Open Journal of Advanced Oxidation Technologies 2018; 21(1)

Access Article ID- 20170008

Article:

Synthesis and Characterization of Photoactive

S-TiO2 from TiOSO4 Precursor Using an
Integrated Sol-Gel and Solvothermal Method
at Low Temperatures
Jerson Mosquera-Pretelt1, María I. Mejía1, 2, Juan M. Marín1, *
1Grupo Procesos Fisicoquímicos Aplicados, Departamento de Ingeniería Química, Facultad de Ingeniería, Sede de
Investigación Universitaria, Universidad de Antioquia UdeA, Calle 70 No. 52-21, Medellín, Colombia.
Doctorado en Ingeniería, Facultad de Ingeniería, Universidad de Medellín, Carrera 87 No. 30-65 Medellín,
2

Colombia.
[email protected]; [email protected]; *[email protected]

Abstract

Photoactive S-titanium dioxides (S-TiO2) were synthesized from TiOSO4 as only Ti and S precursor using an integrated sol-gel
and solvothermal method at low temperatures (200 °C – 250 °C). The effect of the synthesis conditions (molar ratios of
water/TiOSO4 and solvent (ethanol)/TiOSO4 as well as temperature, < 250 °C) of the applied method in the properties and the
photoactivity of the synthesized materials was evaluated through Box-Behnken experimental design. The prepared
photocatalysts were characterized by X-ray diffraction (XRD), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS),
X-ray fluorescence (XRF), attenuated total reflectance - Fourier transform infrared spectroscopy (ATR-FTIR), ultraviolet – visible
diffuse reflectance spectroscopy (UV/vis-DRS), scanning electron microscopy (SEM), transmission electron microscopy (TEM),
porosity and BET surface area analysis. The photocatalytic activities of the synthesized S-TiO2 materials were determined by the
photodegradation of formic acid. Results indicated that the integrated sol-gel and solvothermal method at low temperatures led
to obtain mesoporous, crystalline and photoactive S-TiO2 materials, crystallized as anatase phase with UV and visible light
absorption for all synthesis conditions. All synthesized S-TiO2 materials showed high activity in formic acid photodegradation,
which was associated on their degrees of crystallinity, particle sizes and sulfur contents, being higher in the materials
synthesized with the temperature of 250 °C. Material synthesized with molar ratio water/TiOSO4 of 4.0, molar ratio
ethanol/TiOSO4 of 15 and T = 250 °C, showed the highest photocatalytic activity, a crystallite size of 42.34 nm, surface area of
35.77 m2/g, sulfur content of 0.818 wt %, high UV and visible radiation absorption and band gap of 3.03.

Keywords

Titanium Dioxide; Titanium Oxysulfate; Sol-gel Method; Solvothermal Method; Photocatalysis; Formic Acid
Received: January 19, 2017; Revised: May 22, 2017; Accepted: July 11, 2017

Introduction

TiO2 is a semiconductor extensively used in photocatalytic processes due to its easy acquisition, photostability, no
toxicity and low price [1]. However, low efficiencies in TiO2 photocatalytic applications are associated with the fact
that some of the electron-hole pairs, generated during excitation of TiO2, are unused due to the marked
recombination mechanism [2]. In addition, the radiation absorption by TiO2 is limited to UV light, which restricts

Journal of Advanced Oxidation Technologies 2018; 21(1) www.jaots.net

DOI: 10.26802/jaots.2017.0008
Journal of Advanced Oxidation Technologies 2018; 21(1)

its use in other types of irradiation of the electromagnetic spectrum. For these reasons, improving the optical,
structural and morphological properties of this semiconductor has been subject of considerable researches [3].

Works reported in the literature are being focused on different aspects, one of them is the evaluation of different
titanium precursors to synthesize TiO2 with better photocatalytic properties [4-6]. Another aspect has been the
employment of several synthesis methods. These methods include the development of new synthesis routes or the
modification of existing ones [7], such as microemulsion [8], chemical vapor deposition [9], co-precipitation [10],
sol-gel [11] and solvothermal [12, 13]. A third strategy is the doping of the photocatalyst as a way to modify the
optical properties of TiO2, either due to the decrease in band gap or the delay of the recombination phenomenon
[14].

The development of S-TiO2 materials has been studied using various titanium and sulfur precursors as well as
different methods of synthesis with the purpose of extending the optical response of TiO2 in the UV light range and
even in the visible light range. For the synthesis of these materials, the sol-gel method has been employed using
titanium tetraisopropoxide and sulfuric acid [15-18] as well as titanium butoxide and thiourea [19]. Titanium
tetraisopropoxide also has been employed with thiourea by the precipitation [20] and sol-gel [21] methods, and by
hydrolysis with sulfuric acid at room temperature [22]. S-doped TiO2 nanorods using metallic Ti and thiourea
through the oxidant peroxide method was synthesized [23] as well as from titanium tetrachloride with sulfuric acid
using the precipitation method [24] and with thiourea through the ionic liquid [25]. All these methods led to the
production of S-TiO2, however an additional S source and high temperature treatments or long process times were
required.

