Influence of Phases Content on Pt/TiO2, Pd/TiO2 Catalysts for Degradation of 4-Chlorophenol at Room Temperature

1CICATA Altamira, Instituto Politécnico Nacional, Km 14.5, Carretera Tampico-Puerto Industrial Altamira, 89600 Altamira, TAMPS, Mexico 2División Académica Multidisciplinaria de Jalpa de Méndez, Universidad Juárez Autónoma de Tabasco, Carretera Estatal Libre VHS-COM Km 27+000 s/n Ranch. Ribera Alta, 86205 Jalpa de Méndez, TAB, Mexico 3CIITEC, Instituto Politécnico Nacional, Cerrada de Cecati s/n, Colonia Santa Catarina, 02250 Azcapotzalco, DF, Mexico 4Instituto Tecnológico de CiudadMadero, Avenida Primero deMayo s/n, Colonia LosMangos, 89440CiudadMadero, TAMPS,Mexico 5Division Académica de Ingenieŕıa y Arquitectura, Universidad Juarez Autónoma de Tabasco (UJAT-CICTAT), Carretera Cunduacan-Jalpa Km 1, Colonia La Esmeralda, 86690 Cunduacán, TAB, Mexico


Introduction
The current world population growth has provoked a big pollution problem for the overexploitation of natural resources and continuous use of toxic substances applied in the production of many materials.In this way, chlorophenols (CPs) are extensively used in the synthesis of pesticides, herbicides, and dyes and as preservative agents for wood, paints, fibers, and leather [1].Most of the CPs have been listed as priority pollutants due their acute toxicity, resistance to biodegradability, and strong bioaccumulation potential [2].Therefore, removal or degradation of hazardous materials and contaminants from wastewater, ground, or surface water is a significant global challenge; many methods including biological, chemical, or physical methods have been developed to make the water free of toxic chemicals and pathogens [3].
The use of semiconductors as photocatalysts has emerged as promising method in the area of catalysis [4].Among the semiconductors used, titanium dioxide (TiO 2 ) is one of the most suitable material due to it showing four crystal structures: anatase, rutile, brookite, and TiO 2 () [5,6].The functional properties depend on the phase and combining them could achieve synergic effects [7,8].In addition, TiO 2 shows high photocatalytic activity, good photostability, nontoxicity, and low price.However, its main drawbacks in effective commercial applications are the agglomeration of ultrafine powders, which prompt an adverse effect on the catalyst performance as well as its band gap (3.2 eV) that avoids the use of sunlight [9].In this way, many strategies have been developed to overcome these limitations, such as synthesis of compounds with crystal phase combinations [10][11][12] and noble metal doping [4,9].Also, some researches have reported the effect of tridoped and quaternary doped in anatase-TiO 2 nanophotocatalysts [13,14].
It has been showed that the incorporation of noble metals (Ag, Ni, Cu, Au, Rh, Pt, and Pd) [9,15] into the TiO 2 lattice improves the photocatalytic properties of this compound.Pt and Pd ions have shown excellent efficiency in diverse applications, such as hydrogenation of aromatic hydrocarbons, electrooxidation of small organic molecules with high sulfur [16], CO reduction [17], and in the degradation of several contaminants [18].On the other hand, it is important to consider the preparation procedures as well as the optimal conditions to achieve high quality catalyst.So far, the solgel method has demonstrated to be an effective technique for the TiO 2 lattice ion incorporation [16].Our previous report has evidenced that at 400 ∘ C the addition of different Pd amounts can modulate the photocatalytic performance of the Pd-doped TiO 2 materials [19].The present study was thus undertaken in order to investigate the structure and catalytic performance of Pt-doped TiO 2 composites and compared with the Pd-doped TiO 2 at 500 ∘ C. The effect of phase composition with temperature was also analyzed.The conditions used during the catalysts synthesis favored the incorporation of the noble metals into the anatase lattice and the different amounts of noble metal allowed obtaining diverse percentages of anatase and rutile mainly.The photocatalytic activity of the compounds was evaluated in the degradation of 4-chlorophenol under UV light.

