www.elsevier.com/locate/apsusc

Applied Surface Science 253 (2007) 7019–7023

Growth of TiN films at low temperature

L.I. Wei, Chen Jun-Fang *

School of Physics and Telecommunication Engineering, South China Normal University, Guangzhou 510631, China

Received 13 December 2006; received in revised form 7 February 2007; accepted 8 February 2007

Available online 15 February 2007

Abstract

Thermodynamic analysis on growth of TiN films was given. The driving force for deposition of TiN is dependent on original Ti(g)/N(g) ratio

and original partial pressure of N(g). TiN films were deposited by ion beam assisted electron beam evaporation system under suitable nitrogen gas

flow rate at 523 K while the density of plasma varied with diverse discharge pressure had been investigated by the Langmuir probe. TiN films were

characterized by means of Fourier transform infrared absorption spectrum (FTIR), X-ray diffraction (XRD) and observed by means of atom force

microscopy (AFM). The results of these measurements indicated preferential TiN(1 1 1) films were deposited on substrate of Si(1 0 0) and glass by

ion beam assisted electron beam evaporation system at low temperature, and it was possible for the deposition of TiN films with a preferential

orientation or more orientations if the nitrogen gas flow rate increased enough. Sand Box was used to characterize the fractal dimension of surface

of TiN films. The results showed the fractal dimension was a little more than 1.7, which accorded with the model of diffusion limited aggregation

(DLA), and the fractal dimension of TiN films increased with increase of the temperature of deposition.

# 2007 Elsevier B.V. All rights reserved.

PACS : 47.54.Jk; 68.65.�k

Keywords: Titanium nitride (TiN); Driving force; Fractal dimension; Temperature of deposition

1. Introduction

Titanium nitride (TiN) is a promising material for many

excellent properties, such as its high hardness, high melting

temperature, high thermal and electrical conductivity, and good

resistance to corrosion, abrasion, and oxidation. These

properties of TiN are extremely important in a wide range of

technological applications [1–3].

TiN films have been studied intensively for some years. The

information on TiN films deposited under high temperature is

available in many literatures [4–8]. Pihosh et al. got TiN films by

rf magnetron sputtering below 1073 K [9]. Gerlach et al. obtained

TiN films by reactive evaporation at 1023 K [10]. Kodanbaka

et al. got TiN films by dc magnetron sputtering at 1030 K [11].

Patsalas et al. obtained TiN films by reactive magnetron

sputtering [12]. But there are few reports on TiN films deposited

under temperaturewhich was lower than 673 K, and there are few

reports on thermodynamic analysis on growth of TiN films, and

TiN films are scarcely deposited by electron beam evaporation

system.

* Corresponding author.

E-mail address: tolwwt@163.com (C. Jun-Fang).

0169-4332/$ – see front matter # 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.apsusc.2007.02.028

Since Mandelbrot brought forward the theory of fractal [13],

it was known that the fractal dimension could be used to

describe complex coarse surface. Meakino believed that the

surface of film far from equilibrium was a kind of self-similar

fractal structure and could be described with the fractal

dimension [14].

2. Thermodynamic analysis on growth of TiN films

In our system, we divided the process of growth of TiN films

into following stages:

(1) M

ixed gas and excited particles were delivered to the

deposition area;

(2) R

eactant molecules that remained gas phase were diffused

onto substrate surfaces;

(3) C

hemical reactions occurred between absorbed molecules

or between absorbed molecules and gas molecules, at the

same time desired deposition particles migrated on

substrate surface and combined into crystal lattice.

From the phase diagram of N–Ti system, we know there can

be two reaction as following near the vapor–solid interface when

L.I. Wei, C. Jun-Fang / Applied Surface Science 253 (2007) 7019–70237020

temperature is lower than 3223 K.

TiðgÞ þ NðgÞ ! TiNðsÞ (1)

2TiðgÞ þ NðgÞ ! Ti2NðsÞ (2)

Here, g and s denote gas phase and solid phase.

