Visible light activated photocatalytic behaviour of rare earth modifiedcommercial TiO2

D.M. Tobaldi a,*, R.C. Pullar a, A. Sever Skapin b, M.P. Seabra a, J.A. Labrincha a

aDepartment of Materials and Ceramic Engineering/CICECO, University of Aveiro, Campus Universitario de Santiago, 3810-193 Aveiro, Portugalb Slovenian National Building and Civil Engineering Institute, Dimiceva 12, SI-1000 Ljubljana, Slovenia

1. Introduction

The contamination of natural water reserves, air (indoor and

outdoor) and soil is currently one of the major global environmen-

tal concerns. Therefore, to further the sustainable development of

modern society, there is an urgent need for advances in green

technologies to achieve environmental remediation.

Environmental catalysis is one of the most studied processes in

order, not only to reduce, but also to prevent the causes of

pollution. Amongst these catalytic processes, photocatalysis is

expected to be one playing an important role in the 21st century’s

efforts in reducing environmental pollution [1]. Titanium dioxide

(TiO2) is the most investigated photocatalyst; the great interest in

TiO2 can be related to the work by Fujishima and Honda, published

in 1972 [2]. This described the photo-assisted electrolysis of water

upon irradiation of a single-crystal TiO2 electrode (with a Pt

counter-electrode), with photons of energy greater than the band

gap of TiO2. However, this was not strictly catalysis, the reaction

not being thermodynamically feasible without the photons – it

was actually an energy storage reaction that can be termed

photogalvanic [3]. The first reported works rigorously dealing with

‘‘photocatalysis’’ are those by Doerfler and Hauffe, published in

1964 [4,5]. In any case, we are still experiencing today an

enormous boom in this field of research, with a great number of

publications concerning photocatalysis appearing over the past 20

years. Photocatalysis can be described as that phenomenon in

which a material (a semiconductor) modifies the rate of a reaction,

via the action of light having a suitable wavelength [6,7]. When a

semiconductor is irradiated with photons having energy higher

than, or equal to, its energy band gap (Eg), an electron (e�) is able to

migrate from the valence band to the conduction band, leaving a

hole (h+) behind. Such a photo-generated couple (e�–h+) is able to

reduce and/or oxidise a pollutant adsorbed on the photocatalyst

surface [8].

As a photocatalytic material, TiO2 is chemically inert and non-

toxic; the reactions take place at mild operating conditions (e.g. a

low level of solar or artificial illumination, room temperature (RT)

and atmospheric pressure); no chemical additive is necessary;

Materials Research Bulletin 50 (2014) 183–190

A R T I C L E I N F O

Article history:

Received 16 April 2013

Received in revised form 2 October 2013

Accepted 17 October 2013

Available online 26 October 2013

Keywords:

A. Oxides

B. Phase transitions

B. Optical properties

C. Raman spectroscopy

D. Catalytic properties

A B S T R A C T

A commercial TiO2 nanopowder, Degussa P25, was modified with several rare earth (RE) elements in

order to extend its photocatalytic activity into the visible range. The mixtures were prepared via solid-

state reaction of the precursor oxides, and thermally treated at high temperature (900 and 1000 8C), with

the aim of investigating the photocatalytic activity of the thermally treated samples. This thermal

treatment was chosen for a prospective application as a surface layer in materials that need to be

processed at high temperatures.

The photocatalytic activity (PCA) of the samples was assessed in gas–solid phase – monitoring the

degradation of isopropanol (IPA) – under visible-light irradiation. Results showed that the addition of the

REs lanthanum, europium and yttrium to TiO2 greatly improved its photocatalytic activity, despite the

thermal treatment, because of the presence of more surface hydroxyl groups attached to the

photocatalyst’s surface, together with a higher specific surface area (SSA) of the modified and thermally

treated samples, with regard to the unmodified and thermally treated Degussa P25. The samples doped

with La, Eu and Y all had excellent PCA under visible-light irradiation, even higher than the untreated

Degussa P25 reference sample, despite their thermal treatment at 900 8C, with lanthanum producing the

best results (i.e. the La-, Eu- and Y-TiO2 samples, thermally treated at 900 8C, had, respectively, a PCA

equal to 26, 27 and 18 ppm h�1 – in terms of acetone formation – versus 15 ppm h�1 for the 900 8C

thermally treated Degussa P25). On the other hand, Ce–TiO2s had no significant photocatalytic activity.

� 2013 Elsevier Ltd. All rights reserved.

* Corresponding author. Tel.: +351 234 370 041.

E-mail addresses: david.tobaldi@ua.pt, david@davidtobaldi.org (D.M. Tobaldi).

Contents lists available at ScienceDirect

Materials Research Bulletin

jo u rn al h om ep age: ww w.els evier .c o m/lo c ate /mat res b u

0025-5408/$ – see front matter � 2013 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.materresbull.2013.10.033

volatile organic compounds (VOCs) [9,10], and even very recalci-

trant and persistent pollutants, can be degraded [11]. Moreover,

TiO2 heterogeneous photocatalysis is, at the same time, efficient in

green chemistry, in fine chemicals, and in advanced oxidation

processes (AOP) [12,13].

