SURfACE

&COATINIiS
HGHNOJDGY
ELSEVIER Surface and Coatings Technology 74-75 (1995) 168-172

The role of ion implantation in industrial sputtering processes for

tribological applications
Franz-Josef Korber
ABB Calor Emag Scha/tall/agell AG, Bro\\lll Boveri Str. 30, 63457 Hallall, Germally

Abstract

Since the early 1970s ion implantation into metal has been used successfully in the field of tribology. Because of the availability
of high-current ion sources for gas ions, the most widely applied process is the implantation of nitrogen ions into various kinds of
steel. Apart from direct implantation, ion-beam-assisted processes (IBAD) and ion beam mixing processes (IBM) are becoming
increasingly important. In most cases the IBM process is performed with high energy implanters whereas the IBAD processes are
performed with low energy ion sources in combination with electron beam evaporators within one process chamber. In addition
to these established processes, the present paper presents the combination of high current medium energy ion beams with high
rate magnetron sputtering. The ion beam is used for the pretreatment of the substrates, for assisting the growing film, for mixing
of thin film layers and for post-treatment of reactively sputtered nitrides. The processes have been performed in a prototype
implanter for industrial use applying a radio frequency ion source with slit-shaped ion optics which fit the demands of modern
sputter sources with respect to beam size and dose rate. Evaluation of the films has been carried out using REM, SNMS, Rockwell
C, Scratch test and pin on disc test. It is shown that medium-energy linear ion sources can be used sucessfully and simultaneously
together with magnetron sputter sources within one process chamber. This can be done with already existing highly productive
batch-type coating machines and also with in-line coaters by adapting the ion source so that it can be used in place of an
additional magnetron source. The combination of magnetron sputter sources with simultaneously operating ion sources allows a
highly controllable process, which fits the demands of high quality coatings for the future. This applies especially to the coating of
polymers and ceramics.

Keywords: Implantation; Sputtering; Ion beam mixing; Tribology; Calotest

1. Introduction medium energy ion beams. The processes have been

performed within one process chamber in different
In the family of plasma-assisted vapour deposition implantation and coating steps.
processes for tribological coatings, the sputtering process
seems to be the most versatile process for future demands
[1]. The biggest advantages are flexibility with respect 2. Description of the implantation-coating facility
to the coating material, the possibility of the synthesis
of hard nitrides, carbides and oxides in single, multi, Fig. 1 shows the arrangement of the substrate carrier
and gradiend layers [2]. The sputtering technique allows and the two sputtering sources (magnetrons 0150 mm)
free design of the size and position of the sputter sources, [7J mounted in the process chamber of the high-current
which can also be used as PECVD [3J or arc-evaporator ion implanter described in Re£. [8]. The ribbon-shaped
sources [4]. ion beam enters the cubic process chamber (w x d x h:
In contrast to this, ion beam assisted deposition 1000 mm x 700 mm x 600 mm) vertically and centrally.
(IBAD) techniques on the other hand have been mainly The two independent sputtering sources, which are
used in combination with electron beam evaporators arranged on the left and the right side of the ion beam,
[5]. Thin films for ion beam mixing (IBM) processes can be inclined freely. The subtrates are fixed on a biased
have been electron beam evaporated or have been x-y table or on a special substrate carrier entering the
prepared in special coating facilities outside the implant- chamber through a load lock from the left side. A
ation facility [6]. The present paper deals with the rotating fixture with 20 axles can also be mounted on
improvement of magnetron sputtered coatings by using the x-y table.

0257-8972/95/$09.50 © 1995 Elsevier Science SA AU rights reserved

SSDJ 0257-8972(95)08362-6
 F-J. Korber/SlIIface and Coatings Technology 74-75 (1995) 168-172 169

Fig. 1. Process chamber with water cooled x-y table, load lock
substrate carrier and two independent sputtering cathodes (courtesy
MAT GmbH, Dresden).

