Tribological study of vanadium-based alloys ion implanted at low energy and high temperature

Tribological study of vanadium-based alloys ion implanted at low energy and high temperature

Vacuum 67 (2002) 543–550 Tribological study of vanadium-based alloys ion implanted at low energy and high temperature J.A. Garc!ıaa,*, R. Rodr!ıgueza...

277KB Sizes 0 Downloads 2 Views

Vacuum 67 (2002) 543–550

Tribological study of vanadium-based alloys ion implanted at low energy and high temperature J.A. Garc!ıaa,*, R. Rodr!ıgueza, R. Sa! ncheza, R. Mart!ıneza, M. Varelab, D. Ca! ceresb, * b, I. Vergarab, C. Ballesterosb A. Munoz b

a AIN-Centro de Ingenier!ıa Avanzada de Superficies, Cordovilla, 31191 Pamplona, Spain Universidad Carlos III, Dept. F!ısica, Avda. Universidad 30, 28911Legan!es, Madrid, Spain

Abstract This paper reports the improvements achieved in the tribological performance of V 5 at% Ti vanadium-based alloy ion implanted with nitrogen at low energy. Nitrogen ion implantation was carried out at 4001C with a current density of 1 mA/cm2 and an accelerating voltage of 1.2 kV. Nanoindentation tests showed hardness increases by a factor close to 3. Ball-on-disk tests showed a decrease in friction coefficient. The thickness of the active implanted layer estimated from the nanoindentation tests is close to 500 nm. A clear correlation between the improvement in the tribological properties: hardness, friction and wear and the implanted layer has been observed. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Ion implantation; Vanadium alloy; Hardness; Tribology

1. Introduction Vanadium-based alloys are amongst the most promising candidates for structural material for the fusion power devices. Despite its good properties under irradiation, low activation energy for neutrons and not long-life products [1–3], vanadium-based alloys show a poor tribological performance and wear resistance. The addition of Ti improves the mechanical properties of V alloys, but recent results suggest nucleation of precipitates assisted by Ti segregation on defects, even at low-Ti contents [4,5]. Ion implantation of light ions, specially nitrogen ions, into metallic materials has been subject of *Corresponding author. Tel.: +34-948-421-101; fax: +34948-421-100. E-mail address: [email protected] (J.A. Garc!ıa).

active research in the last years. Ion implantation at high energies (50–200 keV), improves dramatically the mechanical and tribological properties [6–9]. Recently, low-energy-high temperature ion nitriding has proved to be an effective tool to improve the tribological properties of different steels [10,11]. Two processes have been suggested as responsible for the improvement in the tribological properties in N implanted materials: N solid solutions and nitrides formation [12,13]. For both processes, the high diffusivity of N in the metallic matrix is exploited. At low-energy implantation, the ion bombardment is assumed to enhance the nitrogen mobility, and it can be associated with some of the elements of the alloy [14]. In order to obtain deep layers, high-flux and high-temperature are used in low-energy implantation. In this paper, a preliminary study has been made about the improvement of the tribological performance

0042-207X/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 2 0 7 X ( 0 2 ) 0 0 2 4 6 - 4

544

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

observed for V 5 at% Ti alloys N implanted at low-energy (1.5 keV) and elevated temperature (4001C). These improvements affect hardness, friction and wear.

Berkovich indenter was used. All tests were performed at room temperature. Each specimen was tested using the continuous stiffness measurement technique developed by Oliver and Pethica [15] at maximum load of 160 mN.

2. Experimental

2.5. Friction

2.1. Sample preparation

Friction measurements were made employing a ball-on-disk tribometer FALEX 320 PC, which allows recording the evolution of the friction coefficient. This equipment has a humidity controller to any chosen humidity between 5% and 95%. The test parameters were 50% of relative humidity, linear speed of 0.04 m/s and applied loads of 25, 50, 75 and 100 g. The wear tests were carried out against a 1/800 WC ball and the calculated contact pressures were: 0.4, 0.5, 0.6 and 0.65 GPa.