The region of Ti2N in the phase diagram of N–Ti system is

much less than that of the region of TiN. In order to simplify the

model, we assume that reaction (1) is the dominant reaction to

be discussed in this study. The equilibrium constant expression

of reaction (1) is:

KP ¼1

PTiPN

(3)

The following limitation can be obtained from reaction (1).

P0N � PN ¼ P0

Ti � PTi (4)

Here, P0N and P0

Ti denote original partial pressure of

corresponding element respectively, PN and PTi denote

equilibrium partial pressure of corresponding element respec-

tively. The driving force of reaction (1) can be obtained from

expression (3) and (4):

DP ¼ P0N � PN ¼ P0

N �P0

Nð1� xÞ þffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiP0

Nð1� x2Þ þ 4=KP

p2

¼ P0N

2

�1þ x�

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffið1� xÞ2 þ 4

KPP02N

s �(5)

Here, x ¼ P0Ti=P0

N, denotes original Ti(g)/N(g) ratio. The

equilibrium constant of reaction (1) is the following expres-

sion:

DG0 ¼ �RT ln KP (6)

We can obtained DG0 from the following expression

[15,16]:

DG0 ¼ Aþ BT ln T þ CT þ DT2 (7)

Here, A = �83,750, B = �11.91, C = 123.7, D = �4.7 � 10�4.

Fig. 1 shows the driving force for deposition of TiN, DP,

which is as a function of original Ti(g)/N(g) ratio for several

original partial pressure of N(g) at 523 K. The driving force

Fig. 1. The driving force for deposition of TiN, DP, which is as a function of

original Ti(g)/N(g) ratio for several original partial pressure of N(g) at 523 K.

depends on original Ti(g)/N(g) ratio and original partial

pressure of N(g) which is influenced by a nitrogen gas flow rate

indirectly in our system. DP decrease with the decrease

of P0N, and we can see that the value of DP will be very low if the

value of P0N is too low. So, TiN films can’t be deposited easily

unless the value of P0N is high enough. As shown in Fig. 1, when

P0N remains constant, DP will increase with the increase of

original Ti(g)/N(g) ratio up to some value, then it becomes

constant too.

Fig. 2. The FTIR spectra of deposited TiN films on glass under diverse nitrogen

gas flow rate. (a) 1#: 31.1; (b) 2#: 37.7; (c) 3#: 41.2.

Fig. 4. The XRD spectra of deposited TiN films under diverse nitrogen gas flow

rate.

L.I. Wei, C. Jun-Fang / Applied Surface Science 253 (2007) 7019–7023 7021

3. Experimental

TiN films were deposited by means of electron beam

evaporation of a Ti target (99.99% pure) in the atmosphere of N

2(99.99% pure) which was assisted with Ar ion beam from an

ion source. Glass and Si(1 0 0) were chosen as substrate. For

preparing the deposition of TiN films, the axial and radius

distribution of plasma density had been investigated by the

Langmuir probe, which had a magnitude of 108 cm�3. The

density of plasma varied with diverse discharge pressure had

also been investigated. Therefore, appropriate plasma region

(radius located between 50 and 150 mm), and suitable

discharge pressure (less than 0.1 Pa) were obtained. The

pressure of the chamber must be less than 0.1 Pa so that the E

type gun can work. The deposition of TiN films on glass was

performed for 240 s under a pressure of 0.09 Pa in reaction

chamber when the nitrogen gas flow rate was 31.1, 37.7 and

41.2 sccm, samples of which are named from 1# to 3#. The

deposition of TiN films on Si(1 0 0) was performed for 240 s

under a pressure of 0.09 Pa in reaction chamber when the

nitrogen gas flow rate was 15.1, 31.1, 37.7 and 41.2 sccm,

samples of which are named from 4# to 7#, the temperature of

substrate during deposition was 523 K. The deposition of TiN

films on Si(1 0 0) was performed for 240 s under a pressure of

0.09 Pa in reaction chamber when the nitrogen gas flow rate

was 30.2 sccm, samples of which are named from 8# to 10#, the

temperature of substrate during deposition was 300, 400 and

523 K. The electron beam of 250 mA was chosen.

4. Results and discussion

The Fourier transform infrared absorption spectrum

(FTIR) and atom force microscopy (AFM) were measured

on samples 1# to 3# to determined related structural

characteristics.

