Catal Lett (2015) 145:1272–1280 DOI 10.1007/s10562-015-1529-0
Pd-Promoter/MCM-41: A Highly Effective Bifunctional Catalyst for Conversion of Carbon Dioxide Yingquan Song1 • Xiaoran Liu1 • Linfei Xiao1 • Wei Wu1 • Jianwei Zhang1 Xuemei Song1
•
Received: 25 January 2015 / Accepted: 6 April 2015 / Published online: 29 April 2015 Ó Springer Science+Business Media New York 2015
Abstract As on today, the conversion of carbon dioxide into methanol has been considered as one of the top research priorities all over the world. In this work, a series of supported Pd catalysts was prepared by impregnation method with using Pd (NO3)2 as a precursor and employed as a catalyst in the synthesis of methanol from carbon dioxide. In this process, the surface area and pore volume of support affected the Pd dispersity and catalytic activity of supported Pd catalyst. At the same time, the catalytic performance was enhanced when the alkaline earth oxide CaO was used as a promoter, and the effect of the loading of Pd and promoter was investigated. The 12.1 % of CO2 conversion and the 65.2 % of selectivity for methanol were obtained over the catalyst Pd6Ca3/MCM-41 at 9 h onstream. Graphical Abstract CO2 3H2
O
O
OH
H H H H HH
Pd
Pd CaO
CH3 H H
O
OH
Pd
CaO
CH3OH + H2O
C O
O
CaO
H H H H H
Pd
C O
OH OH
CaO
H H H H H
Pd
O
C
H O
CaO
OH Pd CaO
Keywords Carbon dioxide Methanol Supported Pd catalyst Promoter Hydrogenation & Wei Wu
[email protected] 1
International Joint Research Centre for Catalytic Technology, Key Laboratory of Chemical Engineering Process and Technology for High Efficiency Conversion, College of Heilongjiang Province, School of Chemistry and Material Sciences, Heilongjiang University, Harbin 150080, People’s Republic of China
123
1 Introduction Global warming caused by increasing atmospheric carbon dioxide (CO2) concentration and the depletion of fossil fuels is becoming the focus of worldwide attention. On-site
Pd-Promoter/MCM-41: A Highly Effective Bifunctional Catalyst for Conversion of Carbon Dioxide
catalytic conversion of CO2 to liquid fuels or chemicals, given a steady supply of the CO2 substrate and excess energy, is a promising route that may offer a solution to these important environmental and energy issues. In particular, installation of chemical reactors for the hydrogenation of CO2 to methanol could be an attractive supplement to coal-fired power stations [1–4], iron refineries and incinerators, which not only offers the abatement of CO2 on site but also produces transportable fuel from this reaction. Over the past decades, numerous catalysts have been developed for the hydrogenation of CO2 to CH3OH [5]. The Cu/ZnO–Al2O3 catalyst is typically used to synthesize methanol on an industrial scale from mixtures of CO/CO2/ H2 (syngas) [6, 7] and the yield of CH3OH was enhanced when the Cr2O3 [8], Ga2O3 [9], ZrO2 [10] or CeO2 [11] was added to this type of catalyst as a promoter. However, these catalysts are easily susceptible to sulfur poisoning. As a result, the high potential of metal Pd has been exploited as an active catalyst for the hydrogenation of CO2 to form CH3OH, with Pd-based catalysts reported to be more effective than the conventional Cu-based catalysts [12]. In 1995, Fujitani reported that a Pd/Ga2O3 catalyst was prepared and used in the synthesis of CH3OH from CO2 hydrogenation [13], which can well compete with the classical Cu/ZnO catalyst. Tsang et al. found that the strong metal-support interaction between Pd and the polar (002) surface of the plate form of Ga2O3 facilitated electron transfer compared with