The V–Ti alloys with Ti concentration of 5 at% were produced from 99.9% pure V and 99.5% pure Ti by repeated arc melting in a high-purity He atmosphere. Prior to ion implantation, the surfaces were polished down to 1 mm diamond paste and then the alloys were annealed at 1573 K for 6 h in an oil-free vacuum of 103 Pa or less. 2.2. Ion nitriding The samples were N-implanted by means of an ID 2500 Kaufmam source at an accelerating voltage of 1.2 kV and a beam current of 1 mA/ cm2. The beam diameter was 7.5 cm and the total ion implantation time was 1 h. During ion implantation, samples were maintained at 4001C, using a thermocouple attached to the back of the samples to measure their temperature. The implanted dose was 4  1019 cm2, and taking into account the special behaviour of the Kaufman sources the implanted species were N+ at 1.2 keV + + and N+ 2 at 0.6 keV. The ratio of N –N2 ions was about 55–45. 2.3. Microhardness Microhardness measurements were made using a Fischercope H100VP microindenter. A Vickers indenter was used and four maximum loads were tested: 2, 5, 10 and 25 mN. For each sample, 10 indentations were done in order to obtain a statistical average.

2.6. Surface topography and wear The wear coefficient after ball-on-disc tests and the removed volume in the wear track was measured by using an optical profilometer WYCO RST 500 [16]. This equipment allows measuring the surface topography with resolutions better than 3 nm [17–19]. The wear rate K (loss volume/ sliding length*load) was measured in six different zones along the wear scars of the samples and the results were calculated averaging these six single measurements. 2.7. Scanning electron microscopy and X-ray analysis Scanning electron microscopy (SEM) and X-ray microanalysis (EDX) were carried out in a Philips XL 30 microscope equipped with a DX-41 EDAX spectrometer. X-ray diffraction (XRD) spectra were obtained in a Philips XPert diffractometer. 3. Results and discussion

2.4. Nanoindentation 3.1. XRD Nanoindentation experiments were made with a Nanoindenter Iis (Nano Instruments, Inc., Knoxville, TN) mechanical properties microprobe. A

XRD spectra were employed to obtain a lattice parameter of the unimplanted alloy of 0.3050 nm

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

545

hardness after N implantation. The corresponding hardness evolution with depth for the implanted and reference samples is represented in Fig. 1b. The thickness of the layer modified by N implantation can be estimated from Fig. 1b, ranging between 300–500 nm. Even at a final load of 2 mN, the indentation depth goes beyond the 10% of the implanted layer thickness, which implies that the measured value is a mixture between the layer and the substrate hardness [20–22]. The maximum values measured in the microindentation tests where found at 2 mN of final load, reaching increases in hardness bigger than 162%. Table 1 summarises the microindentation results for the implanted and unimplanted samples.

in good agreement with the reported data for this composition [4]. No significant changes have been observed for the implanted samples. No evidence of nitride formation is observed probably due to the random orientation of the TiN precipitates. 3.2. Microindentation and nanoindentation tests Implanted and unimplanted samples exhibited the same qualitative behaviour when indented at the same maximum indentation loads of 2, 5, 10 and 25 mN. Fig. 1a shows the load/unload curves for the implanted and unimplanted reference sample at 2 mN of maximum indentation load. The lower depth of indenter penetration for the implanted sample indicates a noticeable increase in micro-

Before Implantation 0.18 0.16

Implanted

Depth (microns)

0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0.25

0.50

0.75

1.00

(a)

1.25

1.50

1.75

2.00

0.6

0.7

2.25

Load (mN) 6000 5500

2

Universal Hardness (N/mm )

5000 4500

Implanted

4000 3500 3000 2500 2000 1500 1000

Before Implantation

500 0 0.0

(b)

0.1

0.2

0.3

0.4

0.5

0.8

Depth (microns)

Fig. 1. (a) Microindentation load–unload curves for the implanted and reference sample. (b) Microhardness profile for the implanted and reference sample.

LOAD (mN)

120

After implantation

100 80 60 40 20 0 0

(a)

14007100 24007300 19007170 1500750 5.771.1 Implanted

1000

1500

2000

DISPLACEMENT (nm)

6 5

After implantation

4 3 2

Before implantation

1

15007110 16007180 15007125 1300760

500

7

HARDNESS (GPa)

— 0.25-0.9 0.25-0.9 0.42-0.9 —

70728 1.4E-13 2.7E-13 3.0E-13 2575 — 0.9 0.9

Before implantation

140



0.9

180 160

-20

Unimplanted 1.870.3

Ranm Rqnm 2 mN

5 mN

25 mN

100 mN

0.4 0.5 GPa GPa

0.6 GPa

0.65 GPa

0.25 0.5 N N

0.75 N

1N

Roughness Wear coefficient (m2/N) at load of Friction coefficient at contact pressure of Nanoindentation Universal hardness (H.U. MPa) (GPa) at final load of Sample

Table 1 Summary of the microhardness and tribological results

2.6 E-13 1.4E-13 60722 85735

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

546

0 0

(b)

500

1000

1500

2000

CONTACT DISPLACEMENT (nm)

Fig. 2. (a) Nanoindentation load–unload curves for the implanted and reference sample. (b) Hardness profile for the implanted and reference sample.

