Indian Journal of Chemistry Vol. 53A, April-May 2014, pp. 530-534
Catalytic tri-reforming of glycerol for hydrogen generation M Ashwani Kumara, Ch Venumadhava, T V Sagara, M Surendara, N Lingaiaha, G Nageswara Raob & P S Sai Prasada, * a
CSIR-Indian Institute of Chemical Technology, Hyderabad 500 007, India Department of Chemistry, Andhra University, Visakhapatnam 530 003, India.
b
Email:
[email protected] Received 30 January 2014; revised and accepted 18 February 2014 Tri-reforming of glycerol has been carried out, for the first time, on CeO2, ZrO2 and CeO2-ZrO2 supported 10 wt.% Ni catalysts. The catalysts have been characterized by nitrogen adsorption, X-ray diffraction, temperature programmed reduction, scanning electron microscopy, Raman spectroscopy and X-ray photoelectron spectroscopy and tested for glycerol tri-reforming at atmospheric pressure and 673-1073 K. Ni/CeO2-ZrO2 has shown 95% of thermodynamic limit for hydrogen production, at 873 K. Also, the catalyst shows good stability. Keywords: Catalysts, Supported catalysts, Nickel catalysts, Ceria, Zirconia, Glycerol, Tri-reforming, Synthesis gas, Hydrogen production
Global warming due to CO2 release into the atmosphere and the depletion of fossil fuel sources has focussed the attention of the world on the development of technologies for hydrogen as a green fuel. Glycerol is an important biomass resource for hydrogen production due to its renewable nature and availability at cheap rates from biodiesel production. Among the several methods proposed, the value addition of glycerol by hydrogen generation is important since it also improves the economics of biodiesel production1,2. Steam reforming, dry reforming, auto-thermal reforming and aqueous phase reforming3 are the notable methods for the generation of hydrogen from glycerol. Though steam reforming is the most common technique, it offers lower hydrogen selectivity4 and operates at high temperatures5. Dry reforming makes use of CO2, but requires high energy input due to endothermicity and provides syngas of lower H2/CO ratio. During autothermal reforming, the external heat requirement is compensated by the exothermic combustion5. However, the high operating temperatures lead to catalyst deactivation by coking. Cortright et al.6 have developed the aqueous phase reforming process wherein hydrogen is produced from alcohols and sugars. This process operates at low temperature (543 K) but requires high pressure (60 bar) in comparison with steam reforming. To get higher hydrogen selectivity, precious metal catalysts
like Pt have to be used. Cheaper catalysts like Ni lead to severe deactivation due to change in oxidation state of the metal7. Tri-reforming is a new concept carried out in the presence of CO2, H2O and O2. It is a judicious combination of steam reforming, dry reforming and partial oxidation, wherein the endothermic heat requirement due to steam and dry reforming reactions is compensated by the exothermicity of the partial oxidation. This methodology offers the advantages of conducting the reaction under thermoneutral conditions and providing syngas of higher H2/CO ratio. Kang et al.8 tested nickel based catalysts for tri-reforming of methane with CO2. Izquierdo et al.9 studied tri-reforming of biogas. Song & Wei10 carried out both thermodynamic analysis and experimental investigations of tri-reforming of methane using Ni based catalysts. Several thermodynamic analyses have also been carried out on glycerol reforming, dry auto-thermal reforming11 being notable among them. The optimum operating conditions for achieving high hydrogen yield have been reported. To the best of our knowledge, no detailed study has been reported either on the thermodynamic analysis or on catalyst development for the tri-reforming of glycerol. Herein, Ni was selected as a catalyst for reforming because of its high capability towards C-C bond breakage, low cost and easy availability. It also
KUMAR et al.: CATALYTIC TRI-REFORMING OF GLYCEROL FOR HYDROGEN GENERATION