Characterization Methods.
Powder XRD data for the thermal treated samples were acquired on a Bruker D8 advance machine with a Cu K radiation and equipped with a Linxeye detector.Data were collected in the 2 range 20-100 ∘ with a step size of 0.017 ∘ and a step time of 21.4 s.Rietveld refinement was performed for further analysis using TOPAS 3.0 software and the representation of the refined structures was realized by the Crystalmaker 2.5 software.Phase characterization was also realized by Raman spectra measurements using a Spex 1043 double monochromator (Edison, New Jersey) with 514.5 nm line of argon laser (lexel laser) at a power level of 40 mW (800-100 cm −1 ).Morphological features of the as-prepared samples were observed with the scanning electron microscopes (FEI QUANTA 650 FEG SEM) operated at 5.5 kV and JEOL JSM-6300, whereas the atomic structure was performed by high resolution transmission electron microscopy (HRSEM/STEM) using a JEM-2200F equipment at 200 kV of accelerating voltage.The BET method applied in the nitrogen adsorption isotherms obtained from a Quantachrome Autosorb-3B apparatus was used to obtain the specific surface area, pore size distribution, and pore volume of the powders.Prior to the adsorption, the compounds were desorbed under high vacuum at 300 ∘ C during 3 hours.The surface chemical states of the compounds were analyzed by X-ray Photon Spectroscopy technique using a VG Escalab 220i XL spectrometer after exciting the samples by a nonmonochromatized Mg K line at 1253.6 eV.The band gap values of the ceramics were obtained using a spectrophotometer Varian Cary 100 UV-Vis.

Photocatalytic Evaluation.
The photocatalytic activities of the composites were estimated by decomposition of 4chlorophenol suspension at room temperature under visible light irradiation.The experimental procedure was realized following the same procedure previously reported [19].