Some studies have used the precursor TiOSO4, which lead to establish synthesis routes of S-TiO2 with titanium and
sulfur precursors in a single compound, simplifying the synthesis method because the interaction stage between
the titanium precursor and the sulfur precursor is not required, and lead to obtain mesoporous TiO2. These studies
report that a portion of the sulfates is chemically absorbed at the TiO2 surface, forming complexes that have a
stabilizing effect of the anatase phase, which inhibit the conversion to the rutile phase, and improve the
photocatalytic activity of titanium dioxide [26-28] through a narrowing of the band gap or a higher visible light
absorption [15-17]. The methods used to obtain the S-TiO2 materials with a high degree of crystallinity from
TiOSO4 precursor required calcination processes or treatments at elevated temperatures, which increases the time
of synthesis and energy demand. Among these methods are the thermal hydrolysis method [5, 29] and composite
template route [30] with calcination temperatures between 300 and 900 °C, and the thermal treatment method
between 500 and 800 °C [31].
This study presents the synthesis of mesoporous and photoactive S-TiO2 from TiOSO4, as only Ti and S precursor,
employed an integrated sol-gel and solvothermal method that uses low temperatures to obtain materials with a
high degree of crystallinity and without requiring calcinations stages, which make it faster, efficient and
environmentally friendly. Furthermore, the integrated sol-gel and solvothermal method could facilitate the control
of properties such as morphology, size and sulfur composition, and in consequence could improve their radiation
absorption and the photocatalytic activity [32]. Different synthesis conditions (molar ratios of water/TiOSO4 (RWT)
and ethanol/TiOSO4 (RET) as well as temperature, < 250 °C, (T)) were evaluated through Box-Behnken
experimental design [33] to determine their effect on the properties and the photoactivity of the synthesized S-TiO2
materials. The samples were characterized by analyzes of XRD, Raman, XPS, XRF, ATR-FTIR, UV/vis-DRS, SEM,
TEM, porosity and BET surface area. Photocatalytic activities of the synthesized S-TiO2 materials were confirmed
by the degradation of formic acid.

Experimental and Methods

Synthesis of S-TiO2 Nanoparticles

S-TiO2 samples were prepared via an integrated sol-gel and solvothermal method at low temperatures. Titanium
oxysulfate (TiOSO4, 15% wt, Aldrich) and Mili-Q water were slowly added under continuous stirring for 2 hours at
room temperature in 60 mL of ethanol (C2H5OH, 99%, Merck) to obtain a gel. Subsequently, the gel was transferred
 Journal of Advanced Oxidation Technologies 2018; 21(1)

into a Teflon lined stainless steel reactor (autoclave) and it was heated during 2 hours at the selected low
temperature (Table 1). The solid obtained was filtered using 0.45 µ m filter and washed several times with distilled
water until pH = 4 was reached. Finally, the samples were dried at 100 °C for 1 hour and ground.
The hydrolysis and condensation reactions were performed with different amounts of water and ethanol based on
the quantity of titanium oxysulfate, represented by RWT (molar ratio water/TiOSO4) and RET (molar ratio
ethanol/TiOSO4). Table 1 shows the conditions used in each synthesis according to Box-Behnken experimental
design.
TABLE 1 BOX-BEHNKEN EXPERIMENTAL [33] DESIGN FOR SYNTHESIS OF S-TIO2 NANOPARTICLES CONSIDERING THREE VARIABLES (RWT, RET AND T)
AND THREE LEVELS IN EACH VARIABLE.

Sample RWTa RETb T (°C)c

S1 3.0 10 225
S2 4.0 10 225
S3 3.0 20 225
S4 4.0 20 225
S5 3.0 15 200
S6 4.0 15 200
S7 3.0 15 250
S8 4.0 15 250
S9 3.5 10 200
S10 3.5 20 200
S11 3.5 10 250
S12 3.5 20 250
S13 3.0 10 200
S14 3.5 15 225
S15 4.0 20 250
a molar ratio water/TiOSO4
b molar ratio ethanol/TiOSO4
c synthesis temperature

Characterization
X-ray diffraction (XRD) analyses were performed using a diffractometer Rigaku Minifles, Cu Kα radiation and 2θ
from 10° to 70°. Raman analyses were obtained in Horiba Jobin Yvon, model Labram HR 800 Microraman. BET
surface areas, adsorption-desorption curves and porosity of materials were determined using N2 physisorption at
77 K in a Micromeritics GEMINI VII/708 with a degassing temperature of 423 K. The total content of sulfur in the
samples was determined by X-ray fluorescence spectrometry (XRF) in a PANalytical Axios sequential
spectrophotometer with a rhodium source. Ultraviolet - visible diffuse reflectance (UV/vis-DRS) spectra were
obtained using a spectrophotometer Evolution 600 from Thermo Electron Corporation, with diffuse-reflectance
sphere accessory; a Labsphere spectralon was used as reference. Attenuated total reflectance - Fourier transform
infrared (ATR-FTIR) spectra was achieved in an IR-Prestige-21/8400S spectrometer provided with attenuated total
reflectance attachment. The spectrum was taken in the range of 4500–500 cm−1. X-ray photoelectron spectroscopy
(XPS) analysis was achieved in a SPECS spectrometer equipped with a PHOIBOS analyzer and equipped with
aluminum X-ray sources. The energy of the analyzer was 60 eV. Scanning electron micrographs (SEM) were
obtained in a scanning electron microscope (JEOL JSM 6490LV) provided with a BS-SE-detector using image
analysis. Transmission electron micrographs (TEM) were obtained in a transmission electron microscopy (Tecnai
G2 F20) using a FP 5360/22 HAADF detector for 120/200 kV. The statistical analysis was determined using the
experimental data and the Statgraphics Centurion XVII computer program.

