- Email: [email protected]
Vacuum arc filtered metal plasma application in hybrid technologies of ion-beam and plasma material processing
Surface & Coatings Technology 201 (2007) 8596 – 8600 www.elsevier.com/locate/surfcoat
Vacuum arc filtered metal plasma application in hybrid technologies of ion-beam and plasma material processing I.B. Stepanov a,⁎, A.I. Ryabchikov a , N.A. Nochovnaya b , Y.P. Sharkeev c , I.A. Shulepov a , I.A. Ryabchikov a , D.O. Sivin a , S.V. Fortuna d a Nuclear Physics Institute, Lenina ave. 2a, Tomsk, 634050, Russia All-Russian Scientific Research Institute of Aviation Materials, Radio str. 17, Moscow, 105005 Russia Institute of Strength Physics and Materials Science, Russian Academy of Sciences, 2/1 Pr. Academicheskii, Tomsk, 634021, Russia d Siberian Group of Chemical Enterprises, 1 Kurchatov st., Seversk, Tomsk Region, 636000, Russia b
c
Available online 12 March 2007
Abstract In this paper the latest results obtained for (Ti, Al)N, TiSiB coating deposition using a hybrid technology are presented. This technology includes several approaches. For sample pretreatment the repetitively pulsed high-current ion-beam and high-frequency short-pulsed metal plasma immersion ion surface treatment were used. An intermediate layer between coating and sample surface was formed by metal ion implantation using a high-current vacuum ion-beam and plasma source “Raduga-5”. The vacuum arc microdroplet-filtered metal plasma coating deposition was realized under ion mixing using high-frequency short-pulsed plasma immersion ion technology. It was experimentally shown that for deposition of TiAlN coatings, mono (Ti, Al) and multi-element (TiAl) cathodes can be used. Phase and elemental composition, structural state, and mechanical properties of (Ti,Al)N and TiSiB coatings were observed, and they are presented in this paper. © 2007 Elsevier B.V. All rights reserved. Keywords: DC vacuum arc; Microdroplet-filtered metal plasma; Coating deposition; Hybrid technologies
1. Introduction DC vacuum arc discharge (VAD) has been used for metal plasma coating deposition technologies for more than 30 years. Advantages of vacuum arc plasma include high-level ionization, elemental purity, and high directed velocity of the ions [1]. Simultaneously, the presence of microdroplets in a plasma stream significantly restricts dc VAD application for formation of such prospective coatings as (TiAl)N, TiC, CrN, (TiC)N, (TiSi)N, ZrN, MoS2, DLC, etc. New possibilities for further investigations in the field of dc VAD plasma use are promoted by development of high-efficiency ion-beam sources, metal plasma microparticle filtering system, and new methods for execution of ion-beam material treatment [2].
⁎ Corresponding author. Tel.: +7 3822 423963; fax: +7 3822 423934. E-mail address: [email protected] (I.B. Stepanov). 0257-8972/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2006.09.320
This paper is devoted to research of (TiAl)N, TiSiB coatings deposition using hybrid technologies of ion-beam and dc VAD metal filtered plasma treatment. 2. Equipment and investigation methods The schematic of experimental set-up is presented in Fig. 1. The experimental set-up comprised three dc vacuum arc evaporators (VAE). The evaporators were equipped with plasma filters (PF) of the shutter-type for microdroplet filtering [3,4]. For ion-beam treatment, the “Raduga-5” high-current ion-beam and plasma source was used. The ion source operational principle is based on microdroplet-filtered metal plasma generation with repetitively pulsed ion-beam extraction [5]. The set-up was equipped with a high-frequency negativevoltage generator for execution of high-frequency short-pulsed metal plasma immersion ion implantation or coating deposition [6]. Reactive gas was supplied through a leak fixed in the vacuum chamber.
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
8597
Fig. 1. The schematic of experimental set-up.
