Hydrogen production from glycerol steam reforming ...

Catalysis Communications 77 (2016) 83–88

Contents lists available at ScienceDirect

Catalysis Communications journal homepage: www.elsevier.com/locate/catcom

Short communication

Hydrogen production from glycerol steam reforming over molybdena–alumina catalysts Gheorghiţa Mitran a,⁎, Octavian Dumitru Pavel a, Mihaela Florea a, Daniel G. Mieritz b, Dong-Kyun Seo b a Laboratory of Chemical Technology and Catalysis, Department of Organic Chemistry, Biochemistry & Catalysis, Faculty of Chemistry, University of Bucharest, 4-12, Blv. Regina Elisabeta, 030018 Bucharest, Romania b School of Molecular Sciences, Arizona State University, Tempe, AZ 85287-1604, USA

a r t i c l e

i n f o

Article history: Received 14 September 2015 Received in revised form 16 December 2015 Accepted 25 January 2016 Available online 26 January 2016 Keywords: Glycerol Hydrogen Molybdena–alumina Steam reforming

a b s t r a c t The glycerol steam reforming was investigated on alumina supported molybdena catalysts (with 2, 5 and 12 wt.%) prepared by the sol–gel method and gel combustion. The catalysts were characterized by XRD, BET, UV–VIS, DRIFT, SEM and TEM. The catalytic performances were studied at 400–500 °C, steam to glycerol molar ratio between 9:1 and 20:1 and feed flow rate 0.04–0.08 ml/min. The conversion is directly proportional to molybdena loading, while the hydrogen selectivity has reached greater value on catalyst with 2% MoO3. The optimum ratio steam to glycerol for reforming is 15:1 and for decomposition in syngas 9:1 and the ratio 20:1 favors water gas shift reaction. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Biomass is an important source of energy and provides raw materials for the chemical industry. The advantage for the energy obtained from biomass, compared with fossil fuels which are the traditional sources of energy, consists in reduction of greenhouse gas emissions [1]. Lately, a special importance was given to the use of compounds obtained from renewable biomass sources for hydrogen production. The glycerol valorification is of particular interest since it comes from renewable resources and is a suitable bio-renewable substrate for hydrogen production being preferred over the fossil fuels [2]. The hydrogen is an energy carrier, a renewable and clean fuel, and a raw material for ammonia production [3] and Fischer–Tropsch synthesis [4]. The usual methods for hydrogen production are based on catalytic reforming of hydrocarbons. In the past years, the hydrogen obtained by glycerol reforming has been widely studied. The convenience of using glycerol for H2 production comparatively with ethanol, another source of hydrogen, consists in the fact that glycerol contains more number of moles of H in its chemical structure. The glycerol steam reforming was deeply investigated over supported transition metals, such as Ni, Co, Pt, and Pd. Ni/Al2O3 was proved to be a good catalyst for this reaction but presents some drawbacks: is susceptible to deactivation by carbon deposition; the hydrogen production is strongly affected by reaction ⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (G.ţ Mitran).

http://dx.doi.org/10.1016/j.catcom.2016.01.029 1566-7367/© 2016 Elsevier B.V. All rights reserved.

temperature, and increases with temperature increasing [5,6]. The catalytic activity was improved by the addition of Co [7] and Sn as a promoter and CeO2–MgO–Al2O3 as support [8]. Likewise, CeO2 has been used as a promoter as well as a support for Ni-based catalysts as a result of its distinctive oxygen storage capacity, which determines the presence of highly active oxygen, making the catalyst more active for this reaction [9]. Recently, nanomaterial oxides with particular morphology were synthesized and used as catalysts for glycerol steam reforming, as for example noble metals containing materials, Ir/La2O2CO3 [10]. Pt-based catalysts supported on Al2O3, SiO2, MgO, TiO2 [11], Pt/C, and Pt-Re/C [12] were also studied in this reaction and were found to be more effective. Nevertheless, such catalysts are not common in industrial applications because of their cost and limited availability. Molybdena supported on different substrates such as alumina, silica, titania, and zirconia is used as a catalyst for selective oxidation reactions [13–15], olefin metathesis [16,17], and dehydrogenation [18]. The higher reductibility of molybdenum makes this a good catalyst for water–gas shift reaction [19]; MoOx could be reduced by CO (with CO2 generating) and re-oxidized by H2O forming H2. Furthermore, the use of alumina at support provides a high dispersion of molybdena and has an advantage in the prevention of Mo oxo-species aggregation. From the above reasons, in the present work molybdenum was selected as an active phase to be included in catalyst formulations for glycerol steam reforming. Herein, we report the results obtained for the glycerol steam reforming over molybdena nanomaterials supported on alumina. No references were found in the literature for this reaction with molybdena catalysts. The effects of operating conditions including temperature,