Photocatalytic Tests

Photocatalytic activity of materials was evaluated on degradation of formic acid (HCOOH, 98%, Merck) (FA).
Journal of Advanced Oxidation Technologies 2018; 21(1)

Experiments were performed in a borosilicate reactor, which was placed on a plate with constant stirring. UV
irradiation lamp Nippon 3U (26 W), emission range 300 nm - 400 nm with a maximum in 350 nm or visible
irradiation lamp ElectroXX YGP-20 (20 W), emission range 400 nm - 700 nm with main maxima in 484 nm and 542
nm, were used. The lamps were located at the top of the reactor, inside of a concave bell shape of aluminum in
order to irradiate the solution with a homogenous intensity in all points. Experiments were made at room
temperature. The intensity of incident radiation on the solution was measured with a Solar Light Co. PMA 2200
photometer and it was determined in 80 W/m2 (UVA light) and 60 W/m2 (visible light). For photocatalytic
experiments, 30 mg of photocatalyst was dispersed in 60 mL of an aqueous solution of 200 mg/L FA. The
suspension was magnetically stirred in the dark for 1 h, at which the adsorption equilibrium was reached. The
progress of the photocatalytic degradation of FA was monitored by conductivity measurements (HQ40d
multiparameter, HACH) every 60 minutes until four hours of photodegradation. Photolysis of FA was performed
during the same four hours.

Results and Discussion

XRD and Raman Analysis

Figure 1 shows the XRD diffractograms of the synthesized S-TiO2 nanoparticles. All materials crystallized as
anatase phase showing the characteristic XRD peaks of this phase at 25.281°, 37.800°, 48.049°, 53.890°, 55.060° and
62.688°, corresponding to the planes (101), (004), (200), (105), (211) and (204), respectively [14, 34]. The synthesis
method did not lead to the formation of the rutile or brookite phases in the prepared materials, since the most
characteristic peaks of those phases were not found [34-35]. This behavior is attributed to the temperatures used in
the integrated method that are low to contribute to the formation of such phases, however, they are enough to
obtain materials with a high degree of crystallinity in anatase phase without requiring calcination processes as in
other synthesis methods like thermal hydrolysis [5], sol-gel [36] and thermal treatment [31].

FIG. 1 XRD PATTERNS OF THE SYNTHESIZED S-TIO2 MATERIALS

The temperature was the synthesis condition with higher influence on the degree of crystallinity of the synthesized
S-TiO2 materials. S-titanium dioxides synthesized at 250 °C (S7, S8, S11, S12 and S15) showed peaks higher and
narrower indicating a higher formation of TiO2 crystals in anatase phase and a reduction of the amorphous
material compared to the samples obtained at lower temperatures. Materials synthesized a 200 °C as S5, S6, S9, S10
and S13 samples, showed peaks with lower intensity and wider. These differences in the materials obtained with
variations in temperature of only 50 °C reveals the impact of the integrated sol-gel and solvothermal method in the
S-TiO2 crystallization, given the particular environment generated by the thermal process in the closed system. For
detailed verification of the crystallinity degree of the synthesized S-TiO2 materials, Raman analysis was performed.
This analysis is precise to determine crystallinity and microstructure; the Raman peaks are shown lower and wider
in samples with small imperfections [37]. The Raman spectrum of anatase single crystal has six allowed modes in
144 cm-1 (Eg), 197 cm-1 (Eg), 399 cm-1 (B1g), 513 cm-1 (A1g), 519 cm-1 (B1g), and 639 cm-1 (Eg) [38].
 Journal of Advanced Oxidation Technologies 2018; 21(1)

The main Eg Raman mode, which appears at 144 cm-1, is defined in all synthesized S-TiO2 materials, indicating that
the anatase phase is present in all of them. However, the peaks show different intensities and full widths at half
maximum (linewidths) revealing small differences in the degree of perfection or crystallinity of the materials,
where the higher and narrower peaks correspond to the samples synthesized at the temperature of 250 °C (Figure
2a). These differences are also observed when analyzing the mode Eg at 197 cm-1 (Figure 2b), which does not appear
in the Raman spectrum of the materials that were synthesized using the temperature of 200 °C, indicating possibly
some imperfections in their crystalline net compared with those obtained at temperatures of 225 °C and 250 °C [37].

FIG. 2 RAMAN SPECTRA OF THE EG MODE IN THE SYNTHESIZED S-TIO2 MATERIALS. A) MAIN EG MODE, B) EG MODE AT 197 CM-1

Raman peaks associated with the brookite [39] or rutile [40] phases are not observed. In addition, no peaks
associated with TiOSO4 (238 cm-1 and 782 cm-1) were observed indicating that the precursor was completely
decomposed. This last fact shows how the process of the integrated method contributes to reach highly energetic
states because this method led to the complete TiOSO4 decomposed even at 200 °C while under atmospheric
pressure it is necessary to reach temperatures close to 600 °C [31].
Figure 2a shows a shift in the main Eg Raman mode of anatase, which varies between 143 cm-1 and 147.5 cm-1,
where the samples that were synthesized at 200 °C showed larger frequency shift towards higher wavenumbers,
followed by samples that were synthesized at 225 °C. Zhang et al. attribute this behavior to a combined mechanism
involving both phonon confinement and non-stoichiometry effects and possible interfacial vibrations that increase
with the decreasing in crystallite size contributing to asymmetric broadening of Raman peak [37]. This behavior
can also be inferred from Figure 3 that shows an increasing of linewidth of the main Raman Eg mode with a
decreasing in the used synthesis temperature (Figure 3a), which represent materials with local lattice imperfections
[38], and a reduction in the crystallite size (Figure 3b) [41].