The elemental composition of the ion-alloyed surface layer was analyzed by Auger electron spectroscopy (AES). The phase composition and structural state were investigated by transmission electron microscopy (TEM). Phase analysis was carried out by identification of the microdiffraction patterns. Additionally, the phase composition was studied by X-ray diffraction (XRD). The change in hardness across the ion-implanted layer depth was investigated using a Nano-Hardness Tester “NHT-S-AX000X” (CSM). The Vickers nanoindenter load was in the range of 15–280 mN. The tribological properties were investigated by High Temperature Tribometer “THT-S-AX000” (CSM). Wear studies were carried out by the “ball-on-disc” method. Surface morphology was inspected using the 3D Profilometer “Micromeasure 3D Station” (Stil). 3. Ti(Al)N coatings deposition The samples underwent mechanical polishing, involving abrasive paste and chemical cleaning in an ultrasonic bath. Sample surface pretreatment in the vacuum chamber was executed under the regime of microdroplet-filtered Ti plasma generation combined with high-frequency short pulse metal plasma immersion ion implantation (HFSPPI3). This regime allows us to decrease on-surface arcing which is typical of dc bias potential formation and, thus, get rid of the mechanism of microcrater formation on sample surfaces. The bias voltage amplitude (Ub) of up to 4 kV together with the pulse repetition rate of up to 440 kHz and pulse duty factor of 0.66 allow us to effectively pretreat, activate and heat the surface, including effective sputtering of thin films formed between voltage pulses. In the case of decrease in pulse repetition rate and bias voltage pulse amplitude, plasma flow onto the target exceeds
the flux of atoms sputtered from the target surface. This fact provides conditions for ion mixing of the substrate–coating boundary with formation of a transition layer. The sample temperature was in the range of 450–550 °C. TiN coating formation was executed at bias potential amplitude of (from − 300 to − 2500 V, with reactive gas pressure (N2) of 2 · 10− 2 Pa. In contrast to widely used methods of coating deposition using magnetron sputtering systems [7,8] or vacuum arc evaporators with composite or compound cathodes [9–12], in our experiments the coating formation regime using two VDEs with Al cathode and one VAE with Ti cathode was realized. Coatings were deposited in the reaction gas (N2) environment. The ratio of Ti and Al plasma flows depended on VAD current (Id). Fig. 2 (a) and (b) shows the results of investigation of element composition of TiN and TiAlN coatings. The stoichiometric composition of deposited TiN (Ti ∼ 45%; N ∼ 45%) and TiAlN (Ti ∼ 29.5%; Al ∼27%; N ∼ 39%) coatings corresponds to the data on the diagram. Thick transition layers between the coating and substrate show that the regime of ion assistance was accompanied by intense diffusion processes near the sample surface. The results of diffraction electron microscopic phase analysis of TiN coatings showed that when the bias potential was in the range from − 750 to −2500 V, multi-phase structures composed of δ-titanium nitride (TiN) with grain size of 110–150 nm, respectively, were formed. When Ub was equal to − 300 V, ϵtitanium nitride (TixNy) and δ-titanium nitride phases was observed in the coatings; however, the size of grains of the former phase was half the size of grains of the TiN phase. The second type of coatings is characterized by formation of a solid solution (Ti1 − x, Alx)N with prominent growth structure. The average size of formed phases is 54 nm.
8598
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
Fig. 2. Element distribution into the depth of the deposited coating: a — TiN, b — (TiAl)N, c — TiSiB.
Fig. 3 shows the results of morphological investigation of the deposited coatings as images of TiN and (TiAl)N film surfaces. The presented data confirm that use of PF results in a decrease
of surface roughness by several times. For (TiAl)N coatings formed without PF, surface roughness equals tens of microns. This phenomenon is attributed to the presence of a large amount of microdroplets in the Al plasma flow (up to several tens of percents). For this case, the droplet size can be equal to more than 100 μm. The use of PF for Al plasma filtering allows us to decrease the content of microparticle fraction in the stream by several orders of magnitude, which permits us to deposit coatings with a roughness not more than 0.64 μm. Comparative analysis of TiN coating morphology shows that decrease in Id and increase in Ub allows us to significantly decrease the surface roughness of TiN coatings. This effect can be explained by an increase in plasma ion component transmission efficiency in PF electrodes and a decrease in microparticle fraction in the arc plasma flow at the expense of decrease in Id, as well as due to reflection of droplets from the negative potential near the sample surface [13] the number of the reference is changed and sputtering of microedges on the coating surface. Fig. 4 presents the results of investigation of (TiAl)N coating hardness as a function of the ratio of Ti and Al plasma concentrations. According to the data, coating with the hardness HV = 3670 kg/mm2 was obtained at Id = 80 A. Data from the elemental analysis show that the measured hardness corresponds to optimal stoichiometric composition of the coating. The data presented in Fig. 4 confirm that use of PF allows us to increase the TiN coating hardness by 40%. It was also discovered that an increase in bias potential in the range from − 300 to −2500 V results in a decrease in coating hardness by 15%–20%. The observed effect can be explained by the increase in phase formation temperature and consequent increase in phase size. The results of nanohardness measurement show that for coatings formed with the use of
Fig. 3. Profile of the coating surface: a — TiN without PF, b — TiN with PF (Id = 125 A, Ub = 500 V), c — (TiAl)N, d— TiSiB.
