E L S E V I E R Surface and Coatings Technology 79 (1996) 119 130

.COATIN6S i ilNOLO T

Deposition of a ceramic coating on a thermoplastic polymer by atmospheric plasma and laser cladding

S. Ayrault a,b,c, A. Chateauminois b, j.p. Soulier a, D. Tr6heux b, A.B. Vannes b,c a Laboratoire d'Etudes des Matdriaux Plastiques et des BiomatOriaux, URA CNRS 507, Universit~ Lyon I,

69622 Villeurbanne Cedex 2, France b Laboratoire MatOriaux-MOcanique-Physique, URA CNRS 447, Ecole Centrale de Lyon, BP 163, 69131 Ecully Cedex, France

CALFETMAT, INSA de Lyon, 69621 Villeurbanne Cedex, France

Received 9 August 1994; accepted in final form 16 January 1995

Abstract

Alumina powder was deposited on a thermoplastic polymer using an atmospheric plasma and laser (C02,)~ 10.6 gm). The results have shown that a mechanical anchorage of the ceramic exists on the surface of the polymer, regardless of the technique used or different ceramic structures. In addition, the morphology of the poly(ethyleneterephthalate) changes during the treatments and the polymer crystallizes. Hence, different techniques were used to investigate the morphological changes (crystallization and amorphization) by X-ray analysis (wide-angle X-ray diffraction) and to observe the structural modifications by microindentation hardness testing. In this way, knowledge of the morphology of the material after laser cladding or plasma spraying enables us to characterize the mechanical properties of the samples.

Keywords: Plasma spraying; Laser cladding; Ceramic coating; Polymer

1. Introduction

Despite huge development of polymers, we can now see certain limits for using these materials at high temperatures or under difficult tribological conditions. Thus, to protect polymer surfaces, further improvement can be achieved by using a specific ceramic hard coating on the surface. Various applications are already known or could be developed, such as protheses, plastic moulds, composites, etc.

For this purpose, ceramic powder can be deposited using an atmospheric plasma or laser spraying method to obtain a thick coating on the polymeric materials. These techniques are usually employed to treat metallic materials, [ 1-5] and are more rarely used with polymers [ 6 -9 ] . The thermal characteristics of polymers are poor for these high energy techniques. Hence, thermal degra- dation is very likely to occur for these materials.

The purpose of this study was to deposit a ceramic coating on polymer surfaces using two techniques. For this purpose, poly(ethyleneterephthalate) (PET) was selected with alumina powder. Owing to the fact that the polymer is heated at high temperatures with these two techniques, the material 's morphology can change. Hence, to determine changes in the morphology and the

0257-8972/96/$15.00 © 1996 Elsevier Science S.A. All rights reserved SSDI 0257-8972(95)02432-8

microstructure of the polymers, X-ray analysis and microindentation hardness testing were used. The effect of the atmospheric plasma and the laser beam on the polymer structure are of special interest for the final materials characteristics. The morphological changes are important to master because they have a strong effect on the mechanical properties and the dimensional stability of the substrate.

2. Experimental details

2.1. Materials

Commercial PET was supplied by Rh6ne Poulenc, in the form of pellets, with the referential name 92 T. The number average molar mass of these pellets was M : = 20 000 g mol-1 , with a polydispersity index of Mw/Mn =

2 [103. PET specimens were obtained by injection moulding.

Prior to injection, the pellets were dried at 140 °C for 4 h to avoid polymer hydrolysis during melting. A conventional parallelepiped mould with cross- section 5 mm × 25 mm was used. The mould was water cooled.

To perform measurements of the absorption of the

120 S. Ayrault et al./Smface and Coatings Technology 79 (1996) 119 130

laser radiation by the polymer, PET cast films 300 gm thick were also prepared by extrusion through a coat- hanger die and subsequent quenching on a roll mill. Measurements were also made with a power meter behind the film during the laser treatment. With the Beer-Lambert law defined by I = Io e x p ( - ~z) where :~ is the absorption coefficient, I the power measured behind the film, I o the initial mean power and z the thickness of the sample, the absorption coefficient for the polymer can be determined.

