INSTRUMENTATION Revista Mexicana de Fısica S59 (1) 90–94 FEBRUARY 2013

Interface description using computational methods and tribological characteristicOF TiN/TiC films prepared by reactive pulse arc evaporation technique

D.M. Devia Narvaeza,b, J.M. Gonzalez Carmonac, and A. Ruden MunozcaUniversidad Nacional de Colombia-sede Manizales, Campus La Nubia, Colombia.

e-mail: dmdevian@utp.edu.cobUniversidad Tecnologica de Pereira-La Julita, Colombia.

cUniversidad del Valle, Edificio 349, espacio 1003-Ciudad Universitaria, Colombia.

Received 30 de junio de 2011; accepted 9 de septiembre de 2011

The TiN/TiC bilayers have been deposited by Plasma Assisted Physical Vapor Deposition Technique (PAPVD) - Reactive Pulsed Arc.The coatings were analyzed by X-Ray Photoelectron Spectroscopy (XPS) and X-Ray Diffraction (XRD). From the signal treatment of thenarrow XPS spectra and the XRD diffraction patterns, the formation of TiN (titanium nitride), TiC (titanium carbide) and TiCN (titaniumcarbonitride) was confirmed, with fm-3m spatial group, corresponding to the FCC phase of the synthesized compounds. The multilayer wassimulated using Density Functional Theory (DFT) by the Unrestricted Hartree Fock method. Charge distributions and electron total densitywere obtained; finding bond formation at the interphase, electrical neutrality and system stability. Anomalies in the corners of the structuresdue to edge effect, simulation ideality and the no internal tension inclusion, intrinsic to the growing, are observed. The ball on disc tribometerwas used to measure the friction and wear coefficient to verify the interface formation.

Keywords: Electronic density; interphase; tribology.

Los recubrimientos de TiN/TiC fueron depositados por la tecnica de Deposicion Fısica en Fase de Vapor asistida por Plasma (PAPVD)por Arco Pulsado. Los recubrimientos fueron analizados por Espectroscopia de fotoelectrones de Rayos X (XPS) y Difraccion de RayosX (XRD). El tratamiento de la senal del espectro angosto XPS y los patrones XRD confirmaron la presencia de TiN (Nitruro de Titanio),TiC (Carburo de Titanio) y TiCN (Carbonitruro de titanio), con grupo espacial fm-3m, correspondiente a la fase FCC de los compuestossintetizados. La multicapa fue simulada usando Teorıa de Funcionales de Densidad (DFT) usando el metodo de Hartree Fock no restringido.Se realizaron analisis de la distribucion de carga y la densidad total electronica, encontrando la formacion de enlaces en la interfase delsistema, neutralidad electronica y estabilidad del sistema. Se observan anomalıas en las esquinas de las estructuras debidas a efectos deborde, idealidad de la simulacion y la no inclusion de esfuerzos intrınsecos a la formacion del material. Se utilizo el metodo de Bola en Discopor medio de un tribometro para observar la friccion y el coeficiente de desgaste para verificar la formacion de la interfase.

Descriptores: Densidad electronica; interface; tribologıa.

PACS: 31.15.E-; 62.20.Qp

1. Introduction

In the multilayer films the interaction between the differ-ent interphases of the materials, determinates the adherence,therefore, plays a main role in the different industrial appli-cations to improve the coating performance [1,2]. Multilayercoatings grown by high-energetic deposition techniques al-ways struggle with high atomic mixing of both adjacent coat-ing materials. One of these high-energetic deposition tech-niques is the Cathodic Arc Physical Vapor Deposition [3].The applications include cutting tools, use in thermal barri-ers and corrosion resistance in metallic structures [4,5]. TheTiC applications have raised a great investigative interest, es-pecially to understand the compound structural and superfi-cial properties [6,7], to allow the control of these propertiesin function of the technique and deposition parameters. TheTiN film, which is in contact with substrate, has the functionof promote adherence, reducing interface tensions [8]. TheTiC layer is used to give high hardness to the system, lowfriction coefficient and wear resistance [9]. The aim of thispaper is to observe the interphase behavior trough the forma-tion of a TiCN layer and the tribological characteristics of thecomplete set of compounds.