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
8599
Fig. 5. Change in friction coefficient depending on the number of rotations. Fig. 4. Change in hardness depending on coating deposition regimes.
microdroplet-filtered plasma, high magnitudes of Young's modulus are typical. The data on coating adhesion strength were obtained on the basis of scratch test results. Research results show that delamination of the TiN coating deposited without PF and at constant bias potential begins at loading equal to 0.81 N. A surface image made near the indenter track shows that the structure is porous, which reveals high inner tensions in the material. In the case of PF and pulsed negative bias potential application, the loading at which coating delamination begins increases up to 1.17 N, which confirms an increase in coating adhesion strength by 70%. For this case a high-intensity combined ion pretreatment, surface heating and activation, formation of a thick transition layer between the substrate and coating, and ion assisted coating deposition were used. A complex of these factors allowed us to form a (TiAl)N coating from dc VAD plasma; coating delamination was observed at critical loading of 1.13 N. For this case, cracking lines characteristic of materials with high internal tension are absent near the indenter track. In was experimentally found out that with the increase in Ub in the range from − 300 to − 2500 V, we observed an increase in adhesion strength of TiN coatings by 20%. The highest values of adhesion strength were registered in
Fig. 6. Wear of TiN and (TiAl)N coatings depending on deposition regimes.
the range from − 300 to − 750 V. The effect can be attributed to decrease in compressing residual tensions resulting from increase in temperature, ion treatment of the surface, increase in formed phase sizes, etc. Fig. 5 shows the results of investigation of the friction coefficient of TiN and (TiAl)N coatings. The data confirm that if we deposit TiN coating onto HSS steel, the surface friction coefficient decreases by 2.5 times — with or without PF. At the first stage of testing we observed a sharp increase in friction coefficient, which can be explained by “rubbing” of samples because of surface roughness — the same was noticed for TiN coatings deposited with PF. Due to low roughness, the effect of sample “rubbing” is absent at the initial stage, and more dense and homogeneous coating structure causes additional decrease in friction coefficient by 40%. Influence of the proposed treatment regimes on surface morphology modification can be seen in the diagram of Fig. 6. The diagram is normalized by unit with respect to wear intensity of the HSS initial sample. The presented data show that TiN coatings deposited using the conventional technology are characterized by a double increase in wear resistance. At the same time, use of PF allows us to additionally increase steel wear resistance by 5 times. The data obtained on wear intensity
Fig. 7. Influence of the TiSiB coating with a thickness of 10 μm onto VT6 sample fatigue strength (sample temperature is 450 °C, loading frequency is 2800–3000 Hz).
