www.elsevier.com/locate/surfcoat

Surface & Coatings Technolog

Photocatalytic performance of plasma sprayed TiO2�ZnFe2O4 coatings

Yi Zeng *, Juntao Liu, Wei Wu, Chuanxian Ding

Shanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai dingxi road 1295, 200050, P.R. China

Received 12 May 2004; accepted in revised form 8 August 2005

Available online 9 September 2005

Abstract

A novel TiO2�ZnFe2O4 coating has been prepared by plasma spraying. The effects of spraying parameters and the composition of

powders on the microstructure, surface morphology and photo-absorption of plasma sprayed coatings were studied. The photocatalytic

efficiency of the as-sprayed coatings was evaluated through the photo mineralization of methylene blue. It was found that TiO2 coatings can

decompose methylene blue under the illumination of ultraviolet rays, and the degrading efficiency is improved with an increase in the content

of FeTiO3 in the coatings. However, the presence of large amount of ZnFe2O4 compound would substantially lower the photocatalytic

efficiency of the TiO2�ZnFe2O4 coatings for the unfavorable photo-excited electron–hole transfer process.

D 2005 Elsevier B.V. All rights reserved.

Keywords: Plasma spraying; Photocatalytic activity; TiO2; ZnFe2O4

1. Introduction

Heterogeneous photocatalysis, a new wastewater treat-

ment and water purification technique, has been a fast

growing research area in the past decade [1,2]. Among all

oxide semiconductors that have been applied, titanium

dioxide is the most promising one for its high stability

against photo-corrosion, favorable band-gap energy (photo-

activity) and low cost. However, for such applications

titania exhibit at least two disadvantages: poor efficiency in

the conversion of solar energy and difficulty in reclaiming

when used as powder.

There are generally two main crystal phases in TiO2

photocatalyst, anatase and rutile. It is commonly believed

that anatase is the active phase in photocatalytic reactions

[3,4]. Pure rutile normally shows weak activity in contrast

with anatase [5]. However, it has been also realized that the

band-gap of anatase (approximately 3.2 eV) means that the

electron can only be excited from the valence band (VB) to

the conduction band (CB) by the high energy UV light

irradiation with a wavelength no longer than 385 nm. This

0257-8972/$ - see front matter D 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.surfcoat.2005.08.006

* Corresponding author. Tel.: +86 21 52413108; fax: +86 21 52413107.

E-mail address: zengyi@mail.sic.ac.cn (Y. Zeng).

limits the application of sunlight as an energy source for the

photocatalysis. Recently, many studies have been devoted to

the extension of the photoresponse and improvement of the

photoactivity by ion implantation and adding the other

semiconductor such as WO3, ZnO, Al2O3, Fe3O4, etc. [6–

13]. Particularly, narrow band-gap semiconductor has been

paid more consideration. Because the overlap of different

band could promote the excitation of the valence band

electrons, ZnFe2O4, with a relatively narrow band-gap [14]

(ca. 1.9 eV), has been used as a novel photocatalyst.

Valenzuela et al. [15] have reported that ZnFe2O4 has some

photocatalytic activity. Yuan and Zhang [16] have synthe-

sized TiO2�ZnFe2O4 nanocomposite, which exhibits better

photoactivity than pure TiO2 nanomaterials.

Generally, the photocatalytic performance increases with

the increase of specific surface. Therefore, micro-powders

are often applied as photocatalysts because the specific

surface area is larger than that of the membrane. But in

practical application, micro-powder is very difficult to

reclaim after photocatalytic reaction. In order to avoid this

technical problem, a number of methods have been used to

form TiO2 films, including wet chemical processing (such as

sol–gel, screen printing) and vapour deposition techniques

(e.g. CVD, PVD), etc. But the coating formation speed and

bonding strength are very low and it is difficult to produce

y 200 (2005) 2398 – 2402

Table 1

Plasma spraying parameters

Coating (a) (b) (c) (d) (e)

Component TiO2+3 wt.% ZnFe2O4 TiO2+3 wt.% ZnFe2O4 TiO2+15 wt.% ZnFe2O4 TiO2+30 wt.% ZnFe2O4 Pure P25

Plasma gas [SLPM] 50 50 50 50 50

Arc current [A] 600 400 400 400 400

Arc voltage [V] 36 35 35 35 35

Power [kW] 21.6 14 14 14 14

Spraying distance [mm] 80 80 80 80 80

Y. Zeng et al. / Surface & Coatings Technology 200 (2005) 2398–2402 2399

large surface coatings by chemical process such as sol–gel

and CVD. In contrast with these methods, plasma spraying

technique is an economical and versatile fabrication process

to produce large surface coatings. The coatings’ thickness,

texture and bonding strength can be controlled though

spraying parameters, powders and substrate state [17], etc.