The alumina powder used was provided by Metco (designated 101 B-NS). This is a crushed grey powder with a hexagonal crystallography structure and a melting temperature at 2010 °C. The ceramic composition was as follows: alumina, 96.15%; titanium dioxide, 2.53%; silicon dioxide, 0.48%; iron oxide, 0.04%; other oxide, 0.8%. The grain sizes were between 15 and 53 ~tm. This choice was made to provide a good flow in the lateral nozzle during the laser deposition.

2.2. Ceramic Deposition Techniques

2.2.1. Atmospheric Plasma Spraying To melt the ceramic powder particles injected into the

gun, the thermal spray devices used a plasma flame commercialized by SNMI [ 11 14]. The molten powder particles were sprayed by the plasma flame onto the polymer substrate, where they rapidly solidified. The powder injection system was based on a M E T C O 4MP device. Multiple sprays on the polymer substrate were necessary to obtain a homogeneous deposit.

To avoid overheating of the polymer surfaces, the specimens were rotated at 30 rev rain -1 during the treatment and were cooled by argon jets. Therefore, these conditions were equivalent to a linear speed of 0 .4m -1 s. A movement was also imposed onto the plasma flame in the x and z directions. Any overheating was indicated by bubble formation and the appearance of a brownish area on the polymer surface. The spraying parameters are given in Table 1 [15,16]. In this study, only the effects of the number of deposition runs were considered. A run can be defined as a one-way passage of the plasma flame over the sample.

2.2.2. CO 2 laser deposition The laser cladding of the polymer was performed by

projection of the alumina powder onto the polymer

Table 1 Parameters for the plasma projection

Powder rate Projection length Plasma gas rates

Intensity Rotating speed

42 g min 1 160 mm Ar, 301min 1 H2, 11 1 min- 1 6 mA 20-50 rev min-

surface simultaneously irradiated by a CO2 laser beam. The powder injection system was based on a Plasma Technik Twin 10 C device. The powder was injected through a lateral nozzle located 100 mm above the specimen surface, with an angle of 25 ° between the nozzle and specimen surface. The powder introduction rate was set to 4 g min - 1.

The laser source was a CILAS (C! 4000) continuous wave COz laser which has been described elsewhere [5,17]. The system was operated with a mean power of 180 W and a beam energy with a gaussian distribution. The beam focus point was located above the specimen surface, so that the laser beam diameter on the specimen surface was 37 mm (the beam diameter was measured by the impact on thermal paper). The resulting specific mean power was 1 7 W c m 2, as defined by the ratio P/~zr 2, where P is the laser power and r the laser spot radius on the material.

The specimens were mounted on a numerically con- trolled x y table and scanned under the laser head during the treatment. The coating was carried out using two displacement schemes, in scheme 1, successive runs were performed on a single track; in the second scheme, scheme 1 was successively repeated on three parallel tracks to deposit the ceramic particles over the entire specimen surface. In this case of scheme 2, the homo- geneity of the deposit was found to be optimized when the tracks were spaced by 7.5 mm. Scheme 1 was used to analyze the thermal effects of the laser beam in the polymer, without taking into account the thermal transfers that resulted from the overlaying of the runs.

Both the effects of the scanning speed and the number of runs have been investigated. The scanning speed of the samples under the laser beam ranged from 0.05 to 0.1 m s -1. Scanning speeds lower than 0.05 m s I were found to induce polymer overheating.

2.3. X-ray Diffraction

Wide-angle X-ray diffraction (WAXD) was used to investigate the morphological changes (crystallization and amorphization) which could occur in the polymer during plasma or laser treatment. The analysis was carried out on a Siemens D 500 diffractometer for 20=3 c' 50 °. Cu Kc~ radiation (0.154 nm) was used.