The maximum applicable beam voltage during the

coating operation in the lO- z Pa range (IBAD) was
60 kV and without simultaneous sputtering it is 150 kY.
The main advantages of the d. ion source [9,10J are
the low working pressure (10-z Pa), the low power
consumption (100-400 W d. power), the ribbon-shaped
beam pattern (300 mm length) and the high current
density (up to 1 mA cm -Z) for gas ions (N, Ar).
The coating sources have been operated up to coating
speeds of 2 nm S-1 with up to 1.5 mA cm- z substrate- (b)
bias current at -100 V bias voltage. T;
con~nlr. AT""
[~.] 100 C, c;-
lin \
90 A
1\ A f\ 1\
\
3. Processes so
70

3.1. Ion beam mixing ofsputtered lIlultilayers of TiAI-Cr 60 ~AI

50 If ,( 11

40 Ti
An alternating multilayer TiAl-Cr system was chosen 30

in order to investigate the mixing efficiency and range 20

as function of the process parameters such as ion species, 10

vv V V V V V V \.J
ion dose and substrate temperature. Fig. 2. shows the LEYBOLD AG "" 500 1000 1500 2000 2500
lime rsiOOOlin
multilayer cross-section, the calotest and the SNMS (c)
depth profile of the samples after sputter deposition. The
profiling of the 24 layers, 40 nm thickness each, shows Fig. 2. Cross fracture (a), Calotest (b) and SNMS depth profile of a
TiAl-Cr multilayer used for ion beam mixing experiments.
a good depth resolution. After ion beam mixing with
argon ions (100keV Ar+, dose=7 x 1016 cm -z, T s =
200 QC) a clear mixing effect can be observed (Fig,. 3(a)).
The signals of the three elements are shifted together strongly on the temperature of the sample during
within the first four layers indicating the formation of implantation. For T s = 100 QC the nitrogen distribution
gradient alloy TiAlCr. The range of the mixing effect is is uniform and gaussian, and the concentration profiles
about two times the ion range and can be explained by of the metallic layers is not influenced (Fig. 3(b)). With
collision cascades and radiation effects [11]. a high temperature applied during implantation (Ts =
After high dose implantation with nitrogen (100 keY 500 QC), the nitrogen concentration has increased in the
N+jN z +, dose=2xlO 17 cm- Z ) the result depends TiAl layers and has decreased in the Cr layers.
170 F-J. KorberJSlllface and Coatings Tecllllology 74-75 (1995) 168-172

100 200 300 <00 500 600 700

(a)

i
t~c~"tr. ~
11" ,

(a)
::1\
60i ;~
50 1; .
I
'0 1I

200 300 400 500 600 800 900

b)

oncent,..,
I • I
Un

(b)

Fig. 4. Cross grinding section (Calotest) of a reactively sputtered TiN

600 700
coating prepared in 100 layers with (a) and without (b) subsequent
Ur.e
nitrogen implantation.
(c)

Fig. 3. SNMS depth profile of a TiAl-Cr multilayer after ion beam

mixing: (a) 100 kV argon implantion; (b) 100 kV nitrogen implantation masked completely. This sequence was repeated 100
at low temperature; (c) 100 kV nitrogen implantation at high times, leading to a total film thickness of 1 JlID. The
temperature. calotest of the implanted coating shows a dense and
clean cross-grinding section (Fig. 4(a)), while the unim-
Furthermore A1 has diffused towards the first er layer planted coating is porous, and the film thickness is
and to the surface (Fig. 3(c)). reduced owing to delamination of the film. This result
can be explained by the lack of any bias current during
3.2. Ion beam mixing of TiN pe/formed in an alte/'l1ating coating. The coefficient of friction was measured with a
coating-implantation mode rotating ball on disc tester to be Jl=0.2 for the implanted
and J..l=0.6 for the unimplanted reference specimen
HSS steel discs were sputter cleaned during im- (testing conditions reported in Ref. [12 ]).
plantation (25 keY N+ /N2+, dose=2 x 10 17 cm- 2 , T s =
300°C) and then reactively sputtered with 10 nm TiN 3.3. (TiAI)N films obtained by ion implantation during
at a coating speed of 1 nm s -1. No bias voltage was sputtering (lBAD)
applied during coating. After each coating step, half of
the sample was implanted (40 kV N+ /N2+, dose= For the IBAD process the sputtering cathode was
2 x 10 15 cm - 2, dose rate = 2 x 10 14 / cm 2 s -1) whereas alligned at 45° to the vertical ion beam at a distance of
for the second half of the sample the ion beam was 180 mm from the substrates (Fig. 1: cathode on the right
 F-J. Korber/SlIIface and Coatings Technology 74-75 (1995) 168-172 171

