Professional Documents
Culture Documents
1 s2.0 S0301679X16303735 Main
1 s2.0 S0301679X16303735 Main
Tribology International
journal homepage: www.elsevier.com/locate/triboint
A R T I C L E I N F O A BS T RAC T
Keywords: Molecular dynamics simulations (MD) was carried out to investigate the surface nano-tribology of grown Ta
Molecular dynamics sputtered thin films on stainless steel with and without TaN interlayer. Ta and TaN films were deposited by
Roughness magnetron sputtering technique at different temperatures. Argon and nitrogen mixture gas was used to create
Sputtering TaN interlayer. Using AFM technique, the surface of sputtered thin films was characterized. MD simulation
Tantalum
results were validated by experimental results and a good consistency was observed the two techniques. The
AFM characterization
effects of TaN interlayer and process parameters were investigated. The results showed that the surface
roughness can be remarkably improved by using a TaN seed layer. Higher sputtering powers at lower
temperatures provide the best condition to produce smoother Ta films.
1. Introduction tantalum films based on physical vapor deposition (PVD). The surface
roughness of deposited films by magnetron sputtering is, therefore,
Recently tantalum based films have been extensively studied and affected by the process parameters such as glancing angle of deposition
used owing to their excellent properties such as corrosion resistance, [22], pressure and power [23], temperature [24,25] and so on. Due to
high conductivity, biocompatibility, high melting point, high ductility the limitations of experimental conditions, it is difficult to understand
and high strength at high temperatures [1–5]. They can potentially be and control simultaneously all affecting process parameters. However,
applied in proton exchange membrane fuel cells (PEMFCs), biocom- molecular dynamics (MD) simulations is an interesting method which
patible coating in vitro and vivo [5,6]. Ta thin films are currently used helps to overcome abovementioned limitations. MD simulations is a
in semiconductor industry for microelectronic devices [7–9] and powerful tool for examining the atomic scale structure of deposited
mechanically protecting coatings [10,11]. However, Ta films, which atomic layers [26] which has been used for first time in tantalum
are deposited via sputtering, have not the same properties (such as sputtering process with TaN seed layer.
surface roughness, porosity, grain size and etc.) as the bulk material In the present work, MD simulations of Ta deposited thin films
due to differences in their processing techniques and structures. were carried out onto both AISI 316L (SS) substrate and TaN seed
Surface roughness of deposited thin films is practically one of the layer under different deposition process conditions. Actual sputtered
most important factor in bio-compliant coatings on medical implants Ta thin films were also applied onto similar surfaces under the same
[12,13], friction coefficient of coating [14,15], the interfacial contact conditions, by magnetron DC sputtering technique, to validate MD
resistance (ICR) in PEMFCs system [16], adhesive strength of nano- simulations results. The effects of power, temperature and TaN seed
scale contacts [17] and components for water wetting properties [18]. layer were studied on the roughness of the resulted sputtered thin
Furthermore, as reported by Maeng [19] and Evgeny [20], the increase films. In order to identify the affecting parameters on the roughness
of surface roughness of deposited films may lead to increase of variations, MD simulations were employed to calculate surface atom
corrosion rate which significantly affects both protein adsorption and diffusion rate and interfacial stress under various conditions.
cellular response in biomedical applications [12]. Indeed, reducing
surface roughness is generally interested in the fabrication of thin films 2. Simulation method
[21].
Magnetron sputtering is a well-established process to produce Molecular dynamics simulation was carried out to investigate the
⁎
Corresponding author.
http://dx.doi.org/10.1016/j.triboint.2016.10.010
Received 29 August 2016; Received in revised form 7 October 2016; Accepted 8 October 2016
Available online 11 October 2016
0301-679X/ © 2016 Elsevier Ltd. All rights reserved.
M. Nikravesh et al. Tribology International 105 (2017) 185–192
effect of tantalum nitride seed layer on growth mechanism of tantalum effects of processing parameters.
thin film on the AISI 316L stainless steel (SS) substrate under different After the simulation process completed, the mean surface rough-
sputtering powers and substrate temperatures. Fig. 1 shows a three- ness of the samples, RMS , were calculated using Eq. (6) [32,33].
dimensional model of Ta atoms depositing onto a substrate with the N
dimensions of 140*20*20 Å. The motions of all mobile adatoms and ∑i =1 (Zi − Z )2
RMS =
substrate atoms are governed by the Newton's second law. They are N (6)
determined by the direct integration of the classical Hamiltonian
where the index i runs over all atoms along the X (or Y) direction. N is
equations of motion using Velocity-Verlet method. The deposition
the total number of atoms, while Z is the mean surface position in the
process is simulated using constant micro-canonical ensemble (NVE)
surface normal, z, direction.
with constant number of particles (N), the system's volume (V) and the
The surface diffusion rate of adatoms and interfacial stress between
total energy in the system (E). Each depositing tantalum atom enters
sputtered films and substrate were calculated. The average horizontal
the simulating box from a random points, x and y coordinates, and it
displacement of Ta adatoms was the base of measuring the interface
will rest at a point on surface with minimum energy.
diffusion rate.
