Laser-assisted ECR-plasma-CVD of a-C:H films

ELSEVIER

Diamond and Related Materials 5 (1996) 420-424

D IAMOND AND RELATED MATERIALS

Laser-assisted ECR-plasma-CVD of a-C:H films S.M. Metev a, M. Ozegowski a, K. Meteva a, G. Sepold ~, T.V. Kononenko b V.I. Konov b E.D. Obraztsova b, S.M. Pimenov b, V.G. Ralchenko b, A.A. Smolin b "BIAS-Bremen Institute of Applied Beam Technology, Klagenfurter Str. 2, 28359 Bremen, Germany b General Physics Institute, Russian Academy of Sciences, 38 Vavilov Str., 117942 Moscow, Russia

Abstract UV laser irradiation has been found to influence significantly the growth process of a-C:H films and to result in a modification of the film properties. Up to 2 gm thick, defect-free films have been synthesized. The optical and mechanical film properties are changed as a function of the laser energy density. Especially noteworthy is a significant increase in the film microhardness. At 60 GPa it reaches a value which is twice as high as that observed for unirradiated films. The laser-induced structural transformations in a-C:H films have been studied by Raman spectroscopy. Keywords: ECR-plasma-CVD; Excimer laser; a-C:H film; Raman spectroscopy

1. Introduction

2. Experimental details

Today, various methods exist, in particular plasmaCVD and ion beam techniques, for the deposition of diamond-like carbon (DLC) films [1]. Unfortunately, certain unsolved problems involving the synthesis of these films limit their application. An improvement in the adhesion and a decrease in the internal stress without the degradation of other useful properties, such as the hardness, chemical inertness and wear resistance, are typical problems. The aim of this work was to overcome these problems by incorporating an additional, independent, energy source (UV laser beam) into a conventional ECR-plasma-CVD apparatus to influence favourably the properties of the films during the deposition process. It is known that, due to the fast heating-cooling processes, post-growth irradiation of DLC films with short laser pulses results in partial crystallization and the appearance of diamond or graphite nanocrystals embedded in an amorphous matrix [2]. These phase transformations and the change in the hydrogen content influence substantially the film properties, but are restricted to near-surface areas [-2-4]. UV laser irradiation during the growth process is expected to give additional possibilities to control the synthesis of DLC films, e.g. the deposition of films with uniform modified properties across the whole thickness. In particular, a reduction in the internal stress in the films is expected.

The main features of the experimental set-up are shown schematically in Fig. 1. The a-C:H films were synthesized in a microwave ECR downstream reactor, produced by Roth & Rau Oberfl~chentechnik GmbH. The substrates (steel, aluminium or silicon plates) were placed on a water- cooled substrate holder. This could be negatively biased up to 700 V. The microwave power was varied in the range 80-800 W at a frequency of 2.46 GHz. Acetylene and argon gas were supplied to the processing chamber using mass flow controllers. The total gas pressure was in the range 0.005-0.01 mbar. Several quartz windows were installed for in situ laser irradiation and film growth monitoring. A K r F excimer laser (wavelength, 248 nm; pulse duration, 20 ns; energy up to 300 m J; repetition rate up to 50 Hz) was used to irradiate the growing film during deposition. In situ reflection interferometry of the growing film was performed with an H e - N e laser and provided information on the film thickness, refractive index and absorption coefficient at 630 nm. The structure and morphology of unirradiated and laser-processed films were analysed by scanning electron microscopy (SEM, Zeiss DSM 940) and Raman spectroscopy (Jobin Yvon S-3000, microRaman equipment). The Raman spectra were measured using excitation by an argon ion laser operated at 488 nm. Additional information about the film surface relief and film thickness was obtained by profilometric

0925-9635/96/$15.00 © 1996 Elsevier Science S.A. All rights reserved SSD1 0925-9635(95)00460-2

S. M, Meter et aL/Diamond and Related Materials 5 (1996) 420-424 Magnetron

Circulator

Tuner

Waveguide

lll

/

III

N

Magnetic coil

Antenna

~

Q u a r t z cup

[__,__

-1

...... ~

J

1

Excimer laser

~

:

Vacuum chamber J Plasma stream

laser

He-Ne

-

..-

!

Photodiode

Filter

421

-~-

/

f

/

-JJJ

'

/

Substrate Substrate holder ~

L V

I, Gas

T o v a c u u m pump

DC bias v o l t a g e

Fig. 1. Schematic diagram of the ECR-plasma-CVD apparatus.

measurements (Dektak3ST). The microhardness of the films was determined by a nanoindenter (Nano Instruments).

