Titanium and hafnium based coatings in this work were sputtered on steel substrates in a r.f. Sputtering unit with a diode arrangement. The results for hardness, deposition rate, internal stress, adhesion, structure, phase analysis and wear show that the properties of HfNx coatings are less influenced by the deposition process than the coatings of TiNx.
Original Description:
Original Title
Comparison of Fundamental Properties of r.f.-sputtered TiNx and HfNx Coatings on Steel Substrates
Titanium and hafnium based coatings in this work were sputtered on steel substrates in a r.f. Sputtering unit with a diode arrangement. The results for hardness, deposition rate, internal stress, adhesion, structure, phase analysis and wear show that the properties of HfNx coatings are less influenced by the deposition process than the coatings of TiNx.
Titanium and hafnium based coatings in this work were sputtered on steel substrates in a r.f. Sputtering unit with a diode arrangement. The results for hardness, deposition rate, internal stress, adhesion, structure, phase analysis and wear show that the properties of HfNx coatings are less influenced by the deposition process than the coatings of TiNx.
E L S E VI E R Surface and Coatings Technology 74 75 (1995) 135-142
Compari son of fundamental properties of r.f.-sputtered Ti Nx and Hf N x
coati ngs on steel substrates G. Berg, C. Friedrich, E. Broszeit, K.H. Kl oos Institute of Materials Science, Technical University of Darmstadt, Grafenstrafle 2, D-64283 Darmstadt, Germany Abstract Titanium and hafnium based coatings in this work were sputtered on steel substrates in a r.f. sputtering unit with a diode arrangement. By feeding nitrogen in addition to the inert processing gas argon, TiN x and HfN x coatings were deposited under reactive sputtering conditions. Deposition pressure, nitrogen flow in the chamber and bias voltage on the substrates were systematically varied to get optimized coating systems. The coatings were investigated on their fundamental properties in order to compare titanium and hafnium based films, sputtered under identical process conditions. The results for hardness, deposition rate, internal stress, adhesion, structure, phase analysis and wear show that the properties of HfNx coatings are less influenced by the deposition process than the coatings of TiNx in the field of selected deposition parameters. The behavior of HfN x is comparable to TiNx, owing to a very similar crystal structure of the materials, but the conclusion of this present paper the well known TiNx has to be preferred for tribological applications in mechanical components. Keywords: PVD coatings; Sputtering; Coating properties; Titanium nitride; Hafnium nitride 1. Introducti on Ha r d coat i ngs deposi t ed by means of physi cal va por deposi t i on ( PVD) are well est abl i shed in wear pr ot ect i on in addi t i on t o ot her appl i cat i ons, e.g. har d coat ed cut t i ng t ool s wi t h l ong lifetime [ 1] . I n this cont ext Ti N films are well known for qui t e hi gh har dness and sui t abl e adhesi on whi ch are i mpor t a nt in the field of t ri bol ogy. Thi s l eads t o ext ended use of Ti N for t he wear pr ot ect i on of machi ne c ompone nt s and tools. For ma ny years Ti N has been a s t andar d coat i ng for a wide r ange of subst r at e mat eri al s. Si mi l ar pr oper t i es t o Ti N can be expect ed by usi ng coat i ngs based on met al s si mi l ar t o t i t ani um, for exampl e haf ni um I - 2- 4] . Ref. [ 5] r epor t s very hi gh har dness of Hf N coat i ngs of over HV 6000. Besides this, Oakes r epor t s in Ref. [ 6] l owest wear for Ht N coat ed cut t i ng t ool s. For this r eason the basi