In order to measure the hardness of the outer surface, nanoindentation tests were carried out on unimplanted and implanted samples. Fig. 2a shows the load–unload curves for the implanted and unimplanted samples at a maximum load of 160 mN and Fig. 2b shows the hardness results for the unimplanted and implanted samples. An increase of more than a factor 3 in the hardness values is observed for the implanted samples. The maximum hardness for the implanted layer is of 5.771.1 GPa, clearly higher than the 1.870.3 GPa of the unimplanted sample. The thickness estimated from Fig. 2b for the layer modified by N implantation is of 500 nm, in good agreement with the value estimated from the microindentation tests. Nanoindentation results indicate that

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

547

Fig. 3. (a) Shows the friction coefficient evolution for the implanted samples at the different contact pressures: (A) 0.65 GPa, (B) 0.6 GPa, (C) 0.5 GPa, (D) 0.4 GPa, and (E) 0.6 GPa for the reference sample. (b) Friction coefficient evolution for the implanted and reference sample at 0.6 GPa.

the N-implanted layer is inhomogeneous, obtaining the maximum hardness at a thickness close to 35 nm. Ti easily forms the nitride phase and can improve the hardness. In the case of steels this

fact can occur at Ti concentrations below 1% [13]. The presence of Ti in the alloy, the elevated temperature of the implantation (4001C) and the observed increase in hardness suggest the formation of nitride precipitates.

548

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

3.3. Friction and wear tests The ball-on-disk test shows a decrease in the friction coefficient, m; for the implanted samples at all levels of load and contact pressure. An example of the evolution of m versus the number of cycles, for the implanted and the reference sample at a contact pressure of 0.6 Gpa, is shown in Fig. 3b. Fig. 3a shows the evolution of m for four different levels of contact pressure, 0.65, 0.6, 0.5 and 0.4 GPa. Three different periods in the measured friction coefficient are observed. At the first stage below 800 cycles, the implanted samples showed a lowfriction coefficient associated to the mild wear period, m ¼ 0:220:42: This behaviour is only present in the implanted samples and could be attributed to some kind of surface contamination produced during the implantation process. At the intermediate period, ‘‘rising’’ period or transition period, after 800 cycles, the friction coefficient rises to the value of the unimplanted samples, m ¼ 0:9: Although the number of cycles of the first period is the same for the different loads, the number of cycles of the second period depends strongly on the applied load, see Fig. 3a, where the curve corresponding at 0.4 GPa saturates at 40 000 cycles. At the third period the friction coefficient measured was m ¼ 0:9 and this is associated to a severe wear regimen. This behaviour is the same for the implanted and unimplanted sample. The three periods observed in the evolution of m suggest a variation in composition along the implanted layer. The sharp transition between the ‘‘rising period’’ and the final period, characteristic of an implanted layer, is associated to a homogeneous thickness for the implanted layer. The results are summarised in Table 1. Samples have been analysed by SEM and EDX. N-detection by EDX is difficult, mainly due to the presence of the L-lines of V and Ti. In order to improve the N detection, referring to the alloy components, low electron beam energy below 1.5 keV has been used. In these conditions it is possible to detect the presence of N in the implanted samples. No N has been detected analysing the wear track after the rising period, as expected from the friction coefficient measurement.

The measurement of the wear coefficient was very difficult, due to the irregularities in the track. A wear decrease of more than a factor 2 was obtained for a load of 1 N. The observed wear decrease is expected after the hardness increase measured in micro- and nanoindentation tests. Moreover, low-energy-high-temperature nitrogen implantation into vanadium alloys modifies strongly the surface topography. Table 1 also gathers the roughness parameters for the implanted and unimplanted samples, and it can be observed a big increase in roughness. The modification of the surface topography can be appreciated in Fig. 4. In the implanted samples a contrast associated to the grain boundaries was observed indicating a preferential sputtering induced by low-energy N implantation. Ar+ ion sputtering at similar energies of 1.2 keV allows modifying the surface roughness [24]. The achievement of surface structures has been explained on the basis of the interplay between the sputtering processes, which depends on the local curvature, and the surface diffusion smoothening processes, which depend on the local temperature. Both processes are not material dependent [23]. Surface topography affects the friction coefficient and could be partially responsible for the mild wear period observed.