Crystallinity values xc were determined as the ratio of the crystalline area to the total area observed between 3 ° and 50 ° (20) after substraction of the background [18,19]. To investigate the crystallization gradients through the treated bars, various depths of the bars were removed to obtain smooth flat surfaces on which to carry out XRD. The polymer was removed by polishing the bars in the longitudinal direction. The depths ana- lysed were set at 1 and 2 m m below the irradiated surfaces. It should be noted that, using the atomic composition of the PET, the depth analysed can be

S. Ayrauh et al./Surface and Coatings Technology 79 (1996) 119 130 121

estimated to be close to 1 mm for 20=50 ° with 100% X-ray diffracted and 400 ~tm for 80% X-ray diffracted. As a result, the crystallinity values obtained at different depths from the surface are averaged over a distance of about 1 mm from the surface analysed.

2.4. Microhardness

In recent years, investigations of the microhardness properties have emerged as a physical method which can provide quantitative information about changes in polymeric materials [20 23]. It is now well established that microindentation hardness testing is a technique that is sensitive to morphological and structural changes in semicrystalline polymers. It has been shown that the hardness increases with the volume fraction of spheru- lites and can be defined by

H = O s p h ( ~ -~ H a ( l - - •)

where Hsp h and Ha are the microhardness values of spherulites and interspherulitic ("amorphous") material, respectively, and % is the volume fraction of spherulites [ 18,24]. Therefore, microhardness experiments were per- formed to investigate the changes induced in the surface morphology by the plasma and laser treatments.

Microhardness measurements were carried out on the treated surfaces using a Leitz microhardness tester with a Vickers diamond indenter. The microhardness H (MPa) was calculated from the indentation diagonal value d (in millimetres) according to the expression

H = kP/d 2

where P is the applied force (in newtons), d is the mean diagonal length of the indentation (in millimetres) and k is a geometrical factor equal to 1.854 [25]. A load of 1 N was employed and a loading cycle of 15 s was used.

Because of the difficulty of accurately measuring the indentation diagonal in the ceramic deposit, Vickers measurements on the laser or plasma cladding have not been carried out.

0.30

0

uJ 0.15

<

(a)

Z 60 90 120 150 180 210 240 270 300

0.30 l

;t I.U

0.15

N. .A < r.D

Cb)

L 30 60 90 120 150 180 210 240 270 300

Fig. 1. DSC analysis of the PET specimens (10 °C min 1): (a) amor- phous polymer; (b) crystalline polymer.

from the glassy state, which results in an exothermic peak located at 130 °C;

(3) the melting of the crystalline polymer, which corres- ponds to an endothermic peak at 242 °C.

During polymer cooling, a crystallization peak is observed between 218 and 140 °C, depending on the cooling rate, which varied from 5 to 40 °C rain 1.

To investigate the occurrence of each of these trans- formations during laser or polymer interaction, experi- ments were first carried out without ceramic treatment. A high scanning speed (0.1 m s 1) was selected for these

3. Results and discussion

3.1. Laser treatments

3.1 .I. Treatment without ceramic deposition During CO 2 laser exposure, energy is transferred as

heat into the substrate and can induce some morphologi- cal transformations [26-28]. From the analysis of the differential scanning calorimetry (DSC) trace (Fig. 1) of the "as cast" PET specimens, the following transitions can be identified during polymer heating: (1) the glass transition temperature Tg at 72 °C; (2) the crystallization of the amorphous PET regions Fig. 2. Cross-sections of the thermally modified area of the PET.

122 S. Ayrault et al./Surface and Coatings Technology 79 (1996) 119-130

experiments, to delay the thermal degradation of the sample.

For single-track treatment (scheme I), a progressive whitening of the PET bars, resulting from further crystal- lization, was noticed during subsequent laser runs of the laser beam on the track. Observation of cross-sections of the samples revealed the development of a thermally affected zone which tended to occupy the overall depth when the number of laser runs increased (Fig. 2). To analyze these morphological changes on a more quanti- tative basis, XRD measurements and microhardness testing were performed.