c mechani cal and t r i bol ogi cal pr oper t i es of haf- ni um based ni t ri des in compar i s on t o t i t ani um based ni t ri des are the subject of t hese i nvest i gat i ons. The coat i ngs in this pa pe r were pr oduced under i dent i cal sput t er i ng condi t i ons in or der t o give al so an i mpr essi on of the different deposi t i on behavi or of Ti Nx and Hf Nx usi ng the same deposi t i on process. Hf and Ti .... f ound in the same gr oup of the per i odi c 0257-8972/95/$09.50 1995 Elsevier Science S.A. All rights reserved SSDI 0 2 5 7 - 8 9 7 2 ( 9 5 ) 0 8 2 9 9 - 9 t abl e of el ement s - - show, for exampl e, the same hexago- nal cryst al st r uct ur e and f or mal oxi dat i on states, whereas the densi t y as well as ot her pr oper t i es of the pur e met al , like the at omi c mass, differ. In or der t o have a br oa d field for the i nvest i gat i ons and the results, the deposi t i on par amet er s of ni t r ogen flow, bi as vol t age and pr essur e were changed. These mai nl y affect the f undament al pr oper t i es of sput t er ed PVD coat i ng syst ems [ 7] . The pr oper t i es shown in the paper were exami ned wi t h c ommonl y used met hods of surface anal ysi s and t r i bol ogy [ 8] . 2. Experi ment al detai l s 2.1. Sputtering system and parameters.for coating deposition The coat i ngs were pr oduced by means of a commer ci al r.f. sput t er i ng uni t (Alcatel SCM 601). The gl ow dis- char ge was suppl i ed by a gener at or of 13.56 MHz wi t h spl i t t abl e power in or der t o per f or m sput t er cl eani ng of t he subst r at es as well as r.f. bi as sput t eri ng. All coat i ngs were pr oduced wi t h the s ame geomet r i cal condi t i ons in t he sput t er i ng unit, the same r.f. power 136 G. Berg et al./Surjace and Coatings Technology 74 75 (1995) 135 142 and an equal time of sputtering. Also the cleaning of the substrates, made of ball beari ng steel (SAE 52100; DI N 1.3505) and casting steel ( DI N 1.0443), using ul t rasoni cs and special cleaners based on al cohol as well as sput t er cleaning in pur e ar gon at mospher e ( Us = - 6 0 0 V and p = 1.5 Pa for 20 min) was unchanged. A t hi n i nt ermedi at e l ayer of the pur e met al was deposi t ed fol l owed by a thin gradi ent coat i ng of ni t rogen up t o the desired film paramet ers. The field of deposi t i on par amet er s is described by the ni t rogen flow beginning with 0 cm 3 min 1 up t o 7.5 cm 3 mi n- 1 which allows several phases t o be examined. The var i at i on of the bias vol t age has changed from floating pot ent i al (approx. +2 0 V), gr ound pot ent i al t o a nega- tive bias vol t age of - 40 V. The pressure of the worki ng gas increased onl y from 1.0 Pa t o 2.0 Pa. values shown l at er on were t aken under l oadi ng condi - tions. In this case elastic and plastic def or mat i on are measur ed t oget her. Ther ef or e the values for t he hardness are not so high as the values represent i ng onl y the plastic deformat i on. 2.2.4. X-ray diffractometry X-ray diffraction was used t o st udy the phases and t o measure i nt ernal stress on the surface. The equi pment was a 0- 20- goni omet er (Siemens D 5000) with Eul er cradl e for ~b-rotation, with grazi ng i nci dent -angl e 7, secondar y monoc hr oma t or and scintillation count er for peak det ect i on. The i nt ernal stress was measur ed by the sinZ@ met hod and the peak posi t i ons were eval uat ed by using the cent er of gravi t y met hod. 2.2. Characterization of coatings It is necessary t o define some abbrevi at i ons for the nitrides of t i t ani um and hafnium. Keepi ng in mi nd t hat all coat i ngs pr oduced were mul t i phase coat i ngs (except the coat i ngs based on the pur e metals) Ti N and Hf N demonst r at e t hat the layers are domi nat ed by the stoi- chi omet ri c phase. Ti N x and HfNx always means a field of mul t i phase coatings based on t i t ani um and hafni um nitrides with different met al t o ni t rogen rat i o (ot her aut hor s use e.g. Ti - N or Ti (N)). 