the other non-polar surfaces, yielding a higher activity for methanol from CO2 hydrogenation [14]. Subsequently, Pd catalysts supported on ZnO [15], Al2O3 [16], La2O3 [17], CeO2 [18] and Nd2O5 [17] were prepared and used in the selective hydrogenation of CO2 into CH3OH. It well known that alkali and alkaline earth metals are effective promoters for enhancing the CH3OH synthesis activity of Pd/SiO2 [19]. Thus, on the basis of the above experimental facts, Pd is found to be a good catalyst for CO2 hydrogenation. As a result, the development of catalysts for efficient and selective CO2 hydrogenation, which has been one of the major challenges in catalysis, may be achieved through the use of Pd. Recently, zeolites have attracted much attention as a catalyst support because of their ordered channel structures, narrow pore size distributions, and high surface areas and pore volumes; in addition, it is beneficial to form and stabilize small metal nanoparticles inside mesopores. With the use of Pd as a catalyst, the stable and small Pd particles practically could enhance the yield of CH3OH. Thus, the activity enhancement for CH3OH formation would be expected when small Pd particles and metal oxide nanoparticles are incorporated inside the mesopores of the mesoporous materials. In the present work, Pd particles supported on mesoporous materials were prepared and used
1273
in the hydrogenation of CO2 to from methanol, and the reaction conditions were optimized. In addition, the effects of different promoters on the hydrogenation of carbon dioxide were investigated.
2 Experimental 2.1 Materials MCM-41 and SBA-15 were purchased from Nankai University catalyst Co. Ltd., and SiO2 was obtained from Jiding Longhua Chemical Co. Ltd. Pd (NO3)2 was supplied by Shanxi Kaida Chemical Engineering Co. Ltd., and other chemical were purchased from the commercial market. 2.2 Preparation of Catalysts With using aqueous solutions of Pd (NO3)2 as a precursor, supported Pd catalysts were prepared by a pore filling incipient wetness impregnation method. Before preparation, the incipient pore volumes of each support were measured by water. The concentrations of Pd in the precursor solutions were adjusted so that Pd loading of the catalysts prepared using different support materials was kept constant. Next, the impregnated sample was dried at 383 K for 12 h in an electric oven and then calcined in an electric furnace at 673 K for 3 h and labeled as Pd/Sup.. To prepare the promoted catalysts, a series of catalysts were prepared by coimpregnation of aqueous solutions of Pd (NO3)2 and second metal salts (the molar ratio of Pd to the second metal is 2.0). All the samples were dried at 333 K for 12 h and then calcined at 673 K for 3 h; these samples were labeled as PdxMex/2/Sup.. 2.3 Catalyst Characterization 2.3.1 H2 Chemisorption The dispersion of palladium was measured via H2 chemisorption using an Autochem II TPD/TPR equipped with a TCD detector (Micromeritics, America). The catalyst samples were previously reduced at 673 K for 1.0 h under hydrogen atmosphere and then outgassed at 723 K for 2 h using argon. Next, the catalyst samples were allowed to cool down to room temperature, and then 10 vol% H2/Ar pulses were injected at regular interval. The realtime concentrations of all the effluent gases were monitored by the TCD detector at the same time. The dispersity of Pd was calculated assuming H/Pd = 1 chemisorption stoichiometry. The diameter of the Pd particles (dp) was then estimated from the dispersion data assuming a hemispherical morphology.