The photocatalytic reaction of TiO2 is activated by UVA light,

although titania is transparent for most of the visible radiation

region, since it is a wide band gap semiconductor material (Eg = 3.2

and 3.0 eV for anatase and rutile, respectively). This means that the

photocatalytic reaction is exploited by only 3–5% of the solar

spectrum [14]. Several different paths were followed in attempts to

create visible-light activated photocatalysts. With the aim of

extending the absorption edge of TiO2 into the visible region, this

was using doped/modified with non-metal atoms [15,16], transi-

tion metals [17], or, more recently, noble-metal atoms [18,19].

Anionic doping often had a detrimental effect on the photocatalytic

activity (PCA) of TiO2 under UV-light irradiation, because of an

enhancement of the charge recombination [20], and also because

such TiO2 has a low thermal and chemical stability [21]; moreover,

an excessive nitridation treatment might remove the doped N

atoms from the TiO2 matrix, leading to a decrease in the visible-

light PCA [22].

In this paper we report the synthesis, via solid-state reaction, of

visible-light activated photocatalysts, made from a commercial

TiO2 powder (Degussa P25) modified with several rare earth

elements (RE = cerium, lanthanum, europium and yttrium), and

thermally treated at temperatures as high as 900 and 1000 8C, in

order to investigate the photocatalytic activity mechanism (PCA) of

the RE-modified-TiO2. Degussa P25 has been previously modified

with REs, in order to investigate its PCA under UV-light exposure

[23,24], but this is the first time that such a high temperature

thermal treatment (essential for co-processing with some materi-

als), together with exclusively visible light irradiation, is assessed

in this system. REs were chosen as modifying agents because, due

to the transitions of 4f electrons, the optical absorption of titania is

increased, thus improving its visible-light response, and promoting

the separation of photo-generated electron–hole pairs [25].

2. Experimental

2.1. Sample preparation and characterisation

Samples were prepared via solid-state reaction of the two

crystalline end-members, according to the stoichiometry: Ti1�x-

RExO2 where RE = Ce, Eu, La, and Y, and x = 0, 0.01, and 0.025 mol.

Degussa P25 TiO2 powder was used as the titania source (hereafter

designated as P25), while reagent-grade CeO2, Eu2O3, La2O3, and

Y2O3 (all supplied by Aldrich) were used as RE precursors. Powders

were admixed, and wet ground in a rotary ball mill (30 min with

deionised water and sintered zirconia balls). Mixtures were dried

in an oven (105 8C for 12 h), ground in an agate mortar, and then

thermally treated; the thermal treatment was performed in an

electric oven with static air atmosphere, the heating rate was

200 8C h�1, and two maximum temperatures were used (900 and

1000 8C), with a soaking time at the maximum temperature of 4 h,

followed by natural cooling. Pure TiO2 samples, used as a reference,

were referred to as P25 followed by a number indicative of the

maximum temperature reached (i.e. P25-900 for Degussa P25 TiO2

powder thermally treated at 900 8C). For TiO2–RE mixtures, the

symbol of the RE chemical element followed by the amount of RE

cation inserted – with the molar content expressed as a percentage

– was used, followed by the temperature (i.e. sample with 0.01 mol

of Europium added to P25 titania, and calcined at 900 8C, is referred

to as: Eu1-900).

The phase composition of the starting titania powders, as well

as that of the TiO2–RE mixtures, was obtained via X-ray diffraction,

using a Rigaku Geigerflex (JP); the patterns were collected in the

15–808 2u range (0.028 2u s�1 step-scan, and 5 s/step). A semi-

quantitative estimate of the likely amount of anatase/rutile ratio,

as well as the secondary phases – in the TiO2–RE mixtures – was

achieved by adopting the generalised RIR method [26,27],

assuming the absence of amorphous phase in the examined

powders. By contrast, the as-received P25 was assumed to be –

after Ohtani et al. [28] – a mixture of anatase/rutile/amorphous

phase, that is approximately: 78/14/8; having a specific surface

area of �50 m2 g�1. The morphology of the samples was

investigated by FE-SEM (Hitachi, SU-70, JP), equipped with an

energy dispersive X-ray spectroscopy (EDS) attachment (Bruker

Quantax 400, DE).

Optical spectra of the samples were acquired on a Shimadzu UV

3100 (JP), in the UV–Vis range (200–800 nm), with 0.02 nm step-

size, and using BaSO4 as reference. The diffuse reflectance (R1) was

converted into the absorption coefficient a, using the Kubelka–

Munk equation: a ¼ ð1 � R1Þ2 � 2R�11 [29]. The Eg of the samples

was calculated using the Tauc plot. This method assumes that the

absorption coefficient, a, of a semiconductor can be expressed as:

(ahn)g= A(hn � Eg), where A is a material constant, h is Plank’s

constant, n is the frequency of the light, Eg is the energy band gap of

allowed transitions, and the power coefficient g is characteristic of

the type of transition. For nanoscale semiconductor materials, the

value of g is accepted to be equal to 0.5, since for such materials,

the transition is assumed to be indirectly allowed [30]. Hence,

plotting [F(R1)hn]0.5 against hn, one can obtain the energy band

gap of the semiconductor material, from the x-axis (a = 0)

intercept of the line tangent to the inflection point of the curve.