84

G.ţ Mitran et al. / Catalysis Communications 77 (2016) 83–88

steam to glycerol molar ratio, flow rate of liquids and carrier gas on glycerol steam reforming were tested. The material characterization was performed by XRD, N2 adsorption, UV–VIS–NIR spectroscopy, DRIFT spectroscopy, SEM and TEM microscopy. 2. Experimental 2.1. Catalyst preparation The catalysts used for glycerol steam reforming, MoO3/Al2O3, were prepared by the sol–gel method and gel combustion. Aluminum nitrate (Al(NO3)3·9H2O from Tunic) was dissolved in water (10 wt.%) and the solution was stirred at 60 °C for 30 min [20]. Citric acid (C6H8O7·H2O, from Silal, 99.5% purity) was added in the solution, with a molar ratio citrate-to-nitrate of 0.5. Then aqueous (NH4)6Mo7O24·4H2O (Fluka Analytical) solutions were added, so as to get a percentage of 2, 5, and 12% MoO3 on alumina. The excess of water was slowly evaporated at 80 °C to obtain a gel. The gel was rapidly heated at 200 °C. Finally, all the prepared materials were calcinated in flowing air at 400 °C for 2 h and 550 °C for 2 h. 2.2. Catalyst characterization The crystalline phases of the catalysts were investigated by the X-ray diffraction (XRD) method. XRD patterns were obtained with a Philips PW3710 type diffractometer equipped with a Cu Kα source (λ = 1.54 Å), operating at 50 kV and 40 mA. They were recorded over the 5–70° angular range with 0.02° (2θ) steps and an acquisition time of 1 s per point. Data collection and evaluation were performed with PCAPD 3.6 and PC-Identify 1.0 software. The surface areas of the catalysts were measured from the adsorption isotherms of nitrogen at − 196 °C using the BET method with a Micromeritics ASAP 2020 sorptometer. The samples were first outgassed at 300 °C for 4 h in the degas port of the adsorption apparatus. The pore size distribution curves were calculated using desorption branch of the isotherms with the Barrett–Joyner–Halenda (BJH) method. The UV–VIS–NIR spectra were recorded using a UV3600 UV–VIS spectrophotometer with Shimadzu ISR-3100 integrating sphere attachment having an angle of incident light 0–8°, wavelength range of 220– 2600 nm, and two light sources: D2 (deuterium) lamp for the ultraviolet range and WI (halogen) lamp for the visible and near-infrared range. UV–VIS–NIR spectrophotometer has three detectors, consisting of a PMT (photomultiplier tube) for the ultraviolet and visible regions and InGaAs and cooled PbS detector for the near-infrared region. The spectra were recorded in the range of 220–2600 nm (the switching wavelength of the lamps is between 282 nm and 393 nm) with a wavelength step of 2 nm, having the slit width of 8 nm. The UV–VIS spectra were measured using samples diluted with extra pure barium sulfate (purchased from Nacalai Tesque). DRIFT investigation of the catalysts was carried out using a Thermo Nicolet 4700 spectrometer, with the following parameters: 300 scans and 400–4000 cm−1 scan range at a 4 cm−1 resolution.

