THERMOCHEMICAL TWO-STEP WATER-SPLITTING USING PEROVSKITE OXIDE FOR SOLAR HYDROGEN PRODUCTION

Transcription

1 Proceedings of the Asian Conference on Thermal Sciences 7, st ACTS March 6-, 7, Jeju Island, Korea ACTS-P577 THERMOCHEMICAL TWO-STEP WATER-SPLITTING USING PEROVSKITE OXIDE FOR SOLAR HYDROGEN PRODUCTION Yuta Sugiyama, Nobuyuki Gokon *, Hyun-Seok Cho, Selvan Bellan, Tsuyoshi Hatamachi, Tatsuya Kodama Graduate School of Science and Technology, Niigata University, Japan Center for Transdisciplinary Research, Niigata University, 85 Ikarashi -nocho, Nishi-ku, Niigata 95-8, Japan Dep. of Chem. & Chem. Eng., Faculty of Engineering, Niigata University, Japan Presenting Author: * Corresponding Author: ABSTRACT Perovskite oxide (LaSr(Al, Cr)MnO-δ) powder was examined as a redox material on a thermochemical two-step water-splitting for converting high-temperature heat into chemical fuels. The perovskite oxide was synthesized by using a modified pechini method, and calcined in air atmosphere during 6 hours at temperature of 5-4 C. Firstly, a reactivity of LaSrAlMnO-δ powder was studied for oxygen release and hydrogen production from water via the thermochemical two-step water-splitting cycle of thermal reduction (TR) temperature of 5 C and waterdecomposition (WD) temperature of C. The amount of oxygen and hydrogen, its stoichiometry, and reproducibility of oxygen evolution and hydrogen production were evaluated for different chemical composition of the perovskite oxide. Finally, LaSrCrMnO-δ powder was prepared and tested in order to increase hydrogen productivity. The reactivity and reproducibility of oxygen evolution and hydrogen production were evaluated for different Cr doping amount of the perovskite oxide. The oxygen evolution increased with increasing Mn content of the perovskite oxide. The thermally-reduced LaSrAlMnO-δ powder suffered from high-temperature melting and coagulating of the material during the T-R step. The hydrogen production for LaSrCrMnO-δ powder was enhanced in comparison to LaSrAlMnO-δ powder. KEYWORDS: Hydrogen production, Thermochemical water-splitting cycle, Perovskite oxide, High temperature solar heat, Concentrated solar radiation. INTRODUCTION Development of technology for secure and efficient production of hydrogen is a topical issue pertaining to establishment of the hydrogen economy. Abundant but intermittent solar radiation is the largest renewable energy resource on earth, and tireless efforts have been devoted to investigating the conversion of energy from solar radiation into chemical fuels such as hydrogen. Thermochemical water-splitting cycles can achieve attractive and environment-friendly production of hydrogen and oxygen from the dissociation of water by the provision of high temperature based on concentrated solar radiation []. Thus, solar-driven thermochemical water-splitting cycles are regarded as prospective permanent solutions for the production of hydrogen from water. This approach enables utilization of renewable technologies that are independent of fossil fuels via exploitation of the incomparable magnitude and availability of solar resources []. The thermochemical two-step water-splitting cycle using a metal oxide as a redox pair is generally represented as follows [-5]:

2 M xo y M xo y- + /O (Thermal Reduction or TR step) () MxOy- HO MxOy+ H (Water Decomposition or WD step) () where M is a metal, MxOy is the corresponding metal oxide, and MxOy- is the reduced oxide. In the first step, oxygen is released and evolved from the metal oxide when the oxide is thermally reduced at high temperature. In the second step, the reduced oxide is oxidized by reacting with steam at relatively low temperature, thereby generating hydrogen. The oxide is then recycled to the first step. In general, prospective materials for use as redox materials for the thermochemical cycles can be divided into two categories: metal oxide/metal (e.g., ZnO(s)/Zn(g)), and metal oxide/metal oxide [-5]. The latter case can be further categorized into subcategories, i.e., a single multivalent metal (e.g., FeO4(s)/FeO(s), MnO4(s)/MnO(s)), or the more recently tested CeO(s)/CeO- (s), and multivalent mixed metals (ferrites, doped ceria, hercynite, Perovskites). Solidgas phase transitions occur during the TR step of metal oxide/metal redox cycles, whereas in the redox cycles of metal oxides, the species remain in the solid phase [-5]. Perovskite oxide is known to remain in the condensed state during the two-step reaction, and the perovskite phase maintains its structure across a wide range of content of reduced species,. This leads to formation of oxygen vacancies that promote the transport of oxygen atoms through the material. In contrast to ceria, where vacancies are intrinsically formed at low temperatures, perovskite oxides reported in literatures induced extrinsic vacancies by ionic doping [6-9]. Recently, perovskite oxides-based on (La, Sr, Ca)MnO - and their alumina solid solutions were reported as a potential material of solar thermochemical water-splitting cycle [6, ]. Generally, it was well-known that the redox performances for perovskite oxides provide high extents of thermal reduction and low levels of water-oxidation of thermochemical cycle. In this study, we experimentally screened a reactivity of the perovskite oxides-based on (La, Sr)(Al,Mn)O - for the thermochemical two-step water-splitting cycle. The perovskite oxides for promising chemical composition were improved by substituting Al ions into Cr ions in the perovskite oxide to enhance a rate of hydrogen production during WD step. The redox performances were experimentally evaluated in this study.. EXPERIMENTAL. PREPARATION OF REDOX MATERIALS Perovskite oxide (LaSr(Al, Cr)MnO-δ) powders were synthesized by a wet process using a modified pechini method involving the polymerization of the complex precursor from an aqueous mixed solution of La(NO) 9HO (purity 99.9 %) with Al(NO) 6HO (purity 99.9%), Sr(NO) (purity 98 %), Mn(NO) 9HO (purity 98 %), and Cr(NO) 9HO (purity 99.9 %). These reagents were purchased from Wako Pure Chemical Industries ( Ltd, and used without further purification. Deionized water free of oxygen and CO was prepared by passing N through the deionized water for a few hours. The reagents above were dissolved in the deionized water at appropriate concentrations. Ethylene glycol and citric acid were added in the aqueous solution, thus resulting in the formation of the complex precursor. The solution was stirred in a beaker, and heated at 8 ºC for h and subsequent 7 ºC for.5 h while the stirring process was maintained during the heating. A gel formed from the solution was collected, and dried in air stream in an oven at 8 ºC for h, thus resulting in the formation of the oxide powder. The dried powders were then calcined at 5-4ºC for 6 h in air, before performing the hightemperature cyclic reactions. The synthesized oxide powders were characterized by X-Ray Powder Diffraction (XRD) (MX Labo, MAC Science) using CuK radiation for identification of the phases formed. Table summarizes the abbreviations used for the chemical compositions Table Summary of abbreviations used for the different chemical compositions. Chemical composition La. Sr.7 Al. Mn.7 O -δ La.4 Sr.6 Al.4 Mn.6 O -δ La.5 Sr.5 Al.5 Mn.5 O -δ La.6 Sr.4 Al.6 Mn.4 O -δ La.7 Sr. Al.7 Mn. O -δ La.4 Sr.6 Al.6 Mn.4 O -δ La.6 Sr.4 Al.4 Mn.6 O -δ La.7 Sr. Al. Mn.7 O -δ La.8 Sr. Al. Mn.8 O -δ La.9 Sr. Al. Mn.9 O -δ LaMnO -δ La.7 Sr. Cr. Mn.8 O -δ La.7 Sr. Cr. Mn.7 O -δ La.7 Sr. Cr.4 Mn.6 O -δ La.7 Sr. Cr.7 Mn. O -δ abbreviation LSAM77 LSAM4646 LSAM5555 LSAM6464 LSAM77 LSAM4664 LSAM6446 LSAM77 LSAM88 LSAM99 LM LSCM78 LSCM77 LSCM746 LSCM77