In order to find out the possible occurrence of OH groups and/or

water adsorbed on photocatalyst’s surface, FT-IR analysis was

performed using a Bruker Tensor 27 (DE). The measurements were

carried out over the wavenumber range of 4000–350 cm�1, and the

powders (2 mg) were mixed with KBr (200 mg, to give 1 wt% of

powder in the KBr disks), and pressed into thin pellets. Prior to

carrying out the FT-IR analysis, the pellets were stored in an oven at

120 8C for 30 min. Raman spectra of the samples were acquired on

a HR 800 (Horiba Jovin Yvon, JP), equipped with a 532 nm laser as

the excitation source (DPSS Ventus laser). The specific surface area

(SSA) of the prepared samples was evaluated by the Brunauer–

Emmett–Teller method (BET) (Micromeritics Gemini 2380, US),

using N2 as the adsorbate gas.

2.2. Evaluation of photocatalytic activity (PCA)

The PCA of the prepared powders was evaluated in gas-solid

phase, by monitoring the degradation of isopropanol (IPA) and the

subsequent formation of acetone, with FT-IR spectroscopy [31,32].

Dyes were not used because they can be excited by visible-light

irradiation, and consequently act as a sensitiser, with electron

injection from the photo-excited dye to the photocatalyst [33,34].

Hence, this electron transfer may destroy the regular distribution

of conjugated bonds within the dye molecule, causing its

decolourisation, but not its mineralisation [35]. The device

employed for the gas-solid phase tests, operating in batch

conditions, is shown in Fig. 1. It is composed of a cylindrical

reactor (1.4 L in volume), covered by quartz glass and connected by

Teflon tubes to the FT-IR spectrometer, the whole system was

hermetically sealed; the PCA experiments were assessed at

23 � 2 8C. The light source was a 300 W Xenon lamp (Newport Oriel

Instruments, US); visible-light irradiation was achieved by way of a

filter at 400 nm. Light intensity was approximately zero in the l range

of 300–400 nm, and 160 W m�2 in the l range of 400–800 nm (inset

in Fig. 1).

The samples were prepared in the form of a thin layer of powder,

with a constant mass (about 50 mg), and thus approximately

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190184

constant thickness, in a 6 cm diameter Petri dish; the working

distance between the Petri dish and the lamp was 6 cm. The

relative humidity in the reacting system was kept constant in the

range 25–30% by means of a flow of air passing through molecular

sieves, until a pre-defined humidity was attained. Each experi-

ment was performed by injecting 5 mL of IPA (�1100 ppm in gas

phase) into the reacting system through a septum; the total

reaction time was set at 24 h. At the beginning of each

photocatalytic experiment, when IPA had been injected, the lamp

was kept switched off, so as to achieve an IPA adsorption–

desorption equilibrium with that powder. The lamp was then

switched on after that adsorption–desorption equilibrium had

been accomplished. The photocatalytic activity of the powders

was assessed according to the value of the rate constant

representing the oxidation of IPA into acetone. The concentrations

of IPA and acetone in the reacting system, before and after

switching on the lamp, were continuously checked by monitoring

the calculated areas of their characteristic peaks, at 951 cm�1 and

1207 cm�1, respectively, using a FT-IR spectrometer (Perkin Elmer

Spectrum BX, US) in situ. The analysis of the data was established

on a two-step oxidation mechanism: in the first step, IPA degrades

into acetone; in the second step, acetone transforms into further

oxidation products. The first step is usually considered a zero-

order reaction – with a rate constant (k1) – while the second step

can be assumed to be a first-order reaction [31,32]. In the present

context we are merely interested in the relative change of the rate

constant k1, which is an estimate of the photocatalytic activity.

Therefore, the PCA was evaluated as the rate constant of the initial

acetone formation, because at RT, the photocatalytic oxidation of

IPA to form acetone is fast, whereas the subsequent oxidation of

acetone to CO2 and H2O is slower [36]. Over very short times, the

slope is proportional to the formation rate constant. Being the first

intermediate of IPA degradation, the formation of acetone is a

process that can easily be distinguished from the subsequent

degradation process occurring during IPA photo-oxidation [37].

The PCA measurements of P25-REs were repeated in duplicate in

all cases, and the values reported were calculated as an average

value. The experimental error of the method employed was found

to be in the range of 0.5 ppm h�1.

3. Results and discussion

3.1. X-ray diffraction

The addition of RE did not delay the anatase-to-rutile phase

transition (ART) in all cases, at the thermal treatment temperatures

used. Notwithstanding, a delaying effect by REs on the ART with

P25 TiO2 has actually been reported by Lin and Yu (CeO2, La2O3, and

Y2O3, at a concentration of 0.5 wt%, with calcination temperatures

of 650 and 700 8C) [23], and by Du et al. (CeO2 and La2O3, at a

concentration of 2 wt%, thermally treated at 800 8C) [24]. The XRD

patterns all showed only the rutile phase of TiO2, the only

exception being sample La1 thermally treated at 900 8C, which still

contained a small amount of anatase (1.8 wt%) amongst 98.2 wt%

rutile (Table 1, inset in Fig. 2). As recently reported by Ermokhina

et al. [38], as lanthanum is more electropositive than titanium,

causing a partial transfer of electrons from the La–O bonds to the

Ti–O bonds, leading to a strengthening of the Ti–O bonding, in the

anatase structure. Here, we believe that La3+ ions occupied

interstitial positions in titania lattice – because of ionic radii

issues, see below – thus favouring the slow grain growth rate of

anatase [39]. This was due to the segregation of La3+ at the anatase

grain boundaries, thus increasing the diffusion barrier at the

titania–titania grain interface, hindering the grain growth process.