3. Results and discussion

2.3 2.2

E4 o ~o o

The growth rate and properties of a-C:H coatings obtained without laser treatment were dependent on various parameters (microwave power, gas pressure, composition of the gas phase, bias voltage). The most important parameter for the film properties was the applied bias voltage U. As an example, plots of the refractive index and absorption coefficient (at 630 nm) of the films vs. the bias voltage are presented in Fig. 2. For small voltages U < 150 V, the deposited films showed better optical transmittance, smaller refractive index and were relatively soft, i.e. typical of polymer-like films. For higher bias values, the absorption increased, the refractive index was close to n~2.2, the films were much harder and had better adhesion (as revealed by scratch tests). The typical final film thickness in our experiments was 500-2000 nm; the deposition rates were in the range 20-100 nm rain -1. Above a thickness of about 700 nm, more defects were observed; severe damage appeared with increasing film thickness probably due to the high internal stresses in the films. The laser-assisted modification of a-C:H films during plasma-CVD depended on the irradiation parameters and process conditions. The K r F laser energy density was varied between 50 and 150 mJ cm -2, the pulse

2.1

x o

2,0 .-->

3

1.9

o2 .o <

J refractive index

n,"

1.8 1.7

0

, 0

I 100

,

I 200

~

I 300

,

I 400

~

I 500

,

I 600

Bias voltage IV] Fig. 2. Refractive index and absorption coefficient of a-C:H films vs. bias voltage at 630 nm (microwave power, 320 W; Ar and CzHz gas flow, 40 sccm).

repetition rate was kept constant at 1 Hz and the irradiation period was equal to the deposition time. Laser processing of soft films ( U < 1 5 0 V) did not lead to significant changes in the film structure. Irradiation of hard a-C:H films grown at U = 600 V resulted in remarkable changes in the film surface morphology and film structure, when the energy density E exceeded a threshold of about 60 mJ cm-2. Above an energy density of 120 mJ cm -z, laser irradiation led to damage of the a-C:H films. Fig. 3 shows a scanning electron micrograph of an a-C:H film on steel with a thickness of 1.3 gm and a laser-modified zone on the left side. The non-irradiated film has many defects and is nearly completely destroyed,

S.M. Meter et aL/Diamond and Related Materials 5 (1996) 420-424

422

Fig. 3. Scanning electron micrograph of an a-C:H film grown on steel with a laser-modified zone on the left side ( E = 9 0 mJ cm z; film thickness outside the laser-irradiated spot, d = 1.3 gm).

whereas the laser-modified film is very smooth and has no observable defects. In the surface profile in Fig. 4, an additional feature is observed. The laser-irradiated film has a higher quality with a roughness similar to that of the substrate and a greater thickness than the non-irradiated film. An increase in the deposition rate due to additional direct laser stimulation of the deposition process (laser-CVD) is principally possible [5], but under our experimental conditions this effect is negligible, as revealed in pure laser-CVD experiments without a microwave plasma. The higher film thickness in this case is due, in our opinion, to a decrease in the film density in the laserirradiated area. This assumption agrees with the observed reduction in the refractive index of laserprocessed films from about n = 2.2 to n = 2.0 as measured interferometrically during the deposition process. It should be noted that similar effects of density variation

laser irradiated nonirradiated

E >-

L

0

I 1000

2000

X [gm] Fig. 4. Surface profile of an a-C:H film grown on steel outside and inside the laser-irradiated spot ( E = 90 mJ cm-2).

were also observed for some DLC films with postgrowth laser irradiation [-6] and for a-C:H films synthesized with the bombardment of highly energetic ions [ 7]. Raman spectroscopic investigations of the films show that the laser radiation during the deposition process influences significantly the structure of the films. In Fig. 5 (a), Raman spectra are presented at different deposition times. They demonstrate the beginning of a graphitization process. To estimate the degree of laser-induced graphitization, we used the relative integral intensity ID/I ~ of tWO Raman spectral peaks. The "G" peak corresponds to the fundamental FZg mode of graphite at 1580cm -1 and the "D" peak is defined by the disorder-induced A 1 mode of graphite at 1350 cm -1. It can be seen that the D band appears first as a shoulder of the G band for t = 2 min, and becomes more and more resolved with increasing deposition time (irradiation dose). Our investigations show an increase in the ratio ID/I ~ with increasing deposition time from 0.98 ( t = 2 min) to 1.79 (t=21 min). Simultaneously, the G band shifts to higher frequencies. According to Refs. [6] and [8], we may interpret the observed structural transformations as the beginning of a laser- induced graphitization process, at which separate crystallites with a size of about 20 A appear in the amorphous matrix. Fig. 5(b) demonstrates the effect of laser energy on the Raman spectra of the modified a-C:H films (growth duration, 20 min). The spectrum of the untreated film has one wide asymmetrical band at 1560 cm -1, which is typical of an amorphous-type structure. Lasermodified films show again two bands at 1350 cm -1 (D peak) and 1580cm -1 (G peak). With higher energy density, the G peak decreases slightly, but the general features of all the spectra are nearly the same independent of the value of the laser energy density. This can be explained by a threshold character of the beginning of the graphitization process at a certain energy density. On reaching the threshold, the graphitization process may depend on the irradiation dose rather than on the laser energy density. The structural changes (graphitization) in the irradiated area are assumed to be due to the thermal action of the laser radiation. The partial transition from the disordered and stressed amorphous structure to the crystalline graphite structure ("annealing") leads to relaxation of the internal stresses of the film. This is probably the reason for the higher film quality even at high film thicknesses (2 gm or more). At the same time, the lower density of graphite contributes to a decrease in the overall mean density of the film, as indicated by the higher film thickness in the irradiated area and the lower refractive index. A very interesting result was obtained by measuring the microhardness of the films. The microhardness was determined at different laser energy densities for every sample inside and outside the laser-irradiated spot (see