promotes the water gas shift reaction, thus increasing hydrogen production12. In the present investigation, we have studied tri-reforming of glycerol to produce hydrogen, over Ni catalysts supported on CeO2, ZrO2 and CeO2-ZrO2. The non-stoichiometric thermodynamic analysis has also been carried out using the ASPEN PLUS software under the conditions similar to that of the catalytic process for obtaining the maximum achievable hydrogen yield. The results obtained by the experimental evaluation of catalysts are compared with those of the theoretically estimated values. The stability of the best catalyst during the time-on-stream analysis is also reported. Materials and Methods Preparation of catalysts
CeO2 and ZrO2 supports were prepared following the procedure reported by Rossigonol et al.13 Aqueous solution of Ce(NO3)3.6H2O and ZrOCl2.8H2O (Sigma Aldrich, 99.00%) were used for precipitation in the presence of NH4OH solution, maintaining the pH at 9.5. After repeated washings to free the cake of chloride ions, it was dried at 393 K for 12 h and subjected to calcination in air at 873 K for 4 h. Co-precipitation method was used to prepare CeO2-ZrO210 (Ce:Zr atomic ratio 3:1). Briefly, 10% Ni was deposited on the supports (CeO2, ZrO2, CeZrO2) by wet impregnation using aqueous Ni(NO3)2.6H2O (SD Fine Chem. Ltd, Mumbai, 99.00%) solution of required concentration. The solid mass was dried at 393 K for 12 h, ground into fine powder and calcined in air at 1143 K for 6 h. Since the Ni content is constant at 10%, the catalysts are represented as Ni/CeO2, Ni/ZrO2 and Ni/CeO2-ZrO2 hereafter. Characterization of catalysts
The BET specific surface areas of the catalysts were estimated by N2 adsorption at liquid nitrogen temperature using Quadrasorb-SI V 5.06 instrument (Quantachrome Instruments Co., USA). XRD patterns of catalysts were recorded on a Rigaku Multiflex (Rigaku Corporation, Japan) diffractometer using Ni filtered Cu Kα radiation, with a scan speed of 2°/min in the 2θ range of 2°–80° at 30 kV and 15 mA. TPR profiles of the catalysts were generated using a homemade apparatus using 10 vol% of H2 and balance Ar gas mixture. A quartz micro-reactor holding the catalyst was heated at a rate of 10 K/min from 313-1143 K and the effluent was analyzed by a thermal conductivity detector (TCD) attached to a gas
531
chromatograph (Varian, CP-3800). SEM analysis was carried on a Hitachi S-520 electron microscope at an accelerated voltage of 10 kV. Samples were mounted on aluminium stubs using a double adhesive tape. Confocal micro-Raman spectra were recorded at room temperature in the range of 200–1200 cm-1, using a Horiba Jobin-Yvon Lab Ram HR spectrometer with a 17 mW internal He–Ne (Helium–Neon) laser source of excitation wavelength of 632.8 nm. The catalyst samples in powder form (about 5–10 mg) were loosely spread onto a glass slide below the microscope for measurements. XPS measurements were performed on a Kratos Axis 165 (Shimadzu, Japan) photoelectron spectrometer, using Mg Kα radiation (1253.6 eV). Uniform coatings of the powder sample were made on a carbon tape in order to circumvent possible charging of the powered samples. All the spectra were corrected with respect to the C1s binding energy of 284.6 eV. The spectra obtained were de-convoluted using the CASA XPS program. Reforming of glycerol
The reforming reaction was performed in a fixed bed vapour phase microreactor operating at atmospheric pressure, taking 300 mg of the catalyst suspended between two quartz plugs. The catalyst was pre-reduced in hydrogen (50 mL/min) at 923 K for 2 h. Aqueous glycerol solution (30% v/v) was introduced into the reactor at a rate of 5.6 mL/h using a HPLC pump. Carbon dioxide (50 mL/min) and oxygen (10 mL/min) were fed to the reactor and the reaction was carried out in the temperature range of 673-1073 K. The liquid product, after separation in a gas-liquid separator, was analyzed by a FID-fitted gas chromatograph (Shimazdu-GCMS, molecular sieve column) and the gas phase was analyzed chromatographically (Varian, CP-3800, Carboxen1000 column), using a thermal conductivity detector. The material balance made across the reactor was 95±5%. Results and Discussion The surface areas of CeO2, ZrO2, CeO2-ZrO2 were found to be 20, 22 and 25.3 m2/g, respectively, the mixed oxide showing higher surface area as also observed by Roh et al.14 This increase in surface area may be due to substitution of the low radius (0.86 Å) Zr4+ ion in the high radius (1.09 Å) Ce4+ ion in the lattice and also due to the inhibition of the sintering process during calcination15. Impregnation of Ni reduces the surface area (Table 1) due to