FIG. 3 LINEWIDTH OF THE MAIN RAMAN EG MODE OF THE SYNTHESIZED S-TIO2 MATERIALS ACCORDING WITH THE USED
SYNTHESIS TEMPERATURE (A) AND WITH THE CRYSTALLITE SIZE (B)
Journal of Advanced Oxidation Technologies 2018; 21(1)

Crystallite Size Analysis

The crystallite size (L) of samples was calculated using the Debye-Scherrer equation (eq. 1) at λ = 0.15418 nm, K =
0.9, β = full width at half maximum, 2θ = 25.281°, and the values are shown in Table 2. An increase in the synthesis
temperature led to a growth in the crystallite size. Therefore, for two samples with same RWT = 4.0 and RET = 15,
but with different temperature 200 °C (S6) and 250 °C (S8), the crystallite size found was L = 25.41 nm and L = 42.34
nm, respectively. This is consistent with the results of other authors using different methods of synthesis, sol-gel
[16] or composite template route [30], where an increase of temperature led to an increase in crystal size of the
anatase phase.
TABLE 2 CRYSTALLITE SIZE (L) (NM), BET SURFACE AREA (SBET) (M2/G), BAND GAP (BG) AND % S (XRF) (WT %)

Sample(T) La SBET BGb %S

S1(225) 31.46 61.32 3.06 1.330

S2(225) 25.73 106.39 3.03 1.010
S3(225) 38.44 59.03 3.08 1.070
S4(225) 32.69 71.59 3.11 1.100
S5(200) 24.19 88.95 3.09 1.210
S6(200) 25.41 87.05 3.03 1.300
S7(250) 42.97 31.17 3.06 0.586
S8(250) 42.34 35.77 3.03 0.818
S9(200) 22.73 84.50 3.02 1.130
S10(200) 28.77 68.54 2.99 1.050
S11(250) 41.94 27.54 3.10 0.741
S12(250) 44.05 31.86 2.92 0.689
S13(200) 25.00 79.80 3.07 1.260
S14(225) 34.80 27.22 2.96 0.684
S15(250) 46.22 25.98 3.04 0.545
a obtained by Scherrer equation
b obtained by Kubelka-Munk function

Anatase particle sizes from 2 nm to 7 nm were obtained from titanium oxysulfate, water and ethanol, applying a
larger RWT > 200 and a different synthesis method consisted of a reflux of the components for 2 to 8 h, at room
temperature [42]. The integrated sol-gel and solvothermal method yielded greater crystallite sizes and high degree
of crystallinity using lower amounts of water for hydrolysis, shorter time of synthesis, and low temperatures 
250 °C.

(1)

Although the main effect on the crystal growth of the samples is given by the temperature, an analysis at the same
temperature indicates that the crystal growth showed a dependence of the RET ratio, enhancing with the increase
of RET, Figure 4. This behavior can be attributed to the amount of solvent because its properties such as dielectric
constant and saturate vapor pressure [15] could contribute to control the molecular mobility or atomic diffusion
during the different stages of synthesis. Therefore, for two samples with the same RWT = 3.0 and T = 225 °C, but
with different RET, 10 (S1) and 20 (S3), the crystallite size found was L = 31.46 nm and L = 38.44 nm, respectively.

An analysis of variance with the crystallite size as response factor and the used synthesis conditions (RWT, RET
and temperature) as variables, showed that the variables with statistical significance were temperature and RET,
which have a probability value (P-value) [43] lower than 0.05, corroborating the above analysis. This statement also
can be evidenced in Figure 5, which shows the standardized Pareto chart [43], performed to determine the effects
of each variable and their interactions on the response factor, which are considered significant when exceed the
vertical reference line. The statistically significant effects were due to RET and temperature variables, since their
bars past the vertical reference line reaching a confidence interval of 95%, and additionally both effects are positive
 Journal of Advanced Oxidation Technologies 2018; 21(1)

contributions to the crystallite size. The response variable was affected to a lesser degree by the variable RWT and
the interactions between the variables (RWT-RET, RWT-T, RET-T), which did not have a statistically significant
effect. This result is according with the experimental where slight changes in crystallite size was due to changes in
RWT, which could be attributed that water quantity used in the synthesis were higher than required by
stoichiometry for complete hydrolysis, avoiding the presence of un-hydrolyzed precursor and their steric
hindrance that could prevent crystallization of the anatase phase [44].

FIG. 4 CRYSTALLITE SIZE OF THE SYNTHESIZED S-TIO2 MATERIALS OBTAINED WITH DIFFERENT RET AND TEMPERATURES

FIG. 5 STANDARDIZED PARETO CHART OF THE STATISTICAL ANALYSIS OF THE CRYSTALLITE SIZE OF THE SYNTHESIZED S-TIO2
MATERIALS

The statistical analysis generated a model to describe the response factor according with the applied variables to
predict in our study the crystallite size in function of RWT, RET and T, eq. 2, where the values of the variables are
specified in their original units. The R-squared statistic indicates that the model has 96.9272% of adjustment of
crystallite size. The adjusted R-squared statistic, which is more suitable for comparing models with different
numbers of independent variables, is 91.3961%.
Crystallite size = -141.07 + 52.7425*RWT + 3.57275*RET + 0.1592*T - 6.73*RWT^2 - 0.002*RWT*RET - 0.037*RWT*T -
0.0415*RET^2 - 0.00786*RET*T + 0.000976*T^2 (2)

The crystallite sizes reached in this work for all synthesized materials are larger than those obtained by other
authors, who obtained materials with crystallite sizes less than 20 nm, using the sol-gel method [36] and thermal
hydrolysis [5], and calcinations between 300 °C and 600 °C with the titanium oxysulfate as precursor. This behavior
could be attributed to the integrated sol-gel and solvothermal method that allowed obtaining crystals of anatase
more defined at lower temperatures thanks to the atmosphere generated in the closed system. Additionally, the
Journal of Advanced Oxidation Technologies 2018; 21(1)

crystallite sizes achieved in this work are similar to the reported ones for samples that required production times
around 60 hours and were calcinated at temperatures upper than 800 °C [36]; this indicated that the integrated
method significantly reduces energy consumption and reduces the time required for the S-TiO2 synthesis.