8600
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
without removal of wear debris show that the presence of wear debris in the indenter track, later used as an abrasive, results in a sharp increase in wear intensity. However, for this case wear resistance decreases 4 times compared to the initial state. Moreover, we did not notice any dependence of coating wear resistance on the sample bias potential. 3. Formation of TiSiB coatings The peculiarity of TiSiB coating formation regime is the use of the “Raduga-5” ion source at the stages of sample pretreatment, heating, and formation of the substrate–coating transition layer. The VAE of the “Raduga-5” source was equipped with a TiSiB composite cathode. VT6, VT8, and VT9 plates were used as target material. The samples underwent mechanical polishing and electrolytic polishing. Sample ion treatment was executed at the following parameters of the “Raduga-5” ion source: accelerating voltage Uacc = 20 kV, jb = 5–10 mA/cm2, ion-beam pulse duration of 400 μs, pulse repetition rate of 200 pps. The pressure in the vacuum chamber was P = 6.5 · 10− 3 Pa. An additional VAE combined with the “Raduga-5” ion source was used to increase coating deposition rate. Fig. 2c shows element concentration distribution profiles in the coating and into target depth. The presented data show that a coating with the thickness of 6 μm was formed on the coating surface. Elemental composition can be evaluated as TiSiB (Ti ∼ 40%; Si ∼ 43%; B ∼ 6%). The XRD method revealed the presence of TiB2 and Ti5Si3 phases in the coating composition. Due to high-intensity ion treatment, the formed substrate– coating transition layer depth reaches 1.5 μm. The diffusion mechanism of doped layer formation influences the smooth character of the dopant concentration decrease with depth in the target. Structural changes in target material were recorded at depths that exceed those of doped layers many times. Formation of the amorphous layer was observed at a depth of up to 25 μm at coating thickness of 5 μm. The results of TiSiB coating surface morphology investigation are presented in Fig. 3d. The presented data show that in the case when the plasma is filtered from microdroplets, a coating with a roughness of not more than 0.7 μm is formed on the substrate surface with a roughness of 0.64 μm. No microdroplets are observed on the coating surface in this case. The presence of a deep modified layer can explain the multiple increase in fatigue limit after repeated sample loading. The results of investigations executed at the Moscow Aviation Institute are presented in Fig. 7.
One can assume that the observed effect is connected with both change in the fatigue crack origin mechanism (for coated samples, the mechanism starts as a subsurface one) and change in the fatigue failure mechanism. For Ti alloys in initial samples the failure is accompanied by formation of facets of the quasichip type. In the case of coated samples, more small disoriented facets with subsequent formation of repeated chips are formed. In the coated area, the fracture has a smooth structureless surface, which is typical for fracture of amorphous, roentgen–amorphous or quasi-amorphous materials. The fracture velocity in the layer is slowed down, however, in case of layer embrittlement the fracture takes place according to the divergent scars schematic. Investigations of TiSiB coating erosion resistance were executed under sample treatment with quartz sand with a mean particle size of 80–120 μm. Particle velocity equaled 200 m/s, collision angle was 90°. At a coating thickness of 10 μm, formed in conditions of high-frequency short-pulsed ion assistance on VT8 samples, we observed an increase in the erosion resistance up to 20%. For VT6 alloy, the erosion resistance of the samples increased by 2.5 times. At the same time, with an increase in testing period up to 100 s a monotonic increase in erosion resistance of the coated samples in comparison with their initial state was observed. References [1] J.M. Lafferty, Vacuum Arcs. Theory and Application, Wiley, New York, 1982. [2] R.L. Boxman, V.N. Zhitomirsky, Rev. Sci. Instrum. 77 (2006) 1. [3] A.I. Ryabchikov, I.B. Stepanov, Patent RU 2097868 C1, 1998. [4] A.I. Ryabchikov, I.B. Stepanov, Rev. Sci. Instrum. 69 (1998) 893. [5] A.I. Ryabchikov, I.B. Stepanov, S.V. Dektjarev, O.V. Sergeev, Rev. Sci. Instrum. 69 (1998) 810. [6] A.I. Ryabchikov, I.A. Ryabchikov, I.B. Stepanov, Vacuum 78 (2005) 331. [7] W.-D. Munz, J. Vac. Sci. Technol. A4 (1986) 2717. [8] Shengrui Jiang, Dongliang Peng, Xueying Zhao, Liang Xie, Qiang Li, Appl. Surf. Sci. 84 (1995) 373. [9] H. Freller, H. Haessler, Thin Solid Films 153 (1987) 67. [10] J.R. Roos, J.P. Celis, E. Vancoille, et al., Thin Solid Films 193/194 (1990) 547. [11] Y. Tanaka, T.M. Gur, M. Kelly, et al., J. Vac. Sci. Technol., A, Vac. Surf. Films 10 (1992) 1749. [12] P. Holubar, U.M. Jilek, M. Sima, Surf. Coat. Technol. 133/134 (2000) 145. [13] P.M. Shanin, N.N. Koval, A.V. Kozyrev, I.M. Goncharenko, J. Langner, S.V. Grigoriev, Proceedings of the 5th Conference on Modification of Materials with Particle Beams and Plasma Flows, Tomsk, Russia, 2000, p. 438.