INDIAN J CHEM, SEC A, APRIL-MAY 2014
532
Table 1—BET surface area, Ni 2p3/2 binding energy and average NiO particle size of catalysts Catalyst Ni/CeO2 Ni/ZrO2 Ni/CeO2- ZrO2 a From XRD
BET surface area (m2/g)
BE of Ni (eV)
Avg. NiO particle sizea (nm)
3.5 3.7 4.3
854.79 855.91 855.63
32 34 36
blockage of support pores by the metal precursor. High temperature treatment also may have led to the decrease in surface area, an observation similar to that of Perez-Hernandez et al.16 In the XRD patterns (Fig. 1) cubic NiO diffraction peaks are observed at 2θ = 37.12, 43.08, 62.83 and 76.59º as also reported by Song et al.10 Ni/CeO2 shows peaks at 28.34, 32.79, 47.39, 56.16, 58.93 and 69.36°, representing the cubic fluorite structure of CeO2 corresponding to its (111), (200), (220), (311), (222) and (400) planes, respectively. ZrO2 exhibits peaks at 28.2 and 31.5° related to the monoclinic phase dominating the tetragonal phase appearing at 30.2 and 35.2º. Ni/CeO2-ZrO2 shows broader peaks due to lower crystallinity. A slight shift in the peak positions of CeO2 observed in the case of CeO2-ZrO2 indicates the incorporation of Zr in Ce locations. Clearer peaks of NiO are observed in Ni loaded mixed oxide supported catalyst, possibly due to marginal increase in the particle size, as shown in Table 1. Figure 2 shows the TPR patterns of the catalysts with variation in the extent of interaction between NiO and the supports. Literature reveals that CeO2 is only partially reduced10 up to 1123 K. The addition of ZrO2 to CeO2 enhances the reducibility of CeO2. Ni/CeO2 displays two peaks with Tmax at 732 K and 1103 K, corresponding to the reduction of NiO to Ni0 and the partial reduction of Ce4+ to Ce3+, respectively. Ni/ZrO2 exhibits one peak at 713 K and two shoulders at 740 and 829 K, as also observed by Roh et al.17 The first one indicates the reduction of NiO to Ni0 and the other two represent the reduction of the Ni species with various degrees of interaction with ZrO2. The reduction peaks of NiO/CeO2 appear at lower temperatures as compared to those of NiO/ZrO2, suggesting weaker interaction of NiO with CeO2 than with ZrO2. The TPR profile of Ni/CeO2-ZrO2 appears to be a little complicated. Roh et al.17 proposed widening of reduction range when Ni is supported on CeO2-ZrO2 as compared with Ni/CeO2. The first reduction peak appears at a much lower temperature than Ni/CeO2, indicating that a part of the NiO exists
Fig. 1—XRD patterns of the catalysts. [1, CeO2; 2, Ni/CeO2; 3, ZrO2; 4, Ni/ZrO2; 5, CeZrO2; 6, Ni/CeO2-ZrO2. (#): NiO].
Fig. 2—TPR profiles of the catalysts. [1, Ni/CeO2; 2, Ni/ZrO2; 3, Ni/CeO2-ZrO2].
in the well dispersed state on the support and is easily reducible, as revealed by the appearance of Ni crystallites in the XRD pattern of the corresponding sample. The ambiguity in the reduction behaviour at higher temperature arises because of the possible simultaneous reduction of NiO and CeO2, as explained by Song & Wei10. The peak appearing at 1118 K may be assigned unambiguously to the reduction of Ce+4 to Ce+3. The Ni-support interactions in Ni/CeO2-ZrO2 appear to be quite different from those of the others18. The SEM micrographs in Fig. 3 indicate the particle agglomeration in the cases of Ni/CeO2 and
KUMAR et al.: CATALYTIC TRI-REFORMING OF GLYCEROL FOR HYDROGEN GENERATION
533
Fig. 4—Effect of temperature on hydrogen production. [1, Ni/CeO2; 2, Ni/ZrO2; 3, Ni/CeO2-ZrO2; 4, Theoretical data].