Scanning electron microscopy (SEM) studies were performed on dry sample deposited carbon tape by using an FEI XL-30 Environmental SEM with 5–15 keV electrons. Before analysis, using a Denton Desk II sputter-coater, the samples were coated with an Au–Pd (60–40) target for 140 s to deposit a coating of ~12 nm. For transmission electron microscopy (TEM) and high-resolution TEM (HR-TEM) studies, samples were prepared by dry-grinding and dusting on to TEM grids. The images were collected on a JEOL 2010F at an accelerating voltage of 200 kV. 2.3. Catalytic reaction The experimental equipment used for glycerol steam reforming is a fixed-bed quartz reactor (12 mm internal diameter) that operates at atmospheric pressure. In the reactor was injected an aqueous solution (with water to glycerol molar ratios 9:1, 15:1 and 20:1) by a pump with feed flow between 0.04 and 0.08 ml/min. The evaporation was carried out in the first third of the reactor. The nitrogen, used as carrier gas, was introduced with feed flow of 30 ml/min and the mass of the catalyst was 0.1 g. The reaction products were analyzed with a Thermo Finnigan Gas-Chromatograph equipped with a thermal conductivity detector (TCD) with an alumina column and a flame ionization detector (FID) with a CTR I column. 3. Results and discussion 3.1. Catalyst characterization Crystalline phases in the catalysts were characterized by XRD. The diffractograms are presented in SI 1. The diffraction lines corresponding to alumina support were detected for all samples studied. No diffraction lines corresponding to molybdenum were detected, showing that the molybdena species are amorphous and well dispersed on the support surface. From the diffraction patterns, the amorphous character of the powders is evidenced for 5 and respectively 12MoAl, when for 2MoAl appears a partially crystallized transition alumina phase. It is observed that peaks at ca. 37°, 46° and 66.8° attributed to the planes (111), (100) and (110) of γ-Al2O3. BET surface area for all catalysts and for support are presented in Table 1, and a decrease of surface area from 73.7 to 39.9 m2/g with increasing of Mo contents from 2% to 12% is observed; this behavior could be attributed to pore occlusion. This behavior is in good agreement with the literature data [21]. In the SI 2 the N2 sorption isotherms and pore size distributions for studied catalysts are shown. The samples present typical type IV N2 sorption isotherms with H3 and H4 hysteresis loops and exhibits spherical cavities that evaporate completely in their condensate at lower pressure due to a percolation phenomenon and cylindrical pores [22]. Sample with 2% molybdena presents hysteresis loop at P/P0 0.4–0.9, while for 5 and 12MoAl it appears at 0.2–0.9 and is assigned with the presence of mesoporosity. The surface density of Mo (Table 1), defined as the number of molybdenum atoms present per unit specific surface area

Table 1 Physico-chemical characteristics of the catalysts. Catalyst

Al2O3 2MoAl 5MoAl 12MoAl a

Specific surface area (m2/g)

232 73.7 42.8 39.9

Oxygen in balance.

Total pore volume (cm3/g)

0.13 0.10 0.03 0.03

Average pore width (nm)

5 2.9 3.3

Chemical composition (wt.%) by EDXa

Chemical composition (at.%) by EDXa

Mo

Al

Mo

Al

– 1.78 5.08 10.95

– 59.32 32.64 46.59

– 0.40 0.99 2.54

– 47.30 22.64 38.41

Average pore diameter (nm)

Surface density (Moat./nm2)

2.7 4.9 5.9 4.0

– 1.14 4.89 12.58


85

Surface density Mo atoms=nm2 wt:%Molybdena loading∙6:023∙1023 ¼ Molecular Weight of Molybdena∙100∙Surface Area

Species of Mo have a high tendency to form a monomolecular layer on alumina surface without the formation of bulk MoO3 or bulk Al2(MoO4)3 [23]. The limit for monolayer coverage of the alumina support was reported at ~5Mo/nm2 by experimental considerations [24]. The UV–VIS absorption spectra of the catalysts are presented in SI 3. All samples display an absorption band in the 350–400 nm region commonly attributed to oligomeric species and octahedrally coordinated to Mo+6 (Oh) [25]. A shift of absorption band towards higher wavelength at increasing of molybdena loading could be attributed to an increase of oligomer weight. The DRIFT spectra of the samples are described in Fig. 1. The band situated at 1035 cm−1 corresponds to symmetric and antisymmetric vibration of Mo_O bonds in polymeric surface molybdena species [26] and is more pronounced for 12MoAl suggesting that on this catalyst polymeric species prevails. The bands in the region 415–490 cm−1 correspond to bridging stretching modes of Mo–O–Al indicating a strong interaction of molybdena with Al2O3 support [27]. The absence of Mo_O stretching corresponding to tetrahedral MoO− 4 species bonded to alumina surface that appears at 950–955 cm−1 confirms the presence of polymeric species, as it was observed by UV–VIS spectroscopy. The band situated at 1635 cm−1 is suggested to be the result of the bending vibrational mode of water absorption [28] The amount of molybdena on alumina supports was determined by SEM–EDX analysis. Micrographs of MoAl catalysts are shown in Fig. 2 and molybdena loading in Table 1. Numerous aggregates appear on the catalyst surface and the distribution of grain was more uniform for 12MoAl. The molybdena content (wt.%) is very close to calculated values (2, 5 and respectively 12%). From the table, the larger pore

Fig. 2. SEM micrographs of MoO3/Al2O3 catalysts calcinated at 550 °C (a — 2MoAl, b — 5MoAl, c — 12MoAl).