123
1274
2.3.2 N2 Physical Adsorption The N2 adsorption–desorption characterization of the samples (t-plot pore volume, BET surface area) and the measurements of the Pd dispersion were performed using an Autosorb-1-C/TCD/MS unit (Quantachrome Instruments). Prior to the adsorption measurements, the samples were degassed under vacuum for nearly 12 h at 573 K. The specific surface areas were estimated by using the B.J.H. approach. The mesoporous surface area and pore volume were obtained by applying the t-plot formalism to the desorption branch of the isotherms. 2.4 Catalytic Performance of Supported Pd Catalyst The CO2 hydrogenation activity and selectivity of the prepared catalysts were investigated under pressurized conditions using a high-pressure fixed bed reactor system. The apparatus was equipped with an electronic temperature controller for a furnace, a tubular reactor with an inner diameter of 9 mm, thermal mass flow controllers for gas flows and a back-pressure regulator. 1 g of catalyst was placed in the reactor along with inert quartz sands above and under the catalyst. All the catalysts were reduced in a flow of hydrogen atmosphere at 623 K for 1 h before the reaction. After the H2 reduction, the temperature was reduced to 523 K, and then the feed gas CO2/H2/Ar (H2:CO2 = 3:1) was flowed into the system at the pressure of 3.0 MPa for the activity evaluation. All the products were introduced in the gaseous state and analyzed using a gas chromatograph (GC). Ar, CO, CH4 and CO2 were analyzed using a GC equipped with the thermal conductivity detector (TCD) and a column of activated charcoal, and CH3OH was analyzed using another GC equipped with the flame ionization detector. Ar was used as the internal standard for the GC/TCD analyses.
3 Results and Discussion 3.1 Physico-Chemical Properties of the Catalyst The H2 chemisorption technique was used to estimate the dispersion of the Pd (0) nanoparticles in the reduced catalysts. The palladium dispersity, cubic crystallite size and metallic active sites are presented in Table 1 for Pd/Sup. and Pd-Me/Sup. The results indicated the highest Pd (0) dispersity (25.23 %) was obtained among the catalysts investigated when using MCM-41 as a supporter, and the decreasing order of the dispersity is Pd4/MCM-41 [ Pd4/ SBA-15 [ Pd4/SiO2 (Table 1, entries 1–3). Based on the results, the ordered channel structures, narrow pore size distribution, high surface area and pore volume with
123
Y. Song et al.
supporter were beneficial to decreasing the cubic crystallite size and increasing Pd dispersity and the number of metallic active sites [20]. In the literature, metal oxides have been reported to enhance the activity of the Pd (0) catalyst in the hydrogenation of CO2 to form methanol [16, 21, 22]. Therefore, a series of the metal oxide-doped Pd catalysts were prepared; the data of H2 chemisorption for these catalysts are listed in Table 1. From Table 1, when CaO, K2O or Ga2O3 was added as a promoter, the Pd dispersity was increased (entries 4, 6, and 7); however, the Pd dispersity was decreased when La2O3 was added as a promoter (entry 5). At the same time, the Pd dispersity was decreased from 27.53 to 17.95 % with increasing dosage of Pd from 2 to 8 % (Table 1, entries 4, 7–9). The physical properties of the supported Pd catalysts were investigated using the N2 physisorption technique; the results of the physical property characterization are shown in Fig. 1. Also, Table 2 summarizes the BET surface area and mesopore volume values