Plasma immersion ion implantation of pure aluminium at elevated temperatures

Plasma immersion ion implantation of pure aluminium at elevated temperatures

Nuclear Instruments aad Methods in Physics Research B 127/128 (1997) 873-878 Beam-tions with Yater&sls& Atoms EISEVIER Plasma immersion ion implant...

532KB Sizes 13 Downloads 73 Views

Nuclear Instruments aad Methods in Physics Research B 127/128 (1997) 873-878

Beam-tions with Yater&sls& Atoms


Plasma immersion ion implantation of pure aluminium at elevated temperatures C. Blawert Institurfiir

[email protected]

*, B.L.


und Werkstoffrechnik. Technische Universiriir Clausthal. AgricolastraJe D-38678 Clausthal-Zellerfeld, Germany


Abstract A thin nitride layer containing high amounts of oxygen is formed on the surface of A1995 treated by Plasma Immersion Ion Implantation (PHI) at temperatures up to 500°C. Below this layer, nitrogen is found in a diffusion zone several microns thick but does not form nitrides. The thickness of the diffusion zone is dependent on the treatment temperature. Microhardness measurements reveal a decrease in hardness due to a loss of work-hardening. In spite of this annealing, both the wear and corrosion


are improved.

Keywords: Plasma Immersion Ion Implantation; Aluminium; Aluminium nitride; Wear

1. Introduction Aluminium nitride is a material that has many potential industrial applications in surface engineering because it has good mechanical properties, such as high hardness and stiffness, while it also has interesting physical properties, such as high thermal conductivity, low electrical conductivity and useful optical properties. For many applications, a coating of AlN can be applied on a substrate by reactive sputtering or by other physical or chemical vapour deposition techniques. In applications where an aluminium substrate is desirable, a surface layer of AlN can be produced by plasma nitriding [l-S] or nitrogen ion implantation (6-251. For tribological applications, good adhesion between the AlN layer and the substrate is required to prevent delamination. In this context, ion implanted layers or layers with a concentration gradient (such as are produced by plasma nitriding) on an ahnninium substrate are advantageous. At temperatures around 5OO”C,thick AlN layers can be produced by plasma nitriding [l-5] while at room temperature or low temperatures ( < 200°C) [6- 12,14,1625] and higher temperatures up to 500°C [7,8,13,15,20,23], thin AlN layers can be produced by ion implantation. If the temperature is increased during conventional ion implantation, outward diffusion of nitrogen is observed [8,13,21].

* Corresponding author. Fax: [email protected]


5323-9385-15; email:

However, by analogy with results obtained in steels, it may be possible to prevent the outward diffusion of nitrogen using plasma immersion ion implantation (PI111 [28]. In this way, implantation and nitriding may be combined to produce thick AlN layers which can substantially improve the tribological behaviour of aluminium. The aim of this work was to study the effects of PI11 treatment on commercially pure aluminium (A199.5) at temperatures up to 500°C with as little softening as possible. The results presented in this paper suggest a new approach to the PI11 treatment of aluminium alloys in general.

2. Experimental details Square specimens (30 mm X 30 mm> were cut from a 5 mm thick sheet of A199.5. The specimens were polished to a 1 pm diamond finish and cleaned in an alcohol ultrasonic bath before treatment. The PI11 treatments were carried out in the ANSTO Mark 1 system at TU Clausthal, details of which appear elsewhere [27]. A working pressure of 1.4 X 10m3 mbar nitrogen was used and the plasma was maintained by 288 W of rf power at 13.56 MHz. Specimens were treated at temperatures of 300, 400 and 500°C for a time of three hours using pulses of -40 kV amplitude and a pulse length of 100 us. The treatment temperature was controlled by varying the frequency of the I-N pulses. Consequently, the implanted dose increased with temperature.

0168-583X/97/$17.00 0 1997 Published by Elsevier Science B.V. AlI rights reserved PI1 SOl68-583X(97)00065-7



C. Blawert, B.L. Mordike/Nucl.

Instr. and Meth. in Phys. Rex B 127/128

Structural changes in the modified layer were investigated using cross sections for optical and Scanning Electron Microscopy @EM). Glancing angle X-ray diffraction (Siemens DSOO)with Cu Ko radiation was used to determine the phases present in the modified layer. Elemental depth profiles of nitrogen and oxygen were obtained by Sputtered Neutral Mass Spectroscopy (SNMS). Additional information about the consistency and thickness of the layer was obtained by Auger analysis. The surface hardness was measured with a Leitz Durimet microhardness tester with a Vickers indenter at loads between 0.05 and 3 N (5, 2.5, 50, 100 and 300 g). The wear performance of the modified layers was assessed with a Teer scratch tester using a constant load of 3 N applied to a 5 mm diameter 10006 ball oscillating on the surface of the specimen at a maximum speed of 0.05 m/s. Simple corrosion tests (DIN 50 905) were performed using a 3% NaCl solution adjusted to a pH value of 1.5 with HCl. The specimens were alternately exposed for 22 h to the NaCl/HCl solution and 2 h to air over a period of 3 d. During their time in air, the specimens were weighed and inspected to monitor changes in their appearance.