3 described hereafter. Note that LaSr(Al,Mn)O-δ are referred to as written throughout the paper.. ACTIVITY TEST OF REDOX MATERIALS FOR THERMOCHMICAL TWO-STEP WATER-SPLITTING CYCLE The powdered materials of LaSr(Al, Cr)MnO-δ were tested for activity in the two-step water-splitting thermochemical cycle under the same reaction conditions. Approximately g of the powder material was packed into a platinum crucible ( mm in diameter and 7 mm in depth) and mounted on the ceramic bar in a quartz reaction chamber (SSA E45, Ulvac Rico) with an inner diameter of 45 mm (Fig. (a). In the TR step, the powder materials were heated to 5ºC within min using an infrared furnace (RHL VHT E44, Ulvac Rico) while passing N gas (purity %) through the reactor at a flow rate of. dm min at normal state. The temperature of the powder material was controlled using an R-type thermocouple in contact with the platinum crucible. After heating at a constant temperature of 5ºC for min, the powder material was cooled to room temperature. An aliquot of the effluent gas from the reaction chamber was fed through a capillary tube and introduced into a residual gas analyzer mass spectrometer (RG P, Ulvac) to determine the chemical composition of product gases, and to measure the production rate of oxygen. Variations in oxygen partial pressure in the product gases were measured with respect to reaction time during the TR step, and the amounts of oxygen evolved were determined from these variations. The mass spectrometer was equipped with a standard oxygen gas tank that can supply oxygen gas at a constant flow rate of. - mol/s. This oxygen releasing equipment was used to calibrate the relationship between partial pressure and the amount of oxygen evolved. The thermally reduced powder material was pulverized using a mortar and pestle and then packed into a quartz tube reactor (Fig. (b)) with an inner diameter of 7 mm in order to perform the WD step. A gas mixture containing HO and N was introduced into the reactor. This mixture was produced by bubbling N gas at a flow rate of 4 Ncm min through a glass tube into distilled water at 8 or 95ºC. The partial pressures of water vapor in the resultant gas stream were estimated to be 5 and 84%, respectively, by considering the water vapor pressure at bar at these temperatures. The powder material was heated to the given WD temperatures in the ºC range within min in an infrared furnace (RHL E45P, Ulvac-Rico), and the WD step was performed for 6 min. The temperature of the powder material was controlled using a K-type thermocouple in contact with the materials bed located inside the reactor. To determine the amount of hydrogen produced during the WD step, the effluent gas was collected in a bottle by the water displacement method, and the gas compositions were determined using a gas chromatograph (GC 8A, Shimadzu) with a thermal conductivity detector. The powder material was then characterized using XRD. After the WD step, the powder material was again pulverized using a mortar and pestle, mounted in the platinum crucible, and placed in the quartz reaction chamber (Fig. (a)) to repeat the TR step. The TR and WD steps were alternately repeated up to a maximum of three cycles. The powdered material was pulverized after each step in order to examine oxygen and hydrogen productivities at each cycle and to evaluate the reactivity of the powdered material throughout the cyclic reaction. The LaSr(Al, Cr)MnO-δ powders appeared to be sintered at high temperature, and became a porous pellet of fine sintered particles after the TR step at 5ºC. The materials could be pulverized using a mortar and pestle, so that the subsequent WD step to generate hydrogen could be performed. The materials did not appear to be sintered at the temperatures of ºC during the WD step, but remained as a powder. Fig. Experimental setup for (a) TR and (b) WD steps of the thermochemical two-step water-splitting cycle using LaSr(Al, Cr)MnO-δpowders.