2. Experimental Setup

TiC/TiN coatings were deposited using noncommercialPulsed Arc PAPVD [10] technique on AISI 304 substrates,with titanium target 99.99% purity. In Table I the depositionparameters are shown, obtaining∼372 nm and 41 nm overallthickness and roughness. For X-ray diffraction characteriza-tion a Bruker D8 AXS Diffractometer was used. It has anX-ray source of Cu-Kα, with λ=1.5406A wave length anda secondary graphite monochromator. Diffraction patternswere taken under parallel beams with a grazing incidence of2◦ and a scanning range of 2θ between 30◦ and 80◦. XPS a-

TABLE I. TiN/TiC bilayers deposition parameters

Conditions TiN TiC

Atmosphere N2 CH4

Pw (mbar) 1.7 3.0

Charge Voltage 300 280

Number of pulses 4 4

Ts (◦C) 110◦ C 110◦C

INTERFACE DESCRIPTION USING COMPUTATIONAL METHODS AND TRIBOLOGICAL CHARACTERISTIC OF TiN/TiC. . . 91

FIGURE 1. (a) Narrow spectra for Ti2p in the TiN/TiC bilayer. (b)Narrow spectra for C1s in the TiN/TiC bilayer. (c) Narrow spectrafor N1s in the TiN/TiC bilayer.

nalysis was performed in a Thermo Vg Scientific ESCALAB250 XPS/ISS, with X-ray source of monochromatic Al-Kα.The system was simulated through the software Gaussian98 [11]; the duration of the simulation was 45 hours, 20 min-utes, 54 seconds. Density functions convergence criteria was

10−5. The friction coefficient was measured using a tribome-ter ball on disc CSEM and the software Tribox 2.9 was usedto calculate the wear coefficient (Al2O3 ball, 1N load, 10 m/sspeed,∼23◦C, %50 relative humidity and 100 m testing dis-tance), and the worn track area was measured by profilometerXP 2 Ambios Technology.

3. Results and Discussion

XPS Results

Figure 1a, 1b and 1c shows the XPS narrow spectra for ele-ments of TiN/TiC bilayers (Ti2p1/2, Ti2p3/2, C1s and N1s),the coating was deposited at 110◦C substrate temperature.The signal was deconvolucionated to obtain the binding ener-gies for TiC [12], TiCN [13], CN [14] and Ti-N [15], the val-ues of these energies are shown in Table II. Figure 1b showsthe peak corresponding to C-C bond, which is formed by non-reactive molecules of the gas (CH4) in the deposition of TiC.

TABLE II. XPS obtained energies for the elements in the TiN/TiCbilayers.

Element Bond Binding energy (eV)

Ti 2p1/2Ti-C-N 462.4

Ti-C 460.5

Ti 2p3/2Ti-C 454.4

Ti-C-N 456.3

C1sTi-C 282

C-C 285

N1sTi-N 396,9

C-N 398,7

FIGURE 2. XRD diffraction patterns for the TiN/TiC bilayers, ob-serving (111), (200) and (311) peak landslide for 110◦C substratetemperature.

Rev. Mex. Fis. S59 (1) (2013) 90–94

92 D. M. DEVIA NARVAEZ, J. M. GONZALEZ CARMONA, AND A. RUDEN MUNOZ

XRD Results

In Fig. 2 the diffraction pattern for bilayers coating is ob-served. The pattern shows orientation on the crystallographicplanes corresponding to FCC phase of the compounds pre-viously analyzed by XPS, TiN, TiC, TiCN, in the directions(111), (200), (220) and (311). Previously, Yamasaki [16] andA. Devia [17] presented these guidelines as for bilayers ofTiN/TiC. The diffraction peaks overlap, is due to the samefactor structure of the analyzed compounds that crystallizedin the same space group fm-3m, in addition, their lattice pa-rameters are similar [18,19], as reported by Kim [20] andDevia [21]. The diffraction pattern of the multilayer peaksobserved coincide with those reported in the database ISDD#064904 [18], which corresponds to a diffraction pattern withlow residual stresses, this effect is due to the growth temper-ature because it presents the best atomic diffusion gradient,decreasing the grain boundaries and increasing the coatingcrystallinity [22].