Results and Discussion
Figure 1 shows XRD patterns of the sol-gel noble metal (Pt or Pd) doped TiO 2 catalysts using different molar percentages (1, 3, and 5) and calcined at 500 ∘ C. It can be seen that the undoped TiO 2 sample (Figures 1(a) and 1(b)) is featured by a combination of two structural phases, anatase and rutile.The diverse peaks match well with the crystallographic cards ICDD 89-4921 and ICDD 70-7347, for anatase and rutile structures, respectively.The XRD patterns for synthesized samples with 1 mol-% of Pd (Figure 1(a)) showed that palladium was almost totally incorporated into the anatase lattice where a weak peak of PdO was detected.At this composition is also evident that Pd addition at 1 mol-% caused an effect of stabilization for the anatase phase, inhibiting the formation of the rutile phase.The increment of Pd at 3 and 5 mol-% saturated the anatase lattice and as a consequence the PdO (ICCD 43-1024) and rutile phases were formed.In agreement with other works, the Pd addition enhances the anatase phase stabilization modifying the crystal growth [20].However, such stabilization is misplaced after the saturation of the TiO 2 lattice in the anatase structure.On the other hand, Pt addition showed poor tendency to stabilize the anatase phase (Figure 1(b)).It can also observe the apparition of quite weak PtO 2 peaks (ICCD 73-2370), even with low amount of dopant [21].These results point out that the Pt nanoparticles Journal of Nanomaterials could not be incorporated in the oxide matrix or interstitial sites; however, they led to obtaining catalysts with significant variation in the structure [22].
X-ray diffraction patterns were adjusted using Rietveld refinement to confirm and quantify the phases in the compounds.This method uses a theoretical structure modeling of the possible individual phases in the material to fit the whole powder diffraction pattern using a suitable analytical profile fitting function [23,24].It is important to consider crystallographic data (lattice parameters, fractional coordinates, occupancy, and temperature factor, etc.) of the diverse possible structures in the compounds as well as instrumental parameters [24].However, estimating the volume contents of individual phases in a compound is a real challenge, due to the fact that multiphase materials could present partially or completely overlapped reflections [25].The possible TiO 2 phases in the catalysts (TiO 2 (), anatase, rutile, Pd, PdO, Pt, and PtO 2 ) were considered to fit the experimental profiles using the most suitable pseudo-Voigt analytical function [24,26] modified with the Thompson-Cox-Hasting model.The quality of the Rietveld analysis is based on the -factors ( wp ,  exp ,   ), viewing graphically the observed and calculated patterns as well as ensuring that the model is chemically plausible [27][28][29].The  wp factor is the more valuable due to the fact that it does not depend on the absolute value of the intensities; but it depends on the background.In this sense, the weighted profile ( wp ) follows directly from the square root of the quantity minimized, scaled by the weighted intensities [27,28]: where  (obs) is the observed intensity at the th step,  (calc) is the calculated intensity at the th step, and   is the weight.The best possible  wp is called expected profile ( exp ).It reflects the quality of the obtained data in the measure of the XRD patterns (statically counts) [27,28]: where  is the number of observed data,  is the number of parameters to refine,  (obs) is the observed intensity, and   is the weight.The average of these values is  2 or goodness of fit (GoF) [27][28][29]: During the refinement process  2 starts out large when the model is poor and decreases as the model produces better agreement with the data [27]; when  2 is close to one, it can be considered that the values derived from least squares variance-covariance matrix are accurate.The results of the phase quantification of the as-prepared samples are summarized in Table 1.The diverse compounds displayed GoF values lower than 1.5 and graphically the observed and calculated patterns showed good fitting and minimum differences (Figures 2(a)-2(g)).
Thus, considering that  wp ,  exp , and  2 have no statistical basis but are very valuable as a measure of refinement quality, we assume that a reasonable fitting was achieved [23,27].It is worth mentioning that the low values of GoF in the Rietveld refinement were considering the TiO 2 () phase in the analysis.TiO 2 () structure is usually obtained  [19] as well as their close position with the rutile phase.Our previous studies [19] indicated that the synthesis conditions favored the TiO 2 () production explaining its residual content in the compounds (<3 vol-%).On the other hand, the Rietveld results of the 1% Pd/TiO 2 compound (Table 1) identified contributions of PdO phase in the XRD pattern.The low percentage of this phase (<1 vol-%) suggests that the majority of palladium was incorporated into the anatase lattice.Rietveld refinement of Pd/TiO 2 and Pt/TiO 2 was used to obtain the anatase structure models and typical assemblies are shown in Figures 3(a)-3(c).For comparison, the Rietveld analysis of undoped TiO 2 structure is also showed in the figure.The structures of Pd-doped anatase phase (Figure 3(b)) and Pt-doped anatase phase (Figure 3(c)) showed variations in the lattice parameters and O-Ti-O bond distances compared with undoped anatase phase.These modifications suggest that Pd 4+ and Pt 4+ ions incorporate into the anatase structure, because Ti 4+ (0.605 Å), Pd 4+ (0.615 Å), and Pt 4+ (0.625 Å) have similar ionic radii [32].It is known that some dopants stabilize the anatase phase by occupying interstices, thereby distorting the anatase lattice and restricting the lattice contraction involved in the transformation to rutile [6].Besides, the closer values of Ti 4+ and Pd 4+ ions could explain the higher amount of quantity of anatase phase in these compounds.Finally, it is possible to observe that anatase phase is present in all the specimens showing diverse combinations with other phases (since ∼30 to ∼93 vol-% content) which could result in materials with better photocatalytic properties [10][11][12].
An estimation of the crystal size was also obtained with Rietveld refinement and the results are shown in Figures 4(a)-4(c).From this figure, it seems that the crystal dimensions varied according to the structural phase.The highest crystal sizes were found for the undoped TiO 2 samples, ∼67 nm for anatase and ∼50 nm for rutile, whereas monoclinic beta-TiO 2 displays a value of ca. 5 nm.In general, except for this last structural phase, the crystal size of anatase and rutile reduces with the Pd or Pt addition (Figures 4(a) and 4(b)) in the range of 30-60 nm, approximately.This effect has been previously reported, indicating that metal ions doped titania provoked a reduction in the nucleation process [33].Another feature to highlight in Figure 4(c) is that longer amounts of Pd provoke secondary phases formation (Pd and PdO) with an important crystal size (∼232 nm); whilst independent of the quantity of Pt added to TiO 2 , the crystallite size of the formed secondary phases are lower than 10 nm.The results pointed out that the synthesis parameters and preparation method more than noble metal content play a key role in the composition of structural phases and crystal sizes [34,35].
Figures 6(a)-6(i) display the morphology of the catalysts obtained by SEM and HRSEM techniques.The undoped TiO 2 compound (Figure 6(a)) contains irregular shape particles, which are aggregates of tiny crystals of TiO 2 phases [4].The sizes observed in the noble metal doped TiO 2 samples match well with the crystallite size calculated from Rietveld analysis indicating that a good fitting was achieved.In general, it can be observed that Pt-doped samples displayed lower particle size than that presented by Pd-doped specimens.For example, 3% Pt/TiO 2 samples showed uniform Pt particles with sizes ∼12 nm (Figure 6(c)), whereas 3% Pd/TiO 2 (Figure 6(b)) specimen exhibited agglomerates with variable diameter sizes between 15 and 400 nm; similar results were found with the other molar compositions (Figures 6(d)-6(i)).These results suggest that due to accepting species on the TiO 2 surface