Vacuum arc filtered metal plasma application in hybrid technologies of ion-beam and plasma material processing I.B. Stepanov a,⁎, A.I. Ryabchikov a , N.A. Nochovnaya b , Y.P. Sharkeev c , I.A. Shulepov a , I.A. Ryabchikov a , D.O. Sivin a , S.V. Fortuna d a Nuclear Physics Institute, Lenina ave. 2a, Tomsk, 634050, Russia All-Russian Scientific Research Institute of Aviation Materials, Radio str. 17, Moscow, 105005 Russia Institute of Strength Physics and Materials Science, Russian Academy of Sciences, 2/1 Pr. Academicheskii, Tomsk, 634021, Russia d Siberian Group of Chemical Enterprises, 1 Kurchatov st., Seversk, Tomsk Region, 636000, Russia b
c
Available online 12 March 2007
Abstract In this paper the latest results obtained for (Ti, Al)N, TiSiB coating deposition using a hybrid technology are presented. This technology includes several approaches. For sample pretreatment the repetitively pulsed high-current ion-beam and high-frequency short-pulsed metal plasma immersion ion surface treatment were used. An intermediate layer between coating and sample surface was formed by metal ion implantation using a high-current vacuum ion-beam and plasma source “Raduga-5”. The vacuum arc microdroplet-filtered metal plasma coating deposition was realized under ion mixing using high-frequency short-pulsed plasma immersion ion technology. It was experimentally shown that for deposition of TiAlN coatings, mono (Ti, Al) and multi-element (TiAl) cathodes can be used. Phase and elemental composition, structural state, and mechanical properties of (Ti,Al)N and TiSiB coatings were observed, and they are presented in this paper. © 2007 Elsevier B.V. All rights reserved. Keywords: DC vacuum arc; Microdroplet-filtered metal plasma; Coating deposition; Hybrid technologies
1. Introduction DC vacuum arc discharge (VAD) has been used for metal plasma coating deposition technologies for more than 30 years. Advantages of vacuum arc plasma include high-level ionization, elemental purity, and high directed velocity of the ions [1]. Simultaneously, the presence of microdroplets in a plasma stream significantly restricts dc VAD application for formation of such prospective coatings as (TiAl)N, TiC, CrN, (TiC)N, (TiSi)N, ZrN, MoS2, DLC, etc. New possibilities for further investigations in the field of dc VAD plasma use are promoted by development of high-efficiency ion-beam sources, metal plasma microparticle filtering system, and new methods for execution of ion-beam material treatment [2].
⁎ Corresponding author. Tel.: +7 3822 423963; fax: +7 3822 423934. E-mail address: [email protected] (I.B. Stepanov). 0257-8972/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2006.09.320
This paper is devoted to research of (TiAl)N, TiSiB coatings deposition using hybrid technologies of ion-beam and dc VAD metal filtered plasma treatment. 2. Equipment and investigation methods The schematic of experimental set-up is presented in Fig. 1. The experimental set-up comprised three dc vacuum arc evaporators (VAE). The evaporators were equipped with plasma filters (PF) of the shutter-type for microdroplet filtering [3,4]. For ion-beam treatment, the “Raduga-5” high-current ion-beam and plasma source was used. The ion source operational principle is based on microdroplet-filtered metal plasma generation with repetitively pulsed ion-beam extraction [5]. The set-up was equipped with a high-frequency negativevoltage generator for execution of high-frequency short-pulsed metal plasma immersion ion implantation or coating deposition [6]. Reactive gas was supplied through a leak fixed in the vacuum chamber.
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
8597
Fig. 1. The schematic of experimental set-up.