Fig. 3. TEM micrographs of MoO3/Al2O3 catalysts (a — 2MoAl, b — 5MoAl, c — 12MoAl).

Fig. 1. The DRIFT spectrum of the alumina supported molybdena catalysts.

(molybdenum atoms per nm 2 ), was calculated by the following equation:

86


diameters were obtained at molybdena addition, which proves that molybdenum can act as a pore expander but only up to 5%. The distribution of molybdenum species in the samples is investigated by TEM (Fig. 3). The 12MoAl sample seemed to have mostly sharp and thin pieces, while 2MoAl had more spherical particles.

3.2. Glycerol steam reforming The catalytic performance of MoO3/Al2O3 catalysts was investigated at temperatures between 400 and 500 °C, steam to glycerol molar ratio 9:1–20:1, feed flow between 0.04–0.08 ml/min and nitrogen feed flow 30 ml/min. The partial pressure of reactants ranged from 3.2 to 6.5 kPa for glycerol, 54.7 to 69.1 kPa for water and from 26.3 to 41.7 for nitrogen. The H2, CO, CO2 and CH4 are obtained by glycerol steam reforming (1), glycerol decomposition (2), water gas shift (3) and by methanation (4) reactions: [29] C3 H8 O3 þ 3H2 O→3CO2 þ 7H2

ð1Þ

C3 H8 O3 →3CO þ 4H2

ð2Þ

CO þ H2 O↔CO2 þ H2

ð3Þ

CO þ 3H2 ↔CH4 þ H2 O:

ð4Þ

In the liquid products, acetaldehyde, acetone and 1,2-propanediol were detected, therefore, we study the glycerol total conversion Xt and glycerol to gaseous product conversion Xg.

The total conversion (Xt), gaseous product conversion (Xg), product selectivities (Si), turnover frequency (TOF) and space time yield for H2 (STY) were calculated based on the following equations: Xt ð%Þ ¼

nC3 H8 O3 in nC3 H8 O3 nC3 H8 O3 in

Xg ð%Þ ¼

nCO2 þ nCO þ nCH4 100 3 nC3 H8 O3 in

out

100

The gaseous product selectivity is expressed as [30]: Si ð%Þ ¼

Ni NH2 þ NCO2 þ NCH4 þ NCO 100

where NH2, NCO2, NCH4, and NCO correspond to produced moles of H2, CO2, CH4, and CO TOF ðs 1Þ ¼

specific activity ðmol=g hÞ 100 MMoO3 3600 ðsÞ wtð%ÞMoO3

STY ðmol=g hÞ ¼

FXS m 104

where F = feed rate (mol/h), X = conversion, S = selectivity in H2, and m = weight of catalyst (g).

Fig. 4. Glycerol conversion and gaseous product selectivities as a function of temperature: a — 2MoAl, b — 5MoAl, c — 12MoAl (feed flow rate 0.06 ml/min, molar ratio steam to glycerol 15:1, Pgly = 4.2 kPa, PH2O = 62.5 kPa, PN2 = 33.3 kPa), d — TOF and STY for hydrogen (500 °C).


87

3.3. Effect of reaction temperature

3.4. Effect of steam to glycerol ratio

Fig. 4(a, b, c) indicates the conversion and the selectivity for H2, CO and CO2 over Mo/Al catalysts at 400, 450 and 500 °C. The main product for all catalysts was hydrogen followed by CO and CO2, while methane was detected in minor amounts proving that reaction of methanation did not occur in the conditions of the present research. The hydrogen selectivity was dependent on the temperature as illustrated in Fig. 4. Hydrogen selectivity was increased as a function of reaction temperature up to 500 °C. Regarding the conversion, the tendency observed in Fig. 4(a, b, c) showed that liquid phase conversion decreases and gas phase conversion increases with increasing temperature. The increasing of temperature enhances the conversion to gas products avoiding secondary reactions (e.g. dehydration, dehydrogenation and hydrogenolysis) that lead to the coke deposition. The glycerol conversion varies directly proportional with the molybdenum content while the hydrogen selectivity is inversely proportional. The increase of hydrogen and the CO2 selectivities with temperature are determined by the endothermic nature of steam reforming reaction [31]. Fig. 4d represents the turnover frequency and the space time yield in hydrogen for all catalysts at 500 °C, feed flow rate 0.06 ml/min and at a molar ratio steam to glycerol 15:1. The turnover frequency is observed to be a strong dependence on specific surface area, decreases with surface area decreasing. The space time yield is directly proportional with Mo surface density. The catalyst with higher Mo density on surface could provide more active Mo sites and consequently delivers higher efficiency for hydrogen production.