the 460 cm−1 band22-24. Peaks at 216, 334, 344, 473 cm-1 indicate the monoclinic ZrO2 phase. The Ni 2p3/2 spectrum of pure NiO25 exhibits a binding energy of 853.6 eV (Supplementary Data, Fig. S2). The binding energy values of Ni 2p3/2 (854.8, 855.9, 855.6 eV) shown in Table 1 are higher than that of pure NiO, indicating strong interaction of Ni with the supports. In the case of CeO2-ZrO2, a slight shift in the Ni peak towards lower side as compared to that of the CeO2 supported catalyst indicates that the Ni species is relatively easily reducible in CeO2-ZrO2 than in Ni/CeO2, as also observed in the TPR patterns. There is a decrease in peak intensity representing increase in particle size for the same loading25, which is in accordance with the results of XRD. Production of hydrogen
Fig. 3—SEM images of the catalysts. [(a) Ni/CeO2; (b) Ni/ZrO2; (c) Ni/CeO2-ZrO2].
Ni/ZrO2, whereas a fine dispersion of Ni is seen on CeO2-ZrO2, supporting the data obtained by TPR analysis. The Raman peaks with their maxima at 302, 335, 346, 380, 476 and 615 cm-1 (Supplementary Data, Fig. S1) represent the monoclinic phase of ZrO2, supporting the evidence obtained from that of XRD analysis19. The intense band of CeO2 at 460 cm−1 is attributed to the F2g mode of materials with its cubic fluorite structure20,21. Ni/CeO2-ZrO2 shows a broad band at 557 cm−1 due to the Zr-O bond22 in addition to
Since it is not possible to directly compare the present results, with those available in the literature as tri-reforming is not reported yet, thermodynamic analysis is adopted to arrive at the maximum possible hydrogen yield. The RGIBBS module of ASPEN PLUS software analyses the chemical system by minimizing the Gibbs free energy. Glycerol, water, oxygen and carbon dioxide constitute the reactants while hydrogen, carbon monoxide, methane and solid carbon are taken as the products, fixing the inlet temperature as 298 K. The theoretically calculated and the experimentally observed values of hydrogen, the desired product in reforming reaction, are compared in Fig. 4. It is clear that the hydrogen yield increases with temperature up to 873 K and on further
INDIAN J CHEM, SEC A, APRIL-MAY 2014
534
Table 2—Flow rates of the products at the reactor outlet at 873 K and the total carbon accumulated during the reaction Catalyst
Ni/CeO2 Ni/ZrO2 Ni/CeO2-ZrO2
Carbon Rate of production of products (mol/h/g cat) accumulation (mg) H2 O2 CO CH4 CO2 0.238 0.251 0.267
0 0 0
0.076 0 0.626 0.0783 0 0.623 0.0846 0 0.617
5.4 2.8 2.3
increase, it decreases. The maximum yield obtained varied in the order: Ni/CeO2-ZrO2 > Ni/ZrO2 > Ni/CeO2. Thus, Ni/CeO2-ZrO2 exhibits the best performance giving 95% of the theoretically possible hydrogen yield at 873 K with 100% conversion of glycerol. The higher performance of the mixed oxide supported catalyst may be due to the existence of a part of easily reducible NiO in the well dispersed state on the support, as observed in the TPR patterns of the catalysts. Particularly, the support CeO2-ZrO2 favours the water gas shift reaction, increasing the yield of H2, as reported by Maria et al.31 Time-on-stream studies at 873 K on Ni/CeO2-ZrO2 shows stable performance up to 25 h of evaluation. The rates of hydrogen and CO obtained are 0.267 and 0.084 mole/h/g cat, respectively. The H2/CO ratio is also high with a value of 3.16. It can also be observed from Table 2 that the amount of carbon accumulated is also much less in the case of Ni/CeO2-ZrO2 as compared to Ni/CeO2 and Ni/ZrO2 catalysts. Conclusions Tri-reforming of glycerol offers complete conversion of glycerol and 95% of theoretical hydrogen yield was achieved. Ni/CeO2-ZrO2 offers better yields than Ni/CeO2 and Ni/ZrO2 due to availability of Ni in dispersed state with the additional advantage of the support catalyzing the water gas shift reaction. The catalyst is also stable under reaction conditions. Supplementary Data Supplementary data associated with this article i.e., Figs S1 and S2, are available in the electronic form at http://www.niscair.res.in/jinfo/ijca/IJCA_53A(4-5) 530-534_SupplData.pdf. Acknowledgement The authors gratefully acknowledge the financial support by Department of Science and Technology, New Delhi sanctioned under Project No. 100/IFD/6777/2010.