Apart from with cerium, which does not form a stable mixed

titanate compound in air, La1-900 was also the only sample which

did not contain evidence of a reaction between the RE and TiO2

forming a RE-titanate compound (Fig. 2). The absence of a solid

solution between RE and TiO2 at the low temperature of 900 8C is

reasonable, due to the significant difference between the effective

ionic radii of REs and Ti4+ ([6]Ce4+ = 0.87 A, [6]Eu3+ = 0.95 A,[6]La3+ = 1.03 A, [6]Y3+ = 0.90 A; [6]Ti4+ = 0.61 A [40]). At 1000 8C,

or with 2.5 mol% La3+ at 900 and 1000 8C, the lanthanum titanate

La4Ti9O24 was detected in significant amounts. The pyrochlore

Fig. 1. Scheme of the reactor utilised for the photocatalytic tests; in the inset is reported the spectral irradiance of the Xenon lamp, at the working distance, with filter at

400 nm.

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190 185

phases Eu2Ti2O7 and Y2Ti2O7 existed in all Eu and Y doped samples.

This observed behaviour is also in good agreement with the

respective phase diagrams. According to Iwasaki [41], Mizutani

et al. [42], and Skapin et al. [43], Eu2O3, Y2O3, and La2O3 reacted

with TiO2 to form the RE-titanium oxides Eu2Ti2O7, Y2Ti2O7 and

La4Ti9O24, respectively. In the TiO2–Y2O3 system observed here,

together with rutile and the Y2Ti2O7 intermediate, there was also a

very small amount (1.4–0.6 wt%) of the other end-member/

starting material, Y2O3, which did not react with TiO2. At a higher

temperature, 1000 8C, the amount of Y2O3 decreased, with greater

Y2Ti2O7 formation. Neither the Eu nor La doped samples retained

any of the RE oxide starting materials. As expected, the TiO2–CeO2

system shows no reactions between the two end-members, at both

the temperatures used here.

3.2. FT-IR and Raman analysis

The FT-IR spectra of the samples are depicted in Fig. 3a–c. The

Ti–O–Ti vibration band, in the range of 400–600 cm�1 [44], is

observed in all the FT-IR spectra. The peak at �1620 cm�1

corresponds to the bending vibrations of O–H, and the broad band

at around 3450 cm�1 was attributed to the surface adsorbed

hydroxyl groups [45]. The sharp bands that are in the region 2750–

3000 cm�1 are attributable to the surface adsorbed water [31].

The thermal treatment at 900 8C of the undoped P25 led to a

decrease in the intensity of the bands at 3450, and 2750–

3000 cm�1, suggesting that the terminal hydroxyls are removed

from the surface of TiO2. On the other hand, the TiO2–RE mixtures,

thermally treated at 900 8C (Fig. 3b and c), still showed the

presence of those bands (more intense in the La-, and Eu-doped

samples, not present in Ce-doped samples), indicating that the RE

modification resulted in more surface hydroxyl groups on the

photocatalysts, even after the thermal treatment. It has been

suggested that RE metal ions with a 3+ oxidation state, create a

charge imbalance in TiO2. Hence, more hydroxide ions would be

adsorbed on the surface, for a charge balance. These hydroxide

ions can accept holes to form hydroxyl radicals, oxidising the

adsorbed molecules [23]. CeO2–TiO2, not having that charge

Table 1

Phase composition, energy band gap (Eg), and specific surface area (SSA) of the samples.

Sample Phase composition (wt%) Eg (eV) SSABET (m2g�1)

Anatase Rutile RE–TiO2 RExOy

P25-900 – 100 – – 3.01 2.0

Ce1-900 – 98.0 – 2.0 2.91 2.1

Ce2.5-900 – 94.1 – 5.9 2.90 3.3

Eu1-900 – 98.7 1.3 – 3.03 8.6

Eu2.5-900 – 91.7 8.3 – 3.04 9.1

La1-900 1.8 98.2 – – 3.04 11.5

La2.5-900 – 88.5 11.5 – 3.04 12.0

Y1-900 – 98.2 1.8 – 3.03 8.1

Y2.5-900 – 94.8 3.9 1.4 3.03 8.1

P25-1000 – 100 – – 2.98 0.5

Ce1-1000 – 98.0 – 2.0 2.84 <0.1

Ce2.5-1000 – 94.8 – 5.2 2.89 0.4

Eu1-1000 – 96.6 3.4 – 3.01 2.8

Eu2.5-1000 – 91.2 8.8 – 3.01 3.1

La1-1000 – 94.8 5.2 – 3.02 4.6

La2.5-1000 – 88.4 11.6 – 3.03 4.4

Y1-1000 – 97.6 2.4 – 3.00 2.2

Y2.5-1000 – 93.8 5.6 0.6 3.01 2.3

Fig. 2. XRD patterns of the RE-modified samples thermally treated at 900 8C. The inset represents a magnification of the La1-900 sample (framed area), in the 24.5–29.58 2u

range, in order to highlight the anatase (1 0 1) reflection. A and R are symbols, standing for: anatase and rutile, respectively; *, y, z, � symbols indicate: CeO2, Eu2Ti2O7, Y2O3,

and Y2Ti2O7, respectively.

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190186

imbalance, has considerably fewer hydroxyl groups adsorbed on

the surface (cf Fig. 3b and c).