In MD simulation, the Morse potential is applied for expressing the
The stresses were defined by a computation that computes the
interaction between both substrate atoms and adatoms. The Morse
symmetric per-atom stress tensor for each atom in a group. The tensor
potential is written as [27]:
for each atom has 6 components and is stored as a 6-element vector in
U (rij ) = D[e−2α(rij − r0)−2e−α(rij − r0)] (1) the following order: xx, yy, zz, xy, xz, yz. A virial contribution produced
by a small set of atoms is assigned in equal portions to each atom in the
Where U is a pair potential energy function, rij is the distance between set [34].
atoms i and j, r0 is the nearest atomic distance at equilibrium, D is the It is required to mention that, before the simulation process of
cohesion energy and α is a material fitted parameter that depends on sputtering, substrate (stainless steel and TaN) had been constructed,
the binding tension energy of the material and the bulk modulus. For using the information of elements, crystal structures and lattice
the different atomic species interaction, more parameters are needed in constants related to the desired structure.
the Morse function. These parameters can be calculated using the
Lorentz–Berteloth mixing rules using the following equations [28]: 3. Materials and method
DA − B = (DADB )1/2 (2)
In the present work, discs of AISI 316 L stainless steel (SS) with
1 dimensions of 30 mm in diameter and 8 mm in thickness were used as
αA − B = (αA + αB )
2 (3) substrates. The disc surfaces were polished with different grades of
polishing papers to obtain smooth mirror like surfaces. The polished
2
r0A − B = (σAσB )1/2 +ln ( ) discs were subsequently cleaned in acetone and methanol ultrasonic
αA − B (4) bath to remove any surface contamination.
2 Ta was deposited by magnetron sputtering with Ta target 99.95%.
σA, B = r0A, B−ln ( ) Argon with the purity of 99.99% was used as plasma gas and nitrogen
αA − B (5)
with the purity of 99.99% was used as reactive gas to create TaN
The Morse parameters used in this research are listed in Table 1 sputtered seed layer. Sputtering processes were applied under base
[28,29]: pressure of 5×10−6 Torr and the working pressure of 5×10−3 Torr.
Considering both computational efficiency and accuracy of results, Tantalum thin films were deposited with and without TaN interlayer
the computational time step was set 1 fs. 3600 Ta atoms were under different sputtering conditions of temperature and power as
deposited during simulation runs. The substrate temperatures were presented in Table 2:
set from 25 to 300 °C and the incident energies were set from about Intermediate sputtered films were applied on some samples, as
0.08–2 eV. shown in Table 2, by nitrogen leaks into the sputtering chamber before
The simulation processes were performed using the large-scale final Ta film deposition. After applying the interlayers, with interrup-
atomic/molecular massively parallel simulator (LAMMPS) [30] and the tion of nitrogen flow without breaking vacuum, Ta sputtering operation
post analyses were visualized by OVITO [17,31]. stage was performed.
MD simulation method can be systematically performed to analyze Morphological investigations and roughness measurements were
the growth process, morphology, surface diffusion and interfacial stress performed using atomic force microscope (AFM) VEECO and subse-
of the deposited thin films in details in order to study the simultaneous quently the prepared images were analyzed by Gwyddion software.
Fig. 1. Simulation model of the deposition process, (a) on stainless steel and (b) on TaN interlayer; (Gray: Fe in SS substrate; White: O in Cr2O3; Green: Cr in Cr2O3; Light blue: Ta
Adatoms; Red: N in TaN and Blue: Ta in TaN). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
186
M. Nikravesh et al. Tribology International 105 (2017) 185–192
Table 1
The value of Morse parameters of each pair used in simulation.
(a) (b)
0.08
(eV)
0.31
(eV)
0.70
(eV)
1.24
(eV)
1.94
(eV)
Fig. 2. The cross section of the Ta thin films deposited at room temperature and different incident energies (eV) onto (a) SS surface and (b) TaN interlayer.