The elemental composition of the ion-alloyed surface layer was analyzed by Auger electron spectroscopy (AES). The phase composition and structural state were investigated by transmission electron microscopy (TEM). Phase analysis was carried out by identification of the microdiffraction patterns. Additionally, the phase composition was studied by X-ray diffraction (XRD). The change in hardness across the ion-implanted layer depth was investigated using a Nano-Hardness Tester “NHT-S-AX000X” (CSM). The Vickers nanoindenter load was in the range of 15–280 mN. The tribological properties were investigated by High Temperature Tribometer “THT-S-AX000” (CSM). Wear studies were carried out by the “ball-on-disc” method. Surface morphology was inspected using the 3D Profilometer “Micromeasure 3D Station” (Stil). 3. Ti(Al)N coatings deposition The samples underwent mechanical polishing, involving abrasive paste and chemical cleaning in an ultrasonic bath. Sample surface pretreatment in the vacuum chamber was executed under the regime of microdroplet-filtered Ti plasma generation combined with high-frequency short pulse metal plasma immersion ion implantation (HFSPPI3). This regime allows us to decrease on-surface arcing which is typical of dc bias potential formation and, thus, get rid of the mechanism of microcrater formation on sample surfaces. The bias voltage amplitude (Ub) of up to 4 kV together with the pulse repetition rate of up to 440 kHz and pulse duty factor of 0.66 allow us to effectively pretreat, activate and heat the surface, including effective sputtering of thin films formed between voltage pulses. In the case of decrease in pulse repetition rate and bias voltage pulse amplitude, plasma flow onto the target exceeds
the flux of atoms sputtered from the target surface. This fact provides conditions for ion mixing of the substrate–coating boundary with formation of a transition layer. The sample temperature was in the range of 450–550 °C. TiN coating formation was executed at bias potential amplitude of (from − 300 to − 2500 V, with reactive gas pressure (N2) of 2 · 10− 2 Pa. In contrast to widely used methods of coating deposition using magnetron sputtering systems [7,8] or vacuum arc evaporators with composite or compound cathodes [9–12], in our experiments the coating formation regime using two VDEs with Al cathode and one VAE with Ti cathode was realized. Coatings were deposited in the reaction gas (N2) environment. The ratio of Ti and Al plasma flows depended on VAD current (Id). Fig. 2 (a) and (b) shows the results of investigation of element composition of TiN and TiAlN coatings. The stoichiometric composition of deposited TiN (Ti ∼ 45%; N ∼ 45%) and TiAlN (Ti ∼ 29.5%; Al ∼27%; N ∼ 39%) coatings corresponds to the data on the diagram. Thick transition layers between the coating and substrate show that the regime of ion assistance was accompanied by intense diffusion processes near the sample surface. The results of diffraction electron microscopic phase analysis of TiN coatings showed that when the bias potential was in the range from − 750 to −2500 V, multi-phase structures composed of δ-titanium nitride (TiN) with grain size of 110–150 nm, respectively, were formed. When Ub was equal to − 300 V, ϵtitanium nitride (TixNy) and δ-titanium nitride phases was observed in the coatings; however, the size of grains of the former phase was half the size of grains of the TiN phase. The second type of coatings is characterized by formation of a solid solution (Ti1 − x, Alx)N with prominent growth structure. The average size of formed phases is 54 nm.
8598
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
Fig. 2. Element distribution into the depth of the deposited coating: a — TiN, b — (TiAl)N, c — TiSiB.
Fig. 3 shows the results of morphological investigation of the deposited coatings as images of TiN and (TiAl)N film surfaces. The presented data confirm that use of PF results in a decrease
of surface roughness by several times. For (TiAl)N coatings formed without PF, surface roughness equals tens of microns. This phenomenon is attributed to the presence of a large amount of microdroplets in the Al plasma flow (up to several tens of percents). For this case, the droplet size can be equal to more than 100 μm. The use of PF for Al plasma filtering allows us to decrease the content of microparticle fraction in the stream by several orders of magnitude, which permits us to deposit coatings with a roughness not more than 0.64 μm. Comparative analysis of TiN coating morphology shows that decrease in Id and increase in Ub allows us to significantly decrease the surface roughness of TiN coatings. This effect can be explained by an increase in plasma ion component transmission efficiency in PF electrodes and a decrease in microparticle fraction in the arc plasma flow at the expense of decrease in Id, as well as due to reflection of droplets from the negative potential near the sample surface [13] the number of the reference is changed and sputtering of microedges on the coating surface. Fig. 4 presents the results of investigation of (TiAl)N coating hardness as a function of the ratio of Ti and Al plasma concentrations. According to the data, coating with the hardness HV = 3670 kg/mm2 was obtained at Id = 80 A. Data from the elemental analysis show that the measured hardness corresponds to optimal stoichiometric composition of the coating. The data presented in Fig. 4 confirm that use of PF allows us to increase the TiN coating hardness by 40%. It was also discovered that an increase in bias potential in the range from − 300 to −2500 V results in a decrease in coating hardness by 15%–20%. The observed effect can be explained by the increase in phase formation temperature and consequent increase in phase size. The results of nanohardness measurement show that for coatings formed with the use of
Fig. 3. Profile of the coating surface: a — TiN without PF, b — TiN with PF (Id = 125 A, Ub = 500 V), c — (TiAl)N, d— TiSiB.