2.2.1. Deposition rates by the ball crater method The thickness of har d coat i ngs influences di rect l y the possible removal of the surface under t ri bol ogi cal l oad- ing. To anal yze the thickness of t hi n films and to get the deposi t i on rates rel at ed to special deposi t i on paramet ers, the ball cr at er met hod [-8] allows pr ompt and sufficiently precise results t o be obt ai ned. A cr at er is gr ound i nt o the coat ed specimen by a steel ball of defined diameter, wet t ed by a di amond suspension on an al cohol base. The concent ri c circles are measur ed by mi cr oscopy and the thickness is cal cul at ed f r om these data. 2.2.5. Scratch testing For the eval uat i on of adhesi on, scrat ch testing was selected. A commerci al CSEM Revetest scrat ch t est er was used with a Rockwel l C i ndent er whi ch was dr awn over the coat i ng surface with increasing force. Dur i ng scrat chi ng on the one hand frictional force bet ween di amond and t he coat i ng and on the ot her hand acoust i c emission were cont i nuousl y moni t or ed. The results were t hen verified by opt i cal mi croscopy. 2.2.6. Plate on cylinder wear test Li near wear and friction coefficient were studied using a plate on cyl i nder t ri bomet er. The coat ed plates made of casting steel were tested against gr ound cylinders made of SAE D2 ( DI N 1.2379). Nor mal force, testing speed and sliding di st ance were varied, while dr y friction at an ambi ent t emper at ur e of 23 + 3 C and relative humi di t y of 50%_+6% were par t of the fixed test condi t i ons. A di st ance sensor moni t or ed the l i near wear and a force sensor i ndi cat ed the friction coefficient. The results concer ni ng linear wear coul d be verified by pr of i l omet r y plots of the tested coat i ng systems. 2.2.2. Morphology and thickness of the coatings by SEM The mor phol ogy of the coat i ngs was anal yzed by means of scanni ng el ect ron mi cr oscopy (Zeiss DSM 962). The specially pr epar ed specimens for these investi- gations were br oken and the fract ure surfaces were inspected. The thickness of the coat i ngs was also mea- sured with SEM l ooki ng at right angles t o the fracture. 2.2.3. Dynamic ultra microhardness testing The hardness was measur ed with a dynami c ul t ra mi cr ohar dness t est er (Shi madzu DUH 202) using a Vickers di amond, a force of 100 mN, a l oadi ng speed of 6.75 mN s 1 and a resting t i me of 10 s. The hardness was anal yzed as a funct i on of force and depth. In cont r ast t o the commonl y used Vickers hardness, the 3. Res ul t s 3.1. Influence of nitrogen flow on coating properties 3.1.1. Color The first r emar kabl e influence of the ni t rogen flow on the propert i es of the investigated har d coat i ngs is shown by the col or of the films. Whereas Ti Nx coat i ngs st ay gol den when the Ti N phase is built, the Hf N x coat i ngs onl y show a gol den surface for a small range. At hi gher flow rates for ni t rogen the films became dar k grey, in cont r ast t o the pale grey in the range of l ower ni t rogen values. G. Berg et aL /Surface and Coatings Technology 74-75 (1995) 135-142 137 1 6 0 1 4 0 1 2 0 1 0 0 8 0 6 0 4 0 2 0 0 d e p o s i t i o n r a t e Deposition rate calculated from thickness [ n m l m i n ] ~ ~ values measured by bal l crater testing h e x a g o n a l ~ I I . " ~F~7" * r ~? II , . i c e HfN =.