(a)
Fig. 5. Cross fracture of a (TiAl)N layer implanted simultaneously
with nitrogen during sputter deposition.

hand side) Before the coating process started, the HSS

substrates were implanted (40 kV N+ /N2+, dose =
5 x 10 17 cm -2) in order to sputter clean the surface and
to increase the substrate hardness. The coating was
performed by reactive sputtering from a TiAI (50:50)
target at a coating speed of 0.8 nm s -1. The bias current
was 0.5 mA cm -2 at -100 V, and the temperature of
the sample was calculated to be 400°C. The beam
parameter during nitrogen implantation was
0.02 mA cm -2 N+ /N2+ at an acceleration voltage of
40 kV, leading to an ion/atom ratio of 0.05, which is in
accordance with the process parameters used in Refs.
[13,14]. For one half of the sample the ion beam was
completely masked to give a reference coating. The Fig. 6. Wear tracks of reactively sputtered CrN" coatings with (a) and
orientation of the (TiAI)N film of both coating zones is without (b) subsequent nitrogen implantation.
clearly (111) with a lattice constant of 0.417 nm for the
IBAD film and 0.4225 nm for the sputtered only film. indica tes ma terial transfer from the ball to the disc after
The cross-sections of both coating zones show a dense only 5 m sliding distance (Fig. 6(b)), whereas for the
and fine-grained structure (Fig. 5). The measured values implanted disc the wear track after 100 m is smooth,
for the lattice constants are in accordance with the without any transfer of material from the counterpart.
results reported in Ref. [15]. The same positive results have even been obtained while
testing both conditions against a Si 3 0 4 ball. Because no
3.4. Implantation of reactively sputtered erN coatings major structural or mechanical changes in the bulk of
the coating occur after low-dose low-temperature nitro-
Especially for machine parts, CrN~ coatings provide gen implantation, this result must be explained by
good tribological behaviour for the coated parts as well changes in the near surface chemistry or a modification
as for the uncoated counterparts. The coefficient of of surface layers as explained for sputtering effects in
friction in combination with a WC ball in a dry sliding Ref. [16]. The results are in accordance to the effects
rotating ball-on-disc tester (test conditions in Ref. [12]) obtained by nitrogen implantation into electroplated
varies between 0.2 and 0.6. To improve this situation, chromium films [17].
samples with a high value ofJ.1=0.5 have been implanted
with a comparably low dose of nitrogen (100 kV
N+ /N2+, dose = 5 x 10 16 cm -2) in order to provide 4. Conclusion
short process times. The coefficient of friction was low-
ered to 0.2 after implantation. The wear track after a The examples for the different applications of ion
rotating ball-on-disc test of the unimplanted specimen implantation in combination with sputtering techniques
172 F-J. Korber/Slllface alld Coatillgs Techllology 74-75 ( 1995) 168-172