(1997) 873-878

shown in Fig. 3. A treatment temperature of 300°C appears to be the lower limit for any significant diffusion of nitrogen. After treatment at 500°C nitrogen was found at depths beyond 8 pm. Associated with the nitrogen diffusion is an inwards diffusion of oxygen. In the 300°C and 400°C specimens, the surface concentration of nitrogen and oxygen together is less than that required to form stoichiometric AlN and it may be that carbon is also embedded in the AlN layer. More than 33 wt% nitrogen, necessary for stoichiometric AlN, was only observed in the 500°C specimen. 3.2. Hardness, wear and corrosion The specimens were treated in the as-received condition which means that they had been hardened by cold working. Treatment at elevated temperature resulted in


3. Results 3.1. Structure and elemental depth projling Typical SEM micrographs from the surface and cross sections of treated specimens are shown in Fig. 1. With increasing treatment temperature, the surface roughness increased and the colour of the specimens changed from dark gold (300°C) to black (SOOOC).Fig. la shows that the surface is completely covered with small nodules which are most probably the result of sputtering and redepositing of Al, AlN and impurities. This morphology can also be seen in cross section, Fig. lb, which also shows a thin surface layer of AlN. This layer was observed on all of the specimens. Precipitation of nitrides with an hcp structure is evident in the diffraction patterns obtained by glancing angle XRD and shown in Fig. 2. The splitting of the aluminium peaks, which is more obvious at higher detector angles, is a result of Cu Kl3 radiation which is not completely filtered out by the post-specimen monochromator used in the detector. The amount of nitride precipitation increases with increasing temperature. The existence of a coherent layer of AlN on the surface was confirmed by Auger analysis. The measured Auger peaks were compared with those obtained from commercially available AlN powder (manufacturer: H.C. Starck). In the first 10 nm, only AlN peaks were observed without any metal Al peaks. Right at the surface, peaks corresponding to impurities of oxygen and carbon were detected. The nitrogen and oxygen depth profiles obtained by SNMS after a treatment of 3 h at various temperatures are


Fig. 1. Micrographs obtained from (a) the surface and (b) cross sections of A199.5 specimens treated for three hours at 40 kV and 500°C ( + electroless nickel-plating).

C. Blawerr, B.L. Mordike / Nuci. Instr. and Meth. in Phys. Res. B 127/ 128 (1997) 873-878


3h 500°C 3h 400°C 3h 300°C untreated Fig. 2. Glancing angle XRD patterns (incident angle of 3”) obtained from pure A199.5 and specimens treated for three hours at 40 kV at different temperatures.

annealing with a subsequent

loss in hardness as shown in

Fig. 4. A similar decrease in hardness was observed for specimens heated in a conventional furnace. An increase in the surface hardness due to the formation of the AlN layer is not sufficient to compensate for this loss. An attempt was made to convert the free nitrogen in the lattice to AlN by secondary heat treatments up to 600“C but no change in hardness was observed.

In spite of the decrease in hardness, the oscillating pin tests revealed an increase in wear resistance. The coefficient of friction and the wear volume were both reduced and, provided the AlN surface layer remained intact, only slight wear occurred. After the breakthrough of the layer, severe adhesive and abrasive wear occurred, similar to that observed for the untreated material, Fig. 5.

Optical micrographs of the surfaces of treated and untreated specimens after corrosion testing in NaCl/HCl solution are shown in Fig. 6. The coherent AlN layer was able to protect the Al against the severe pitting corrosion seen on the untreated specimens. As the corrosion tests continued, dissolution of the AlN layer was observed by a

sputtering time (s)


Fig. 3. SNMS element depth profiles from specimens treated for three hours at (a) 40 kV, 300°C and (b) 40 kV, 500°C.

Fig. 4. Microhardness as a function of load obtained from specimens treated under various conditions.