= Ti TIN0.3 ~ "=---~-- - ~ " - - - ~ I ~ " 3 9 . 9 - - . . . . . . ~ nm/ mi n cubi c l at t i ce ~ I system: Alcatel SCM 601 ~ V Ti N =.. I 4 3 5 s u b s t r a ~ e t seperation: 65 mm ~ nm/ mi n power: 750 W; pressure: 1.5 Pa bias voltage: -20 V; time 160 min i w w 0 1, 2 2 3 4 7 , 5 n i t r o g e n f l o w [ c m3 / mi n ] Fi g. 1. I nf l ue nc e of n i t r o g e n f l ow on t he d e po s i t i o n r at e. 3.1.2. Deposition rate The deposition rate is strongly influenced by the sputtered metal first of all; the rates of Hf based coatings are much higher then the rates of Ti based coatings. Fig. 1 shows that the thickness values for Hf based coatings are two to five times higher than for TiNx layers of the studied parameter field. According to the phases of the analyzed coatings, the deposition rate decreases clearly when the stoichiometric TiN and HfN phases with cubic lattice are detected. The hexagonal lattice of the pure metals and the following substoichio- metric nitrides at low nitrogen flow do not influence the deposition rate in a characteristic way. 3.1.3. Structure and morphology The phases analyzed by X-ray diffraction show the hexagonal crystal lattices of the pure metals Ti and Hf. The cubic phases TiN and HfN only exist at exactly defined, rather high nitrogen flow rates. At these flow rates of 3 cm 3 mi n- 1 for TiN and 4 cm a mi n- 1 for HfN the best crystallinity is also found. The JCPDS standard files sorted to the measured systems of coatings at low nitrogen flow show further substoichiometric nitride phases for both metal-nitrogen compounds correspond- ing to TiNo,3 (ASTM card 41-1352) and HfNo, 4 (ASTM card 40-1277). Tetragonal phases like Ti2N were not detected by X-ray diffraction. The reason could be that they occur only at a special deposition temperature, which was not strictly controlled. Increasing nitrogen flow leads to coatings showing an extremely dense and fibrous structure as shown in Fig. 2 for the case of HfN x. The morphology of HfN x films looks similar to that of TiN x, whereas the coating of the pure hafnium appears more dense. defined by the phases. In these investigations a compres- sive stress state is always detected. Lowest stress was detected for the pure metals Ti (approx. 0 N mm -2) and Hf ( - 1000 N mm-2). Increasing nitrogen flow during the deposition process - - restricted to a single phase - - leads to higher internal stress values probably caused by additional nitrogen located in the lattice. Generally stoichiometric phases show lower compressive stress values than nonstoichiometric phases. 3.1.5. Adhesion Very important for any application of hard coatings are the results on adhesion of the coatings. With respect to the fact that the coatings were produced with the same intermediate layers, the varying of the nitrogen flow rate shows clearly that the coatings with stoichio- metric phases are best. As pointed out in Fig. 3, nonstoi- chiometric films studied in the range of the deposition parameters could not reach this optimum. 3.1.6. Hardness The hardness of PVD coatings, depending on the deposition parameter nitrogen flow, seems to be mainly influenced by the phases detected with XRD. Highest hardness values (note that the hardness is measured under loading conditions) were detected by analyzing the stoichiometric phases TiN and HfN and also coatings with lower nitrogen flow rates. It should also be kept in mind that the internal stress conditions influence these results. HfN ( DUH= l 1 5 7 Nmm -2) reaches a higher hardness than TiN (DUH = 815 N mm-2}. The hardness is, besides the deposition rate of the sputtering process, the only fundamental property which is better for HfN than for