XRD profiles through the thickness of the "as-cast"

specimens are presented in Fig. 3(a). During the treat- ment, an increase in the initial crystallinity rate can be noted, regardless of the depth considered. However, this increase occurred to different extents and at different rates, depending on the sections analysed. After a rapid increase in the crystallinity rate during the first eight laser runs, a steady state was obtained on the specimen surface. However, a continuous increase in the crystal- linity rate was detected at 1 and 2 mm below the sample surface. This increase occurred at a slower rate than for the specimen outer layer and no steady state was reached for the number of laser runs considered.

The microhardness profiles reported in Fig. 4(a) were

50

a)

v

40

30

20

10

oT o

i I I

5 10 15 20

Number of r u n s

50

4O

30

= 2o

10

I I

0 5 10 i

15 20

Number of r u n s

Fig. 3. Changes in the crystailinity rates vs. the number of laser runs in the case of polymer treated without ceramic deposition: (a) scheme 1; (b) scheme 2; [] , at the surface; 4l,, 1 m m from the surface; I , 2 m m from the surface.

S. Ayrault et aL/Surface and Coatings Technology 79 (1996) 119-130 123

a) . ,c =

230

220

210

200

190

180

170

160

150

140

130

120

110

1 O0 0

crystalline polymer

i !

10 20

initial a.mrp_hou.2 _

polymer

!

30 40

Width (mm)

b)

230

220

210

200

crystalline polyn~,r

190

180

170 r ~

160

15o i l

140

k initi~tl

120 ~ , , ' m S ~ - • w "a-a.~.~.,.iZl~ ~er 11o

I oo ! ! !

0 10 20 30 40

Width (mm)

Fig. 4. Microhardness analysis of the laser-treated polymer (without coating, v=0.1 m s- l ) : (a) scheme 1; (b) scheme 2; [], 2 runs; e , 4 runs; II, 6 runs; O, 10 runs.

measured across the laser track. Data corresponding to the "as-cast" polymer and to the crystallized polymers are also provided as baselines. These results revealed a strong dependence of the surface hardness on the number of laser runs. During the first six laser runs, a progressive increase in the hardness was noted, with a maximum located at the centre of the track. A drastic change in the shape of the microhardness profile then occurred between six and 10 laser runs. Two pronounced maxima appeared on each side of the track, whereas a small decay in the microhardness occurred in the centre of the track.

An explanation for the changes in the XRD patterns and microhardness profiles may be proposed from laser absorption measurements [29]. Assuming an exponen- tial decay of the laser beam intensity in the polymer, absorption measurements were made on both cast (amorphous) films and crystallized films. The corre- sponding absorption coefficients have been found to be roughly independent of the PET morphology: amor- phous PET, ~ = 1.2 x 10 - 2 I l m - 1 ; crystalline PET, c~= 1.4 × 10 -2 pm -1. The values obtained also show that the energy of the laser beam is absorbed over a thin layer on the specimen surface (90% of the incident

124 S. Ayrault et al./Surface and Coatings Technology 79 (1996) 119 130

power is absorbed at about 200 gm below the surface). As a result, the polymer heating in the inner layers of the material does not result from direct laser/polymer

a)

b)

Crystalline ~ id ~ Hv

Fig. 5. Diagram of the morphology changes which occur in the polymer during the laser treatment of the treated track: (a) for a small number of runs; (b) for a high number of runs.

interactions but is the consequence of the diffusion of a thermal front from the outer layers.

During the subsequent laser runs of the beam on the treated track, the following sequence of processes can be supposed to occur (Fig. 5).

