; (2) shearing of fibrils from the UHM~JPE bearing; and (3) sub-surface fatigue in the UHMWPE 1. The method involves solvent lmmerslon of the UHMldPE t to remove short chalns of polyethylene ~t or near the surface of the ~, and to swell 15 and ~L~y~l.en thg subsurface of the ~ , t. The method also involves treatlng the metal alloy spherlcal ball to create a metal-s~ e l..~eLr3ce whlch permlts flrmer ~h~5~0n of D~C and thereby reduces the coeff~r1~nt of frlctlon of sald surface. Although ths methods of the present lnventlon are partlcularly useful in 20 UL ~ paedic appllcatlons, the methods also can be used to trezt simllar ~.s used in other appllcatlons.
DETAI~ED DES~RIPTION OF THE INVENTION
The present lnventlon has two aspects. The flrst aspect i8 801Vent 1 ~ nn ~Leat t of the UmlWPE ~ t . The second WO 95/26169 2 1 8 6 ~ 6 1 P~ ~

aspect 18 a method for creating strong adhesion of the Dr~c co~tin to thQ spherical ball.
SOLVENT IMMERSION OF Ul{l~SWPE COMPONNT
Solvent immerslon accordlng to the present lnvention may be S used to treat any UHMvlPE - , ~ L that will be exposed to frictlon during use in order to lncrease the llfe of such a . ~ L.
Solvent lmmersion particularly is ugeful ln cc"~.~e.;~lon with medlc~l devices, and most partlcularly with medical devices such as total ~olnt replr- ~, which contain UHMWPE bearings that wlll be 10 exposed to friction during use.
Methods of manufacturlng ~8 made of UHMWPE are known.
The process of the present inventlon preferably ls p~c~ ~ after the ~ , t has been formed, but beforQ lnsertlon lnto an end product, such as the metal cup of a total ~olnt prosthesls. The 15 ap~sLcl~us used to lmmerse the UHMWPE , t ls not crltlcal to the present lnventlon. For example, lf deslred, the UHMWPE
,~ ~ may be ~ ~c~ ln the solvent uslng a slmple hand-held ln~
A ~.~r~ solvent for lmmerslon accordlng to the present 20 lnventlon ls dec~l.y.l.u..~ alene (CloH~ rl-y-l-uus), 31so known ~15 Decalln'Y. However, other organic solvents which are capable of dissolvlng short chaln polyethylenes and of _ - 11 1 nSI the ml<;Lua~uu~u~ of the , ~ also may be used. Such solvents lncludo aromatic lI~dLUC8~IJVIIS~ such as benzene, toluene, xylene, 25 and o-dichluLuh~ n~; other allcyclic hydrocarbons, such as 21 ~62~ 1 WO 95/26169 l ~ 7~

cyclnheY=no and tetrahydronsphthalene, also known as Tetralln'Y;
~nd, aliphatic ~yJLuu~Lbùns such as n-paraffin, isoparaffin, and mixtures thereof.
The solvent chosen should be placed in an appropriate 5 contalner for lmmerslon of the UHMWPE ~, t under controlled condltlons of tlme ~nd ~ UL~. The solvent should be heated to a temperature that will ~Y1m~o the dlssolutlon of short polyethylene chains, typically between about 3û-100C. A
,- a~uL~ of between about 30-50C ls pLe~LL~d because 0 ~ /eLatULes above about 50C result ln excessive ~ n~J of the polymer. Once a stable solvent t al uLt: has been reached, the UHMWPE ~ should be 1 ed and retained in the solvent for a perlod sl~ff~a~ont to dlssolve any short chain polyethylenes and to swell the mi~;LU~L-~;-ULe of the t. Thls typlcally should requlre about 30-180 seconds, preferably about 30 second8.
The tlme of immersion generally should de-;Lease as the t , ~uLa of the solvent ls lr.~L~sed.
After the UHMWPE ~ t has been ~ ~ed ln the solvent for an apprOpriatQ perlod of time, the . ~ should be removed 20 from the solvent and allowed to dry. A naked" UHMWPE , L
generally should be drled for about 24 hours at room ~ , a~uLts.
The condltlons under whlch the UHMWPE , ~ ~ 18 dried are not crltlcal however, lf the UHMWPE , t has areas that are dlfflcult to dry, some relatlvely mlld form of heat or alr flow may be helpful to dry the e ~~~.