187
M. Nikravesh et al. Tribology International 105 (2017) 185–192
5. Discussion
188
M. Nikravesh et al. Tribology International 105 (2017) 185–192
(higher incident energy). As indicated in Fig. 7, increasing of incident cement distance of adatoms may affect the resulted sputtered film
energy results in decrease in roughness which is well consistent with structure and its roughness. Fig. 10 illustrates the calculated horizontal
the results presented in Fig. 9. displacement distances of Ta adatoms on SS and TaN substrates by MD
Interface diffusion phenomenon which determines average displa- simulations. The horizontal displacement distance is a measure of
189
M. Nikravesh et al. Tribology International 105 (2017) 185–192
Fig. 6. (a) RMS roughness values and (b) Grain size of deposited Ta films with different process parameters on stainless steel substrate (black bar) and TaN interlayer (red bar),
extracted from AFM images. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 7. Normalized roughness values versus power in the both with and without Fig. 9. Normalized interfacial stress values calculated by MD simulation of sputtered
interlayer conditions. films as function of incident energy with and without interlayer.
Fig. 8. Normalized roughness values versus temperature in the both with and without Fig. 10. Ta adatom horizontal displacement on surface during deposition process in MD
interlayer conditions. simulation as a function of incident energy with and without interlayer.
interface diffusion which depends on temperature, incident energy and to decrease in surface roughness. It is important to note that, the effect
substrate surface condition. As seen in Fig. 10, horizontal displacement of TaN interlayer on increasing of adatom average displacement is
(interface diffusion) increases with increase of incident energy leading evident. It leads to higher surface diffusion and faceted (2D) growth
190
M. Nikravesh et al. Tribology International 105 (2017) 185–192
Acknowledgment
References
191
M. Nikravesh et al. Tribology International 105 (2017) 185–192
pressure and power on composition, surface roughness, microstructure and prepared by reactive pulsed dc magnetron sputtering, J Alloy Compd 509 (2011)
magnetic properties of as-deposited Co2FeSi thin films, Thin Solid Films 558 6400–6407.
(2014) 349–355. [41] I. Petrov, P.B. Barna, L. Hultman, J.E. Greene, Microstructural evolution during
[24] A. Anders, A structure zone diagram including plasma-based deposition and ion film growth, J Vac Sci Technol A: Vac Surf Films 21 (2003) S117.
etching, Thin Solid Films 518 (2010) 4087–4090. [42] J.F. Yang, Z.G. Yuan, X.P. Wang, Q.F. Fang, Characterization of WTaN hard films
[25] L.-C. Chang, Y.-I. Chen, H.-L. Kao, Annealing of sputter deposited nanocrystalline synthesized by direct current magnetron sputtering, Surf Coat Technol 231 (2013)
Cr–Ta coatings in a low oxygen containing atmosphere, Thin Solid Films 520 19–23.
(2012) 6929–6934. [43] A.A. Navid, A.M. Hodge, Nanostructured alpha and beta tantalum formation—
[26] D.B. Graves, P. Brault, Molecular dynamics for low temperature plasma-surface Relationship between plasma parameters and microstructure, Mater Sci Eng A 536
interaction studies, J Phys D: Appl Phys 42 (2009). (2012) 49–56.
[27] B.-H. Chen, Chen Co-K, S.C. Chang, Thin film deposition on a corrugated surface: a [44] M. Kamiko, K. Aotani, R. Suenaga, J. Koo, W. Ko, K. Nose, K. Kyuno, et al., The
molecular dynamics approach, Nucl Instrum Methods Phys Res Sect B: Beam Influence of Ta underlayers on the structure of TiO2 thin films deposited on an
Interact Mater At 260 (2007) 517–524. unheated glass substrate, Appl Surf Sci 258 (2012) 8764–8768.
[28] S.A. Roncancio, D.F. Arias-Mateus, M.M. Gómez-Hermida, J.C. Riaño-Rojas, [45] L. Gladczuk, A. Patel, J.D. Demaree, M. Sosnowski, Sputter deposition of bcc
E. Restrepo-Parra, Molecular dynamics simulations of the temperature effect in the tantalum films with TaN underlayers for protection of steel, Thin Solid Films 476
hardness on Cr and CrN films, Appl Surf Sci 258 (2012) 4473–4477. (2005) 295–302.