Surface Area and Porosity Analysis

The BET surface areas calculated to the materials prepared by the integrated sol-gel and solvothermal method were
lower compared with those obtained by other authors using the thermal hydrolysis and sol-gel methods with
calcinations temperatures until 500 °C [5, 36]. The lower surface areas in materials obtained by the integrated
method can be associated with their higher degree of crystallinity and crystallite size achieved at lower
temperatures.

The surface areas were lower in materials prepared at high synthesis temperatures, due to the growing of the
crystal and the generation of agglomerations of particles that led to decrease the surface. This inverse behavior has
been demonstrated by several authors, in materials prepared using different titanium precursors and methods [5,
16, 19]. In this sense, materials synthesized at 250 °C have areas smaller than 37 m2/g while those prepared at
200 °C have the largest areas, with values greater than 68 m2/g.
Figure 6a shows the adsorption-desorption isotherms of the synthesized S-TiO2 materials S4 (225 °C), S6 (200 °C)
and S8 (250 °C). All materials showed type IV isotherms, which are representative of mesoporous materials.
Hysteresis loop appears at a higher relative pressure for the materials synthesized at 225 °C (S4) and 250 °C (S8)
and their profiles suggest aggregates of particles forming slit-like pores with narrower distribution of relative
uniformity [45]. The isotherm corresponding to the materials synthesized at 200 °C showed a wider hysteresis cycle,
evidencing a slightly more pronounced plateau that probably indicates more complex pore networks consisting of
poorly defined pores due to open and partially blocked mesoporous with a large pore size distribution [45]. The
pore volume of the S-TiO2 materials was reduced in samples synthesized at higher temperatures as can be
observed in Figure 6b, where a reduction of volume pore is produced by the increment in the synthesis
temperature, especially between 200 °C and 225 °C, and particularly a lost in pores with lower widths was
observed. This change is attributed to pore wall crystallization or the downfall of the mesoporous structure due to
particle over-growth [30-31].

FIG. 6 NITROGEN ISOTHERMS (A) AND PORE SIZE DISTRIBUTION (B) OF SYNTHESIZED MESOPOROUS S-TIO2 MATERIALS
PREPARED WITH DIFFERENT SYNTHESIS TEMPERATURE: 200 °C (S6), 225 °C (S4) AND 250 °C (S8)

XRF Analysis

The total amount of remaining sulfur (wt % S) was determined by XRF and is shown in Table 2. Presence of sulfur
is due to the nature of the precursor (TiOSO4). The amount of the incorporated sulfur is associated with the used
synthesis temperature in which the materials synthesized at 250 °C had lower sulfur content. Conversely, materials
synthesized at lower temperature (200 °C) showed higher amounts of sulfur. The literature reports that a high
content of chemically absorbed sulfur (in SO42- ions) reduces the degree of crystallinity of the anatase phase [27].
 Journal of Advanced Oxidation Technologies 2018; 21(1)

However, the presence of sulfur stabilizes the anatase phase to temperatures above of 700 °C [31] as well as an
increase in the synthesis temperature reduces the sulfur content and improves the degree of crystallinity of the
anatase phase [27]. In addition, small amounts of sulfur could be promoting the material absorption in the visible
region and the band gap decreasing [17, 19, 31]. For instance, S8 and S15 materials that have low sulfur content
(0.818 and 0.545, respectively) exhibited higher absorption in the visible range (Figure 7), besides S12 and S14
samples, which present low sulfur contents of 0.689 and 0.684 had low band gap values of 2.92 and 2.96,
respectively. For the synthesized S-TiO2 samples, no strong correlation was observed between the sulfur content
and the reached degree of crystallinity as well as in the generated crystallite size. However a certain trend can be
observed in Figure 8a where the amount of remaining sulfur was higher in materials with higher Raman
linewidths of the main Raman Eg mode, which as indicated above correspond to those with lower crystallinity, and
in Figure 8b where materials with larger crystallite size had lower sulfur contents.

FIG. 7 UV/VIS-DRS SPECTRA OF THE SYNTHESIZED S-TIO2 MATERIALS

FIG. 8 SULFUR CONTENT IN THE SYNTHESIZED S-TIO2 MATERIALS ACCORDING WITH THEIR A) RAMAN LINEWIDTHS OF THE
MAIN RAMAN EG MODE AND B) CRYSTALLITE SIZES

UV/vis-DRS Analysis
Absorption spectra of all synthesized S-TiO2 materials (Figure 7) have a high absorption in the ultraviolet region
(200-400 nm), confirming the presence of titanium dioxide, and absorption in the visible region (400-800 nm)
associated to the sulfur presence [17, 31]. Moreover, the samples synthesized at 250 °C (S7, S8, S11, S12 and S15)
had the highest absorption in the ultraviolet and the visible light ranges, being the samples S8 and S15 the
materials with the highest visible light absorption which correspond to the photocatalysts with lower sulfur
content. This result indicates that although the presence of sulfur allows the absorption of visible radiation, this
Journal of Advanced Oxidation Technologies 2018; 21(1)

absorption is not directly proportional to the sulfur content since the small amounts are those that favor the greater
absorption of the visible radiation.

The Kubelka-Munk function was used to calculate the band gap energies of the synthesized materials and the
values obtained were summarized in Table 2. The band gap energies of the S-TiO2 materials, synthesized by the
integrated sol-gel and solvothermal method, were lower than the 3.2 eV reported to anatase [46], results that are
associated with sulfur content. These values indicate a narrowing of the energy band gap with respect to pure TiO2
anatase, and are similar to those reported for materials synthesized from the same precursor using the thermal
hydrolysis method or thermal treatment method with high calcination temperatures [5, 31]. This result shows how
the integrated method leads to obtain photoactive materials with optical response in the visible light range,
without requiring high calcination temperatures, leading to lower energy consumption.