C. Blawert, B.L. Mordike/Nucl.

Imtr. and Meth. in Phys. Res. B 127/ 128 (1997) 873-878

Fig. 5. Optical micrographs of the wear scars and coefficient of friction of (a) untreated and (b) three hours, 40 kV, 400°C treated Al99.5 after 100 cycles (oscillating wear test, 3 N load, 10006 pin, unlubricated).

from black to grey/gold. The specimen treated at 500°C showed the best corrosion resistance with the AlN layer resisting pitting corrosion attack for more than two days.

colour change

4. Discussion PI11 treatment of A199.5 results in the formation of a nitride layer on the surface. After treatment at 5OO”C,the layer is about 50 nm thick which corresponds to the range of 40 keV N: ions in aluminium. Reflections from hcp AlN were detected in the XRD measurements for all specimens with the intensity of the peaks increasing with treatment temperature. The peaks are broad, implying that the grain size of the AlN is very fine. The concentrations of nitrogen and oxygen near the surface measured by SNMS (except the 500°C specimen) is too low to form the stoichiometric AlN compound, but it is obvious from the Auger analysis that a high amount of carbon exists in the surface layer. Carbon and oxygen can fit in the AIN structure by replacing nitrogen. These impurities account

for the colour of the AlN layer which is colourless if no impurities are embedded in the structure. Turning to the nitrogen distribution, it is surprising that the AlN layer is so thin given that nitrogen is found at much greater depth. This is intriguing since there is a high affinity between aluminium and nitrogen and the solubility of nitrogen in aluminium is low. The fact that AIN is confined to depths which correspond to the implantation range suggests that an additional activation energy, supplied by the implanted ions, is required to form AlN. All nitrogen which is not bound as AlN in the implantation region is able to diffuse inward. This is most likely to occur at high ion doses when all Al atoms are already saturated with nitrogen (or oxygen or carbon). A relatively high mobility of nitrogen in aluminium has also been observed in conventional ion implantation at elevated temperatures, although in this case it led to an outward diffusion of excess nitrogen [8,13,21]. In PI11 treatment of steels, it has been shown that the high concentration of nitrogen and its activity in the plasma can prevent this outward diffusion [28]. It appears that a similar phenomena occurs in aluminium.

C. Blawert. B.L. Mordike / Nucl. Instr. and Meth. in Phys. Res. B 127/128

(1997) 873478


Fig. 6. Optical micrographsof corroded surfaces after one and two days of exposure to a NaCI/HCI solution at a pH-value of 1.5: (a) untreated and (b) PI11treatment at 5OO”C,40 kV, three hours.

One would expect that the high amount of free nitrogen at greater depths would lead to an increase in hardness if it were possible to transform it into AIN. But an additional heat treatment of the implanted specimens in a nitrogen atmosphere at temperatures up to 600°C did not produce any further AlN precipitation and no increase in hardness was observed. This confirms our suspicion that a high activation energy is necessary for AlN formation. We are yet to determine the lower limit for this activation energy since AlN is formed even at 200°C and 20 keV, the lowest N: ion energy used in our experiments. We conject that nitrogen in the collision cascade is in a reactive state and that any excess nitrogen reverts to a more inert (molecular) state before diffusing to greater depths. No increase in hardness was observed on any of the treated specimens. This can be attributed to annealing at the elevated treatment temperatures. Furthermore, the natu-

ral aluminium oxide (AI,O,) on the surface of the untreated specimens is about 20 nm thick 1261 and has a higher hardness than AlN. If this layer is replaced by only a 50 nm thick AIN layer, then no significant hardness increase can be expected. Despite this loss in hardness, the wear resistance is improved by a reduction of the coefficient of friction and a change in the wear mechanism. The AlN layer is able to prevent metal-metal-contact and reduce the tendency for adhesive and abrasive wear, both of which occur on the untreated material. Although the AIN layer is gradually removed and the ball is pressed into the soft aluminium matrix, no wear occurs on the ball as long as the layer remains intact. There is a correlation between the appearance of the modified surface, the amount of AlN and the wear improvement. Best results were obtained after 400°C treatment, in which case the specimens have a smoother V. PLASMAS/LOW ENERGY ION BEAMS


C. Blawert, B.L. Mordike/Nucl.

[email protected] and Meth. in Phys. Res. B 127/ 128 (1997) 873-878

surface with finer morphology than that observed on the 500°C specimens. They also have a higher amount of AlN in the near surface region than the 300°C specimens. The formation of an AlN layer also improves the corrosion resistance with treated specimens corroding slower than untreated specimens and showing no pitting corrosion as long as the AlN remains intact. An improvement in corrosion resistance due to the formation of an AlN compound layer on PI11 treated A199.5 has been observed previously [22,24]. Our results show that the improvement is maintained at higher treatment temperatures.