(1) The laser beam is absorbed at first without melting the surface layer (Fig. 5(a)). As a result of heating above Tg, crystallization from the glassy state occurs on these outer layers, especially in the centre of the track, where the incident energy of the gaussian laser beam is maxi- mized. This results in the observed increases in the crystallinity value and in the microhardness. However, the corresponding exposure time is insufficient to allow significant diffusion of a thermal front through the thickness of the PET bar. Consequently, the temperature of the inner layers remains below Tg and no significant increase in the crystallinity can occur during the first laser runs.

(2) During the subsequent laser runs, heat accumula- tion in the outer layers becomes sufficient to induce surface melting of the polymer (Fig. 5(b)). Therefore, on cooling after each laser run, crystallization occurs from the melted state. The crystallinity value obtained at the end of the treatment will also be dependent on the cooling rate of these outer layers. Thus, relatively low measured values (21%) could be explained by rapid cooling. It can be noted that such a hypothesis is consistent with finite difference modelling of the CO2 laser treatment of PET films, which showed that some amorphization could occur upon cooling of the specimen surface from the melted state [29,30]. During this second

50

=,

40

30

20

10

. . . . I . . . . I . . . . I . . . .

0 5 10 15 20

Number of runs

Fig. 6. Changes in the crystallinity rates vs. the number of laser runs for different scanning speeds of the laser beam (analyzed depth, 2 mm): O,v=0.1ms-1; l l l , v=0.083ms 1 ; i ,v=0 .067ms-1 ; [~,v=0.05ms 1.

S. Ayrault et al./Surfuce and Coatings Technology 79 (1996) 119 130 125

stage, the time is sufficient to allow the diffusion of heat through the cross-section of the bars, and temperatures higher than Tg can be attained in the inner layers and on each side of the laser track on the specimen surface. Consequently, the corresponding crystallinity values increase and the microhardness is enhanced on each side of the track.

All these processes involve heat transfer, so can be supposed to depend on the scanning speed. In Fig. 6, crystallinity rates at 2 m m below the specimen surface are plotted vs. the number of laser runs for different scanning speeds. As might be expected, the increase in the crystallinity rates is delayed at the higher scanning speeds because of shorter exposure times. At 0.05 m s - t, specimen overheating was indicated by the appearance of bubbles and brownish areas under the laser beam. Similar trends were noted on the specimen surface and at 1 mm below the surface.

For parallel-track treatment, laser treatments following scheme 2 were also performed on the "as-cast" PET without any ceramic deposition. XRD profiles (Fig. 3(b)) were obtained in the centre of the treated area. Regardless of the depth considered, the resulting crystallinity rates were higher for scheme 2 than for scheme 1. Moreover, the crystallinity gradient through the thickness was strongly reduced when more than four laser runs were performed. It can also be noted that the induction time for the crystallization decreased for scheme 2. From single-track experiments, the width of the treated zone below the laser beam was estimated to be close to 20 mm. Because the shift between each parallel track was only 7.5 mm, it is obvious that an important recovery of the track takes place.

Therefore, all the above effects can be explained by considering the heat accumulation which occurs in the treated area, as a result of the recovery of the three parallel laser tracks. On the specimen surface, the cooling rate of the polymer from the melted state will be lowered and higher crystallinity values will be achieved. In the inner layers, lateral heat transfer through each track will also enhance the crystallization of the polymer from the glassy state.

The microhardness profiles (Fig. 4(b)) are also different from those obtained from the single-track treat- ment. For a low number of laser runs, a maximum hardness was detected in the middle of the treated area, where the recovery of the different tracks is at a maxi- mum. When the number of laser runs on each track was increased, the microhardness was homogeneous on the overall treated width, as a result of a generalized crystalli- zation from the glassy state. With six laser runs per track, a decrease in the microhardness was detected in the centre of the treated area. This can be supposed to result from some quenching from the melted state, achieved with a high number of laser runs. Above six

laser runs, overheating of the surface tended to induce some measurement errors.