2186~61 95J03'`2~
~ IPEA/US o MAR 1996 After the UHMWPE, ,_ --t has been drled, the UHMWPE
component should be placed in a standard vacuum chamber and exposed to a vacuum of between about 10-l - 10-5 torr, prefersbly about 10-3 torr, for a tlme sufficient to remove residual solvent, preferably 5 at least 8 hours. Thereafter, the UHMWPE component is ready for assembly into an end product and/or for any further treatment~s) that may bQ'required before use, e.g., sterilization.
DLC COATING OF METAL COMPONENT
The method for treating a metal alloy to provide a diamond-lO like coating (DLC) uses ion beam assisted deposition of silicon,followed by deposltion of DLC. This method i8 belleved to form st_ong interatomic bonds across the DLC coating-substrate interface. In order to Icnit the successlve layers of metal-silicon-DLC together effeotively, it is necessary to supply a bond-15 interface for the metal-silicon bond as well as for the sllicon-DLC
bond. Without limiting the present invention, it is believed that the present method achieves this result by forming strong interatomic bonds having a character that is intQ ~ te between the type of bond that exists between the atoms in the metal and the 20 type of bonds in the 8ilicon. Preferably, a metal substrate is used that forms a strongly-cohesive sil ~ri~ -that is, an intermetallic compound in which the bonding is partially metallic and partially covalent. Metal subs~Lal~s that form strongly-cohesive qil~ lQ5 include cobalt, nickel, titanium, zirconium, 25 chromium, molybdenum, Lu--y~ , platinum, and palladium.
~MEND~D SHEtl 2 1 8~
~ WO 95/26169 r~
~ 1 After conventlonal cl e~n1 nçl of the , t to remove superficlal contaminants, such as grease, the, , ~ 18 placed - ln a vacuum chsmber that has been ~Yacu~ed to a base L~esgu ~ of preferably less thân lO-s torr. The ~ then 15 t ' ~ d 5 wlth lons, preferably argon lons, at an energy range between about 10 - lO0 keV, preferably around lO keV. Thi8 ion ~ '- ,' ~
provldes an effective means to remove some of the Ll In~ng orb~ atoms from the surface.
The ~ is heated, praferably to a t , alUL-: of about 10 300C, or, if the materlal ls t - al.uL~: sensltlve, to the highest ~cceptable t, a~ULe acceptable for that material. Sllicon then i8 deposlted onto the _ ~ uslng known means. A preferabls means ls to posltlon the w~rkr~eoe dlrectly over the volatlll2atlon hearth which is maintained at a preferred t ,~ a~uLe of about 750C (1382F), until a preferred coating th~r~neQ5 of between 100-200 nm has been achleved. The th1 ~kn~cc, of the coatlng may be monitored by standard methods, e.g., using the rLe~u~ ,y charge of a quartz crystal oscillator.
The ~ preferably is simult~ne~ Qly bombarded with ~n 20 eneLS~ c beam of ions, preferably argon ions, at ~n energy range between 500 eV to 100 keV, preferably between 10-20 keV, in order to form a layer of metal s~l~o~ at the metal-sllicon i~ rr__e.
Although argon lons are preferred, other suitable ions may be used, such as nitrogen, argon, hy-lL~)yc:l~, silicon, methane, helium, or Wo 95/26169 2 t 8 6 2 ~ l r~ g ^~