So, to investigate the effects of ZnFe2O4 additive to TiO2

when they formed coatings, a series of TiO2–ZnFe2O4

composite coatings were deposited on stainless steel by

plasma spraying technique in our experiment, and the

characters of the coatings were analyzed with SEM, X-ray

diffraction, UV–Vis–NIR spectrophotometer and photo-

catalytic activity evaluation system.

2. Experimental

2.1. Materials

In our experiment, TiO2 (Degussa , P25) and ZnFe2O4

powders were used as raw materials. The TiO2 powder

(P25), which is a commercial benchmarking photocatalyst,

has a relatively large surface area (49 m2g�1) and consists

of anatase and rutile phases in a ratio of about 3 :1. The

average sizes of the anatase and rutile elementary particles

Fig. 1. XRD patterns of as-sprayed coatings (A=anatase, R=rutile, S=spinel,

h=h-TiO2, F=FeTiO3, Z=ZnTiO3).

are 85 and 25 nm, respectively. And the average size of

ZnFe2O4 is about 30 nm. In order to investigate the effect of

ZnFe2O4 additive, the ZnFe2O4 and TiO2 nanoparticles

were mixed with ratio of ZnFe2O4 :TiO2=0.03, 0.15, 0.3.

Polyvinyl alcohol was used as a binder for the plasma

spraying TiO2�ZnFe2O4 powder. The mixed powders were

agglomerated into micro-size powders by spray-dried

technology. The substrate is a 50�20 mm stainless steel

plate.

2.2. Plasma spraying equipment

An A-2000 atmospheric plasma spraying system (Sulzer-

Metco, Switzerland) was used to deposit coatings. Argon

was used as the only plasma gas. The definite spraying

parameters are illustrated in Table 1.

2.3. Characterization techniques

The crystallite phases of the coatings were examined by

D/max 2550V X-ray Diffraction Analysis using Cu-Ka

radiation. The average crystallite sizes of the spinel phase of

ZnFe2O4, and anatase as well as rutile phases of TiO2, were

estimated from the full width at half-maximum of their most

intense diffraction peaks using Scherrer’s formula. The

surface morphologies of the coatings were examined using a

JSM-6700F Field Emission Scanning Electron Microscope.

UV–Vis absorbance spectra of the coatings were recorded

on a Cary 500 Scan UV–Vis–NIR spectrophotometer, and

BaSO4 was used as reference.

The photocatalytic efficiency of the as-sprayed coatings

was evaluated through the photo mineralization of methyl-

ene blue (MB) solution (10 mg/L). The methylene blue

concentration in the solution was determined by a Shimadzu

Table 2

The average crystallite size of the as-sprayed coatings

Coating Average crystallite size [nm]

Anatase Rutile Spinel

(a) 37 122 42

(b) 45 134 39

(c) 31 80 54

(d) 34 160 39

(e) 41 215 a

aNo phase present within the sample.

Fig. 2. The typical surface morphology of TiO2�ZnFe2O4 coatings.

Y. Zeng et al. / Surface & Coatings Technology 200 (2005) 2398–24022400

UV-1601PC UV–Vis spectrophotometer. A mercury lamp

(125 W, peak wavelength is 365 nm) was used as the

photocatalysis lamp.

3. Results and discussion

X-ray diffraction analysis can provide detailed informa-

tion on crystallite structure characteristics (i.e. crystallite

phase formation and phase transformation). Fig. 1 gives the

XRD patterns of as-sprayed coatings. From the XRD

patterns of composite coatings, the peaks of two new phases

(FeTiO3 and ZnTiO3) were observed. This shows that a

Fig. 3. UV–Vis absorption spectra of raw powder and composite coatings.

chemical reaction of ZnFe2O4+TiO2YFeTiO3+ZnTiO3

has taken place at the spraying process owing to its high

temperature. More interestingly, a new TiO2 phase called h-TiO2 [18] appears in the composite coatings. Coating (a)

and coating (b) were prepared by the same feedstock using

different spraying power. The power of (a) is greater than

that of (b). From their XRD patterns, it can be concluded

that the relative content of rutile increase obviously when

the spraying power rise. At the same time, the peak of

FeTiO3 has been sharpened. This indicates that more

ZnFe2O4 have reacted with TiO2. The average crystallite

size calculated using Scherrer_s formula was shown in Table

2. From Table 2, it can be seen that the grain size of rutile

Fig. 4. Photocatalytic performance of as-sprayed coatings.