References 1 2 3 4 5 6 7 8 9 10 11 12
13 14 15 16
17 18 19 20 21 22 23 24 25 26 27 28
29 30 31
Haas M J, Mc Aloon A J, Yee W C & Foglia T A, Biores Technol, 97 (2006) 671. Pathak K, Mohan Reddy K, Bakhshi N N & Dalai A K, Appl Catal A, 372 (2010) 224. Adhikari S, Fernando S D & Haryanto A, Energy Conv Manag, 50 (2009) 2600. Dieuzeide M L & Amadeo N, Chem Eng Technol, 33 (2010) 89. Wang H, Wang X, Li M, Li S, Wang S & Ma X, Int J Hydrogen Energy, 34 (2009) 5683. Cortright R D, Davda R R & Dumesic J A, Nature, 418 (2002) 964. Iriondo A, Barrio V L, Cambra J F, Arias P L, Guemez M B & Navarro RM, Top Catal, 49 (2008) 46. Kang J S, Kim D H, Lee S D, Hong S I & Moon D J, Appl Catal A, 332 (2007) 153. Izquierdo U, Barrio V L, Requies J, Cambra J F, Guemez M B & Arias P L, Int J Hydrogen Energy, 38 (2013) 7623. Song C & Wei P, Catal Today, 98 (2004) 463. Kale G R & Kulkarni B D, Fuel Process Technol, 91(2010) 520. Valentina N, Michela S, Federica M, Alessandro G, Vladimiro DS, Giuseppe C & Giuseppina C, Appl Catal B,111 (2012)225. Rossigonol S, Madier Y & Duprez D, Catal Today, 50 (1999) 261 Roh H S, Potdar H S & Jun K W, Catal Today, 93 (2004) 39. Xu S & Wang X, Fuel, 84 (2005) 563. Perez-Hernandez R, Martinez A G, Palacios J, Vega-Hernandez M & Lugo V R, Int J Hydrogen Energy, 36 (2011) 6601. Roh H S, Jun K W, Dong J S, Chang S E, Park Y & Joe Y I, Mol Catal A, 181 (2002)137. Larimi A S & Alavi S M, Fuel, 102 (2012) 366. Song Y Q, He D H & Xu B Q, Appl Catal A, 337 (2008) 19. Kaspar J, Fornasiero P, Balducci G, Di Monte R, Hickey N & Sergo V, Inorg Chim Acta, 349 (2003) 217. Chen J, Wu Q, Zhang J & Zhang J, Fuel, 87 (2008) 2901. Cabanas A, Darr JA, Lesterb E & Poliakoff M, J Mater Chem, 11 (2001) 561. Coman S, Parvulescu V, Grange P & Parvulescu VI, Appl Catal A, 176 (1999) 45. Zhang L, Weng D, Wang B & Wu X, Catal Commun, 11 (2010)1229. Oh Y S, Roh H S, Jun K W & Baek, Y S, Int J Hydrogen Energy, 28 (2003) 1387. Lukman H, Zahira Y, Manal I, Wan D W R & Ratna S, Chem Papers, 67 (2013) 703. Chao W, Binlin D, Haisheng C, Yongchen S, Yujie X, Xu D, Tingting L & Chunqing T, Chem Eng J, 220 (2013)133. Chao W, Binlin D, Haisheng C, Yongchen S, Yujie X, Xu D, Liang Z, Tingting L & Chunqing T, Int J Hydrogen Energy, 38 (2013) 3562. Chirag D D & Pant K K, Renewable Energy, 36 (2011) 3195. Kitamura S, Su-enaga T, Ikenaga Na-oki, Miyake T & Suzuki T, Catal Lett, 141 (2011) 895. Maria L B, Francisco P, Gerardo F S & Nora N N, Catal Today, 213 (2013) 58.