Fig. 2 shows the FTIR spectra of samples 1# to 3#. All of

them are similar. The sharp absorption peak near 480 cm�1 is

assigned to vibration level of Ti–N bond, while the absorption

peak near 1060 cm�1 is assigned to vibration level of Si–O

Fig. 3. The XRD spectra of deposited TiN films on Si(1 0 0) under the nitrogen

gas flow rate of 37.7 sccm.

bond in glass, and the absorption peak near 2800–3000 cm�1 is

also caused by some components of glass. With increase of

nitrogen gas flow rate, the absorption peak near 480 cm�1

becomes stronger, which denotes more Ti–N bond in films.

Fig. 5. The AFM images of the surface of deposited TiN films on different

substrates. (a) on Si(1 0 0); (b) on glass.

Fig. 6. The AFM images of the surface of TiN films at different temperature of

deposition. (a) 300 K; (b) 400 K; (c) 523 K.

L.I. Wei, C. Jun-Fang / Applied Surface Science 253 (2007) 7019–70237022

XRD and AFM were measured to determined related

characteristics of TiN films (Fig. 3).

Fig. 3 shows the XRD spectra of deposited TiN films on

Si(1 0 0) under the nitrogen gas flow rate of 37.7 sccm. The

Fig. 7. The double logarithmic fractal dimension curve of TiN films at different

temperature of deposition. (a) 300 K; (b) 400 K; (c) 523 K.

L.I. Wei, C. Jun-Fang / Applied Surface Science 253 (2007) 7019–7023 7023

strongest peak is assigned to TiN(1 1 1), and other weak peaks

correspond to diffractions from TiN(2 0 0) and TiN(2 2 2). So,

a preferential orientation of TiN(1 1 1) can be obtained on

Si(1 0 0) substrate under low temperature while the velocity of

deposition is high and the even plasma density of deposition

region is high. This result validates our above thermodynamic

analysis.

Fig. 4 shows the XRD spectra of deposited TiN films on

Si(1 0 0) under diverse nitrogen gas flow rate of 15.1, 31.1, 37.7

and 41.2 sccm. There is no strong peak of TiN under the

nitrogen gas flow rate of 15.1 sccm. There is a strong peak of

TiN(111) under the nitrogen gas flow rate of 31.1, 37.7 and

41.2 sccm. There are two weak peaks of TiN(2 0 0) and

TiN(2 2 2) under the nitrogen gas flow rate of 37.7 and

41.2 sccm. We believe that the nitrogen gas flow rate directly

influence the deposition of TiN films. While the nitrogen gas

flow rate is too low, the molecular free stroke increase and the

concentration of reactant N ion around substrate is too low,

which make it is difficult for the deposition of TiN films with a

preferential orientation. While the nitrogen gas flow rate

increase, there are more chances for collision of N ion and Ti

ion so that N ion and Ti ion change energy continually. So, N

ion and Ti ion combine fully on the substrate, and it is possible

for the deposition of TiN films with a preferential orientation or

more orientations.

Fig. 5 shows two AFM photographs of the surface of

deposited TiN films on Si(1 0 0) and glass under diverse

nitrogen gas flow rate of 37.7 sccm.. The average grain size of

TiN films is about 20 nm. The grains of TiN films on Si(1 0 0)

grow more regularly and densely than on glass.

There are various ways of calculation of the fractal

dimension such as Box-counting, Power Spectra Density and

Sand Box. We used Sand Box to calculate the fractal dimension

of the image of AFM.