3.1.2. Treatment with ceramic deposition Laser treatments with ceramic deposition were per-

formed following the two deposition schemes. Regardless of the scheme considered, the laser cladding resulted in the formation of a ceramic coating on the PET substrate, in addition to crystallization through the thickness of the specimen. From observation of the cross-sections of the treated PET bars, the thickness of the deposit was estimated to be close to 150 gm (Fig. 7(a)). Also, with scanning electron microscopy (SEM) observation, the adhesion of the ceramic particles was revealed by encrus- ration in the softened polymer substrate (Figs. 8(a) and 8(b)). Therefore, mechanical anchorage would play an important part in the adhesion of the ceramic coating. It is clear that the softening of the polymer substrate that result from heating it above its glass transition

(a)

Fig. 7. Cross-sections of the deposit: (a) laser cladding; (b) plasma coating.

126 S. Ayrault et aL /Surface and Coatings Technology 79 (1996) 119-130

(a)

the thermal properties of the polymer and on the laser/ polymer interactions.

It must also be stressed that no change in the morphol- ogy of the ceramic particle was observed after the treatment (Figs. 9(a) and 9(b)). Furthermore, the XRD analysis of the coating also revealed that the crystallo- graphic structure of the ceramic (~ phase) remained unchanged after laser exposure. This may be interpreted as evidence of no melting of the alumina particles during the treatment.

XRD analysis of the polymer substrate was performed at 1 and 2 mm below the surface. The results reported in Figs. 10(a) and 10(b) show that lower crystallinity rates are obtained when ceramic particles are deposited. The induction time for this increase in crystallinity is also delayed. Two effects can be suggested to explain these changes:

(b)

(c)

Fig. 8. SEM analysis of (a) the ceramic powder, (b) the laser cladding and (c) the plasma coating.

temperature or its melting point will play an important role in the ceramic deposition. Therefore, the efficiency of this process can be assumed to be dependent on both

"O t - O

1860

O .

¢-

0

0 (a)

1860 " 0

C 0

c-

0

0

(b)

5 10 15 20 25 30 2e (degrees)

r

35 40

5 10 15 20 25 30 35 40 2e (degrees)

45

~ _ c ; .

45

1860

" 0 ¢..

0

Q .

c"

0 U

1

5 10 15 20 25 30 35 40 45 (c) 28 (degrees)

Fig. 9. XRD patterns of (a) the ceramic powder, (b) the laser cladding and (c) the plasma coating.

S. Ayrault et al./Surface and Coatings Technology 79 (1996) 119 130

50

127

a) __.=

40

30

20

10

50

s s

4.

"V /

/ j "

N u m b e r of runs

2 0

b) l= _=

L)

40 ¸

3 0 -

2 0 -

10

I

I

I

0 5 10 1 20

N u m b e r o f runs

Fig. 10. Changes in the crystallinity rates vs. the number of laser runs in the case of the polymer treated with ceramic deposition: (a) scheme 1; (b) scheme 2; [], 1 mm from the surface without ceramic deposition; ~, 2 mm from the surface without ceramic deposition; i , 1 mm from the surface with ceramic deposition; O, 2 mm from the surface with ceramic deposition.

(1) the absorpt ion of part of the laser beam energy by the ceramic powder injected in the beam, which means the incident energy on the polymer surface will be lowered;

(2) the absorpt ion of an increased part of the beam energy by the ceramic coating progressively deposited on the PET surface.

When the laser/polymer interactions are reduced by these two effects, the heating of the P E T substrate is decreased. Thus, the resulting crystallization from the glassy state is lowered.

3.2. Plasma treatment

3.2.1. Trea tment wi thout ceramic deposition

During plasma spraying, changes in the polymer morpho logy were also induced [31,32]. The energy of the flame was spread over the entire specimen surface and the substrate. The localized development of a ther- mally affected zone cannot be detected from the optical observat ion of specimen cross-sections, as was the case for the laser treatment.