neon, h2vlng an energy between 500 eV to 100 keV, preferably 10-30 keV. The ion-to-atom ratlo should be sufficlent, preferably at least 1 ion to 10 silicon atoms, to form a layer of metal ~ r1de 2~t the metal-slllcon lnterface.
Thereafter, thc ~ ~ - t ls cooled to about ~0C, preferably without removing the ~ t from the vacuum chambQr, and the rll2 ' like carbon tDLC) is deposited, preferably using energetlc ion be~m deposltion techniques. The DLC preferably should be deposited by vapori~ing a ~ u~-or~ such as polyphenyl ether, and 10 ~on~Pn5~n~ the precursor onto the surface of the _ _- t. using known means. At the same tlme, the ~r t should be ~ '~ 13d, ~ither in a contlnuous or lnLt:..u~.ed fsshion, with an enerS~eti¢
beam of ions. Preferable ions are nitrogen, argon, ~.yd..,ge.., ~ilicon, methane, helium, or neon, having an energy between 500 eV
to lOO keV, preferably 10-30 keV. The procedure iQ contlnued untll a th~ nPs~3 of DLC between about 100 nm-lO microns ls achieved.
EXAMPLES
rFUF111-T. EXpFI~TMFUTAL rr~ J cS
r~ Test Marh~n-~
The wear test machine used in the followlng, 1~5 provided a versatlle means for materlal wear testlng. The machlne, which i8 shown in FiSI. l, was capable of applylng user deflned load proflles ln a temperature controlled envlronn2ent. Brlefly, the wear test ~achlne inrlu~ed a reclprocatlng table lO controlled by a drlve ll.
On thc reclprocatlng table 10 was a water bath 12 whlch held the 2 1 86~ 1 Wo 95/26169 , ~ J,~ 7 ~3 plate 14 to be tested. A pln 16, described more fully below, was retained stationary adjacent to or abutting the plate 14 or sample to be tested. The .. ~ ~nd weight exerted on the plate 14 by the pin 16 was controlled by a force I,~ c~ er 18 and a dead weight 20, as described more fully below.
Reci~rocatlna Motion Linear reciprocating motion was a~ hPd by means of the DC servo motor drlve 11 (configured in veloclty mode) dlrectly coupled to a ball screw drlven rall table 10. Stroke length was detPr~T~1np~7 ~hy two magnetlc control swltches mounted on the front of the machlne. A trAre701dAl velocity profile was sent to the motor as a voltage from a Mackintosh IIci equlpped wlth a d~t~
acqulsltlon board and software using a LabVIEW software dev system (Natlonal In,,L, ~, Austin, TX). Thls trnre~o~
velocity proflle wa~ under closed loop control uslng the magnetic control switches as posltlon feedback. The speed of reciprocatlon was controlled by the magnltude of the trAr~701 ~Al voltage slgnal sent from th~ Mac ITci.
Load Prof i les Static loads were applied using ~ead welghts 20 placed on the end of a stainless steel beam 22. The normal force on the pin 16 wa8 calculated to be 1.9 tlmes the weight of the dead weights 20 on the end of the beam 22 . The maximum static normal force was 43 . 7 pounds (194.4 N) on the pln [23 pounds tlo2-3 N) located at the end of the beam 22].

WO 95/26169 2 1 8 6 2 ~ 1 r~

Force ll~du~
Frictlonal force recordings between the pln and the plate were obtained from a pie70electric, voltage mode force ~ C.7,"~..
Since thQ capacity of the force ~ cJ...:t.r was 100 lbs (4g4.8 N), 5 very small force mea~u,~ I s could be inac;~u-a-~. Normal force recordings on the pin 16 were obtalned from a p1 e70~ ctric, voltage mode force ~.àns-lu~er 18. Its capacity also was 100 lbs (444.8 N).
CYcle Count The number of cYcles was counted in two places. On the Mac IIci, a sor~wc,-~ counter was incremented each tlme the left magnetlc control swltch was closed. As a hardware backup ~ rm~ a 8epe--Ate magnet~c swltch (whlch used the same power supply as the control swltches) was located on the back of ths wear 15 machlne to in.;.~ t an electronic counter.
T~ a~u.~
The t ~~ atu-~ of the wear envlronment could be regulated by means of a water bath 12 whlch ~u~.ou~.ded the stalnless steel wear chamber 24. The water bâth 12 was heated by a stalnless steel 0 lmmerslon heater controlled by a temperaturQ controller. Water u.e was measured wlth an RTD probe placed ln the bath ~ .
Slnce lt was r7~9~ r?lhl e for the wesr mschlne to operate overnight and on weekends without superYision, a number of safety 25 -- -n~ were used.

2 1 8626 t Wo 95/26169 Software Control ~he , _ ~er program whlch drove the sy8tem had a n stop swltch" on the screen whlch enabled the program to be msnually stopped at any tlme. Also, the program contlnually sampled from 5 two addltlonal magnetlc stop swltches inslde the reclprocatlng rall table. Therefore, ln the event of the stroke length being ~- ~e~ , one of these switches would close and the program would be stopped. These two stop swltches were located sllghtly outslde of the deslred stroke length of the wear test twhlch was dete~n-~d 10 by the two rnagnetlc control swltches prevlously ~ c~ ed ) .

Ha~h~.Le Control rVocated outslde of these stop swltches were two lndustrlal cnl push button swltches placed in serles between the mllln power source and the motor drives. These switches normally were 15 closed.
If the reclprocatlng table lO traveled too far beyond the deslred stroke length of the wear test, one of the magnetlc stop swltches closed, whlch caused the ~ l program to send zero volts to both DC motors ll, stopplng the wear test. However, a 0 very small voltage stlll was sent to the motors ll from the ~_~er, to which the velocity-conflgured motor ll could respond.
...~Lt:~O~e, the wear chamber 24 contlnued slowly to drlft and soon enyaged one ~n~ 1 stop swltch. This ~ ja, ~ turned off the 25 maln power to the motors ll.