CB

VB

recombination

CB

VB

Ti O2

hv

Hole transfer

Electron transfer

Fig. 5. A possible recombination manner of composite coatings.

Y. Zeng et al. / Surface & Coatings Technology 200 (2005) 2398–2402 2401

grows dramatically. On the contrary, anatase grain size only

has a little change. This may be due to anatase mainly

existing in unmelted particles. And the anatase grains grow

relatively slow in the unmelted particle. Furthermore, the

ZnFe2O4 component in the composite also can inhibit its

growth [16]. Unlike anatase, rutile may mainly crystallize

from melted particles. The rapid increase of the grain size of

rutile is thought to be due to an enhancement of the AYR

transformation in its grain growth [19]. ZnFe2O4 cannot

inhibit the increase of rutile grain size, because they react

with each other in the melted state. So, ZnFe2O4 may also

mainly exist in unmelted particles. There is an irregularity,

the grain size of rutile of coating (c) is less than 100 nm and

its ZnFe2O4 grain size is the biggest one among all the

composite coating. This phenomenon is not well under-

stood. One possible reason is thought to be the growth of

ZnFe2O4 retarding the growth of rutile.

Fig. 2 shows the typical surface microstructures of

TiO2�ZnFe2O4 coatings by plasma spraying. From the

comparison between picture (1) and picture (2) in Fig. 2, it

can be seen that the surface of coating (a) is formed by the

build-up of successive layers of liquid droplets that flatten

and solidify on previously solidified material. At the same

time, the surface of coating (b) is formed by solidified

droplets containing many fine particles adhered on the

surface. The difference between the surface morphologies

may be due to the different thermal histories of the particles

in the plasma flame. The particle temperature increased with

electrical power input. Therefore, more particles have

melted and formed lamellar microstructures in coating (a).

Because of the better melted status, more reaction must have

taken place between TiO2 and ZnFe2O4 in coating (a). This

is consistent with the XRD result. From picture (3) and

picture (4) in Fig. 2, it can be seen that many particles about

100 nm exist on the surface of both coatings. For coating

(a), these nanoparticles exist as conglomeration. But these

particles evenly distribute on the surface of coating (b). In

conclusion, the coatings are not very dense, contain many

holes, and many unmelted or partially melted powders lie on

the surface of lamellae. These phenomena will be of benefit

to increase the specific surface and then improve the

photocatalytic activity of the sprayed coatings. Fig. 3 gives

the UV–Vis absorption spectra of raw powder (TiO2+15

wt.% ZnFe2O4) and composite coatings. In comparison with

raw materials, the spectra of composite coatings are some-

what red shift in the range of 340–370 nm. In the visible

light region, the composite coatings have better photo-

absorption than raw materials. This could be due to the deep

blue of the composite coatings. The deep blue is caused by

the oxygen vacancy in the coatings. Moreover, the existence

of the oxygen vacancy can enhance the photoactivity of the

catalyst [20,21]. The photocatalytic performance was shown

in Fig. 4. It is evident from Fig. 4 that the spraying power

increase leads to a few enhancement of photocatalytic

performance. This could be interpreted that more resultants,

such as FeTiO3 (band-gap 2.58 eV) [22], which has good

photo-absorptive ability in the visible spectral range and

favorable photo-excited electron–hole separation characters

[11,23], have formed as the spraying power increases. The

photoactivity of composite coatings obviously drop down

with the amount increasing of ZnFe2O4 additive. This kind

of phenomenon may result from the unfavorable charge

transfer process to adsorbed substance during light illumi-

nation where excess accumulation of electron and hole

undergoes recombination immediately without taking part in

the photocatalytic reaction, videlicet, ZnFe2O4 acting as a

recombination center. As a possible recombination manner

shown in Fig. 5, when the semiconductor is irradiated, the

electron–hole pairs possibly recombine in three steps. First

step: the electron is excited from the valence band to the

conduction band in the TiO2 bulk by high energy photon,

second step: the electron–hole transfer to the conduction

band and valence band of ZnFe2O4, respectively, and third

step: the electron in the conduction band of ZnFe2O4 jumps

into the valence band and recombine with the hole attended

by the emission of energy.