TiN. 3.1.4. Internal stress The influence of the nitrogen flow rate on the internal stress in the coatings of the first order is primarily 3.1.7. Friction and wear To give a representative impression of the wear proper- ties, Fig. 4 illustrates the results from the plate-on- 138 G. Berg et al./Surjdee and Coatings Technology 74 75 (1995) 135-142 cyl i nder test of l i near wear rat es and friction coefficients for uncoat ed cast i ng steel (subst rat e mat er i al ) and the pur e met al s Ti and Hf in r el at i on t o the st oi chi omet r i c Ti N and Hf N phases. In cont r ast t o uncoat ed friction par t ner s i ncreasi ng friction coefficients are t ypi cal l y f ound for coat ed speci mens up t o a rat i o of nor mal force to t angent i al force of 0.8. Bot h coat i ng syst ems give an excellent wear reduct i on, but in the case of Ti N this is appr oxi mat el y five t i mes mor e t han Hf N. 3.2. Inf l uence o f bias vol t age on coat i ng propert i es 3. 2. i. Col or The col or of the Hf N coat i ngs did not change ma r k- edly under the influence of the bi as vol t age. Onl y Ti N is influenced in col or by appl yi ng no subst r at e bias (reddish brown) and sput t er i ng under fl oat i ng bias condi - t i ons (al most bl ack coat i ng). The col our of Ti N films under moder at e bias vol t ages is dar k gol den wher eas in compar i s on Hf N is pal e gol d for all bias condi t i ons. 3.2.2. Deposi t i on rat e The deposi t i on rat es for Hf N x coat i ngs under the gi ven sput t er i ng condi t i ons are much hi gher t han t hose for Ti Nx. Compa r e d wi t h the rat es of Hf N pr oduced under var yi ng bi as vol t age, the t hi ckness of the films does not change very much. Devi at i ons f r om t he aver age val ue are small ( < 10%). I n cont r ast t o this the depos- i t i on rat e of Ti N coul d be influenced mor e st r ongl y in t he studied r ange of bi as vol t age. The r at e decreases cont i nuousl y wi t h i ncreasi ng negat i ve bi as to hal f the val ue (still a rat e of 17 nm mi n t), when compar i ng deposi t i on under a fl oat i ng pot ent i al with an appl i ed bias vol t age of - 4 0 V. 3.2.4. Int ernal stress The i nvest i gat i ons on i nt ernal stress det ect ed by X- r ay di f f r act omet r y led t o the cl ear results shown in Fig. 5. The bias vol t age har dl y influences Hf N, whereas the i nt ernal stress of Ti N coat i ngs is mai nl y domi nat ed by the bias vol t age. l a) Ib) 3.2.5. Adhesi on The adhesi ve failure whi ch depends upon the bi as vol t age on the subst r at es shows t hat moder at e val ues of bi as lead t o the best results. Moder at e in this case means a r ange of - 1 0 V t o - 3 0 V for bot h coat i ng syst ems studied. Referri ng t o a wor ki ng gas pressure of 1.5 Pa the ma xi mu m was f ound havi ng appl i ed - 2 0 V. Haf ni um based coat i ngs also wi t h respect t o the adhesi on coul d be less influenced t han t i t ani um nitride. Fig. 2. Influence of nitrogen flow on structure and morphology of HfN x coatings: (a) pure hafnium; (b) HfN (N2-4 cm 3 min- i); (c) HfN X (N/= 7.5 cm 3 min 1). (c} G. Berg et al./Surface and Coatings Technology 74-75 (1995) 135-142 139 Results from 4 t~sts by optical analysis Critical load = first point of adhesive failure (failure at the eoating-to-substrate interface) 4 7,5 ni t r ogen f l ow [ cm3/ mi n] Fig. 3. Infl uence of ni t r ogen flow on t he adhesi on. 90 80 70 60 50 40 30 20 10 0 l i near we a r [l~m] la=0475 ii~i~ cast i ng st eel ~ii~i~:~ testing conditions plate-on-cylinder tri bometer load: l kg; speed: 004 m/s; distance: 300 m H=0775 i ~ dry friction at room temperature " N ~i ~i ~ ~:~ stoichiometric coatings ~%~i ~% bias