wo 95/26169 2 1 8 6 2 6 1 ~ 'IL t6 Evaluatlon of w~r Evaluatlon of wear resistance and the assoclated coefflclent of slldlng frlctlon was pt~ under reallstlc enYl~r ~c,l but statlc loading condltlons. Two types of wear were measured.
5 Flrst, the wear Lt~ d by 1088 of the coatlng from the rfuL~ c-~e, whlch was guantlfled ln terms of the number of cycles requlred to expose the /~uL~iL~r~-~, or ma~or portlons thereof. Th~s determln8tlon wa8 mo8t effectlvely made by mlcroscoplc obse~vaLlon~
rather than by welght 1088.
In cases where no coatlng was lnvolved, lt was useful to monltor welght loss; however, lt was lmperative to _te for envlY, Lt,l fluld uptake tgaln ln mass) of the wear sample by uslng a dummy sample presoaked and exposed to the test bath slmultaneously ~longs~ the wear ~rec~ . To ensure that trends observed in the data L~ a~ d a steady state sltuatlon, data was obtalned over a perlod .,~,y.on_l~lng 10~ cycles (c~. s~ n~ to about one year of normal ambulatlon). Later te8t8 completed about 1.25 x 10~ cycles.
Slmul~tlon of Jolnt Envlronment The envlronment of a natural ~olnt (wlth synovlal fluld) was slmulated uslng bovlne serum. Pln/plate materlals were chosen based on a ~udgment that the constralnts of the total ~olnt deslgn causa a 1 oc nl 1 7ed sector of the femoral ball to descrlbe a wide ~ath wlthln the 80cket of the acetabular ~_ t. Eventually, a8 seen ln ~olnt slmulator experlments, a m~ L~:L or more of Wo gs/26 l 69 1 ~ l, tl _ polymer may be worn away, with no measurable lOSs of materlal by the metal or ceramic ball. Therefore, these tests utllized metal, ceramlc, or coated metal pins run agalnst coated or uncoated polymeric plates.
5 PolYethYlenQ
The virgin U~;MWPE that was used in the following ~ , 1PS was obtalned from Westlakes Plastics, Lennl, PA. For ~uL}~Osas of the present study, gamma-ray sterlllzatlon was not pe~r~ ~ ~. Ii such sterilization had been pt~L' ', then the sterilization process 10 should have pco,luced a marginal 1 `O,~ ~ of - 1"Al peLL~Lllla~CI~, a8 a cuns~-~ut~ of cross-linking. 8ecause th~a surf~ce of polyethylene is ~.;L~,~..I.ed easily, all of the surfaces and materials were kept scrupulously clean using a laminar air-flow cabinet. A Struers Rotopol pol 1 ~h~-~ with a Pedamat head was used 15 to polish the samples according to the I~L~JCedUL~ given in Table I:
TABLE I
Polishing o~ Polyethylene 20Step Abrasive Applied Load Duration 1. 400 grit SiC lOON 2 min. repeat once 2. 600 grit SiC lOON 2 min. repeat once

3. 1200 grit SiC lOON 2 min. repeat once

4. 2400 grit SiC lOON 2 min. repeat once 255. 4000 grit SiC lOON 2 min. repeat once 6. l~m Diamond Spray 80N 5 min. repeat until on Pan-w Cloth* s~;Lc,~"l,es are gone ~ur~ ~ ~ c~nLoo~rd ILII- ~loth -d~ or ~

Wo 95/26169 2 1 8 ~ ~ 6 1 P~

The alloy was manufactured by C~ r Tet~hn~ y, Houston, Texas, and had the follow~n~ composition: cobalt 70/chromlum 30, wlth a mlnor sddltlon of molybdenum . Af ter r^~h ~ n 1 na, the 10 m~
dlameter test plns were vv~ouL~d at one end to a radlus of 20 mm

5 and pollsh~ by standard metallurglcal terhn~q~ he ~_UlV~U~
was ~lPs~gnP~l to prevent the edge of the pln from cuttlng lnto the surface of thE~ polyethylene flat.
We~r testlng was peL rVL ' uslng the wear test machlne de8crlbed above. The t C~ was malntalned withln the test lO chamber at 23 1 1C by means of an external water bath 12. ~he chamber cont~lnlng the sample plate 14 was recl~,Lv~;a~d beneath tha statlonary pln 16 at 1 Hz over a sl~dlng distance of 50 . 8 mm per cycle; both sllding speed and dlstance approxlmate that whlch obtalns durlng servlce wlthln a total hlp ~olnt. Cobalt--,l. l~
15 molybdenum and alumlna plns 16 were ~ nPd to a 20 mm radlus ( slmllar to that of typlcal femoral balls ) and pollshed metallogr~rh1cAlly to obtaln a surface flnlsh of less than 0.05 um R,.
Samples were 60aked ln palrg ( wear spec~ plus dummy) for 20 one week ln bovlne serum buf fered wlth sodlum azlde ln a 0 .1 vol .
% solutlon, the latter to prevent bacterlal growth. Su~sequent wear testlng was performed ln the same solutlon, and the soaked control sper~ was used to correct for fluld welght galn ln the wear sample. Samples were welghed perloA1c~Ally durlng testlng, and WO 9~/Z6169 2 l 8 6 2 6 l r~l,u~ ~03~-~