Raman spectra of the samples thermally treated at 900 8C are

shown in Fig. 4a and b. The band at around 144 cm�1,

corresponding to the Eg anatase mode [46], is found to be in

every spectrum (Fig. 4a). This is in contrast to the XRD data, which

only showed anatase to be present in sample La1-900. In this

sample the anatase XRD peak was extremely small (Fig. 3),

estimated as <2 wt%, and if present in lower amounts in other

samples, anatase could be undetectable by XRD – the Raman

technique having a lower detection limit [47]. Although the Raman

technique cannot be used for quantitative analysis, a comparison

between different samples can still be performed [48]. It can be

seen that the anatase peak is much more intense for the La1-900

sample, indicating a higher anatase concentration in this sample

than the others, explaining the absence of anatase peaks in the

other XRD data. The peaks at around 447 cm�1 (Eg), 612 cm�1 (A1g),

and 822 cm�1 (B2g), are assigned to rutile [49]. The broad band,

centred at around 250 cm�1, which is a specific feature of rutile,

has a complex nature. Gotic et al. and Frank et al. [50,51], suggested

that second-order scattering, as well as disorder effects, are

involved in its formation. Furthermore, in the RE-modified

samples, the Raman peak corresponding to the Eg mode of anatase

is only very slightly blue-shifted compared to the unmodified

sample (P25-900), whilst that corresponding to the Eg rutile mode

had a very weak red-shift (Table 2). This slight shifting in the Egmodes of anatase and rutile are barely noticeable, and are not

considered significant enough to show any increase in oxygen

vacancies or deficiency [52], that could be induced by RE

Fig. 3. FT-IR spectra of a) the undoped P25 at RT, and thermally treated at 900 8C; b)

FT-IR spectra of the Ce1, Eu1, La1, and Y1 samples thermally treated at 900 8C; c)

magnification, in the wavenumber range 4000–2800 cm�1, of the samples modified

with 2.5 mol% of REs and thermally treated at 900 8C, in order to emphasise the

surface adsorbed hydroxyl groups.

Fig. 4. a) Raman spectra of the samples, in the wavenumber range of 100–

1000 cm�1. A and R are symbols, standing for anatase and rutile, respectively; b)

Raman spectrum of the Eu1-900 sample, in the wavenumber range of 1000–

3500 cm�1.

Table 2

Raman peak positions of the anatase and rutile Eg vibrational modes.

Sample Raman peak (cm�1)

Anatase (Eg) Rutile (Eg)

P25-900 144.0 445.8

Eu1-900 144.8 445.3

La1-900 144.4 445.6

Y1-900 144.3 445.5

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190 187

substitution [53,54]. Therefore, we can say that the oxygen

stoichiometry of the titania is preserved with the addition of

the RE dopants.

The Raman spectrum of sample Eu1-900, in the wavelength

range 1000–3500 cm�1, is depicted in Fig. 4b. It is interesting to note

here the luminescence emissions of the Eu-modified TiO2, consistent

with Frindell et al. [55], who found these emissions on a titania film

doped with Eu3+, but only when this was excited with a wavelength

above the titania band gap. The emission lines shown are associated

with Eu3+ ions, and they correspond to radiative relaxations from the

level 5D0 to its low-lying multiplets 7FJ (J = 0, 1, 2, 3) [56], indicating

that the Eu3+ emission is associated with a non-radiative energy

transfer process from TiO2 to Eu3+ crystal field states [57]. The

following emissions were found: 5D0 !7F0,

5D0 ! 7F1,5D0 !

7F2,5D0 !

7F3 [56,58], with the strongest emission band – located at

�2500 cm�1 – associated with 5D0 !7F2 transitions. These latter

transitions have been reported to be possible only if Eu3+ ions occupy

a site without an inverse symmetry [55].

3.3. UV–vis spectroscopy

The DRS spectra of the doped powders, thermally treated at

900 8C and 1000 8C, are shown in Fig. 5a and c. All the spectra, at

both the thermal treatment temperatures, consisted of a single

absorption band below approximately 410 nm, ascribed to the

metal-ligand charge transfer in titania [59]. At 900 8C, all the doped

powders showed a slight shift of their absorption edge into the

visible region, due to the presence of rutile (Fig. 5a). Moreover, the

Ce–TiO2 sample showed a tail in the visible region of the spectrum,

that extends up to �450 nm, due to the contribution of the free

CeO2 in that sample [60]. At 1000 8C (Fig. 5c), as before, only the

Ce–TiO2 showed a noticeable shift of its absorption edge into the

visible region, keeping the tail, at around 450 nm.

The Egs of the powders were estimated by the Tauc procedure

(Table 1, and Fig. 5b and d): the curve – resulting from the plot of

the transformed Kubelka–Munk equation [F(R1)hn]0.5 versus the

photon energy (hn) – was fitted adopting a sigmoidal Boltzmann

function. Afterwards, the Eg value was obtained by the x-axis

intercept of the line tangent to the inflection point of that curve.

The resulting ‘‘apparent’’ band gaps of the samples are reported in

Table 1, and were assigned to the rutile phase, since they are

consistent with the expected rutile Eg value, that is 3.02 eV. The

only exceptions, consistent with the DRS spectra, were the Ce–TiO2

samples, whose Egs were shifted into the visible region (2.84–

2.91 eV, cf Table 1), due to the contribution of free CeO2 in the

samples, as detected by the tail in the DRS spectra.