of the samples. From Fig. 1a, the nitrogen adsorption isotherms of Pd catalysts exhibited representative type-IV curves, and a hysteresis loop was observed in the relative pressure range of 0.5 \ P/P0 \ 1.0 on the adsorption–desorption isotherms. This result illustrated that the mesoporous structure was well preserved and not destroyed when the catalysts were calcined. When using the mesoporous molecular sieve MCM41 or SBA-15 as the supporter, the nitrogen adsorption was increased significantly when the relative pressure P/P0 was over 0.9, which was due to the capillary condensation of nitrogen molecules in the ordered channel structures of the molecular sieve. When the supporter was changed to mesoporous SiO2, the capillary condensation of nitrogen molecules was not observed. From Fig. 1b, the nitrogen adsorption isotherms of the catalysts, to which was added a promoter (K2O, CaOGa2O3 or La2O3), still were type-IV curves; however, the hysteresis loops of the supported catalysts changed to type-HIV from HII. In addition, the values of surface area and pore volume of the supported catalysts were decreased (Table 2); this decrease probably occurred because that Pd and the promoter species are dispersed uniformly inside the mesopores of MCM-41 [21]. To investigate the effect of CaO on the Pd/MCM-41 catalyst, the catalysts with a fixed ratio of Pd to CaO were prepared by changing the Pd loading. The results in Table 2 indicated that the surface area and pore volume was decreased to different degrees when the loading of Pd was increased. 3.2 Effects of Support on the Synthesis of Methanol from Carbon Dioxide The synthesis of methanol from the hydrogenation of CO2 was performed using a fixed-bed flow reactor system with a
Pd-Promoter/MCM-41: A Highly Effective Bifunctional Catalyst for Conversion of Carbon Dioxide
1275
Table 1 H2 chemisorption results for samples Entry
Sample
Pd dispersity (%)
Activity metallic site (9102, mmol/g)
Cubic crystallite size (nm)
1
Pd4/MCM-41
25.23
3.70
9.44
2
Pd4/SBA-15
23.02
4.06
8.65
3
Pd4/SiO2
20.04
4.66
7.54 12.09
4
Pd4Ga2/MCM-41
32.16
2.90
5
Pd4La2/MCM-41
15.54
6.01
5.85
6
Pd4Ca2/MCM-41
25.85
3.16
9.72
7
Pd4K2/MCM-41
38.59
2.42
14.50
8
Pd2Ca/MCM-41
27.53
3.39
5.18
9
Pd6Ca3/MCM-41
23.82
3.92
13.44
10
Pd8Ca45/MCM-41
17.95
5.20
13.50
A B
Pd4Ca2/MCM-41 Pd4La2/MCM-41
Volume adsorbed
Volume adsorbed
Pd4 /MCM-41 Pd4/SBA-15
Pd4/SiO2
0.0
0.2
0.4
0.6
P/P
0.8
Pd4K2/MCM-41
Pd4Ga0.5 /MCM-41
0.0
1.0
0.2
0.4
0.6
0.8
1.0
P/P o
o
Volume adsorbed
C
Pd2Ca1/MCM-41
Pd4Ca2/MCM-41 Pd6Ca3/MCM-41 Pd8Ca4/MCM-41
0.0
0.2
0.4
0.6
0.8
1.0
o
P/P
Fig. 1 N2 adsorption-desorption isotherms of samples
123
1276 Table 2 N2 adsorption data of samples
Y. Song et al.
Sample
BET surface area (m2/g)
Mesopore volume (cm3/g)a 0.6682
Pd4/SBA-15
478
Pd4/SiO2
184
0.6204
Pd4/MCM-41
569
0.6951
Pd4K0.5/MCM-41
412
0.5361
Pd4Ga0.5/MCM-41
424
0.6963
Pd4La0.5/MCM-41
163
0.4104
Pd2Ca0.5/MCM-41
168
0.4269
Pd4Ca0.5/MCM-41
185
0.3399
Pd6Ca0.5/MCM-41
164
0.3809
Pd8Ca0.5/MCM-41
160
0.4054
a
Volume adsorbed at P/P0 = 0.99