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
8599
Fig. 5. Change in friction coefficient depending on the number of rotations. Fig. 4. Change in hardness depending on coating deposition regimes.
microdroplet-filtered plasma, high magnitudes of Young's modulus are typical. The data on coating adhesion strength were obtained on the basis of scratch test results. Research results show that delamination of the TiN coating deposited without PF and at constant bias potential begins at loading equal to 0.81 N. A surface image made near the indenter track shows that the structure is porous, which reveals high inner tensions in the material. In the case of PF and pulsed negative bias potential application, the loading at which coating delamination begins increases up to 1.17 N, which confirms an increase in coating adhesion strength by 70%. For this case a high-intensity combined ion pretreatment, surface heating and activation, formation of a thick transition layer between the substrate and coating, and ion assisted coating deposition were used. A complex of these factors allowed us to form a (TiAl)N coating from dc VAD plasma; coating delamination was observed at critical loading of 1.13 N. For this case, cracking lines characteristic of materials with high internal tension are absent near the indenter track. In was experimentally found out that with the increase in Ub in the range from − 300 to − 2500 V, we observed an increase in adhesion strength of TiN coatings by 20%. The highest values of adhesion strength were registered in
Fig. 6. Wear of TiN and (TiAl)N coatings depending on deposition regimes.
the range from − 300 to − 750 V. The effect can be attributed to decrease in compressing residual tensions resulting from increase in temperature, ion treatment of the surface, increase in formed phase sizes, etc. Fig. 5 shows the results of investigation of the friction coefficient of TiN and (TiAl)N coatings. The data confirm that if we deposit TiN coating onto HSS steel, the surface friction coefficient decreases by 2.5 times — with or without PF. At the first stage of testing we observed a sharp increase in friction coefficient, which can be explained by “rubbing” of samples because of surface roughness — the same was noticed for TiN coatings deposited with PF. Due to low roughness, the effect of sample “rubbing” is absent at the initial stage, and more dense and homogeneous coating structure causes additional decrease in friction coefficient by 40%. Influence of the proposed treatment regimes on surface morphology modification can be seen in the diagram of Fig. 6. The diagram is normalized by unit with respect to wear intensity of the HSS initial sample. The presented data show that TiN coatings deposited using the conventional technology are characterized by a double increase in wear resistance. At the same time, use of PF allows us to additionally increase steel wear resistance by 5 times. The data obtained on wear intensity
Fig. 7. Influence of the TiSiB coating with a thickness of 10 μm onto VT6 sample fatigue strength (sample temperature is 450 °C, loading frequency is 2800–3000 Hz).