5. Conclusions Our results indicate that it is not possible to produce a thick AlN layer on aluminium using PI11 with a pure nitrogen plasma. Although nitrogen can be implanted and retained in the aluminium because an inward diffusion takes place, AlN formation occurs only in the collision cascade. The thickness of the AlN layer does not exceed the implantation range. However, the overall aim of improving the wear resistance was achieved, despite there being no increase in hardness. This improvement in wear resistance has not been observed for low temperature PI11 treatment of aluminium. The good corrosion resistance observed at low treatment temperatures is maintained at higher temperatures. These results indicate that there is potential for elevated temperature PI11 treatment of aluminium and its alloys. Changes in the treatment processing are needed, however, to maintain the hardness. Preliminary results for age hardening alloys show that it is possible to combine the solution treatment of most alloys with the PI11 treatment. No delamination of the AlN layers occurs after quenching and, if followed by an appropriate ageing treatment, the hardness of the alloy can be restored. The results of these tests will be published elsewhere. To summarise, it is possible to form a wear and corrosion resistant layer on the surface of pure aluminium by PI11 treatment at elevated temperatures, suggesting an interesting possibility for the surface protection of aluminium and its alloys.

Acknowledgments This work is supported by the Bundesministerium Wr Bildung, Wissenschaft, Forschung und Technologie (BMBF). We thank Mr. Yu for the hardness and wear

results which are part of his diploma thesis “Plasmaimmersionsionenimplantation von Titan und Aluminium”.

References [ 11 T. Jung, Phys. Status Solidi (a) 93 (1986) 479. [2] K.-T. Rie and F. Schnatbaum, Metall. 44 (1990) 732. [3] E.1. Meletis and S. Yan, J. Vat. Sci. Technol. A 9 (1991) 2279. [4] H.-R. Stock, H.-Y. Chen and P. Mayr, Aluminium 70 (1994) 220. [5] H.-Y. Chen, H.-R. Stock and P. Mayr, Surf. Coat. Technol. 64 (1994) 139. [6] D. Schwabe and W. Mader, EuroMat’ (1989) p. 1267. [7] B. Rauschenbach, K. Breuer and G. Leonhardt, Nucl. Instr. and Meth. B 47 (1990) 396. IS] S. Lucas, G. Terwagne and F. Bodart, Nucl. Ins& and Meth. B 50 (1990) 401. [9] G. Terwagne, S. Lucas and F. Bodart, Nucl. Instr. and Meth. B 59/60 (1991) 93. [lo] B. Rauschenbach, Nucl. Tracks Radiat. Meas. 19 (1991) 943. [It] G. Terwagne, S. Lucas and F. Bodart, Nucl. Instr. and Meth. B 66 ( 1992) 262. [ 121 G. Terwagne and S. Lucas, Surf. Coat. Technol. 51 (1992) 368. 1131 S. Lucas and J. Chevallier, Surf. Coat. Technol. 51 (1992) 441. [14] T.R. Jervis, H.-L. Lu and J.R. Tesmer, Nucl. In&. and Meth. B 72 (1992) 59. [15] A.K. Gael, N.D. Sharma et al., Indian J. Phys. 67 A (1993) 75. 1161 K.K. Bourdelle and D.O. Boerma, Mat. Res. Sot. Symp. Proc. 279 (1993) 485. [17] C. Lin, Y. Li, J.A. Kilner et al., Nucl. Instr. and Meth. B SO/81 (1993) 323. [IS] L. Xia, R. Wang, X. Ma and Y. Sun, J. Vat. Sci. Technol. B 12 (1994) 931. [ 191 L. Calvo, A. P6rez-Rodrfguez et al., Nucl. Instr. and Meth. B 84 (1994) 214. [20] K. Hojou, S. Furuno et al., Microsc. Microanal. Microstruct. 6 (1995) 141. [21] U.R. Mharte, D.C. Kothari et al., Mat. Manufac. Proc. 10 (1995) 171. [22] J.S. Ghan, N.W. Cheung and K.M. Yu, Mat. Res. Sot. Symp. Proc. 268 (1992) 377. [23] R. Giinzel, E. Wiesner, E. Richter and J. Steffen, J. Vat. Sci. Technol. B 12 (1994) 927. [24] K.C. Walter, J. Vat. Sci. Technol. B 12 (1994) 945. [25] L. Zhang, J.H. Booske and J.L. Shohet, Mater. Lett. 22 (1995) 29. [26] Aluminium-Taschenbuch (Al. Zentrale Diisseldorf, Aluminium-Verlag Diisseldorf, 15. Aufl., 1995). [27] G.A. Collins et al., Surf. Coat. Technol., in press. [28] R. Hutchings, Mater. Sci. Eng. A 184 (1994) 87.