Deposition of TiO2 Nanoparticles by Means of Hollow Cathode Plasma Jet in DC Regime

R. Perekrestov, P. Kudrna, M. Tichý Charles University Prague, Faculty of Mathematics and Physics, Prague, Czech Republic.

Abstract. TiO2 nanoparticles have been investigated in this work. Nanoparticles were obtained in Ar plasma on monocrystaline Si(111) substrate by means of gas-phase deposition using hollow cathode plasma jet (HCPJ). Oxygen was introduced separately from argon through inlet in a chamber. The material of a cathode is pure titanium. The explanation of nanoparticles growth mechanism is given. Morphology of thin films surface was investigated by the atomic force microscope (AFM). A chemical composition of the thin films was investigated by means of energy-dispersive x-ray analysis (EDX).

Introduction Synthesis of nanoparticles is a basis of nanotechnology nowadays. In a general case particles with a size

from 1 to 100 nm are called nanoparticles. The application of nanoparticles is rapidly growing in nanoscale science and engineering. Such properties of nanoparticles as surface to volume ratio, quantum confinement and surface atom arrangement are of a great importance. It means that by controlling the size of nanoparticle we can obtain new properties of material.

During the last years a lot of techniques for nanoparticles synthesis have been developed. The most widely spread techniques of gas-phase nanoparticle synthesis using evaporation of solid material are: pulsed laser ablation, spark discharge generation, inert gas condensation, ion sputtering. The means of gas-phase methods are based on achieving supersaturation necessary to start the nucleation of nanoparticles [Mark et al., 2003], [Hosokawa et al., 2007].

This article deals with ion sputtering deposition of TiO2 nanoparticles by means of hollow cathode plasma jet. TiO2 nanoparticles find their application in different fields of technology mostly because of their photocatalytic properties. They are used in environmental applications for photocatalytic treatment of wastewater and pesticide degradation. Nanoparticles of TiO2 dissociate the water molecule to produce hydrogen. In industry it is used for dye producing [Gupta et al., 2011]. As for crystalline structure of TiO2 it has three polymorphs: anatase (tetragonal), rutile (tetragonal) and brookite (orthorhombic). Experiments have shown that using TiO2 as photocatalyst is more effective in a form of nanoparticles than bulk material [Gupta et al., 2011].

The technique of HCPJ is based on sputtering of the inner part of the titanium cathode by means of glow discharge [Kudrna et al., 2010]. In comparison with more commonly used magnetron sputtering it has some significant advantages in producing of nanoparticles. A velocity of nanoparticles can be controlled by the flow rate of working gas and subsequently we can control the sizes of nanoparticles by means of adjustment of the flow [Sakuma et al., 2006]. The space inside of hollow cathode serves as a place of nanoparticles nucleation because of a heightened pressure. The last important advantage is possibility of manipulation with the position of plasma jet under vacuum.

Experimental setup The scheme of HCPJ is shown in Fig. 1 [Kudrna et al., 2010]. Sputtering element (cathode) has cylindrical

shape with a hole inside through which the gas is flowing. It is powered by negative DC power supply MDX 500. The upper limit of current and voltage of the power supply is 500 mA and 1200 V respectively. The titanium nozzle is cooled by the water flow via the copper blocks surrounding it. The covering made from the Lava ceramic is used to isolate the cathode from the rest part of the system. Ultra high vacuum chamber is pumped by the turbo-molecular pump down to ultimate pressure of 10–5 Pa. Argon serves as a working gas. Its purity is 99.999 %. We use the load lock for the manipulation with samples. It consists of the small vacuum chamber connected to the main system via two gate valves and it is equipped by the moveable feedthrough with fork at the end which holds the substrate holder and sets up the position of the substrate.

We use Langmuir probe and fibre optic thermometer for plasma and neutral gas diagnostics, but we are not focused on these measurements in this work.

Experimental results A lot of parameters that we are able to control are involved in a process of deposition. The most important

among them are discharge power, pressure, working gas flow rate and cathode–substrate separation. Each of

275

WDS'14 Proceedings of Contributed Papers — Physics, 275–279, 2014. ISBN 978-80-7378-276-4 © MATFYZPRESS

PEREKRESTOV ET AL.: TiO2 NANOPARTICLES PRODUCED BY HOLLOW CATHODE PLASMA JET

them has an influence on such physical quantities as ionization level, temperature, plasma flow dynamics, etc. In this work we were concentrated on the effect of discharge current and argon flow rate. The rest part of parameters we tried to keep constant in order to reveal more apparent dependences. We deposited several samples under different discharge currents and argon flow rates on monocrystalline silicon (111) substrate to investigate how these parameters influence the thin film surface morphology.