From X R D measurements (Fig. 11), it can be noted

128 S. Ayrault et aL,/Surface and Coatings Technology 79 (1996) 119 130

50

40

30

.£ --- 20

10

I I I

5 10 15 20

N u m b e r of r u n s

Fig. 11. Changes in the crystallinity rates vs. the number of plasma runs in the case of the polymer treated without ceramic deposition: ~ , at the surface; , , 1 m m from the surface; i , 2 mm from the surface.

that no significant crystallization gradients appeared through the specimen thickness during the treatment. The crystallization rates are similar to those achieved with the laser treatment. Microhardness measurements (Fig. 12) also revealed a continuous increase in the crystallinity rate of the surface. Because of the difficulty of evaluating the energy input on the specimen surface, a comprehensive treatment of these observations is

clearly beyond simple analysis. However, it must be pointed out that morphological changes in the PET substrate are more homogeneous in the case of the plasma treatment.

3.2.2. Treatment with ceramic deposition Spraying of the ceramic powder resulted in the forma-

tion of a grey coating on the PET surface. The thickness

e =

.= .u

230

220

210

200

190

180

170

160

150

140

130

120

110

100

crystalline polymer

{

initial amorphous _

! !

0 10 20 40

po/ymer

!

3O

N u m b e r o f r u n s

Fig. 12. Variation of the microhardness vs. the number of runs for the plasma-treated polymer without coating.

S. Ayrault et al./Surface and Coatings Technology 79 (1996) 119-130

5O

129

40

30

==

20

10

i ! |

5 10 15 20

N u m b e r o f r u n s

Fig. 13. Changes in the crystallinity rates vs. the number of plasma runs in the case of the polymer treated with ceramic deposition: [], 1 mm from the surface without ceramic deposition; , , 2 mm from the surface without ceramic deposition; I , 1 mm from the surface with ceramic deposition; O, 2 mm from the surface with ceramic deposition.

of this coating was about 250-300 tam (Fig. 7(b)). SEM observations (Fig. 8(c)) showed that a significant change in the particle morphology occurred during the plasma spraying. The porous texture of the coating clearly indicates that alumina particles were melted into the flame. This was confirmed by X-ray analysis (Fig. 9(c)) of the coating, which revealed the appearance of a new

crystallographic structure, in addition to the original corundum structure (c~ phase) [33]. The coalescence of the melted ceramic particles on the polymer substrate can also explain the higher densification of the coating achieved by the plasma treatment compared with that for the laser cladding.

No significant difference in the crystallization rates can be detected between the treatment with or without alumina, except for a limited number of rotations of the sample (Fig. 13). As opposed to the laser treatment, the ceramic deposit plays a limited role as a thermal barrier in the case of plasma treatment.

As a result of the melting of ceramic particles, plasma spraying resulted in a denser coating than that obtained by laser cladding, where the particles are unaffected.

Thermal effects associated with these deposition pro- cesses induce morphological changes in the polymer substrate. During the laser treatment, the local develop- ment of a thermally affected zone was observed under the laser beam. Both crystallization from the glassy state and crystallization from the melted state have been found to occur in the polymer substrate. As a result, changes in the mechanical properties and dimensional stability must be expected during such ceramic depos- ition. During plasma spraying, these morphological changes have been shown to be more homogeneous.

Acknowledgements

Financial contributions of Region Rh6ne-Alpes as well as Sappi company are gratefully acknowledged.

4. Conclusions

The following conclusions may be drawn from this work.

Ceramic coatings can be deposited by laser cladding and plasma spraying on PET, despite the low heat thermal resistance of these substrates. The deposition processes involve the mechanical anchorage of the ceramic particles in the softened polymer surface layers. The coating texture depends on the deposition process.

References

[1] S. Kitahara and A. Hasui, J. Vac. Sci Technol., 1I(4) 1974 747 753.

[2] V. Wilms and H. Herman, Thin Solid Films, 39 (1976) 251 262. [3] M. Zaouali, Th~se E N S A M 1990. [4] Zhang Yaping, Gao Jiacheng, Tan Jifu and Zou Zhirong, Surf.