voltage: -20 V; pressure: 1 5 Pa ~;~ilii ~i, ~t=0 70 ~1 Ti Hf Ti N HfN c o a t i n g s y s t e m Fig. 4. Fri ct i on a nd wear pr oper t i es of t i t a ni um a nd ha f ni um and t hei r st oi chi omet r i c ni t ri de phases. - 8000 I J I / st ress evaluation by sin=~-method from 1 ~ " / ~ ' * (210) peak (Siemens diffractmeter D5000) Young~s modulh E(TiN)=400000N/mm2; E(HfN)=380000N/mmZ; v=0.2 - 7000 E -6ooo z - 5000 - 4000 E -3000 .E - 2000 - 1000 1000 f l o a t i n g g r o u n d -10 -20 -30 -40 bi as vol t age [V] Fig. 5. Infl uence of bi as vol t age on t he i nt er nal st ress of Ti N a nd Hf N. 140 G. Berg et aL/Surface and Coatings Technology 74 75 (1995) 135-142 1600 1400 1200 1000 800 6 0 0 400 200 0 f l o a t i n g u n i v e r s a l h a r d n e s s / I TiN [ N I m ~ ~ deviationsrepresent standard dynamic universal hardness (Shimadzu DUH 202) vickers indenter - load 100mN - speed 6,75mN/sec - after time 10sec hardness under loading conditions i = ! = gr ound -10 -20 -30 Fig. 6. Infl uence of bi as vol t age on t he har dness. -40 bi as vol t age [V] 3.2.6. Hardness The influence of the appl i ed bias vol t age on the hardness values of coatings for Ti N films is well known and can be used t o pr oduce defined propert i es. Fig. 6 illustrates the hardness results by compar i ng mean values of Ti N and Hf N under increasing bias voltage. The first poi nt t o be ment i oned is the hi gher val ue for each Hf based l ayer and the cor r espondi ng fact t hat a specific influence is not possible. The hardness values measur ed under floating pot ent i al condi t i ons are of the same size as the one coat ed at - 4 0 V bias voltage. Ti N coul d mor e easily be influenced and shows cont i nuousl y increasing values for hardness with increasing negative bias voltage. The results for hardness and i nt ernal stress depend on each other. 3.3. Influence of deposition pressure on the coating properties 3.3.1. Deposition rate The deposi t i on rat e was st udi ed as a funct i on of the pressure of the worki ng gas. The ar gon flow was changed keepi ng the ni t rogen flow const ant . Because of the change in part i al pressure of the gases the deposi t i on rat e increases for bot h coat i ng systems with increasing pressure. For Ti N the deposi t i on rat e of 10 nm mi n -1 at 1 Pa doubl es for coat i ngs pr oduced at 2 Pa. 3.3.2. Internal stress Int ernal stress in Hf N was little influenced by the vari at i on of pressure compar ed with Ti N. Fig. 7 displays results from the XRD and the bendi ng beam met hod concerni ng the influence of pressure on the i nt ernal stress of Ti N films. The results f r om the qui t e simple bendi ng beam met hod, which takes the def or mat i on of a thin spring steel subst rat e under the effecting i nt ernal stresses of the coatings, cor r espond clearly with the XRD results, but always give l ower values t han the XRD measurement . 3.3.3. Hardness Har dness decreases for increasing ar gon pressure in the range 1.0 2.0 Pa for Ti N f r om DUH = 1200 N mm -2 to 930 N mm- 2 and for Hf N f r om 1390 N mm = down to 1155 N mm -2. The effect on Ti N is, like all changes in deposi t i on paramet ers, st ronger t han t hat on HfN. 3.3.4. Further effects The st ruct ure of the coat i ngs was not affected mar k- edly by the change of pressure of the worki ng gas. The relative small var i at i on in ar gon pressure also di d not cause changes in color. 