The effect of steam to glycerol ratio on the conversion and gas product selectivity at 500 °C and feed flow rate 0.06 ml/min is presented in Fig. 5(a, b, c). The glycerol conversion and the selectivities of H2 and CO2 increased with steam/glycerol increasing due to decreasing of the glycerol in feed amount and, because the water excess was utilized for the glycerol reforming and for water–gas-shift reactions, confirmed by increasing of CO2 selectivity and decreasing of CO selectivity; the same behavior was observed by other authors in previous papers. The ratio between H2:CO2 and H2:CO is presented in Fig. 5d, and from the figure it can be observed that in the presence of steam in excess a significant fraction of the CO was consumed in the water gas shift reaction. The optimum ratio steam to glycerol for glycerol steam reforming is 15:1, while for glycerol decomposition in syngas the optimum ratio is 9:1 and ratio 20:1 favors water gas shift reaction. The data from the figure denote a product ratio H2:CO2 of 2.14 for 5MoAl which is comparable to the stoichiometric H2:CO2 ratios of 2.33, at the same time; for all catalysts at the molar ratio steam to glycerol 9:1 a product ratio H2:CO is very close to stoichiometric ratio of 1.33 (between 1.1–1.24). 3.5. Effect of the feed flow rate The effect of the feed flow rate at 500 °C and molar ratio steam to glycerol 15:1 on the glycerol conversion and the gaseous products are presented in SI 4. The hydrogen and CO2 selectivities decrease with increasing of feed flow rate from 0.04 to 0.08 ml/min, suggesting that a

Fig. 5. Glycerol conversion and gaseous product selectivities as a function of molar ratio: a — 2MoAl, b — 5MoAl, c — 12MoAl (feed flow rate 0.06 ml/min, t = 500 °C), d — ratio of H2 to CO and H2 to CO2 (500 °C).

88


short contact time of reactants suppresses water gas shift reaction. The results suggest that, at the decrease of feed flow rate, all compounds are steam reformed into H2 and CO2. 3.6. Stability test The product distributions in glycerol steam reforming over 12MoO3/ Al2O3 at 500 °C, molar ratio steam to glycerol 15:1, and feed flow rate 0.06 ml/min as a function of time-on-stream is presented in SI 5. As can be seen, the total glycerol conversion decreases only with less than 5%, while the conversion to gaseous products decreases with 8%, indicating that the conversion to liquid products increases in time. Nevertheless, after 6 h of time on stream the catalyst is stable and no other changes are observed, indicating that our catalysts need a long-term steady state in order to reach the optimal catalytic activity and stability.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

4. Conclusions [16]

Molybdena catalysts supported on alumina obtained by the sol–gel method are tested in the glycerol steam reforming in the 400–500 °C temperature range, steam to glycerol ratio 9–20 and feed flow rate 0.04–0.08 ml/min. Hydrogen is formed as the major product followed by CO and CO2 while the selectivity from CH4 is very low, below 5%. The turnover frequency depends on a specific surface area, while the space time yield of hydrogen is directly proportional with the Mo surface density. Increasing of the reaction temperature and steam to glycerol ratio enhanced both glycerol conversion and hydrogen selectivity. Thus, the alumina supported molybdena samples can be considered a promising catalysts for the glycerol steam reforming, since it was stable, with a good selectivity to hydrogen, and are active at lower temperatures.

[17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]

Acknowledgments This work was supported by a grant of the Romanian National Authority for Scientific Research, CNDI–UEFISCDI, project number PCCA-II-56/2014. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.catcom.2016.01.029.