Fig. 5. a) UV–vis spectra of the RE-modified samples, thermally treated at 900 8C; b) Kubelka–Munk elaboration for the same samples. The dotted line represents the x-axis

intercept of the line tangent to the inflection point of the sample La1-900, e.g. its apparent Eg, calculated with the Tauc procedure; c) UV–vis spectra of the RE-modified

samples, thermally treated at 1000 8C; d) Kubelka–Munk elaboration for the same samples. The dotted line represents the x-axis intercept of the line tangent to the inflection

point of the sample La1-1000, calculated with the Tauc procedure.

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190188

4. Photocatalytic activity

The PCA results in gas–solid phase, under visible light

irradiation, are shown in Fig. 6. The untreated P25 powder had

a PCA – in terms of acetone formation – of 23 ppm h�1. Hurum et al.

[61], using electron paramagnetic resonance (EPR) spectroscopy,

claimed that the visible-light response of P25 is due to the presence

of small rutile crystallites amongst the anatase, their smaller band

gap extended the useful range of PCA into the visible region.

Moreover, the points of contact between these anatase and rutile

crystals allow for rapid electron transfer from rutile to anatase.

Thus, rutile acts as an ‘‘antenna’’ to extend the PCA into visible

wavelengths. Quite obviously, the thermal treatment of the

undoped P25 powder, led to a decrease of its PCA (acetone

formation equal to 15 and 4 ppm h�1, at 900 and 1000 8C,

respectively). This is due to the concomitant size growth of the

titania particles, and the decrease of their SSA. On the contrary, at

900 8C, the RE addition to TiO2 improved the PCA of these

thermally treated samples. The best performing sample was La2.5-

900, with an activity of 32 ppm h�1 of acetone formation. The three

samples La2.5-900, Eu1-900 and La1-900 all had excellent PCA

under visible-light irradiation, even higher than the untreated

reference sample, with these two latter samples having a PCA of 27

and 26 ppm h�1, respectively, in terms of acetone formation –

despite the thermal treatment at a temperature as high as 900 8C.

The only sample that had no PCA was the CeO2–TiO2 one. The

increase of the thermal treatment temperature to 1000 8C led to an

obvious decrease of the overall PCA. At this temperature, samples

possessing the best PCA were La1-1000 and La2.5-1000, with an

activity, estimated as acetone formation, of 8 and 13 ppm h�1,

respectively. Such behaviour can be explained by the high SSA

values (and consequently by their reduced particle size), as shown

in Table 1, and by the FE-SEM images in Fig. 7a–c, comparing the

P25-900, La1-900 and Eu1-900 samples after heating at 900 8C.

The samples with the largest SSA values also had the highest PCA

(cf Table 1 and Fig. 6). As a matter of fact, the samples with the

lowest SSA, based on Ce–TiO2, were also the samples with the

lowest PCA, at both the thermal treatment temperatures, and the

increase of the thermal treatment temperature at 1000 8C led to a

decrease of the SSA, and consequently, to a decrease in the PCA. It is

also believed that a high thermal treatment can induce desorption

of surface oxygen, this leaving an oxygen vacancy and Ti3+ sites on

titania. It was hence supposed that electrons may be excited from

the valence band to the oxygen vacancy states, even with the

energy of visible light [62,63]. Yin et al., proposed that rutile with

small particle size, together with its small band gap value, can be a

visible light activated photocatalyst [64]. Also Orlikowski et al.,

monitoring the oxidation of phenol under visible light exposure,

attributed the small particle size as being a key factor for rutile as

an active photocatalyst [65]. Moreover, as detected by the FT-IR

analysis, these RE modifications resulted in more water/hydroxyl

groups attached to the surface of the thermally treated photo-

catalyst. This enhances the PCA of titania, because it helps generate

chemical oxidative species, such as hydroxyl radicals, that ‘‘switch

on’’ the photocatalytic oxidation [66] (with the only exception of

Ce–TiO2 samples, that basically had no PCA – also consistent with

the work by Lin and Yu [23]). Furthermore, as was reported by Niu

et al. [67] and Shi et al. [68], REs such as Eu, La, and Y can favour the

separation of the photo-generated couple e––h+, and hence prevent

their recombination, increasing TiO2’s PCA. This statement wasFig. 6. Photocatalytic activity of the samples, in gas-solid phase, under visible-light

irradiation.

Fig. 7. FE-SEM micrographs of samples: a) P25-900; b) La1-900; c) Eu1-900; in the

inset is reported an EDX spectrum of this sample (the signal of Al belongs to the

sample-holder).

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190 189

further confirmed by Stengl et al. [25]: they claimed that the 4f

electron transitions of REs strengthened the optical adsorption of

the catalysts, supporting the separation of photo-generated pairs.

A combination of all these features resulted in a marked increase of

the photocatalytic ability of the La, Eu or Y doped titania after

heating to 900 8C, particularly in the case of La. The only inactive

samples were those based on Ce–TiO2.

5. Conclusions

A commercial TiO2 powder, Degussa P25, was modified with a

series of REs in order to investigate the PCA mechanism, under

visible-light irradiation, of samples treated at high temperature.

The mixtures studied were prepared via solid-state reaction of the

precursor oxides, and the products of the synthesis were thermally

treated at 900 8C and 1000 8C. From a mineralogical point of view,

the addition of RE accelerated the ART, at both thermal treatment

temperatures, because of the formation of oxygen vacancies. At the

same time, the RE modification led to more surface hydroxyl

groups on the photocatalysts, even after the thermal treatment.