stainless steel tubular reactor. Initially, the effects of the support on the hydrogenation of CO2 was investigated using 1.0 g of supported Pd catalyst at 240 °C and 3.0 MPa (the feed volume ratio was CO2/H2 = 1/3); the results are shown in Fig. 2. As observed from Fig. 2a, the catalytic activity of the supported Pd catalyst was affected by the nature of the support, and the conversion of CO2 decreased in the following order: Pd4/MCM-41 [ Pd4/SBA-15 [ Pd4/SiO2. This result was primarily related to the Pd active site (Table 1). From Fig. 2a, Pd4/MCM-41 was found to have higher catalytic activity for the hydrogenation of CO2 compared with Pd4/SBA-15 and Pd4/SiO2; also, the conversion of CO2 was decreased slightly with increasing reaction time and the conversion of CO2 remained at 10 % when the reaction time was prolonged to 9 h. The product distribution of CO2 hydrogenation was evaluated, as shown in Fig. 2b–d. Figure 2 demonstrated that the selectivity of methanol increased with prolonging reaction time in the presence of supported Pd catalysts and all of the by-product selectivity (methane and carbon monoxide) were decreased. When the catalyst was changed to Pd/MCM-41, the selectivity for methanol was over 50 % at 9 h onstream, and the lowest selectivity of CO and the highest selectivity for methane were achieved. This result occurred because the smaller Pd nanoparticles were obtained when MCM-41 was used as the support; such smaller Pd nanoparticles are beneficial to activate hydrogen and perform hydrogenation of CO2. 3.3 Influence of the Promoter on the Synthesis of Methanol from Carbon Dioxide In the literature, metal oxide was reported to improve the catalytic performance of the Pd catalyst in the hydrogenation of CO2 [16]. To enhance the activity of Pd/MCM41, it was doped by metal oxide and used as a catalyst in
123
the hydrogenation of CO2 (Fig. 3). When CaO or La2O3 was added in the Pd/MCM-41 catalyst, the conversion of CO2 and the selectivity for methanol were higher than that of the Pd/MCM-41 alone, whereas the selectivity for methane and CO was decreased in the presence of Pd4Ca2O/MCM-41. This result occurred because of the acid– base properties of a metal oxide [16]. CaO, with a moderate basic strength, could adsorb and activate CO2 on the surface of the metal oxide, thereby enhancing the surface concentration of formate. As a consequence, the hydrogenation of formate to methanol was increased (Fig. 4). Moreover, La2O3, with its amphoteric properties, can adsorb CO2 as a formate species on the basic sites and can activate the formate for the hydrogenation to methanol by the acid sites. When the metal oxide was changed to K2O, with a strong basic character, the conversion of CO2 and selectivity of methanol were decreased, and the selectivity of CO was increased. This result occurred because the strong basic nature of K2O caused an excessive stabilization of the formate intermediate; moreover, CO was formed by the decomposed formate. When the promoter was changed to Ga2O3, the conversion of CO2 and the selectivity of methanol were decreased, and the selectivity of methane was increased. This result occurred because Ga2O3, with Lewis acid properties, could increase the rate of formate hydrogenation [23]; however, the formation of adsorbed formate was diminished by the weak nucleophilic character of the surface OH groups. Simultaneously, H2 could be dissociatively chemisorbed by the gallium cations, which benefited the formation of methane [24]. The acid– base properties of a metal oxide are related to the electronegativity of its metal cation; therefore, metal oxides with high cation electronegativity have hardly any effect on the rate of methanol formation or are only modest promoters, while metal oxides with a moderate basic nature or with amphoteric properties strongly increase the methanol formation. Thus, a clear volcano-shaped correlation is
Pd-Promoter/MCM-41: A Highly Effective Bifunctional Catalyst for Conversion of Carbon Dioxide