ATR-FTIR and XPS Analysis

ATR-FTIR results show that all samples synthesized from titanium oxysulfate have a similar spectrum that shown
in Figure 9a (S8 material), indicating that the photocatalyst TiO2 has been obtained. The presence of polymer chains
of Ti-O-Ti bond as a result of condensation is observed between 800 cm-1 - 500 cm-1 [11, 47]. The band between 3600
cm-1 and 3000 cm-1 corresponds to the vibrations of -OH groups on the surface of S-TiO2 materials [11, 47].
Moreover, a band located at 1640 cm-1 is attributed to the bending vibration of superficial adsorbed water
molecules [47]. Approximately at 1125 cm-1 and 1050 cm-1 are located the bands corresponding to stretches of
sulfate and/or sulfite groups [15, 48], which are associated with S-O vibration [16].

The XPS analyzes in Figure 9b corresponding to S8 sample, which has high crystallinity, low sulfur content and
greater absorption of visible radiation, shown the binding energies of Ti 2p3/2 and Ti 2p1/2 at 459.63 eV and 465.4 eV,
respectively. These shifts to higher values of binding energies with respect to those corresponding of pure TiO2
have been reported as a strong interaction between sulfate anion and the titanium cation with an increased positive
polarity on the titanium cation [15, 49]. Figure 9c shows the peak S 2p, where the value between 167 eV and 170 eV
corresponds to S-O bonds in SO42- species, indicating an oxidation state of sulfur (S+6/S+4) [17, 19, 49]. This peak is
often assigned to the replacement of Ti+4 by sulfur ions or the presence of sulfate/sulfite groups coordinated on the
titanium dioxide [15, 50, 51]. This agrees with the infrared analysis where stretches assigned to sulfates and/or
sulfites were found. No peaks were found between 160 eV - 162 eV, assigned to the presence of S-2. According to
some authors [15, 50, 51], this result was to be expected because the ionic radius of S-2 (1.7 angstroms) is much
larger compared to the O-2 radius (1.22 angstroms), requiring more energy to form the Ti-S bond that to substitute
Ti+4 for S+6 or S+4.

FIG. 9 A) ATR-FTIR CHARACTERISTIC SPECTRA OF THE SYNTHESIZED S-TIO2 MATERIALS (S8), B) TI 2P AND C) S 2P XPS SPECTRA OF
S8

SEM and TEM Analysis

SEM images of powders presented in Figure 10 indicate that the materials synthesized by the integrated sol-gel and
solvothermal method reached a spherical shape; however, different particle morphology was found when the
synthesis temperature was modified. Therefore, materials obtained at 200 °C (Figure 10a) show a smooth surface
conformed by agglomerates with sizes ranging from 0.5 µ m to 1 µ m, while materials synthesized at 225 °C (Figure
 Journal of Advanced Oxidation Technologies 2018; 21(1)

10b) have a rough surface in agglomerates of approximately 1 µ m of size. Materials obtained at 250 °C (Figure 10c)
have also a rough surface but the spherical shape of particles begins to change into larger particles of sizes between
1 µ m and 3 µ m due to the growth of TiO2 crystallites and their agglomeration. An important influence of the
synthesis temperature on particle surface and its morphology was evidenced, corroborating the results found in
surface area and crystallite size of the synthesized materials.

FIG. 10 SEM IMAGES OF THE SYNTHESIZED S-TIO2 MATERIALS AT A) 200 °C (S6), B) 225 °C (S4) AND C) 250 °C (S8)

Figure 11 corresponds to TEM micrographs of the synthesized material S8. The tetragonal anatase shape is
presented with some distortions probably caused by the presence of sulfur in the structure like was observed in the
XPS analysis. The crystallite size was around 49.11 nm, which is similar to the results calculated by X-ray
diffraction for the same material (42.34 nm). HRTEM micrograph shows the lattice fringes of the synthesized
S-TiO2 material S8 corresponding for the (101) plane (d = 0.35 nm) of anatase [5, 31] indicating its ordered
crystalline structure. This analysis confirms the results obtained in the XRD and Raman analysis.

FIG. 11 TEM IMAGES (A AND B) AND HRTEM IMAGE (C) OF THE SYNTHESIZED S-TIO2 MATERIAL S8

Photocatalytic Activity

Results of UV photocatalytic degradation of formic acid (FA) showed in Figure 12 confirm that all synthesized
materials are photoactive. The photolysis of formic acid was only 2 %. A trend in results is evidenced where the
best photocatalytic results correspond to the materials synthesized at the higher temperature, which main
characteristic are their high degree of crystallinity showed in the XRD and Raman analyses, larger crystallite sizes
and lower sulfur content.

An analysis of variance applied with the UV photocatalytic degradation of FA as response factor showed that the
variable with statistical significance was the synthesis temperature, which have a P-value [43] lower than 0.05,
corroborating the experimental results. This also can be evidenced in Figure 13a, which shows the standardized
Pareto chart [43] for the performed statistical analysis, where temperature variable reached a confidence interval
(95%) with positive contribution to generate materials that allow a greater UV photocatalytic degradation of FA.
The RET-RET interaction, from the statistical point of view, constitutes a significant adjustment to the predictive
equation model provided by statistical software but does not represent a physical meaning in the synthesis
conditions.
Journal of Advanced Oxidation Technologies 2018; 21(1)

FIG. 12 PHOTOCATALYTIC ACTIVITY OF THE SYNTHESIZED S-TIO2 MATERIALS IN THE DEGRADATION OF FORMIC ACID UNDER
UV RADIATION

On the other hand, Figure 13b shows the effects of the variables independently indicating that the UV
photodegradation of FA increased with the increasing of the synthesis temperature of the S-TiO2 materials, as
stated above. The increase in the RWT synthesis condition also contributes to the obtaining of materials with
greater activities for the degradation of FA, while the RET synthesis condition showed a maximum around 15 to
synthesize the materials that achieved greater photodegradation of FA.