8600
I.B. Stepanov et al. / Surface & Coatings Technology 201 (2007) 8596–8600
without removal of wear debris show that the presence of wear debris in the indenter track, later used as an abrasive, results in a sharp increase in wear intensity. However, for this case wear resistance decreases 4 times compared to the initial state. Moreover, we did not notice any dependence of coating wear resistance on the sample bias potential. 3. Formation of TiSiB coatings The peculiarity of TiSiB coating formation regime is the use of the “Raduga-5” ion source at the stages of sample pretreatment, heating, and formation of the substrate–coating transition layer. The VAE of the “Raduga-5” source was equipped with a TiSiB composite cathode. VT6, VT8, and VT9 plates were used as target material. The samples underwent mechanical polishing and electrolytic polishing. Sample ion treatment was executed at the following parameters of the “Raduga-5” ion source: accelerating voltage Uacc = 20 kV, jb = 5–10 mA/cm2, ion-beam pulse duration of 400 μs, pulse repetition rate of 200 pps. The pressure in the vacuum chamber was P = 6.5 · 10− 3 Pa. An additional VAE combined with the “Raduga-5” ion source was used to increase coating deposition rate. Fig. 2c shows element concentration distribution profiles in the coating and into target depth. The presented data show that a coating with the thickness of 6 μm was formed on the coating surface. Elemental composition can be evaluated as TiSiB (Ti ∼ 40%; Si ∼ 43%; B ∼ 6%). The XRD method revealed the presence of TiB2 and Ti5Si3 phases in the coating composition. Due to high-intensity ion treatment, the formed substrate– coating transition layer depth reaches 1.5 μm. The diffusion mechanism of doped layer formation influences the smooth character of the dopant concentration decrease with depth in the target. Structural changes in target material were recorded at depths that exceed those of doped layers many times. Formation of the amorphous layer was observed at a depth of up to 25 μm at coating thickness of 5 μm. The results of TiSiB coating surface morphology investigation are presented in Fig. 3d. The presented data show that in the case when the plasma is filtered from microdroplets, a coating with a roughness of not more than 0.7 μm is formed on the substrate surface with a roughness of 0.64 μm. No microdroplets are observed on the coating surface in this case. The presence of a deep modified layer can explain the multiple increase in fatigue limit after repeated sample loading. The results of investigations executed at the Moscow Aviation Institute are presented in Fig. 7.
One can assume that the observed effect is connected with both change in the fatigue crack origin mechanism (for coated samples, the mechanism starts as a subsurface one) and change in the fatigue failure mechanism. For Ti alloys in initial samples the failure is accompanied by formation of facets of the quasichip type. In the case of coated samples, more small disoriented facets with subsequent formation of repeated chips are formed. In the coated area, the fracture has a smooth structureless surface, which is typical for fracture of amorphous, roentgen–amorphous or quasi-amorphous materials. The fracture velocity in the layer is slowed down, however, in case of layer embrittlement the fracture takes place according to the divergent scars schematic. Investigations of TiSiB coating erosion resistance were executed under sample treatment with quartz sand with a mean particle size of 80–120 μm. Particle velocity equaled 200 m/s, collision angle was 90°. At a coating thickness of 10 μm, formed in conditions of high-frequency short-pulsed ion assistance on VT8 samples, we observed an increase in the erosion resistance up to 20%. For VT6 alloy, the erosion resistance of the samples increased by 2.5 times. At the same time, with an increase in testing period up to 100 s a monotonic increase in erosion resistance of the coated samples in comparison with their initial state was observed. References [1] J.M. Lafferty, Vacuum Arcs. Theory and Application, Wiley, New York, 1982. [2] R.L. Boxman, V.N. Zhitomirsky, Rev. Sci. Instrum. 77 (2006) 1. [3] A.I. Ryabchikov, I.B. Stepanov, Patent RU 2097868 C1, 1998. [4] A.I. Ryabchikov, I.B. Stepanov, Rev. Sci. Instrum. 69 (1998) 893. [5] A.I. Ryabchikov, I.B. Stepanov, S.V. Dektjarev, O.V. Sergeev, Rev. Sci. Instrum. 69 (1998) 810. [6] A.I. Ryabchikov, I.A. Ryabchikov, I.B. Stepanov, Vacuum 78 (2005) 331. [7] W.-D. Munz, J. Vac. Sci. Technol. A4 (1986) 2717. [8] Shengrui Jiang, Dongliang Peng, Xueying Zhao, Liang Xie, Qiang Li, Appl. Surf. Sci. 84 (1995) 373. [9] H. Freller, H. Haessler, Thin Solid Films 153 (1987) 67. [10] J.R. Roos, J.P. Celis, E. Vancoille, et al., Thin Solid Films 193/194 (1990) 547. [11] Y. Tanaka, T.M. Gur, M. Kelly, et al., J. Vac. Sci. Technol., A, Vac. Surf. Films 10 (1992) 1749. [12] P. Holubar, U.M. Jilek, M. Sima, Surf. Coat. Technol. 133/134 (2000) 145. [13] P.M. Shanin, N.N. Koval, A.V. Kozyrev, I.M. Goncharenko, J. Langner, S.V. Grigoriev, Proceedings of the 5th Conference on Modification of Materials with Particle Beams and Plasma Flows, Tomsk, Russia, 2000, p. 438.