show that, especially for non-optimized coating methods of arc evaporation and unbalanced magnetron sputter-
processes, the quality of the films (cohesion, adhesion, ing, Swf Coat. Tee/lllol., 49 (1991) 161-167.
[5] G. Dearnaley, P.D. Goode, FJ. Minter, AT. Peacock, W.
density, frictional behaviour) can be remarkably Hughes and G.W. Proctor, Ion implantation and ion assisted
increased by applying the ion beam as an additional coating of metal, VacuulII, 36 (1986) 807-811.
tool. Furthermore the mixing processes are an economi- [6] T. Flottmann, Statisches und dynamisches Ionenenstrahlmi-
cal and fast method for the alloying of near surface schefeld und Betriebstests, work supported by the BMIT under
Grant FKZ 13 N 5420/0. P/'Oc. Diillnschichttechllologiell '90,
regions. In contrast to metal ion beams, the mixing
VDI Verlag, Dlisseldorf, p. 354-371.
process has no restrictions regarding the melting point [7] Pamphlet 16-100.02: Cathodes, Leybold AG, Hanau, Germany.
of the applied materials and is much faster. The combina- [8] FJ. Korber, WD. Miinz, H. Ranke, S. Reineck, HJ. Fiif3er and
tion of the available high-current gas ion sources [18,19J H. Oechsner, A newly developed linear ion implanter for
and highly productive sputtering plants [3,4 J gives an industrial applications, Mater. Sci. Ellg., A116 (1989) 205-208.
[9] H.J. Fiif3er and H. Oechsner, A novel d. source for ribbon
important tool to fulfill the requirements for high-quality shaped ion beams,in H.l Fiif3er and H. Oechsner (eds.), Sixth
coatings and controlled processes. The integration of lilt. Con! on Ion and Plasllla Assisted Techniques, Brighton,
medium-energy ion sources into existing industrial May 1987.
sputter coaters can be done easily. [10] H.l Fiif3er, Eine Breitband-hochfrequenzionenquelle fiir ober-
fHichen und Diinnschichtprozesse, Thesis, University of
Kaiserslautern, Germany, 1989.
[11] L.E. Rehn and P.-R. Okamoto, Recent progress in understanding
Acknowledgements ion beam mixing of metals, Nucl. Insl:J·wll. Methods Phys. Res.,
B39 (1989) 104-113.
[12] FJ. Korber, H. Petersein and H. Ranke, Wear reduction and
This work has been supported by the Bundes-
sliding behaviour of various substrates obtained by ion implant-
ministerium fur Forschung und Technologie under Grant ation of nitrogen and carbon ions, Thin Solid Films, 181
13 N 5572. The experimental work has been performed (1989) 505-513.
at Leybold AG, Hanau, Germany. I would like to thank [13] M. Barth, W. Ensinger, V. Hoffmann and G.K. Wolf, Stress and
Dr. Bartella and Mr. Petersein for carrying out the adhesion of chromium and boron films deposited under ion
bombardment, Nucl. Insl:J·unl. Methods Phys. Res., B59/60
analytical work. (1991) 254-258.
[14] F. Yucai, Y. Dawei, L. Wensky and W. Wuntao, The dependen-
cies of the dynamic recoil mixing zone on beam energy;
References irridiation dose; dose rate and temperature of sample, VacuulII,
39 (1989) 255-257.
[15] T. Leyendecker, Uber neuartige Schneidwerkzeugbeschichtungen
[1] lE. Sundgren and H.T.G Hentzell, A review of the present state auf Titan- und Aluminiumbasis, Thesis, University of Aachen,
of the art in hard coatings grown from the vapour phase, J. Vac. 1985.
Sci. Techllol., A4 (5) (1986) 2259-2279. [16] H.M Gabriel and K.H. Kloos, Effects of sputtercleaning on steel
[2] H. Holleck, Designing advanced coatings for wear protection, substrate surfaces, Vakuum Techllik, 33 (1985) 242-245.
Swf Eng., 7 (1991) 137-144. [17] F.J. Korber, Some technical aspects of ion implatation as an
[3] D. Hofmann, W.D Miinz, H. Schiissler and J. Gobel, Production industrial process for surface modification of metal, SlIIf Coat.
and properties of hard carbon films prepared by magnetic film Tee/mol., 59 (1993) 226-233.
assisted deposition process, Fifth lilt. Con! 011 1011 alld Plasma [18] Pamphlet 16-610.02: IQ 300 Linear-illlplantation source, Leybold
Assisted Techlliques, Munich, May 1985. AG, Hanau, Germany.
[4] W.D Miinz, F.lM. Hauzer, D. Schulze and B. Buil, A new [19] lH. Keller and lW Robinson, Ion sources and implantation
concept for physical vapour deposition coating combining the systems, Mater. Sci. Eng., 90 (1987) 423-432.