FIG. 13 STATISTICAL ANALYSIS OF THE UV PHOTOCATALYTIC DEGRADATION OF FA. a) STANDARDIZED PARETO CHART, b)
MAIN EFFECTS PLOT

The statistical analysis generated a model to describe the response factor according with the applied variables to
predict the UV photocatalytic degradation of FA using the synthesized S-TiO2 materials in function of their RWT,
RET and T synthesis conditions, eq. 3, where the values of the variables are specified in their original units. The
R-squared statistic indicates that the model has 91.3899% of the adjustment of the UV photocatalytic degradation of
FA. The adjusted R-squared statistic, which is more suitable for comparing models with different numbers of
independent variables, is 75.891%.

UV photocatalytic degradation FA = -446.625 + 15.0*RWT + 14.725*RET + 2.615*T - 14.5*RWT^2 + 0.2*RWT*RET +

0.4*RWT*T - 0.495*RET^2 - 0.002*RET*T - 0.0078*T^2 (3)

In applying eq. 3 to determine the optimum values to maximize the response variable, in our case the UV
photocatalytic degradation of FA, the statistical software find that its optimal value is 61.6405% and that the
appropriate S-TiO2 synthesis conditions would be RWT = 4.0, RET = 15.176 and T = 250 °C. These synthesis
conditions are similar to those used to obtain experimentally the S8 sample.
The photocatalytic activity was conditioned by the sulfur amount, which was higher in materials with lower sulfur
content. This is in agreement with results found in the literature where the presence of sulfur in low contents
improves the activity of TiO2 [31]. In addition, high sulfur contents may generate a screening effect on active sites
 Journal of Advanced Oxidation Technologies 2018; 21(1)

or recombination centers influencing the photoactivity [50] or may contribute to a higher acid character of the
surface limiting the interaction between the formic acid and the photocatalyst. On the other hand, materials with
highest photocatalytic activity were those of lower surface area (less than 40 m2/g), which can be attributed to their
higher degree of crystallinity.

According with the results, the highest photodegradation (67%) of formic acid was reached with sample S8, which
corresponds to the synthesis configuration RWT = 4.0, RET = 15 and T = 250 °C. This material reached a high degree
of crystallinity, a mesoporous structure with crystallite size of 42.34 nm, high UV and visible radiation absorption,
surface area of 35.77 m2/g and sulfur content of 0.818 wt %. This S-TiO2 material was evaluated in the
photodegradation of formic acid under visible light radiation, for which the reactor was covered with a polyester
filter that avoid the UV radiation and irradiated with a visible irradiation lamp ElectroXX YGP-20 (20 W). The
material exhibited activity under visible light leading to 11.2% of formic acid degradation, result that can be
attributed to its physicochemical properties and principally by the presence of sulfur.

Conclusions

Photoactive S-TiO2 materials were successful synthesized by the integrated sol-gel and solvothermal at low
temperatures method and using TiOSO4 as only Ti and S precursor. The materials were obtained in shorter time,
just five hours, and using low energy, making its synthesis by the integrated method a more useful and
environmentally friendly process. The temperature in the applied integrated method was the synthesis condition
with higher influence on the crystallite size, degree of crystallinity, sulfur content, visible absorption and
photocatalytic activity of the synthesized S-TiO2 materials. The integrated method allows to obtain mesoporous
S-TiO2 materials crystallized as anatase phase with UV and visible light absorption, which are associated with the
sulfur content. The synthesis conditions RWT = 4.0, RET = 15 and T = 250 °C led to obtain the highest photoactive
material, which reached the best properties configuration as high degree of crystallinity and visible radiation
absorption, crystallite size of 42.34 nm, band gap of 3.03, surface area of 35.77 m2/g and sulfur content of 0.818 wt %.
The statistical analysis yielded predictive models that were very close to the experimental results and allowed to
corroborate the influence of the synthesis variables on the crystallite size and the photocatalytic activity of the
synthesized S-TiO2 materials.

ACKNOWLEDGEMENT

This work was supported by Universidad de Antioquia through the project PRG14-1-03.

REFERENCES

[1] Hidalgo MC, Sakthivel S, Bahnemann D. Appl. Catal., A. 2004;277:183–189.

[2] Serpone N, Lawless D, Khairutdinov R, Pelizzetti E. J. Phys. Chem. 1995;99:16655–16661.
[3] Chen L, Zhu J, Liu YM, Cao Y, Li HX, He HY, Dai WL, Fan KN. J. Mol. Catal. A: Chem. 2006;255:260–268.
[4] Loryuenyong V, Angamnuaysiri K, Sukcharoenpong J, Suwannasri A. Ceram. Int. 2012;38:2233–2237.
[5] Ngamta S, Boonprakob N, Wetchakun N, Ounnunkad K, Phanichphant S, Inceesungvorn B. Mater. Lett. 2013;105:76–79.
[6] Colón G, Maicu M, Hidalgo MC, Navío JA, Kubacka A, Fernández-García M. J. Mol. Catal. A: Chem. 2010;320:14–18.
[7] Žunič V, Vukomanović M, Škapin SD, Suvorov D, Kovač J. Ultrason. Sonochem. 2014;21:367–375.
[8] Khomane RB. J. Colloid Interface Sci. 2011;356:369–372.
[9] Hou J, Yang X, Lv X, Huang M, Wang Q, Wang JJ. Alloys Compd. 2012;511:202–208.
[10] Yeh SW, Ko HH, Chiang HM, Chen YL, Lee JH, Wen CM, Wang MC. J. Alloys Compd. 2014;613:107–116.
[11] Tripathi AK, Singh MK, Mathpal MC, Mishra SK, Agarwal A. J. Alloys Compd. 2013;549,114–120.
[12] He F, Li J, Li T, Li G. Chem. Eng. J. 2014;237:312–321.
[13] Vargas X, Tauchert E, Marin JM, Restrepo G, Dillert R, Bahnemann D. J. Photochem. and Photobiol., A. 2012;243:17–22.
[14] Jaiswal R, Bharambe J, Patel N, Dashora A, Kothari DC, Miotello A. Appl. Catal., B. 2015;168-169:333–341.
Journal of Advanced Oxidation Technologies 2018; 21(1)