1277
100 20
B
A 15
CH3 OH selectivity/%
CO2 conversion / %
80 2
10
5
0
2
60
40
20
0
2
4
6
8
0
10
0
2
4
Time / h
8
6
10
Time / h 70
80
C
2
50
CO selectivity/%
60
CH4 selectivity/%
D
60
2
40
40
30
20
20
10 0
0
2
4
6
8
10
0 0
2
Time/h
4
6
8
10
Time / h
Fig. 2 Reactivity of CO2 hydrogenation over the Pd4/Sup. catalysts with varying supporters
established between cation electronegativity and methanol formation. 3.4 Optimized the Loading of Pd To understand the effect of Pd loading when Ca was used as the promoter, the catalysts PdxCax/2/MCM-41 with the fixed molar ratio of Ca/Pd were prepared and used as a catalyst for the CO2 hydrogenation to form methanol (Fig. 5). The results in Fig. 5 indicated that the conversion of CO2 was increased with increasing Pd loading and that the highest selectivity ([50 %) for methanol was over the Pd6Ca3/MCM-41 catalyst. This result occurred because the mechanism was the bi-functional mechanism for CH3OH formation from CO2 over the metal oxide-promoted Pd catalysts, where the metal oxide promoters stabilize the
adsorbed formate (or CO2) species and the metallic Pd dissociates H2 molecules simultaneously [16]. When the Pd loading was 2 %, low conversion of CO2, low selectivity of CH3OH, and high selectivity of CO were obtained. This result occurred because the active sites of Pd were scarce with the low loading of Pd, and the metal active sites cannot keep equilibrated the ratio of formate to H2 [22]. Compared with the 6 % Pd loading, a higher conversion of CO2 was obtained when the loading of Pd was 8 %, while the selectivity of methanol was decreased. The H2 chemisorption data indicated the number of Pd active sites was hardly increasing when the loading was increased from 6 to 8 %, but the CaO loading was increased, and the capacity of chemisorption H2 and the dissociated hydrogen spilt-over the Pd active sites were decreased, all of which resulted in less active hydrogen for hydrogenation of the
123
1278
Y. Song et al. 40
100 Pd4Ca 0.5 / MCM - 41
A
Pd4 La0.5 / MCM - 41 Pd4 K 0.5 / MCM - 41 Pd4 / MCM - 41
CH3 OH Selectivity / (%)
CO2 Conversion / (%)
30
Pd4Ga 0.5 / MCM - 41
20
10
90
Pd4Ca 0.5 / MCM - 41
80
Pd4 K 0.5 / MCM - 41
B
Pd4 La0.5 / MCM - 41 Pd4 / MCM - 41
70
Pd4Ga 0.5 / MCM - 41
60 50 40 30 20 10
0
0
2
4
6
8
0
10
0
2
4
Time / (h)
6
8
10
Time / (h)
100 60
C
Pd4 La0.5 / MCM - 41
80
Pd4 K 0.5 / MCM - 41 Pd4 / MCM - 41
CO Selectivity / (%)
CH4 Selectivity / (%)
D
Pd4Ca 0.5 / MCM - 41
60
40
Pd4Ca 0.5 / MCM - 41
Pd4Ga 0.5 / MCM - 41
40
20
Pd4 La0.5 / MCM - 41
20
Pd4 K 0.5 / MCM - 41 Pd4 / MCM - 41 Pd4Ga 0.5 / MCM - 41
0
0
0
2
4
6
8
10
0
2
4
6
Time / (h)
Time / (h)
Fig. 3 Reactivity of CO2 hydrogenation over the PdxMex/2/MCM-41 catalysts
CO2 3H2
OH
O
O
OH
Pd
CaO
O
CaO
CH3 H H
C
C
Pd
Pd CaO
OH
CH3OH + H2O
Pd CaO
Fig. 4 Mechanism of hydrogenation of carbon dioxide
123
O
O H H H H HH
H H H H H
Pd
O
OH OH
CaO
H H H H H
Pd
O
C
H O
CaO
8
10
Pd-Promoter/MCM-41: A Highly Effective Bifunctional Catalyst for Conversion of Carbon Dioxide
1279
100
40
A
B
35
CH3 OH Selectivity / (%)
80
CO2 Conversion /(%)
30 25 20 15
60
40
10 20 5 0
0
2
4
6
8
0
10
0
2
Time / (h)
6
8
10
Time / (h) 100
60
C
D
50
80
CO Selectivity / (%)
CH4 Selectivity / (%)
4
40 30 20
60
40
20 10 0
0 0
2
4
6
8
10
0
Time / (h)
2
4
6
8
10
Time / (h)
Fig. 5 Reactivity of CO2 hydrogenation and selectivity for different products over the PdxCax/2/MCM-41
formate formed on the surface of CaO; as a result, the low selectivity of methanol and high selectivity of CO were obtained [21].