4. Summary and conclusions Micro- and nanoindentation tests showed a hardness increase in the implanted samples of a factor 1.6 and 3, respectively. This hardness increase is associated with the N implantation, suggesting nitride precipitation and giving approximate information about the active thickness of the implanted layer. Friction tests showed decreases in friction coefficient in the implanted samples at all selected loads. At the beginning of the tests, a non-wear regimen was found with a friction coefficient of about 0.25. This period was maintained for more than 800 cycles at all the tested loads. The different periods observed in the evolution of the friction coefficient suggest that the N-implanted layer is not homogenous. The sharp transition between the

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

549

Fig. 4. 3D profiles showing the surface topography before (a) and after the ion implantation (b). A very rough surface and grains are observed after nitrogen implantation indicating a preferential sputtering.

rising period and the final period indicates a sharp interface and a uniform thickness. Preliminary wear tests seem to point decreases in the wear coefficient at high loads, in good agreement with the hardness increase. Repeated tests are needed to validate this assessment. Low– energy-ion nitrogen implantation induces important changes in the surface morphology and preferential sputtering is observed. From the results presented here, it can be seen that improvement in the tribological properties of

vanadium-based alloys is possible by means of low-energy nitrogen implantation at elevated temperatures (4001C). The thickness of the implanted layer estimated from the nanoindentation tests is close to 500 nm, suggesting some diffusion into the material associated with the high temperature of implantation. Other implantation effects, as nitride precipitation, preferential sputtering or diffusion enhanced by implantation, can be responsible of the hardness, wear and friction changes observed.

550

J.A. Garc!ıa et al. / Vacuum 67 (2002) 543–550

Further studies are needed to investigate the correlation between the structure of the implanted layer and the tribological performance of the vanadium-based alloys.

Acknowledgements Authors would like to thank the support received from CICYT through the project MAT 99-1012.

References [1] Chung HM, Lomis BA, Smith DL. J Nucl Mater 1994;804:212. [2] Gold RE, Harrod DL. Int Metals Rev 1980;25:232–8. [3] Honeycombre RWK. The Plastic Deformation of Metals A. * A, Leguey T, Pareja R, [4] Varela M, Fern!andez B, Munoz Ballesteros C. Philos Mag Lett 2001;81:259–64. * A, Pareja R. J Nucl Mater 1999;279:138. [5] Leguey T, Munoz [6] Garc!ıa JA, Gu!ette A, Medrano A, Labrug!ere C, S!anchez R, Lahaye M, Mart!ınez A. Vacuum 2002;64:341–51. [7] Rodr!ıguez RJ, Sanz A, Medrano A, Garc!ıa JA. Vacuum 1999;52:187–92.

[8] Rom!an E, de Segovia JL, Rodr!ıguez RJ. Vacuum 1994;45:1007–8. [9] Miyagawa Y, Nakao S, Ikeyama M, Saitoh K, Miyagawa S. Nucl Instrum and Meth 1997;B121:340–4. [10] Bacal M, Peyre JP. Appl Surf Sci 1999;138–139:155–8. [11] Bull SJ, Jones AM, McCabe AR. Surf Coat Technol 1996;83:257–62. [12] Byeli AV, Lobodaeva OV, Shykh SK, Kukareko VA. Nucl Instrum and Meth 1995;B103:533–6. [13] Jones AM, Bull SJ. Surf Coat Technol 1996;83:269–74. [14] Williamson DL, Davis JA, Wilbur PJ, Vajo JJ, Wei R, Matossian JN. Nucl Instrum and Meth 1997;B127/ 128:930–934 and references therein. [15] Pethica JB, Oliver WC. Mater Res Soc Symp Proc 1984;24:515. [16] Standar test method for wear testing with a Pin-on-Disk Apparatus ASTM G99-95a [17] Caber PJ, Martinek SJ, Niemann RJ. Proc SPIE 1993;2088:195–203. [18] Wyant JC. Laser Focus World 1993;September: 131–3. [19] Caber PJ. Appl Opt 1993;32(19):3438–41. [20] Farhat ZN, Ding Y, Northwood DO, Alpas AT. Mater Sci Eng 1996;A206:302–8. [21] Loubet JL, Georges JM, Meille G. Micro Technol Mater Sci Eng ASTM STP 1986;889:72–89. [22] Oliver WC, Pharr GM. J Mater Res 1992;7:1564–83. [23] Cuerno R, Barab!asi AL. Phys Rev Lett 1995;74:4746. [24] Gago R, V!azquez L, Cuerno R, Varela M, Ballesteros C, Albella JM. Appl Phys Lett 2001;78:3316.