Gaussian Results

Figure 3a and 3b shows the simulated system, with atomicradii in a 50% scale and the simulation results (14 Ti atoms,13 C atoms in TiC, 14 Ti atoms and 13 N atoms for TiN).Figure 3a shows that the system consists in TiN and TiCcell overlapping in order to simulate the interface betweenthese materials. The simulation in (Fig. 3b) shows manybonds generated at the TiN - TiC interface due to the for-mation of internal stresses as seen in the diffraction patterns

in Fig. 2, these stress are reflected in the change of positionand peak broadening of the diffraction pattern by the forma-tion of TiCN, as shown by the XPS and XRD analysis. Highatomic density at the interface leads to formation and de-struction of bonds with the purpose of stabilizing the system,atomic rearrangement is also appreciated, which is a way ofreleasing tension. In Fig. 3c, shows the Mulliken popula-tion charge analysis, the organization agrees to atomic elec-tronegativity, determining the ceramic character of TiN/TiC.Figure 3d shows the total electron density. There is a uniformsurface with nuclear repulsion energy of 1,178 MeV with astoichiometry C13N13Ti28 (2). Some anomalies are observedin the surface, bond generation and simulation ideality pro-duces a charge unbalance in these atomic sites, generatinghigh electronic densities and polarization in the structure.

Ball on Disc Test

In Fig. 4 and Table III, the Ball on Disk (BOD) test resultsare shown for AISI 304SS, TiN, TiN and TiN/TiC. The steelpresent a high friction due to oxidation process on the contactarea in this test, the TiC monolayer have the same tendencybecause as can be observed, the failure occurred around to8 meters of test due to poor adhesion of TiC on steel substrate,the friction coefficient is similar to the 304 steel, and the highwear coefficient is attributed at the particle product of coat-ing failure acting like abrasion particles on the steel surface,increasing the wear process. The TiN monolayer showed amore stable friction coefficient compared with TiC layer dueto better adhesion, the coating failure was presented around

FIGURE 3. (a) Simulated system with 50% scale atomic radii for the TiN/TiC bilayers. (b) Obtained results with a 50% atomic radii for theTiN/TiC bilayers. (c) Mulliken charge distribution for the TiN/TiC bilayers. (d) Electron total density for the TiN/TiC bilayers.

Rev. Mex. Fis. S59 (1) (2013) 90–94

INTERFACE DESCRIPTION USING COMPUTATIONAL METHODS AND TRIBOLOGICAL CHARACTERISTIC OF TiN/TiC. . . 93

FIGURE 4. Friction Coefficient of monolayers, bilayer and sub-strate tested by Ball on Disc Test. Wear Coefficient=Worn Vol-ume/Load*Distance

TABLE III. Coatings and susbtrate wear coefficient

Material Wear Coefficient (mm3/Nm)

304SS 4.93×10−8

304SS/TiC 3.81×10−8

304SS/TiN 6.11×10−8

304SS/TiN/TiC 1.82×10−7

25 meters and the wear coefficient decreases, in this coat-ing the wear process initially is dominated by asperity pol-ishing surface, after a short period of adhesion of aluminaball is present due to high hardness of TiN coating, fi-nally the ploughing process is presented until coating col-lapse. The TiC/TiN bilayer presented the best performanceon BOD tests. TiN assures adhesion with the substrate, fric-tion coefficient is stable even the coating failure is presented(43 meters), this evidence the formation of interface betweenthese materials corroborating the XRD, XPS and DFT re-sults where a TiCN interface is formed. The bilayer presentsthe same wear mechanism that the observed on TiN mono-layer but crack growing and collapse required more time dueto chemical stability and high mechanical properties of tita-nium carbide and titanium carbonitride, producing low fric-tion and wear coefficient [23], compared with the monolayersand substrate (Table III).

4. Conclusions

TiC, TiCN, C-N and Ti-N elements and their correspond-ing binding energies were identified through XPS. It’s ob-served in XRD patterns the orientation of crystallographicplanes for the FCC phase of the compounds previously ana-lyzed by XPS, TiN, TiC, TICN, in the directions (111), (200),(220) and (311). The simulation showed that the bilayer ischemically stable, the density on the interface increases dueto bonds generation, atomic mobility and the formation ofTiCN compound. The system is electrically neutral accord-ing to the Mulliken population analysis, except at the cornerstructure where there are phenomena of electronic exchange.The total electron density is continuous which proves the sta-bility of the system. The simulation results have been sup-ported by the experimental analysis performed on the bilayerof TiN/TiC. The ball on disc test corroborated the interfaceformation between TiC and TiN and improved the perfor-mance compared with a TiN and TiC monolayer. TiN assuresgood adhesion and TiC low wear and friction coefficient.