This, together with an increase of the SSA – compared to the

unmodified sample – greatly contributed to enhance the visible-

light PCA of the Eu-, La-, and Y-TiO2s (all in the rutile phase, the

only exception being sample La1-900, containing a small amount

of anatase), thermally treated at 900 8C. At this temperature, the

only inactive sample was the Ce-modified one. The increase of the

thermal treatment temperature, 1000 8C, led to a dramatic

decrease of the PCA for all the samples.

Acknowledgements

Authors wish to acknowledge PEst-C/CTM/LA0011/2013 pro-

gramme. M.P. Seabra and R.C. Pullar wish to thank the FCT

Ciencia2008 programme for supporting this work. Maria Celeste

Coimbra de Azevedo (Chemistry Department, University of Aveiro)

is gratefully acknowledged for the FT-IR measurements.

References

[1] R.K. de Richter, S. Caillol, J. Photochem. Photobiol. C: Photochem. Rev. 12 (2011) 1–19.

[2] A. Fujishima, K. Honda, Nature 238 (1972) 37–38.[3] C.H. Langford, Catalysts 2 (2012) 327–329.[4] W. Doerffler, K. Hauffe, J. Catal. 3 (1964) 156–170.[5] W. Doerffler, K. Hauffe, J. Catal. 3 (1964) 171–178.[6] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C: Photochem. Rev. 1

(2000) 1–21.[7] A. Fujishima, X. Zhang, D. Tryk, Surf. Sci. Reports 63 (2008) 515–582.[8] M.R. Hoffmann, S.T. Martin, W. Choi, D.W. Bahnemann, Chem. Rev. 95 (1995) 69–96.[9] A.J. Maira, W.N. Lau, C.Y. Lee, P.L. Yue, C.K. Chan, K.L. Yeung, Chem. Eng. Sci. 58

(2003) 959–962.[10] S. Cao, K.L. Yeung, P.-L. Yue, Appl. Catal. B: Environ. 76 (2007) 64–72.[11] J.-M. Herrmann, C. Duchamp, M. Karkmaz, B.T. Hoai, H. Lachheb, E. Puzenat, et al.

J. Hazard. Mater. 146 (2007) 624–629.[12] J.-M. Herrmann, Catal. Today 53 (1999) 115–129.[13] J.-M. Herrmann, J. Photochem. Photobiol. Chem. 216 (2010) 85–93.[14] Z. Sen, Solar Energy Fundamentals and Modeling Techniques Atmosphere Envi-

ronment Climate Change and Renewable Energy, Springer, London, 2008.[15] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293 (2001) 269–271.[16] D.M. Tobaldi, L. Gao, A.F. Gualtieri, A. Sever Skapin, A. Tucci, C. Giacobbe, J Am.

Ceram. Soc. 95 (2012) 1709–1716.[17] M. Anpo, Pure Appl. Chem. 72 (2000) 1787–1792.[18] D.M. Tobaldi, R.C. Pullar, A.F. Gualtieri, M.P. Seabra, J.A. Labrincha, Chem. Eng. J.

214 (2013) 364–375.

[19] P. Wang, B. Huang, Y. Dai, M.-H. Whangbo, Phys. Chem. Chem. Phys. 14 (2012)9813.

[20] X. Zhang, K. Udagawa, Z. Liu, S. Nishimoto, C. Xu, Y. Liu, et al. J. Photochem.Photobiol. Chem. 202 (2009) 39–47.

[21] M. Kitano, K. Funatsu, M. Matsuoka, M. Ueshima, M. Anpo, J. Phys. Chem. B 110(2006) 25266–25272.

[22] S. Hu, F. Li, Z. Fan, C.-C. Chang, Appl. Surf. Sci. 258 (2011) 182–188.[23] J. Lin, J.C. Yu, J. Photochem. Photobiol. Chem. 116 (1998) 63–67.[24] P. Du, A. Bueno-Lopez, M. Verbaas, A.R. Almeida, M. Makkee, J.A. Moulijn, et al. J.

Catal. 260 (2008) 75–80.[25] V. Stengl, S. Bakardjieva, N. Murafa, Mater. Chem. Phys. 114 (2009) 217–226.[26] R.L. Snyder, in: E. Lifshin (Ed.), X-Ray Charact. Mater., Wiley-VCH Verlag GmbH,

Weinheim, Germany, 2007, pp. 1–103.[27] F. Matteucci, G. Cruciani, M. Dondi, M. Raimondo, Ceram. Int. 32 (2006) 385–392.[28] B. Ohtani, O.O. Prieto-Mahaney, D. Li, R. Abe, J. Photochem. Photobiol. Chem. 216

(2010) 179–182.[29] A.S. Marfunin, Physics of Minerals and Inorganic Materials: An Introduction,

Springer-Verlag, 1979.[30] N. Serpone, D. Lawless, R. Khairutdinov, J. Phys. Chem. 99 (1995) 16646–16654.[31] D.M. Tobaldi, A. Tucci, A.S. Skapin, L. Esposito, J. Eur. Ceram. Soc. 30 (2010) 2481–

2490.[32] M. Tasbihi, U.L. Stangar, A.S. Skapin, A. Ristic, V. Kaucic, N.N. Tusar, J. Photochem.