4. Conclusions

TiO2 and TiO2�ZnFe2O4 coatings were deposited on

stainless steel substrates by plasma spraying. The exper-

imental results show that FeTiO3 and ZnTiO3 have formed

by reaction between TiO2 and ZnFe2O4 owing to the high

temperature of plasma flame, the amount of resultant

increases when the spraying power was enhanced, and the

FeTiO3 compound could improve the photocatalytic activity

of the composite coatings. Moreover, the presence of

ZnFe2O4 may depress the photoactivity for unfavorable

electron–hole recombination in the bulk.

References

[1] M. Barbeni, E. Pramauro, E. Pelizzetti, E. Borgarello, N. Serpone,

Chemosphere 14 (1985) 195.

[2] M.R. Hoffmann, S.T. Martin, C. Wonyong, D.W. Bahnemann, Chem.

Rev. 95 (1995) 69.

Y. Zeng et al. / Surface & Coatings Technology 200 (2005) 2398–24022402

[3] M.V. Rao, K. Rajeshwar, V.R. Vernerker, J. Kubow, J. Phys. Chem. 84

(1980) 1987.

[4] S. Nishimoto, B. Ohtani, H. Kajiwara, T. Kagiya, J. Chem. Soc.,

Faraday Trans. 81 (1985) 61.

[5] Z. Ding, G.Q. Lu, P.F. Greenfild, J. Phys. Chem., B 104 (2000) 4815.

[6] M.R. Dhananjeyan, V. Kandavelu, R. Renganathan, J. Mol. Catal., A

Chem 151 (2000) 217.

[7] P. Bonamali, S. Maheshwar, N. Gyoichi, Mater. Chem. Phys. 59

(1999) 254.

[8] A.K.G. Carlos, F. Wypych, S.G. Moraes, N. Duran, N. Nagata, P.

Peralta-Zamora, Chemosphere 40 (2000) 433.

[9] L. Bayer, I. Eroglu, L. Turker, Sol. Energy Mater. Sol. Cells 62 (2000)

43.

[10] T. Ohno, F. Tanigawa, K. Fujihara, S. Izumi, M. Matsumura, J.

Photochem. Photobiol., A Chem. 127 (1999) 107.

[11] F.X. Ye, A. Ohmori, Surf. Coat. Technol. 160 (2002) 62.

[12] G. Marci, V. Augugliaro, C. Martin, J. Phys. Chem., B 105 (2001)

1026.

[13] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293

(2001) 269.

[14] Z.H. Yuan, W. You, J.H. Jia, L.D. Zhang, Chin. Phys. Lett. 15 (1998)

535.

[15] M.A. Valenzuela, P. Bosch, J. Jimenez-Becerrill, O. Quiroz, A.I. Paez,

J. Photochem. Photobiol., A Chem. 148 (2002) 177.

[16] Z.H. Yuan, L.D. Zhang, J. Mater. Chem. 11 (2001) 1265.

[17] R. Wang, C.H. Henager, J. Electrochem. Soc. 126 (1979) 83.

[18] R. Marchand, L. Brohan, M. Tournoux, Mater. Res. Bull. 15 (1980)

1129.

[19] X.Z. Ding, L. Liu, X.M. Ma, Z.Z. He, Y.Z. He, J. Mater. Sci. Lett. 13

(1994) 462.

[20] G.Q. Lu, A. Linsebigler, J.T. Yates, J. Phys. Chem. 98 (1994) 11733.

[21] I. Nakamura, N. Negishi, S. Kutsuna, T. Ihara, S. Sugihara, K.

Takcuchi, J. Mol. Catal., A Chem. 161 (2000) 205.

[22] D.S. Ginley, M.A. Butler, J. Appl. Phys. 48 (1977) 2019.

[23] N. Smirnova, A. Eremenko, O. Rusina, W. Hopp, L. Spanhel, J. Sol–

gel Sci. Technol. 22 (2001) 109.