Coat. Technol., 58 (1993) 125 127. [5] J.M. PeUetier, M.C. Sahour, M. Pilloz and A.B. Vannes, J. Mater.

Sci., 28 (1993) 5184-5188.

130 S. Ayrault et al./Surface and Coatings Technology 79 (1996) 119-130

[6] A. Nojiri, K. Kumazawa and M. Komatsu, Metal coating on polymer films, French Patent 8908, 264.

[-7] T.C. Walton, Proc. 20th Int. SAMPE Technical Conf., 1988, pp. 357-365.

[8] R.A. Pike, V.M. Patarini, R. Zatorski and F.P. Lamm, Int. J. Adhesiv. •2(4) (1992) 227-231.

[9] M.P. Ducos, Galvano Organo, 598. [-10] C. Gauthier, J. Chauchard, B. Chabert, J.P. Trotignon,

G. Battegay and V. Lamblin, Plastics Rubber Compos. Proc. Appl., 20 (1993) 77-82.

[11] E. Fender, Surf. Coat. Technol., 34 (1988) 1-14. [12] M. Ducos, Tech. l'Ingdn., 1645. [13] J.H. Zaat, Ann. Rev. Mater. Sci., 13 (1983) 9-42. [14] S. Wu, (ed) Polymer Interface and Adhesion, New York, 1982. [15] H. Ajrhourh, Doctorate Thesis, Ecole Centrale de Lyon, 1991. [ 16] Q. Sail, Doctorate Thesis, Ecole Centrale de Lyon, 1991. [17] P. Sallamand and J.M. Pelletier, Mater. Sci. Eng., AI71

(1993) 263 270. [18] F.J. Balta Calleja, J. Baranowska and D.R. Rueda, J. Mater.

Sci., 28 (1993) 6074-6080. [19] A.M. Joly, Thesis, 1970. [20] J. Martinez-Salazar, J. Garcia-Pena and F.J. Balta-Calleja, Poly.

Commun. 26 (1985) 57 59.

[,21] J. Martinez-Salazar and F.J. Balta-Calleja, J. Mater. Sci. Lett., 4 (1985) 324 326.

[-22] F.J. Balta-Calleja, J. Martinez-Salazar and T. Asano, J. Mater. Sci. Lett., 7 (1988) 165-166.

[23] J. Marfinez Salazar, J.C. Canalda Camara, E. Lopez Cararcos and F.J. Balta Calleja, Colleq. Polym. Sci., 266 (1988) 41 46.

[-24] C. Santa Cruz, F.J. Balta Calleja, H.G. Zachmann, N. Stribeck and T. Asano, J. Polym. ScL B: Polym. Phys., 29 (1991) 819-824.

[25] J. Lopez, Polym. Test., 12 (1993) 437-458. [26] E.E. Said-Galiev and L.N. Nikitin, Mekh. Kompoz. Mater., 6

(1992) 723. [27] E.E. Said Galiev and V.V. Korshak, Spectrochim. Acta, 43A (2)

(1987) 223 224. [28] E.E. Said Galley and L.N. Nikitin, Mekh. Kompoz. Mater., 29(2)

(1993) 259-266. [--29] S. Etienne, J. Perez, R. Vassoille and P. Bourgin, J. Phys. III,

France 1, 61 (1991) 1587 1608. [-30] S. Lazare and P. Benet, J. Appl. Phys., 74(8) (1993) 4953 4957. [,31] N. Inigaki, S. Tasaka and K. Hibi, J. Polym. Sci. A: Polym.

Chem., 30 (1992) 1425-1431. [,32] Y. Momose, N. Noguguchi and S. Okazaki, Nucl. Instrum.

Methods B, 39 (1989) 805-808. [33] D. Fargeot, Thesis, l'universit6 de Limoges, 1987.