4. Discussion and conclusions Invest i gat i ons i nt o deposi t i on of har d coat i ngs based on the t i t ani um and hafni um nitrides were made t o st udy whet her the propert i es of HfNx can equal or possibly surpass qualities of Ti N. The research wor k was done f r om the view of a pr oducer of coat i ngs for special demands by varyi ng the basic par amet er s of the reactive sput t eri ng process and the t arget material. The f undament al propert i es of r.f. sput t ered Ti Nx and Hf N x coat i ngs are in many cases similar but not the same. Invest i gat i ons on coat i ngs sput t ered reactively under the same process condi t i ons were made and gave an overvi ew of the range of propert i es which can be affected. For absol ut e values given in Figs. 1 7, Fig. 8 shows with a relative scale the range of results for the studied coatings. Concer ni ng the deposi t i on rat e of the investigated t i t ani um and hafni um based nitrides, the rat e for HI N x is always much hi gher using the same sput t eri ng par ame- G. Berg et al./Surface and Coatings Technology 74-75 (1995) 135 142 141 - 6 0 0 0 - E E p_ - 5 0 0 0 -4000 E - 3000 -2000 -1000 XRD stress evaluation by sin=~-method from (220) peak; Siemens diffractometer D5000 Bending beam method using spring steel, d=0,1mm E(HfN)=3S00OON/mmZ; v=0.2 2870 1 1,5 2 pressure [Pa] Fig. 7. Infl uence of pr essur e on t he i nt er nal st r ess of Ti N. ._E X ~ . > E ~ E 100% :::~ relative m 90% i~i E maximum ~k~ r~ itiv m~ ~ 70*/, ~ so% ~ ~. ' 30% '~ ~ f _ N l - = - i ~ N i ~ deposition r a t e compressive adhesion hardness internal stress Fig. 8. Range of resul t s, ma xi ma test procedure a nd mi ni ma of f unda me nt a l propert i es. ters. Increasi ng ni t r ogen flow, domi nat ed by the phases pr oduced, leads t o a significant decrease in t he deposi t i on rate. A var i at i on of the bias vol t age on the subst rat es at the best st oi chi omet ri c phases show a st rong influence on the Ti N coat i ngs while Hf N was onl y slightly affected. For the appl i cat i on of these coat i ngs on t ool s or machi ne part s t here are t wo i nt erest i ng facts. Hf N coat - ings for t ri bol ogi cal use with mi ni mal coat i ng thickness coul d be pr oduced in a shor t er t i me and the process is not ver y sensible accor di ng t o the process par amet er s e.g. bias vol t age, so the pr oper t i es are easier t o r epr oduce for Hf N t han for Ti N. Also t he hardness is hi gher for the Ht N x coat i ngs st udi ed compar ed with i dent i cal l y pr oduced Ti Nx. Hi ghest hardness values were f ound for slightly overs- t oi chi omet r i c films. Again Ti Nx was much mor e influ- enced t han Ht Nx. Thi s result also poi nt s out t hat sput t eri ng Hf N x is easier because the hardness is al most i ndependent of the deposi t i on paramet ers. The i nvest i gat i ons on i nt ernal stress lead t o similar results. Thi s is caused by the di rect cor r el at i on of i nt ernal stress and hardness. The hafni um based nitrides were nearl y not affected whereas for t i t ani um ni t ri de compres- sive i nt ernal stress increases a lot wi t h bias voltage. In or der t o pr oduce coat i ngs with defi ned high or low i nt ernal stress, we concl ude t hat onl y t i t ani um ni t ri de coul d do this in t he field of par amet er s studied. Accordi ng t o t he adhesi on of the coat i ngs with an i nt ermedi at e l ayer of the pur e met al s t i t ani um or haf- nium, st oi chi omet ri c Ti N and Hf N coat i ngs show highest values. Also moder at e bias voltages ( ~ - 2 0 V) on the subst rat e lead t o the best results. Thi s behavi or is of great i nt erest because the i nt er medi at e layers for all coat i ngs of one met al were identical. So adhesive failure is not domi nat ed by the i nt ermedi at e coating. Compar i ng the identically sput t ered coat i ngs on tita- ni um and hafni um bases, the Ti Nx films are bet t er in all cases. That is very i mpor t ant