cc,...~s~o~ ns associated frlctlon coefficients (ratio of linear force to normal load) were measured during sliding.
The desired test load, and that Used for most experiments, c~ Y~ d to a stress level induced by normal body load9 at a 5 total hip interface. Since the articulating pin-on-plate wear surfaces did not conform to the degree oi' the actual ~. 0~ 918, the reguired equivalent stress loads were lower. In particul~r, using the measured plastic impression within a UHMWPE plate as a measure of contact area, the approximate equivalent ( stress ) load was 33 . 4N .
~ ests generally were run for 106 cycles, with wear defined in terms of cumulative mass (volume) loss. Samples werQ ~ ~nPd by 8~nn1ng electron mi.iLus~ y (SEM) at the cr~nc-lllq1t~n of the test.
501vent Immersion The polished samples were treated with Decalinn' for between 30 second8 at 120C and 3 minutes at 50C, as shoWn ln ~able II.
The higher t ~ aLu~ ~8 resulted in excessive ~ n~ of the polymer. From the following results, it was concluded that it mlght be poQQ~hle to use Decalin'Y at room t n, a~uLt: to achleve 20 ade~uate selective dlssolutlon of the low le~-l;qr welght polyethylene fraction.
For CoCr-DLC agalnst Decalln'Y-treated polyethylene, as w811 a8 for CoCr-DLC agalnst u~ a-~d polyethylene, wear vlrtually was zero for more than one mllllon cycle8; however, ma~or wear was 25 obseLv~d for CoCr agalnst u..~ ~a-ed UHMWPE. In addltlon, lt was Wo 95/26169 2 1 8 6 2 ~

ob~eLvtd that wear of CoCr on hlghly E~ol1~hed UHMWPE only oc~;uL,~d following an lncubstion period of approximately 500,000 cycles, versus the immedlate wear ~ d for CoCr agalnst rough pol~sh~
(commerclal quallty flnlsh) polyethylene. Unro-Lu--aLely, once wear 5 beglns, the we~r can be so rapld that the cumulatlve wear volume8 ior rough and smooth surfaces are essentlally equal by approximately 1. 5 milllon cycles. Therefore, the advc~ e achleved by EY~ h1 nrJ alone is short-lived, and ultimately ~ nr " c_.lU~ Llal, due to the devastating wear rate associated with 10 fibrillar pullout/failure.
~ n both cases, wear surfaces exhibited fibrillar pullout and mlcrofallure. ~hese obs~Lvc,Llons are quantified in Fig. 2, which shows actual wear rate and wear factors (wear volume n~'r~ 7~d for load ) for all cases . These wear rates and wear factors for the 15 CoCr-UHMWPE cases are in close a~L~ t with dat~ gene~sLed under similar condltlons and iepuL L~d ln H. A. McKellop, et al., "Wear Characterlstlc8 of UHMWPE," J. Blomed. Mater. Re8., 12 (1978) 895.
For the mlnlmal wear cases, a groove caused by creep defoL,.,~,Llon wa8 pLc. luced ln the plates, but no welght loss was 20 m~ . At 10~ cycles, sllding contact surfaces of the Decalin'Y-treated UHMWPE had become gently undulating, with fine-scale, very flat deformation (but not wear) marklngs superlmposed on the undulat$ons. Tha orlginal surface was flat (no undulations) and covered with fine-scale polish r~rk~n~. The ~atter ~ul~LLasL~ with 25 the orlglnzl polished surface of u~lLL~a~ed polyethylene, which wa8 Wo 95/26169 2 1 8 6 2 6 1 r~ O~t~i7~;

characterized by flbrlllar ~LLU-;-UL~S 1mllar to those observed durlng equllibrlum wear.
Although llkewls~ cha-a~;Ll:LLzed by zero - _Led we~lr for ~t least 1.25 x 106 cycles, the contact mluLu~oyoyLoylly for smooth-5 psll5h~d DLC agalnst CoCr-DLC had a markedly different ayyeOLo~
Undulatlons were not 80 o~p~L~ t~ and materlal appeared to have been pulled out of the surface. However, hlgh r^gn~Lcatlon study of these blLu~uLe8 lndlcated that they were falrly flat, unllke the sharply-peaked, tensllc-failure, flbrlllar ~Luu~uL~8 observed 10 for u~coa~d CoCr agaLnst TJ~rMWPE. Thls ~ugg~s~.~d that the forum 1 r~ probably dld not yet cuLL~syu.~d to wear ~de~ from the substrate ) .
The results of the test, glven ln Table II, show that after 106 cycles, no flbrlls were present at the wear surface of Decalln~
15 treated polyethylene:

WO95126169 2 f ~ 626 1 P~

TABLE I I

PlnPlate Load ~ N ( cycles ) Wear (N) CoCr-DLC UHMWPE 33 . 4 0 .12 537, 498 ~ O
Decalinn' Pretreat 308 at lOO-C
5 CoCr-DLC UE~MWPE 33 . 4 0 .135 1, 000, 610 ~ O
Decalin'Y
Pretreat 308 at 100 C
CoCr-DLC U~WPE 33 . 4 0.11 1, 000, 000 - o DecalinlY
Pretreat 1808 at 50-C
CoCr-DLC U~MWPE 33 . 4 0 . 08 1, 000, 000 ~ o CoCr UHMWPE 33 . 4 0 . 09 1, 000, 000 maJor ( rough polish ) CoCr U}~WPE 33 . 4 0 .14 1, 000, 000 maJor, (smooth polish) f~l l~t-1n~
~ ncubatlon perlod of -5 x 105 cycles 10 CoCr-DLC UHMWPE 33.4 0.09 1,205,000 ~ O
Similar results were obt3ined at 1. 25 x 10~ cycles.
DLC CoatinQ
Fys~ e 1 A DLC coatlng of approximately 1 mlcron ln ~h~ n~ss was prepared by nitrogen ion bombardment of a polyphenyl ether pL~CuL~uc. The alloy was cleaned in lsopropyl alcohol prior to coatin~. Isopropyl alcohol was chosen because lt leaves few, lf any, resldueq. Wear testlng revealed that, under some ¢iL~ Ldnces, there could be a loss of ~lh~S~on of the coatin~7.
- ExamT le 2 A later batch of four CoCr pins was treated using a bond-coat 5 of sillcon. Sllicon wss chosen because (a) DLC i8 known to adhere better to sillcon that any other ~iub2~L~Le~ whlch ls attributed to strong SiC bonds formed at the ln~éLr~_e, and (b) cobalt and, to a lesser extent, cl.r~ 1 are known to form 8~ q (CoSi, CoSi2, CrSi~) at the int~La_e if the L clLuLe exceed_ about 300C.
10 This solid state reaction is known to be ~nhAnre~l by ion assisted depositlon, probably because energetic ions disrupt the surface oxlde or other barriers to interdlffusion at the metal-slllcon ln~eL raC:~ .
Therefore, a coatlng process was chosen which was an argon 15 lon-assisted depositlon of approximately 100 nm of slllcon at 300C. The sillcon was ~va~oLâted from an electron-beam heated hearth and the ~h~ ~kn~ was controlled by means of a guartz crystsl fllm thl~kn~ss monltor ( Int~ LL IC8 Ltd. ) . DLC was deposited, ln a ~e~L~e run. After lnltlal ion L ' - ~ in the 20 vacuum chamber to remove the L. 1n~ng a~l-orbe~l atoms from the surface, the alloy was heated to a temperature of 300C. The DLC
s achieved was approximately 0.5 microns.
In prolonged wear tests, at a contact pressure of 6.9 MPa against I~MWPE under serum, l.e., load and envlronmental condltions Wo 95/26169 2 1 ~ 6 2 6 I T