S.M. Meter et aL/Diamond and Related Materials 5 (1996) 420 424

423

¢ e-

1000

(a)

1200

1400

1600

1800

Raman shift, cm -1

1000

(b)

1200

1400

1600

1800

2000

Raman shift, cm "x

Fig. 5. (a) Raman spectra of irradiated a-C:H films at different deposition times ( E = 100 mJ cm-2). (b) Raman spectra of a-C:H films irradiated with different laser energy densities (deposition time t = 20 rain).

Fig. 6). It was observed in these measurements that the hardness steadily increases with increasing energy density from 66 to 112 mJ cm -2. In the area irradiated with an energy density of 112 mJ cm -2, the film is almost twice as hard as in the unirradiated area. The hardness of about 60 G P a corresponds to that of nanocrystalline diamond films [9] or of recently synthesized tetrahedral D L C films with a large amount of s p 3 bonds [7]. The enhancement of the hardness in the irradiated film area is in contradiction with the graphitization process. This means that the laser radiation also initiates other effects, which lead to an enhancement of the hardness in the non- graphitized area. These effects may include photolytically induced changes of the hydrogen content, photoexcitation and transitions of sp 2 bonds into s p 3 bonds or synthesis of nanodiamond or subnanodiamond clusters. All these effects cannot be investigated with the 70 unirradiated

60

gY~zz~t" I

irradiated

E50

== e-

V//AI .........

7

4o

a0

o

.o_ ~; 20

t [

•

•

10] 0 60

I.,

70

l

Raman technique used in our experiments. We are currently carrying out systematic investigations of the mentioned phenomena, also including the wavelength dependence, using suitable analytical methods.

4. Conclusions

The irradiation of a-C:H films during the growth process by K r F excimer laser radiation leads to a modification of the structure and is accompanied by a change in the film properties. In particular, a significant increase in the maximum thickness without damage of the films, probably caused by a decrease in the internal stress, is observed. At a laser energy density of 60 mJ cm -2, "jump-like" changes in the film structure occur. Raman spectroscopy has shown the formation of nanocrystalline graphite clusters. At the same time, a significant increase in the microhardness is measured. For this effect, other laser-induced processes, such as a change in the hydrogen content, photoinduced transitions of sp 2 into sp 3 bonds or synthesis of nanodiamond clusters, are proposed. Investigations of these phenomena are currently being carried out using suitable analytical methods and we hope to publish our results in the near future.

I

80

90

1O0

Energy density [mJ/cm z]

110

120

Fig. 6. Microhardness of a-C:H films irradiated with different laser energy densities compared with that of unirradiated films.

Acknowledgements

This work was supported Forschungsgemeinschaft (DFG).

by

the

Deutsche

424

S.M. Metev et al./Diamond and Related Materials 5 (1996) 420-424

References [1] A.H. Lettington, in R.E. Clausing (ed.), Diamond and DiamondLike Films and Coatings, Plenum, New York, p. 481. 1-2] L.C. Nistor, J. van Landuyt, V.G. Ralchenko, T.V. Kononenko, E.D. Obraztsova and V.E. Strelnitsky, Appl. Phys. A, 58 (1994) 137. [3] V.Y. Armeyev0 E.N. Loubnin, V.G. Ralchenko and V.E. Strelnitsky, Appl. Phys. Lett., 58 (1991) 2758. [4] S. Weissmantel, G. Reisse and S. Schulze, Diamond Relat. Mater., 2 (1993) 1379. [.5] M. Danno and M. Hanabuse, Mater. Lett., 4 (1986) 261.

[-6] V.P. Ageev, N.I. Chapliev, T.N. Glushko, T.V. Kononenko, A.V. Kuzmichov, A.A. Smolin, V.E. Strelnitsky, V.F. Dorfman and B.N. Pypkin, Surface Coat. Teehnol., 47 (1991) 269. [-7] S. Sattel, M. Weiler, T. Giessen, K. Jung and H. Ehrhardt, in P. Vincenzini (ed.), New Diamond and Diamond-Like Films, Techna, Faenza, 1995, p. 133. 1-8] N.H. Cho, D.K. Veirs, J.W. Ager, M.D. Rubin, C.B. Hopper and D.B. Bogy, J. Appl. Phys., 71 (1992) 2243. [.9] S.M. Metev, K.B. Meteva, G. Sepold, T.V. Kononenko, V.I. Konov, E.D. Obraztsova, S.M. Pimenov, V.G. Ralchenko and A.A. Smolin, in P. Vincenzini (ed.), New Diamond and DiamondLike Films, Techna, Faenza, 1995, p. 31.