Fig. 6 shows AFM photographs of the surface of samples 8#,

9# and 10#. According to Sand Box, The double logarithmic

fractal dimension curve of TiN films at different temperature of

deposition is shown in Fig. 7. The fractal dimension of the

surface of samples 8#, 9# and 10# are 1.746, 1.749 and 1.757,

respectively with increase of the temperature of deposition,

which is close to 1.7 in the model of diffusion limited

aggregation (DLA) [17]. This result indicates the growth of TiN

films accorded with the model of DLA. The reason why the

fractal dimension of the surface of samples 8#, 9# and 10# are a

little more than 1.7 is believed that out branch of aggregation

become denser since the ability of inter-diffusion of particles

with increase of the temperature of deposition.

5. Conclusions

Thermodynamic analysis on growth of TiN films at 523 K

was carried out in our system. The driving force for deposition

of TiN was dependent on original Ti(g)/N(g) ratio and original

partial pressure of N(g). TiN films can’t be deposited easily

unless the value of P0N is high enough. With the help of analysis,

TiN(1 1 1) films were deposited on substrate of Si(1 0 0) and

glass by means of ion beam assisted electron beam evaporation

at 523 K while the density of plasma varied with diverse

discharge pressure had been investigated by the Langmuir

probe. The results of these measurements indicated preferential

TiN(1 1 1) films were deposited on substrate of Si(1 0 0) and

glass by ion beam assisted electron beam evaporation system at

523 K, and it was possible for the deposition of TiN films with a

preferential orientation or more orientations if the nitrogen gas

flow rate increased enough. Average grain size of TiN films is

about 20 nm. Sand Box was used to characterized the fractal

dimension of surface of TiN films. The results showed the

fractal dimension of surface of TiN films was a little more than

1.7, which accorded with the model of DLA, and the fractal

dimension increased with increase of the temperature of

deposition.

References

[1] W.J. Meng, G.L. Eesley, Thin Solid Films 271 (1995) 108.

[2] D.E. Wolfe, J. Singh, J. Mater. Sci. 34 (1999) 2997.

[3] T.-H. Fang, S.-R. Jian, D.-S. Chuu, Appl. Surf. Sci. 228 (2004) 365.

[4] T. Matsue, T. Hanabusab, Y. Ikeuchia, Vacuum 66 (2002) 435.

[5] M. Hua, H.Y. Ma, J. Li, C.K. Mok, Surf. Coat. Technol. 200 (2006) 3612.

[6] D. Devia, R. Ospina, V. Benavides, E. Restrepo, A. Devia, Vacuum 78

(2005) 67.

[7] H.H. Huang, Hon M.H., Thin Solid Films 416 (2002) 54.

[8] L. Combadiere, J. Machet, Surf. Coat. Technol. 88 (1996) 17.

[9] Y. Pihosh, M. Goto, A. Kasahara, T. Oishi, M. Tosa, Appl. Surf. Sci. 244

(2005) 244.

[10] J.W. Gerlach, T. Hoche, F. Frost, B. Rauschenbach, Thin Solid Films 459

(2004) 13.

[11] S. Kodanbaka, S.V. Khare, V. Petrora, A. Vailionis, I. Petrov, J.E. Greene,

Vacuum 74 (2004) 345.

[12] P. Patsalas, S. Logothetidis, Surf. Coat. Technol. 180 (2004) 421.

[13] B.B. Mandelbrot, Fractal: Form Chance and Dimension, Freeman, San

Francisco, 1977, pp. 1–56.

[14] M.P. Fractal, Scaling and Growth Far from Equilibrium, Cambridge

University Press, Cambridge, 1998, pp. 10–181.

[15] X. Zhuyao, Lin. Li, Thermodynamics of Material, Science Press, Beijing,

2005, pp. 27–31.

[16] X. Zhihong, W. Leshan, Database of Inorganic Thermochemistry, Science

Press, Beijing, 1987, pp. 227–281.

[17] Z. Jizhong, Fractal, Tsinghua University Press, Beijing, 1995, pp. 166–219.