Original idea was to use passive TiO2/Ti2O3/TiO layer on the top of the cathode as the main source of oxygen that appears each time after exposure of the cathode to the air atmosphere or introducing oxygen [Gemilli et al., 2003]. This method has a lot of disadvantages: restriction in thickness, high degree of impurities, chemical inhomogeneity. In a case if we introduce oxygen directly to hollow cathode with argon flow the discharge becomes very unstable because of intensive surface oxidation. To avoid those problems we introduced the oxygen separately from argon through the inlet in the main chamber in the amount of 0.5 sccm.

watercooling

O2

Ti nozzle

- +

DC

coppercooler

А

DC

Ar

ceramicshield

Substrate Figure 1. The scheme of vacuum system for plasma diagnostics and thin films deposition by means of hollow cathode plasma jet in DC regime.

(a)

(b)

(c)

(d)

(e)

(f)

Figure 2. AFM scans of TiO2 nanoparticle layers under different values of discharge currents and argon flow rates: (a), (b), (c) — Id = 400 mA, fAr =100, 200, 300 sccm respectively; (d), (e), (f) — Id = 100 mA, fAr = 100, 200, 300 sccm respectively. Dimensions are given in nanometers.

276

PEREKRESTOV ET AL.: TiO2 NANOPARTICLES PRODUCED BY HOLLOW CATHODE PLASMA JET

We used a relatively high pressure (100 Pa) to create a favorable condition for cluster synthesis. Similarly to conventional systems for synthesis of nanoparticles that use separate pumping of aggregation and deposition chambers we have a pressure gradient between hollow cathode inner volume and vacuum chamber. Thus hollow cathode plays a role of aggregation chamber because of heightened pressure inside. Cross section of hollow cathode being in exploitation for approximately one year shows that sputtering is not uniform and there is a region where the erosion reaches its maximum value of intensity. The position of this region is flow dependent. We assume that a volume in proximity to this region is a place from where the seeds originate. Amount of seeds is current dependent therefore nanoparticles have smaller size in a case of higher currents. Atoms of working gas take an extra energy from the relatively hot and unstable particles to let them continue their growth. It should be noted that working gas flow rate directly influences the plasma dynamics. With increase of the argon flow rate plasma torch becomes longer and obtains well defined columnar shape.

Samples deposited under argon flow rates 200 sccm and higher were in direct contact with plasma. AFM scans of the surface are given in Fig. 2. It is possible to observe that roughness of the film is gradually increasing with argon flow rate. Under higher working gas flows rates nanoparticles have a tendency to form bigger agglomerates. The samples deposited under working gas flow rate 100 sccm are out of the plasma region therefore the temperature of the substrate surface sustained at relatively low value (from 50 up to 100 °C) and plasma influence is negligible. Temperature was measured by means of fiber optic thermometer.

In a first approximation nanoparticles have spherical shape. Using this assumption we measured particle diameters to have an idea about their dimensions. We used approximately 70–100 nanoparticles for determination of the mean size. The dependencies of nanoparticle sizes and root mean square roughness (RMSR) on discharge current and argon flow rate are given in Fig. 3. From the first two graphs it is easy to see that nanoparticle sizes vary in a range from 10 nm up to 20 nm. From the Fig. 3a it is possible to observe an evident decrease of the sizes with the growth of discharge current. At the same time argon flow rate almost has no influence on the nanoparticle size. RMSR of the nanoparticle layers is gradually increasing with growth of both parameters.

There are two main mechanisms of nanoparticle growth in the gas phase — coagulation and single atom attachment [Woodruff, 2007; Eggersdorfer et al., 2013; and Schmidt-Ott, 2010]. In a case when coagulation is dominant process bigger particles “consume” smaller ones and increase their size faster than smaller particles. As a result we have a relatively wide size distribution function. In a second case nanoparticles grow uniformly. Distribution function is narrow even when original seeds have sufficient difference in sizes.

(a)

50 100 150 200 250 300 350 400

8

10

12

14

16

18

20

22

24

Parti

cle

size

[nm

]

Discharge Current [mA]

Argon flow =100 sccm Argon flow =200 sccm Argon flow =300 sccm

50 100 150 200 250 300 350 4001.0

1.5

2.0

2.5

3.0

3.5 Argon flow=100 sccm Argon flow=200 sccm Argon flow=300 sccm

RM

SR [n

m]

Discharge Current [mA]

(b)

100 150 200 250 3008

10

12

14

16

18

20

22

24 I=100 mA I=200 mA I=300 mA I=400 mA

Parti

cle

size

[nm

]

Argon flow [sccm]100 150 200 250 300

1.0

1.5

2.0

2.5

3.0

3.5

4.0

RM

SR [n

m]

Argon flow [sccm]

I=400 mA I=300 mA I=200 mA I=100 mA

Figure 3. Dependences of nanoparticle size and root mean square roughness on: (a) discharge current, (b) argon flow rate.