for the appl i cat i on of these films on machi ne part s or t ool s because the best physical pr oper t i es of a coat i ng system coul d not be used if the adhesi on is not sufficient for the t ri bol ogi cal demands. Fi rst i nvest i gat i ons on the t ri bol ogi cal behavi or of the 142 G. Berg et al./SurJace and Coatings Technology 74-75 (1995) 135-142 studied coat i ngs in a pl at e-on-cyl i nder t r i bomet er show t hat the linear wear is much bet t er for Ti N t han for Hf N al t hough Ti N was found t o have a hi gher friction coefficient. Looki ng at the par amet er s for coat i ng deposi t i on the ni t rogen flow allows a wide rangi ng influence on the coat i ng propert i es. Dependent on the phases pr oduced, deposi t i on rate, mor phol ogy and adhesi on can be vari ed over a wide range. The influence of bias vol t age on the subst rat es is most i mpor t ant on the i nt ernal stress of the coat i ngs and the hardness. Eval uat i ng t he influence of the i nert gas pressure, the change in part i al pressure must be seen in addi t i on t o the influence of pressure of the worki ng gas itself. The charact eri st i c propert i es of the Hf N x coat i ngs coul d be much less influenced by the par amet er s appl i ed duri ng the sput t eri ng process t han the TiNx layers. Thi s must be expl ai ned by the high at omi c mass of the el ement hafni um compar ed t o t hat of titanium. Thi s is i mpor t ant for the appl i cat i on of the coat i ngs because slight differences in sput t eri ng does not lead t o mar ked differences in physical and mechani cal pr oper - ties. Also the high deposi t i on rat e for hafni um ni t ri de coat i ngs is an advant age in rel at i on t o the coat i ngs based on titanium. To opt i mi ze Hf N x coat i ngs furt her investigations shoul d be done by furt her opt i mi zat i on of the deposi t i on paramet ers. At least we must state t hat in these i nvest i gat i ons Ti N coat i ngs showed bet t er basic mechani cal propert i es, especially for the very i mpor t ant adhesi on of the films. Ther ef or e Hf N coul d not subst i t ut e t he funct i on of Ti N due t o wear prot ect i on. For these reasons t i t ani um nitride seems t o be still the bet t er choi ce of har d coat i ng for tribological applications. Acknowledgements Thi s wor k is par t of the Schwer punkt pr ogr amm l onen- und Plasmaoberfl~ichentechnik of the Deut sche Forschungsgemei nschaft . The aut hor s wish t o acknow- ledge financial support . References [ 1] G. Kienel, Vakuumbeschichtung 5, Anwendungen Teil 2, VDI Verlag, Diisseldorf, 1993. [ 2] H. Holleck, Materials selection for hard coatings, J. Vac. Sci. Technol. A, 4(6) (Nov/ Dec 1986) 2661-2669. [31 W. Herr, G. Berg, C. Friedrich, E. Broszeit, and K.-H. Kloos, Grundlageneigenschaften und Verschleil3verhalten von HfB 2- und Hf(B,N)-Schichtsystemen, Mater. Wiss. Werkstofftech., 25 (1994) 175 179. [ 4] J.E. Sundgren, B.O. Johannson and U. Helmerson, Structure and properties of HfN coatings grown by reactive sputtering, Schriftenreihe Wissenschaft und Technik, Bd. 30, Darmstadt, 1986, pp. 169-189. E5] U. Kopacz and H. Jehn, Hdrte- und Haftfestigkeitsmessungen an Nitridschichten auf Schnellarbeitsstahl, Schriftenreihe Wissenschaft und Technik, Bd. 30, Darmst adt , 1986, pp. 215 240. [ 6] J.J. Oakes, A comparative evaluation of HfN, A1203, TiC and TiN coatings on cemented carbide tools, Thin Solid Films, 107 (1983) 159 165. E7] B. Rother, and J. Vetter, Plasmabeschichtungsverfahren und Hartstoffschichten, Verlag for Grundstoffindustrie, Leipzig, 1992. E8] H. Jehn, G. Reiners, and N. Siegel, DI N Fachbericht 39 Charakterisierung dfinner Schichten, Beuth Verlag, Berlin, 1993.