equlvalent to walklng, no ~cnhPRlon or 10s8 of DLC was obserYed after 1 25 milllon reciprocated wear cycles ConcluslQns DLC coated CoCr against polyethylene clearly represents a onnR~ rable ~ a~ over ul ~ua-~d CoCr against polyethylene.
I5~ over, ~L~L-~,~t ~ of the UHMWPE with Decalinn' provldes additional wear resistance. Cleârly, the pre-wear damagQ
-- I have been altered, and it is belleved that wear may be iurther po~ L~. ed and, posslbly, reduced ln rate once~ lt beglns by use of the present methods.
From the .t,~ea-ance of the DLC-UHMWPE cont~ct surface, lt appears that lorA~ A~h~Rlnn takes place as the DLC rlder passes oYer the polyethylene surface, possibly pulling out the soft (lower ler~IAr weight) constltuent on a gradual cycllc basls However, the adheslve contact force is insufficient to f~il the pulled out ml;.u~.u~;Lu.~, at least before 1 25 x 10~ cycles The prewear ~L-u~;~u~s shQwn above do not , ` le the lndlvldual "mountaln peakr LU~OY~R~ assoclated wlth submiv- L~ partlcle p-uluvLlon;
therefore, lt ls po~R~hIe that a different particlQ generatlon -n1 ~m~ hence a dlfferent partlcle slze distributlon, wlll prevall. Slnce osteolytic de~-uvLion of the bone-blomaterl~l inL~.~ac;e is lnduced pr~nn1r~11y by submiv-, ter particles, the productlon of larger average slze particles may result ln reduced osteolytic deaL-u-;LlOn. The poss1h1Ilty for ~ v. ~ seems 25 y~ea~ for DLC against solvent-immersed UHMWPE Here, soft-phAse 2 1 862~ 1 ~ Wo 9s/26l69 P~~
2 v~
flbrillatlon was ellmlnated, there ls a hlgh 111~ hood that wear can be po5~v-~ed further, that the eventual wear r ~ and rate wlll be altered, and that the wear particle slze dl5trlbutlon wlll cv--e~Jv~d to larger partlcles, l.e., on the order of the hard 5 semicrystalline domains. Ilv~evver~ slnce Decalln''l pe,~ tes a s1~n~f1~Ant fractlon of a m~ er lnto the UHMWPE, these effects should perslst ~l.rvuyl.vu ~ the wear process .
One of sklll ln the art wlll reco~n~7s that many If ~m~tlonS
may b~ made to the present lnventlon wlthout departlng from the 10 splrit and scope of the present lnventlon. The ~e~mr~hed hereln ls meant to be lllustratl~e only and should not b~
taken as llmltlng the lnventlon, which 18 deflned ln the followlng clalms .

Claims (24)

we claim:

1. A process for treating an ultra-high molecular weight polyethylene, component to reduce frictional shearing off of fibrils from said component comprising:
immersing said ultra-high molecular weight polyethylene component in an organic solvent for a first amount of time and at a temperature sufficient to dissolve polyethylene fibrils that are susceptible to frictional shearing off during use but insufficient to result in damage due to swelling of said ultra-high molecular weight polyethylene component, said organic solvent being selected from the group consisting of an aromatic hydrocarbon, an alicyclic hydrocarbon, an aliphatic hydrocarbon, and a mixture thereof; and exposing said component to a vacuum of 10-1-10-5 torr for a second amount of time sufficient to remove residual solvent.

2. The process of claim 1 wherein said first amount of time comprises at least 30 seconds.

3. The process of claim 1 wherein said second amount of time comprises at least 8 hours.

4. A method for coaling a metal alloy substrate with diamond-like carbon comprising:

exposing said metal alloy substrate to a vacuum of at least about 10-5 torr;
heating said substrate to a first temperature of about 300°C or, if said metal alloy is temperature sensitive, to a highest temperature acceptable for said metal alloy;
depositing silicon onto said substrate in an amount sufficient to form an inner bonding layer of metal-silicide cohesively bonded to an outer layer of silicon;
substantially simultaneously bombarding said deposited silicon with a first energetic beam of ions at a first energy, a first ion density, and for a first amount of time sufficient to form said inner metal-silicide bonding layer cohesively bonded to said outer layer of silicon;
condensing a diamond-like carbon precursor onto said outer layer of silicon at a second temperature and for a second amount of time sufficient to form a film of precursor molecules on said outer layer of silicon;
substantially simultaneously bombarding said diamond-like carbon precursor with a second energetic beam of ions at a second energy, a second ion density, and for a third amount of time sufficient to form an inner silicon carbide layer cohesively bonded to an outer coating of diamond-like carbon.

5. A method for coating a metal alloy substrate with diamond-like carbon comprising:
providing a substrate comprised of a metal alloy selected from the group consisting of cobalt, nickel, titanium, zirconium, chromium, molybdenum, tungsten, platinum, palladium, and combinations thereof:
exposing said substrate to a vacuum of at least about 10-5 torr;
heating said substrate to a first temperature of about 300°C or, if said metal alloy is temperature sensitive, to a highest temperature acceptable for said metal alloy;
depositing silicon onto said substrate in an amount sufficient to form an inner bonding layer of metal-silicide cohesively bonded to an outer layer of silicon;
substantially simultaneously bombarding said deposited silicon with a first energetic beam of ions at a first energy, a first ion density, and for a first amount of time sufficient to form said inner metal-silicide bonding layer cohesively bonded to said outer layer of silicon;

condensing a diamond-like carbon precursor onto said outer layer of silicon at a second temperature and for a second amount of time sufficient to form a film of precursor molecules on said outer layer of silicon:
substantially simultaneously bombarding said diamond-like carbon precursor with a second energetic beam of ions at a second energy, a second ion density, and for a third amount of time sufficient to form an inner silicon carbide layer cohesively bonded to an outer coating of diamond-like carbon.