277

PEREKRESTOV ET AL.: TiO2 NANOPARTICLES PRODUCED BY HOLLOW CATHODE PLASMA JET

Under the high values of discharge currents and flows nanoparticles surface distribution becomes inhomogeneous. Moreover we observed presence of agglomerate clusters (100–300 nm in diameter) created in a process of coagulation of smaller particles. Most probably that coagulation occurred in a solid state, because of the grain-like morphology of cluster surface. Samples deposited under these conditions have a well defined structure. In the middle of the sample we observed round mirror-like region. The diameter of the region approximately corresponds to diameter of plasma torch (10–12 mm). Around this region is a black matte ring where agglomerate clusters are concentrated mostly. In Fig. 4 two AFM measurements that were carried out in a different place of the same sample are given. These results prove that surface distribution of agglomerate clusters is not uniform and is symmetric with respect to sample center. In a center of the sample agglomerate clusters were not found. Such results could be explained by the influence of intensive plasma flow on the surface of the substrate. Clusters can be mechanically moved from the center to periphery. In addition we observed a rapid fall of neutral gas temperature and ionization level outside plasma column. High inhomogeneity of plasma parameters strongly affects nanoparticles growth process. Flow velocity is decreasing from the plasma column axial center to the edge of plasma. It means that time nanoparticle spends to reach the substrate and growth period are different [Sakuma et al., 2013]. It causes additional inhomogeneity of the sample.

(a)

(b)

Figure 4. AFM scans of the sample deposited under Id = 520 mA and fAr = 300 sccm (a) in ≈ 5 mm from the sample center, (b) in ≈ 7 mm from the sample center. In a middle of the sample agglomerate clusters are absent. Dimensions are given in micrometers.

0 1 2 3 4 5keV

0

2

4

6

8

10 cps/eV

Ti Ti O C

Figure 5. EDX patterns of TiO2 nanoparticle layer.

278

PEREKRESTOV ET AL.: TiO2 NANOPARTICLES PRODUCED BY HOLLOW CATHODE PLASMA JET

In order to investigate quantitative composition of thin films, energy dispersive analysis was carried out. Results are given in Fig. 5. For this measurement we prepared a relatively thick sample (3.5 μm) to exclude the influence of silicon substrate on the measurement and make sure that titanium is completely oxidized. Deposition lasted 20 minutes under discharge current 520 mA. According to the data obtained from EDX patterns the ratio between oxygen (63.38 %) and titanium (32.47 %) approximately correspond to titanium dioxide chemical formula. In addition we found a presence of carbon impurities in a small amount (4.15 %). We assume that carbon is concentrated on the surface of the film mostly. Presence of carbon is caused by exposure of the sample to atmosphere after deposition. From EDX data we can make a conclusion that oxygen flow rate (0.5 sccm) is sufficient for complete oxidation of titanium.

Conclusions TiO2 nanoparticles found their application in various areas of industry and science. HCPJ gives us an

opportunity to obtain nanoparticles from the gas phase and gives us control over the process of deposition. Varying such plasma parameters as discharge current and working gas flow rate it is possible to affect the size of nanoparticles and roughness of the thin film. Under high values of discharge current and working gas flow rate nanoparticles have a tendency to agglomerate in clusters (diameter is more than 100 nm). This transition occurs approximately under argon flow rate 200 sccm and discharge current 400 mA.

A lot of questions are still opened and need to be solved. Investigation of crystalline structure of the nanoparticles are of a great importance. In our plans is also to try DC pulsed power supply.

Acknowledgments. The partial financial support by Czech Science Foundation, grants No. 202/03/H162,

P205/2011/0386 and by Charles University Grant Agency, grant No. 120510 and No. 604612 is gratefully acknowledged. The work belongs to the studies performed in frame of the CEEPUS III project AT-0063.

References Gemelli E.,Camargo N.H.A. Oxidation kinetics of commercially pure titanium, Revista Matéria, v. 12, n. 3, pp. 525–531,

2007. Gupta S. M., Tripathi M., A review of TiO2 nanoparticles. Chinese Sci Bull, 56: 1639−1657, doi: 10.1007/s11434-011-4476-

1, 2011. Hosokawa M., Nogi K., Naito M., Yokoyama T., Nanoparticle technology handbook, ISBN: 978-0-444-53122-3, 2007.

Mark T. S. Vapor-phase synthesis of nanoparticles, Current Opinion in Colloid and Interface Science 8 127–133, 2003. Maximilian L. Eggersdorfer, Sotiris E. Pratsinis, Agglomerates and aggregates of nanoparticles made in the gas phase,

Advanced Powder Technology, 25: 71–90, 2014. Sakuma H., Aoshima H., Ishii K., Size-controlled Growth of Fe Nanoparticles in Gas Flow Sputtering Process, Journal of

Magnetics 11(3), 103–107, 2006. Schmidt-Ott A., Aerosol Methods for Nanoparticle Synthesis and Characterization, in Handbook on Nanophysics, CRC

Press, New York, 2010. Woodruff D.P., Atomic clusters: From Gas Phase to Deposited, The chemical physics of solid surfaces volume 12, ISBN:

978-0-444-52756-1, 2007.

279

Kudrna P., Klusoň J., Leshkov S., Chichina M., Picková I., Hubička Z., and Tichý M., A Study of Plasma Parameters in Hollow Cathode Plasma Jet in Pulse Regime, Contrib. Plasma Phys. 50, No. 9, 886–891, 2010.