Observed Calculated
Observed-calculated Bragg position 20   an enhancement on the electron-hole spatial separation can be expected, improving the photocatalytic properties of this material [4,20].HRTEM and STEM images of the as-prepared catalysts are shown in Figures 7(a)-7(l), where it is seen that undoped TiO 2 sample (Figures 7(a 2).On the other hand, this kind of hysteresis loops is characteristic of solids consisting of aggregates (consolidated) or agglomerates (unconsolidated) of spherical particles and they correspond to the H 2 type because pores are nonuniform in size and shape [42].It is important to mention that 3% Pd/TiO 2 catalyst presents very low values of surface area, which could be attributed to agglomerates in the samples previously observed in the HRSEM.In general, the surface areas obtained in these compounds are slightly low compared with other reports [43,44].
XPS measurements were performed to determine impurities in the samples as well as to evaluate Pd and Pt content.Figures 9(a)-9(c) correspond to selected XPS survey spectra of undoped TiO 2 , 3% Pd/TiO 2 , and 3% Pt/TiO 2 , respectively.In general, the spectra show two features assigned to Ti(IV) titanium in the IV oxidation state of the diverse titania phases [41,45].These peaks (Figure 9(d)) correspond to Ti2p states at BE = 458.8eV (Ti2p 3/2 ) and BE = 464.7 eV (Ti2p 1/2 ).On  the other hand, the Pd signal is quite weak due to the fact that moderate counts were used in this sample; however, they were enough to relate the peak to the Pd3d state (Figure 9(e)) at 336 eV and it is a characteristic pointer of the PdO phase [46].The typical XPS doublets peaks of 4f signal for Pt were observed at biding energies of 74.2 eV and 71.0 eV [47,48].Finally, a quantification of the elements is shown in Table 3, where it is seen the quantification of C, O, Ti, Pd, and Pt; thus, we assumed the obtaining of pure phases; however, it cannot discard the presence of small quantities of carbon impurities formed during sample preparation.The 4-CP main absorption band at 278 nm was used to quantify the concentration of the pollutant in the solution using UV-vis measurements.Figures 10(a) and 10(b) present the degradation measurements of Pd/TiO 2 and Pt/TiO 2 compounds; both systems were compared with the undoped TiO 2 sample and the commercial Degussa P25.The photocatalysis process is based on the energy transfer through the interaction of an excited semiconductor and (in this case) an aqueous solution.This interface generates a local density of charge (electron-hole) that produces an electric field promoting the charge transfer.In this sense, diverse reactions could be produced in the catalyst surface; however, the efficiency in the degradation rate of pollutant is influenced by the active site and the photoabsorption of the catalyst used [49].In this work, it is clear that all the synthesized compounds showed better response than the commercial Degussa P25 that started to degrade until 75 min later.This could be attributed to the photocatalytic activity of titanium dioxide because it is greatly influenced by the crystalline form [38,50] and the combinations of phases between catalysts are different.The best results correspond to 1% Pd/TiO 2 catalyst which eliminates the pollutant almost 8 times faster after 105 min.Apparently, Pd addition provoked the transference of photogenerated electrons from TiO 2 conduction band to the dopant conduction and the holes were accumulated in the TiO 2 valence band reducing electron-hole pair recombination [4,51], thus improving the photocatalytic activity.On the other hand, with the increment of Pd content we expected an enhancement in the photocatalytic response; however, 3% Pd/TiO 2 and 5% Pd/TiO 2 catalysts suffered a slight decrement in the degradation efficiency (∼90%), possibly due to the excess of PdO that masked the TiO 2 surface and reduced the light absorption capability [49].One of the most active metals for photocatalytic enhancement is platinum (Pt) and the capture of electrons by this metal is postulated to produce longer electron-hole pair separation, lifetime hindering the electron-hole pairs recombination and enhancing the transfer of holes and possibly electrons to O 2 adsorbed on the TiO 2 surface [18].However, in this work, the photocatalytic effect of Pt/TiO 2 compounds was slightly different compared with Pd/TiO 2 compounds; nevertheless, our results show a better degradation performance in comparison with the commercial one.Also, even the compounds with Pt content at 3 mol-% and 5 mol-% eliminate 7 times faster the pollutant.From the different compounds evaluated the more efficient catalyst was the one with major content of anatase phase (∼92 vol-%), in agreement with the literature [49,52].
It is worth mentioning that the catalysts show good degradation efficiency at UV light exposure; however, to complement this information it is necessary to know the band gap to determine if these compounds could work under visible light.The catalysts were characterized by reflectance measurements.Figure 11 presents a large absorption band between 1.54 and 3.0 eV (400-800 nm) due to  the O 2− (2p) → Ti 4+ (3d) transitions in the tetrahedral symmetry [53,54] as well as contributions of the noble metals; however, the TiO 2 absorption spectrum was not substantially modified allowing to estimate the band gap, through Kubelka-Munk (K-M or ()) function [55] based on where  is the reflectance and () is proportional to the extinction coefficient () [55][56][57].A modified () function can be obtained by multiplying the function by ℎ], where ℎ is Planck's constant and ] is the light frequency (s −1 ) using the corresponding coefficient ( = 1/2, for indirect allowed transitions) associated with an electronic transition as follows: ( () * ℎ])  .
(5) By plotting this equation as a function of the energy in eV, the band gap (  ) of semiconductor particles can be obtained [37], whereas for practical purposes, the band gap energy for the different samples can be calculated using the following equation: In this equation,  and  are obtained by the linear fit ( = +) of the flat section of the UV-vis spectrum [57].Table 2 presents the obtained band gaps of the compounds.Undoped TiO 2 sample exhibits a band gap value of 3.01 eV, lower value than the commercial Degussa P25 (3.16-3.18)[58,59].On the other hand, with addition of the dopant (Pd and Pt)   value shifts up to 1.67 eV.The reduction of the band gaps could be attributed to additional levels of the Ti3d conduction band formed in the TiO 2 [20].These results point out that these catalysts could also work under other frequencies of the electromagnetic spectrum extending their applications; however, more experiments need to be realized.
These results show that engineering TiO 2 catalysts by a mix of different phases is a research axis of considerable interest, because it enhances the photocatalytic behavior compared with single-phase titania.The sol-gel route has been widely used to synthesize TiO 2 compounds and many researches have been developed about this area.Nevertheless, the route of synthesis followed in this work favored to control the kinetics anatase-rutile phase transformation corroborating the importance of the effect processing conditions to modify crystalline size, grain size, surface area, and morphology as well as phase distribution [31].The synthesized compounds were compared with Degussa P-25, one commercial photocatalyst which is utilized as a reference material in many studies and all the compounds showed better photocatalytic behavior attributed to the Schottky barrier-type charge separation from the deposition of noble metals on TiO 2 .