Acknowledgments

The authors want to tanks Alfonso Devia Cubillos Ph.D (inmemoriam), founder of Laboratorio de Fısica del Plasmafrom Universidad Nacional de Colombia who was our teacherand friend.

Nomenclature

TiN Titanium Nitride

TiC Titanium Carbide

TiCN Titanium Carbonitride

PAPVD Plasma Assisted Physical

Vapor Deposition Technique

XPS X-Ray Photoelectron Spectroscopy

XRD X-Ray Diffraction

FCC Face Centered Cubic

DFT Density Functional Theory

BOD Ball on Disk

Pw Working Pressure

Ts Substrate Temperature

1. L. Major et al., Surface and Coatings Technology.200 (2006)6190-6195.

2. W.H. Kao, Y.L. Su, and S.H. Yao,Vacuum80 (2006) 604-614.

3. A. Devia, Y.C. Arango, and D.M. Devia,Surf. Coat. Technol.201(2006) 2919-2924.

4. Y. C. Zhu, K. Yukimura, C. X. Ding, and P. Y. Zhang,Thin

Solid Films388(2001) 277-282.

5. T. Liskiewicz, S. Fouvry, and B. Wendler,Wear 259 (2005)835-841.

6. N. Liu, W. Yin, and L. Zhu,Materials Science and EngineeringA 445-446(2007) 707-716.

7. I.W.M. Brown and W.R. Owers,Current Applied Physics4

Rev. Mex. Fis. S59 (1) (2013) 90–94

94 D. M. DEVIA NARVAEZ, J. M. GONZALEZ CARMONA, AND A. RUDEN MUNOZ

(2004) 171-174.

8. T.A. Rawdanowicz, V. Godbole, J. Narayan, J. Sankar, and A.Sharma,“The hardnesses and elastic moduli of pulsed laser de-posited multilayer AlN/TiN thin films”Composites: Part B30(1999) 657-665.

9. A. A. Voevodin, M. A. Capano, S. J. P. Laube, M. S. Donley,and J. S. Zabinski,Thin Solid Films298(1997) 107-15.

10. A, Devia, E. Restrepo, B. Segura, Y. Arango, and D.F. Arias,Surf. Coat. Technol.190(2005) 83-89.

11. M. J. Frischet al., Gaussian 98(Revision A.11.2, 2001).

12. D. Briggs and M.P. Seah,John WILLEY & SONS. Vol. 1, (sec-ond edition 1993).

13. D.R. Chopra, G. C. Smith, and S. Kunar,J. Vac. Sci. Technol.B. 10 (1992) 1218-1220.

14. A. Fahlman, C. Nordling, G. Johansson, and K. Hamrin,J.Phys. Chem. Solids30 (1969) 1835-1847.

15. T.J. Moravec, and T.W. Orent,J. Vac. Sci. Technol.18 (1981)226-228.

16. T. Yamasaki, Y.J. Zheng, Y. Ogino, M. Terasawa, T.Mitamura,and T. Fukami,Materials Science and Engineering A350(2003) 168-172.

17. A. Devia, V. Benavides, E. Restrepo, D.F. Arias, and R. Ospina,Vacuum81 (2006) 378-384.

18. ICSD data base, card 064904.

19. ICSD data base, card 043369.

20. J. M. Hong, Y. R. Cho, D. J. Kim, J. M. Baek, and K. H. Lee,Surf. Coat. Tech.131(2000) 547-551.

21. V. Benavides, E. Restrepo, and A. Devia,Physical status solid(c). 2 (2005) 3770-3773.

22. U. Gosele,Diffusion-Induced Defects in Silicon, Encyclopediaof advanced materials, R.J.B.D, M.C. Flemings and S. Maha-jan, (Eds., Pergamon Press, 1994) p. 629-6335.

23. G Zhang, B. Li, B. Jiang, F. Yan, and D. Chen,Applied SurfaceScience255(2009) 8788-8793.

Rev. Mex. Fis. S59 (1) (2013) 90–94