[28] [29] [30] [31]

M.L. Dieuzeide, M. Jobbagy, N. Amadeo, Int. J. Hydrog. Energy 39 (2014) 16976–16982. K. Avasthi, R.N. Reddy, S. Patel, Procedia Eng. 51 (2013) 423–429. J. Andersson, J. Lundgren, Appl. Energy 130 (2014) 484–490. N. Fischer, E. van Steen, M. Claeys, J. Catal. 299 (2013) 67–80. A.M. Douette, S.Q. Turn, W. Wang, V. Keffer, Energy Fuel 21 (2007) 3499–3504. M.L. Dieuzeide, V. Iannibelli, M. Jobbagy, N. Amadeo, Int. J. Hydrog. Energy 37 (2012) 14926–14930. E. Sanchez, R. Comelli, Int. J. Hydrog. Energy 39 (2014) 8650–8655. L.F. Bobadilla, A. Penkova, A. Álvarez, M.I. Domínguez, F. Romero-Sarria, M.A. Centeno, J.A. Odriozola, Appl. Catal. A Gen. 492 (2015) 38–47. S. Shao, A. Wu Shi, C. Ling Liu, R. Zhen Yang, W. Sheng Dong, Fuel Process. Technol. 125 (2014) 1–7. X. Huang, C. Dang, H. Yu, H. Wang, F. Peng, ACS Catal. 5 (2015) 1155–1163. M.E. Sad, H.A. Duarte, Ch. Vignatti, C.L. Padró, C.R. Apesteguía, Int. J. Hydrog. Energy 40 (2015) 6097–6106. Z. Wei, A.M. Karim, Y. Li, D.L. King, Y. Wang, J. Catal. 322 (2015) 49–59. N. Ohler, A. Bell, J. Catal. 231 (2005) 115–130. E. Heracleous, A. Lemonidou, Catal. Today 112 (2006) 23–27. G. Mitran, I.C. Marcu, A. Urda, I. Sandulescu, J. Serb. Chem. Soc. 75 (8) (2010) 1115–1124. Y. Cui, N. Liu, Y. Xia, J. Lv, S. Zheng, N. Xue, L. Peng, X. Guo, W. Ding, J. Mol. Catal. A Chem. 394 (2014) 1–9. X. Li, D. Zhang, X. Zhu, F. Chen, S. Liu, S. Xie, L. Xu, J. Mol. Catal. A Chem. 372 (2013) 121–127. E. Heracleous, A. Lemonidou, J. Lercher, Appl. Catal. A Gen. 264 (2004) 73–80. Y. Lian, H. Wang, Q. Zheng, W. Fang, Y. Yang, J. Nat. Gas Chem. 18 (2009) 161–166. J. Li, Y. Wu, Y. Pan, J. Guo, Ceram. Int. 33 (2007) 361–363. K. Kouachi, G. Lafaye, S. Pronier, L. Bennini, S. Menad, J. Mol. Catal. A Chem. 395 (2014) 210–216. C. Velásquez, M. Ojeda, A. Campero, J. Esparza, F. Rojas, J. Mol. Catal. A Chem. 281 (2008) 126–136. L. Kaluža, Z. Vít, M. Zdražil, Appl. Catal. A Gen. 282 (2005) 247–253. S. Xie, K. Chen, A.T. Bell, E. Iglesia, J. Phys. Chem. B 104 (43) (2000) 10059–10068. P. Ciambelli, D. Sannino, V. Palma, V. Vaiano, P. Eloy, F. Dury, E.M. Gaigneaux, Catal. Today 128 (2007) 251–257. L. Boda, G. Onyestyák, H. Solt, F. Lónyi, J. Valyon, A. Thernesz, Appl. Catal. A Gen. 374 (2010) 158–169. H. Nasser, Á. Rédey, T. Yuzhakova, J. Kovács, J. Therm. Anal. Calorim. 95 (1) (2009) 69–74. Q. Ma, H. He, Y. Liu, J. Environ. Sci. 22 (4) (2010) 555–560. M. Araque, L.M. Martínez, J.C. Vargas, M.A. Centeno, A.C. Roger, Appl. Catal. B 125 (2012) 556–566. A. Seretis, P. Tsiakaras, Renew. Energy 85 (2016) 1116–1126. A. Gallo, C. Pirovano, P. Ferrini, M. Marelli, R. Psaro, S. Santangelo, G. Faggio, V. Dal Santoc, Appl. Catal. B Environ. 121–122 (2012) 40–49.