[29] S.S. Firouzabadi, K. Dehghani, M. Naderi, F. Mahboubi, Numerical investigation of [46] M. Backholm, M. Foss, K. Nordlund, Roughness scaling in titanium thin films: a
sputtering power effect on nano-tribological properties of tantalum-nitride film three-dimensional molecular dynamics study of rotational and static glancing angle
using molecular dynamics simulation, Appl Surf Sci 367 (2016) 197–204. deposition, Appl Surf Sci 268 (2013) 270–273.
[30] S. Venkataraj, H. Kittur, R. Drese, M. Wuttig, Multi technique characterization of [47] F.C. Nascimento, C.E. Foerster, S.L.R. Silva, C.M. Lepienski, C.J.M. Siqueira,
tantalum oxynitride films prepared by reactive direct current magnetron sputtering, C.A. Junior, A comparative study of mechanical and tribological properties of AISI
Thin Solid Films 514 (2006) 1–9. 304 and AISI 316 submitted to glow discharge nitriding, Mater Res 12 (2009)
[31] A. Stukowski, Visualization and analysis of atomistic simulation data with OVITO– 173–180.
the open visualization tool, Model Simul Mater Sci Eng 18 (2010) 015012. [48] X. Chen, Y.-W. Wang, X. Liu, X.-Y. Wang, X.-B. Wang, S.-D. An, et al., Molecular
[32] M.A. Karolewski, Molecular dynamics simulations of the initial stages of sputter dynamics study of the effect of titanium ion energy on surface structure during the
erosion of a metal overlayer system: 2 keV Ar→Cu/Ni(100), Nucl Instrum Methods amorphous TiO2 films deposition, Appl Surf Sci 345 (2015) 162–168.
Phys Res Sect B: Beam Interact Mater At 243 (2006) 6–15. [49] Z.-H. Hong, S.-F. Hwang, T.-H. Fang, Atomic-level stress calculation and surface
[33] L. Zhang, C. Xie, J. Wu, Effect of annealing temperature on surface morphology and roughness of film deposition process using molecular dynamics simulation, Comput
mechanical properties of sputter-deposited Ti–Ni thin films, J Alloy Compd 427 Mater Sci 48 (2010) 520–528.
(2007) 238–243. [50] X.-b. Jing, Z.-l. Liu, K.-l. Yao, Molecular dynamics investigation of deposition and
[34] LaboratoriesSN. LAMMPS Users Manual: Sandia Corporation; 2013. annealing behaviors of Cu atoms onto Cu(001) substrate, Appl Surf Sci 258 (2012)
[35] B. Wu, J. Zhou, C. Xue, H. Liu, Molecular dynamics simulation of the deposition 2771–2777.
and annealing of NiAl film on Ni substrate, Appl Surf Sci 355 (2015) 1145–1152. [51] Y. Jin, W. Wu, L. Li, J. Chen, J. Zhang, Y. Zuo, et al., Effect of sputtering power on
[36] D.A. Jaeger, Interface investigations on titanium nitride bilayer systems, Federal surface topography of dc magnetron sputtered Ti thin films observed by AFM, Appl
Polytechnic of Lausanne, 2012. Surf Sci 255 (2009) 4673–4679.
[37] J.-P. Palmquist, Carbide and MAX-phase engineering by thin film synthesis, [52] N. Muslim, Y.W. Soon, C.M. Lim, N.Y. Voo, Properties of nickel films growth by
Uppsala University, 2004. radio frequency magnetron sputtering at elevated substrate temperatures, Thin
[38] A. Khatibi, Growth and heat treatment studies of Al-Cr-O and Al-Cr-O-N thin films, Solid Films 612 (2016) 82–86.
Linköping University, LiU-Tryck, Linköping, Sweden, 2013. [53] A.R. Rafieerad, A.R. Bushroa, B. Nasiri-Tabrizi, J. Vadivelu, S. Baradaran,
[39] V. Chawla, R. Jayaganthan, R. Chandra, A study of structural and mechanical M. Mesbah, et al., Mechanical properties, corrosion behavior and in-vitro bioac-
properties of sputter deposited nanocomposite Ti–Si–N thin films, Surf Coat tivity of nanostructured Pd/PdO coating on Ti–6Al–7Nb implant, Mater Des 103
Technol 204 (2010) 1582–1589. (2016) 10–24.
[40] T. Elangovan, S. Murugeshan, D. Mangalaraj, P. Kuppusami, S. Khan, C. Sudha, [54] A. Ghasemi, The role of annealing temperature on the structural and magnetic, J
et al., Synthesis and high temperature XRD studies of tantalum nitride thin films Magn Magn Mater 403 (2016) 127–132.
192