Conclusions
In summary, undoped TiO 2 , Pd/TiO 2 , and Pt/TiO 2 catalysts with diverse structural compositions were successfully synthesized using sol-gel method.XRD and RAMAN results indicate that Pd and Pt addition achieved modifying the anatase-rutile transformation obtaining diverse catalysts with a combination of phases.Rietveld analysis revealed the vol-% content of anatase (from ∼30 to ∼93 vol-%) in the catalysts as well as its combination with other phases.HRTEM confirmed the formation of anatase phase and the nanometric size of the catalysts.HRSEM displayed the good distribution of the dopants in the samples.All the compounds showed better response in the degradation of 4-chlorophenol (almost eight times faster in the case of Pd/TiO 2 at 1 mol-%) than the commercial Degussa P25, though their efficiency depends on the Schottky barrier-type separation from the deposition of noble metals on TiO 2 .The results obtained here show that the photoactivity of TiO 2 was enhanced notwithstanding the low value of surface area.The arrangement of phases as well as the Pd and Pt doping induced a blue shift in the band gap energy, which explains the observed catalytic behavior of the samples.Finally, these results are very promising for the use of this type of catalysts in the elimination of other pollutants under other frequencies of the electromagnetic spectrum extending their applications.

Figure 8
presents adsorption-desorption isotherms obtained by the BET technique.Pd/TiO 2 (Figure 8(a)) and Pt/TiO 2 (Figure 8(b)) samples are compared with undoped TiO 2 compounds.All samples exhibit a type IV nitrogen isotherm due to the pore size measured (3.40 −3.71 nm), which is characteristic of mesoporous materials (Table

Table 1 :
Percentages of phases determined and goodness of fit values (GoF) by Rietveld refinement.
[6])/anatase junction is free of stacking faults or defects and required a lowest energy pathway to transform into anatase which is a one-step transition; that is, their structures are related in nature[6].These phases (TiO 2 () and anatase) are difficult to identify only by XRD technique because its reflections are in the 2 range of 27-34 ∘ and 43-45 ∘ (ICCD 35-0088) showing non-well-defined peaks

Table 2 :
BET results and band gap values of the different materials obtained in this study.The reliability of these TiO 2 band gap values could be compromised.