6 . The method of claim 5 wherein said first and second beam of ions comprise ions selected from the group consisting of nitrogen, argon, hydrogen, silicon, methane, helium, neon, and combinations thereof.

7. The method of claim 4 wherein said second beam of ions comprises nitrogen ions.

8. The method of claim 5 wherein said second beam of ions comprises nitrogen ions.

9. The method of claim 4 wherein said first energy and said second energy are between about 10-30 keV.

10. The method of claim 5 wherein said first energy and said second energy are between about 10-30 keV.

11. The method of claim 4 wherein said second temperature is about 80°C.

12. The method of claim 5 wherein said second temperature is about 80°C.

13. The method of claim 7 wherein said second temperature is about 80°C.

14. The method of claim 4 wherein said silicon is deposited onto said substrate to a thickness of between about 100-200 nm.

15. The method of claim 5 wherein said silicon is deposited onto said substrate to a thickness of between about 100-200 nm.

16. The method of claim 6 wherein said silicon is deposited onto said substrate to a thickness of between about 100-200 nm.

17. A metal alloy component produced by a process comprising:
placing a component comprised of a metal alloy selected from the group consisting of cobalt, nickel, titanium, zirconium, chromium, molybdenum, tungsten, platinum, palladium, and combinations thereof, in a vacuum chamber at a pressure of less than about 10-5 torr:
heating said metal alloy component to about 300°C, depositing onto said component a layer of silicon having a thickness of between about 100-200 nm, substantially simultaneously bombarding said component with a beam of argon ions having an energy of between about 10-30 keV;
cooling said component to about 80°C;
condensing a diamond-like carbon precursor onto the surface of said component;
substantially simultaneously bombarding said component with an energetic beam of argon ions at an energy between about 10-30 keV whereby a coating of diamond-like carbon accumulates on said component.

18. A method for making an orthopaedic implant comprising a metal alloy substrate in frictional contact with an ultra-high molecular weight polyethylene component, said method comprising the steps of:
immersing said ultra-high molecular weight polyethylene component in an organic solvent for a first amount of time and at a temperature sufficient to dissolve polyethylene fibrils that are susceptible to frictional shearing off during use but insufficient to result in damage due to swelling of said ultra-high molecular weight polyethylene component, said organic solvent being selected from the group consisting of an aromatic hydrocarbon, an alicyclic hydrocarbon, an aliphatic hydrocarbon, and a mixture thereof;

exposing said component to a vacuum of about 10-3 torr for a second amount of time sufficient to remove residual solvent;
exposing a substrate comprising a metal alloy selected from the group consisting of cobalt, nickel, titanium, zirconium, chromium, molybdenum, tungsten, platinum, palladium, and combinations thereof, to a vacuum of at least about 10-5 torr;
heating said substrate to about 300°C;
depositing silicon onto said substrate to a thickness of between about 100-200 nm;
substantially simultaneously bombarding said deposited silicon with a first energetic beam of nitrogen ions having a first energy of between about 10-30 keV at a first ion density and for a third amount of a time sufficient to form an inner bonding layer of metal-silicide cohesively bonded to an outer layer of silicon;
cooling said substrate to about 80°C;
condensing a diamond-like carbon precursor onto said outer layer of silicon at a second temperature and for a fourth amount of time sufficient to form a film of precursor molecules on said outer layer of silicon;

substantially simultaneously bombarding said diamond-like carbon precursor with a second energetic beam of nitrogen ions having a second energy of between about 10-30 key at a second ion density and for a fifth amount of time sufficient to form a silicon carbide bonding layer cohesively bonded to an outer coating of diamond-like carbon.

19 . The process of claim 1 wherein said temperature is between about 30-100°C.

20. The process of claim 1 wherein said temperature is between about 30-50°C.

21. The process of claim 2 wherein said temperature is between about 30-100°C.

22. The process of claim 2 wherein said temperature is between about 30- 50°C .

23. The process of claim 3 wherein said temperature is between about 30-100°C.

24. The process of claim 3 wherein said temperature is between about 30-50°C.