4 Conclusions In summary, a series of supported Pd catalysts was prepared using the impregnation method with Pd (NO3)2 as a precursor; these catalysts were used for the synthesis of methanol from CO2. In this process, the pore structure of the support could affect the dispersion of Pd; the higher dispersity and higher numbers of Pd active sites were obtained over the MCM-41 with the regular pore structure and the appropriate pore size. The higher conversion of CO2 and selectivity for methanol were given in the hydrogenation of CO2 over the Pd/MCM-41 catalyst. The base–acid property of the promoter of the supported Pd catalyst was a key for the selective hydrogenation of
carbon dioxide. The highest hydrogenation activity of CO2 and selectivity for methanol obtained over the Pd/MCM-41 was promoted by CaO. The levels of 12.1 % of CO2 conversion and 65.2 % of selectivity for methanol were obtained when the hydrogenation of CO2 was performed for 9 h at 523 K and 3 MPa over the Pd6Ca3/MCM-41 catalyst. Acknowledgments This work is supported by the National Natural Science Foundation of China (No. 21411130188), and Program of International S&T cooperation (2013DFR40570).
References 1. Amenomiya Y (1987) Appl Catal 30:57–68 2. Melia´n-Cabrera I, Granados ML, Terreros P, Fierro JLG (1998) Catal Today 45:251–256 3. Li YN, Ma R, He LN, Diao ZF (2014) Catal Sci Technol 4:1498–1512 4. Ganesh I (2014) Renew Sust Energy Rev 31:221–257
123
1280 5. Goeppert A, Czaun M, Jones JP, Prakash GKS, Olah GA (2014) Chem Soc Rev 43:7995–8048 6. Centi G, Perathoner S (2009) Catal Today 148:191–205 7. Behrens M (2014) Angew Chem Int Ed 53:12022–12024 8. Inui T, Hara H, Takeguchi T, Kim JB (1997) Catal Today 36:25–32 9. Słoczynıˇskia J, Grabowskia R, Olszewskia P, Kozłowskaa A, Stoch J, Lachowska M, Skrzypek J (2006) Appl Catal A 310:127–137 10. Arena F, Barbera K, Italiano G, Bonura G, Spadaro L, Frusteri F (2007) J Catal 249:185–194 11. Graciani J, Mudiyanselage K, Xu F, Baber AE, Evans J, Senanayake SD, Stacchiola DJ, Liu P, Hrbek J, Sanz JF (2014) Science 345:546–550 12. Erdo¨helyi A, Pa´sztor M, Solymosi F (1986) J Catal 98:166–177 13. Fujitani T, Saito M, Kanai Y, Watanabe T, Nakamura J, Uchijima T (1995) Appl Catal A 125:L199–L202 14. Zhou XW, Qu J, Xu F, Hu JP, Foord JS, Zeng ZY, Hong XL, Tsang SCE (2013) Chem Commun 49:1747–1749
123
Y. Song et al. 15. Liang XL, Dong X, Lin GD, Zhang HB (2009) Appl Catal B 88:315–322 16. Gotti A, Prins R (1998) J Catal 175:302–311 17. Matsumura Y, Shen WJ, Ichihashi Y, Okumura M (2001) J Catal 197:267–272 18. Ichikawa M, Shikakura K (1979) Shokubai 21:253–255 19. Cabilla GC, Bonivardi AL, Baltana´s MA (2001) J Catal 201: 213–220 20. Koizumia N, Jiang X, Kugaia J, Song CS (2012) Catal Today 194:16–24 21. Collins SE, Chiavassa DL, Bonivardi AL, Baltanas MA (2005) Catal Lett 103:83–88 22. Chiavassa DL, Barrandeguy J, Bonivardi AL, Baltanas MA (2008) Catal Today 133–135:780–786 23. Collins SE, Baltana´s MA, Fierro JG, Bonivardi AL (2002) J Catal 211:252–264 24. Collins SE, Baltana´s MA, Bonivardi AL (2004) J Catal 226:410–421