Photobiol. Chem. 216 (2010) 167–178.[33] M. Vautier, C. Guillard, J.-M. Herrmann, J. Catal. 201 (2001) 46–59.[34] X. Yan, T. Ohno, K. Nishijima, R. Abe, B. Ohtani, Chem. Phys. Lett. 429 (2006) 606–

610.[35] J.-M. Herrmann, Appl. Catal. B Environ. 99 (2010) 461–468.[36] S. Larson, J.A. Widegren, J.L. Falconer, J. Catal. 157 (1995) 611–625.[37] Y. Ohko, A. Fujishima, K. Hashimoto, J. Phys. Chem. B 102 (1998) 1724–1729.[38] N.I. Ermokhina, V.A. Nevinskiy, P.A. Manorik, V.G. Ilyin, V.N. Novichenko, M.M.

Shcherbatiuk, et al. J. Solid State Chem. (2013).[39] C.P. Sibu, S.R. Kumar, P. Mukundan, K.G.K. Warrier, Chem. Mater. 14 (2002) 2876–

2881.[40] R.D. Shannon, Acta Crystallogr. Sect. 32 (1976) 751–767.[41] B. Iwasaki, Bull. Chem. Soc. Jpn. 51 (1978) 3223–3226.[42] N. Mizutani, Y. Tajima, M. Kato, J. Am. Ceram. Soc. 59 (1976), 168-168.[43] S.D. Skapin, D. Kolar, D. Suvorov, J. Eur. Ceram. Soc. 20 (2000) 1179–1185.[44] Y.D. Hou, X.C. Wang, L. Wu, X.F. Chen, Z.X. Ding, X.X. Wang, et al. Chemosphere 72

(2008) 414–421.[45] C. He, B. Tian, J. Zhang, Micropor. Mesopor. Mater. 126 (2009) 50–57.[46] H. Chang, P.J. Huang, J. Raman Spectrosc. 29 (1998) 97–102.[47] C.G. Kontoyannis, N.V. Vagenas, Analyst 125 (2000) 251–255.[48] C. Piccirillo, C.W. Dunnill, R.C. Pullar, D.M. Tobaldi, J.A. Labrincha, I.P. Parkin, et al.

J. Mater. Chem. 1 (2013) 6452–6461.[49] A. Li Bassi, D. Cattaneo, V. Russo, C.E. Bottani, E. Barborini, T. Mazza, et al. J. Appl.

Phys. 98 (2005) 074305.[50] M. Gotic, M. Ivanda, S. Popovic, S. Music, A. Sekulic, A. Turkovic, et al. J. Raman

Spectrosc. 28 (1997) 555–558.[51] O. Frank, M. Zukalova, B. Laskova, J. Kurti, J. Koltai, L. Kavan, Phys. Chem. Chem.

Phys. 14 (2012) 14567.[52] J.C. Parker, R.W. Siegel, Appl. Phys. Lett. 57 (1990) 943.[53] Y. Huo, J. Zhu, J. Li, G. Li, H. Li, J. Mol. Catal. Chem. 278 (2007) 237–243.[54] S. Sadhu, P. Poddar, RSC Adv. 3 (2013) 10363–10369.[55] K.L. Frindell, M.H. Bartl, A. Popitsch, G.D. Stucky, Angew. Chem. 114 (2002) 1001–

1004.[56] M. Pal, U. Pal, J.M.G.Y. Jimenez, F. Perez-Rodrıguez, Nanoscale Res. Lett. 7 (2012) 1.[57] A. Conde-Gallardo, M. Garcıa-Rocha, I. Hernandez-Calderon, R. Palomino-Merino,

Appl. Phys. Lett. 78 (2001) 3436.[58] L. Yu, M. Nogami, Mater. Chem. Phys. 107 (2008) 186–188.[59] F.C. Hawthorne, Spectroscopic Methods in Mineralogy and Geology, Mineralogi-

cal Society of America, 1988.[60] F. Galindo-Hernandez, R. Gomez, J. Photochem. Photobiol. Chem. 217 (2011) 383–

388.[61] D.C. Hurum, A.G. Agrios, K.A. Gray, T. Rajh, M.C. Thurnauer, J. Phys. Chem. B 107

(2003) 4545–4549.[62] M. Xing, J. Zhang, F. Chen, B. Tian, Chem. Commun. 47 (2011) 4947–4949.[63] K.-C. Huang, S.-H. Chien, Appl. Catal. B: Environ. 140–141 (140) (2013) 283–290.[64] S. Yin, H. Hasegawa, D. Maeda, M. Ishitsuka, T. Sato, J. Photochem. Photobiol.

Chem. 163 (2004) 1–8.[65] J. Orlikowski, B. Tryba, J. Ziebro, A.W. Morawski, J. Przepiorski, Catal. Commun. 24

(2012) 5–10.[66] G. Balasubramanian, D. Dionysiou, M. Suidan, I. Baudin, J. Laine, Appl. Catal. B:

Environ. 47 (2004) 73–84.[67] X. Niu, S. Li, H. Chu, J. Zhou, J. Rare Earths 29 (2011) 225–229.[68] H. Shi, T. Zhang, H. Wang, J. Rare Earths 29 (2011) 746–752.

D.M. Tobaldi et al. / Materials Research Bulletin 50 (2014) 183–190190