[15] Han C, Andersen J, Likodimos V, Falaras P, Linkugel J, Dionysiou DD. Catal. Today. 2014;224:132–139.
[16] Nishikiori H, Hayashibe M, Fujii T. Catalysts. 2013;3:363–377.
[17] Han C, Pelaez M, Likodimos V, Kontos A, Falaras P, O’Shea K, Dionysiou DD. Appl. Catal., B. 2011;107:77–87.
[18] Colón G, Hidalgo MC, Munuera G, Ferino I, Cutrufello MG, Navío JA. Appl. Catal., B. 2006;63:45–59.
[19] McManamon C, O’Connell J, Delaney P, Rasappa S, Holmes J, Morris MJ. Mol. Catal. A: Chem. 2015;406:51–57.
[20] Ohno T, Akiyoshi M, Umebayashi T, Asai K, Mitsui T, Matsumura M. Appl. Catal., A. 2004;265:115–121.
[21] Lin YH, Hsueh HT, Chang ChW, Chu H. Appl. Catal., B. 2016;199:1–10.
[22] Yamazaki S, Fujinaga N, Araki K. Appl. Catal., A. 2001;210:97–102.
[23] Bakar S, Riberio C. J. Mol. Catal. A: Chem. 2016;412:78–92.
[24] Hussain S, Khan K, Hussain R. J. Nat. Gas Chem. 2009;18:383–391.
[25] Chen X, Kuo DH, Lu D. Adv. Powder Technol. 2017;28:1213–1220.
[26] Zhang D, Wang JJ. Water Process Eng. 2015;7:187–195.
[27] Zeng F, Luo D, Zhang Z, Liang B, Yuan X, Fu L. J. Alloys Compd. 2016;670:249–257.
[28] Murcia JJ, Hidalgo MC, Navío JA, Araña J, Doña-Rodríguez JM. Appl. Catal., B. 2015;179:305–312.
[29] Yang G, Ding H, Chen D, Ao W, Wang J, Hou X. Appl. Surf. Sci. 2016;376:227-235.
[30] Tian C, Zhang Z, Hou J, Luo N. Mate. Lett. 2008;62:77-80.
[31] Xing Z, Li Z, Wu X, Wang G, Zhou W. Int. J. Hydrogen Energy 2016;41:1535-1541.
[32] He F, Ma F, Li T, Li G. Chin. J. Catal. 2013;34:2263–2270.
[33] Myers RH, Montgomery DC, Anderson-Cook ChM. Response surface methodology: Process and product optimization
using designed experiments, 3rd ed. New Jersey: John Wiley & Sons, 2009:317-320.
[34] You YF, Xu CH, Xu SS, Cao S, Wang JP, Huang YB, Shi SQ. Ceram. Int. 2014;40:8659–8666.
[35] Bellardita M, Di Paola A, Megna B, Palmisano L. Appl. Catal., B. 2017;201:150-158.
[36] Sivakumar S, Pillai PK, Mukundan P, Warrier KGK. Mater. Lett. 2002;57:330–335.
[37] Zhang WF, He YL, Zhang MS, Yin Z, Chen Q. J. Phys. D: Appl. Phys. 2000;33:912–916.
[38] Choi HCh, Jung YM, Kim SB. Vib. Spectrosc. 2005;37:33–38.
[39] Iliev MN, Hadjiev VG, Litvinchunk AP. Vib. Spectrosc. 2013;64:148-152.
[40] Ma HL, Yang JY, Dai Y, Zhang YB, Lu B, Ma GH. Appl. Surf. Sci. 2007;253:7497–7500.
[41] Rajender G, Giri PK. J. Alloys Compd. 2016;676:591-600.
[42] Iwasaki M, Hara M, Ito S. J. Mater. Sci. Lett. 1998;17:1769-1771.
[43] Bei D, Marszalek J, Youan BBC. AAPS PharmSciTech. 2009;10(3):1040-1047.
[44] Wang G. J. Mol. Catal. A: Chem. Ref. Data 2007;274:185–191.
[45] Thommes M, Kaneko K, Neimark KV, Oliver JP, Rodriguez-Reinoso F, Roquerol J, Sing KSW. Pure Appl. Chem.
2015;87(9-10):1051-1069.
[46] Carp O, Huisman CL, Reller A. Prog. Solid State Chem. 2004;32:33–177.
[47] Golobostanfard MR, Abdizadeh H. Physica B. 2013;413:40–46.
[48] Selishchev D, Kozlov D. Molecules. 2014;19:21424–21441.
[49] Raj KAJ, Shanmugam R, Mahalakshmi R, Viswanathan B. Indian J. Chem. 2010;49A:9-17.
[50] Rengifo-Herrera JA, Kiwi J, Pulgarin C. J. Photochem. Photobiol., A. 2009;205:109–115.
[51] Ho W, Yu JC, Lee S. J. Solid state chem. 2006;179:1171-1176.