4 . RESULTS AND DISCUSSION Fig. shows XRD patterns of LSAM77, LSCM78, LSCM77, LSCM746 and LACM77 synthesized as an original material. As seen for LSCM78 (Fig. ), a series of peaks due to the perovskite structure were evident in the XRD patterns at LSAM77 and all LSCM powders synthesized. No new peaks for the LSCM powders were also observed as well as the LSAM powder synthesized. This result means that LSCM powder retains the perovskite structure of the LM (LaMnO)-based oxide as a single phase, indicating that the Sr, Al and Cr additions were incorporated into the crystal structure as a solid solution. Thermochemical two-step water splitting with LSAM powders was conducted to examine the oxygen and hydrogen productivities and reproducibility of the redox cycle. The TR step was performed at a temperature of 5ºC for all of the materials, while the subsequent WD step was conducted at temperatures of ºC. The amounts of oxygen evolution and hydrogen production per gram of material are listed in Table. For the LSAM77, 5.4 and.8 Ncm /g-material of oxygen was evolved on each cycle, while.6 Ncm /g-material of hydrogen was produced. For the LSAM4646, LSAM5555, LSAM6464, LSAM77, the amounts of oxygen releases for TR step of st cycle were., 9.5, 7.6,.5 Ncm /g-material, respectively. The amounts of hydrogen production for WD step of st cycle were.5-.9 Ncm /g-material. The results mean that the series of LSAM powders provides higher oxygen release than stoichiometric H/O as the Mn contents in the LSAM powders increases, indicating that Mn ion is closely related to redox reaction to produce oxygen and hydrogen. However, hydrogen productivity for the LSAM powders is too low. In addition, it was observed for the high Mn contents of the LSAM powders to enhance the melting and sintering of the reduced material. The melting and sintering was alleviated as La and Al contents in the LSAM powders increases. Then, the following series of the LSAM powders (LSAM4664, LSAM6446, LSAM77, LSAM88, LSAM99 and LaMnO(LM) ) which La and Mn contents both increased were synthesized and tested for the thermochemical two-step water-splitting. For the LSAM6446, LSAM77, LSAM88, LSAM99 and LM with high Mn contents, the melting of the reduced material was not observed during TR step at 5 ºC, but was a mild sintering. Thus, the cyclability was improved for the second series of the LSAM powders. As seen in Table, the amounts of oxygen release for the second series of the LSAM powders decreased, while those of hydrogen production increased in comparison to the first series of the LSAM powders. In addition, for the LSAM77, LSAM88, LSAM99, the hydrogen production was enhanced, and the H/O ratio approached to the stoichiometry. The results for the second series of LSAM powders indicate the need to improve hydrogen productivity LSAM77 LSCM78 LSCM77 LSCM746 LSCM77 Perovskite structure La O Fig. XRD patterns of LSCM powders synthesized before reaction. The LSAM77 was a control for the LSCM powder. Table Oxygen and hydrogen productions for LSAM powders Sample Evolved amount of H₂ and O₂/Ncm³g ¹-material st nd rd Ave. H₂/O₂ LSAM77 O₂ H₂ LSAM4646 O₂ H₂ LSAM5555 O₂ H₂ LSAM6464 O₂ H₂.5 -. LSAM77 O₂ H₂ O₂ LSAM H₂ LSAM6446 O₂ H₂ LSAM77 O₂ H₂ LSAM88 O₂ H₂ LSAM99 O₂..9.8 H₂.6.. O₂ LM. H₂

5 In order to improve the hydrogen productivity of LSAM powder, we newly synthesized a series of LSCM powders (LSCM78, LSCM77, LSCM764, LSCM77) which all Al ions were completely substituted into Cr ions in the perovskite structure. Thermochemical two-step water splitting with LSCM powders was conducted to examine the oxygen and hydrogen productivities and reproducibility of the redox cycle. The amounts of oxygen release and hydrogen production were plotted against cycle number (Fig. ). In comparison to the LSAM77, the LSCM77 shows that the amounts of oxygen release were almost the same through the cycles, but those for hydrogen production evidently increase. Thus, the Cr substitution for the LSAM powders impacts on hydrogen productivity with thermochemical two-step water-splitting cycle. Among the LSCM powders synthesized, the LSCM78 and LSCM77 were superior in the viewpoint of oxygen and hydrogen productivities and cyclicity. (a) Oxygen release (b) Hydrogen production Amount of O production (Ncm/g-material) LSCM78 LSCM77 LSCM764 LSCM77 LSAM77 st nd rd Cycle number Amount of H production (Ncm/g-material) Fig. Cyclic reactivity of LSCM and LSAM powders (a) for oxygen evolution of the st rd cycles during TR step at 5 ºC; (b) for hydrogen production of the st rd cycles during WD step at ºC LSCM78 LSCM77 LSCM764 LSCM77 LSAM77 st nd rd Cycle number Intensity (cps) original TRst WDst TRnd WDnd TRrd WDrd Fig. 4 XRD patterns of the original LSCM78, the solid materials obtained after the TR step and that formed after the subsequent WD step. Fig. 4 shows the XRD patterns of the original material, the material obtained after the st rd TR step, and the material obtained after the st rd WD step using LSCM78. A series of peaks due to the perovskite structure were initially apparent in the spectra of all LSCM powders. In addition, new series of peaks due to the doping Cr element were not observed in the XRD patterns of the original samples. These results indicate that all LSCM powders comprises a Cr-substituted perovskite structure without second phase. After the TR step of the first run, the peaks due to the LSCM powder with the perovskite structure were slightly shifted to lower diffraction angles compared to the peaks of original LSCM78, while no new peak was observed in the XRD pattern. These variations of the series of peaks indicate that non-stoichiometric phase was formed from the LSCM78 powder in the TR step. 5

6 After the WD step of the first run, the peak due to the LSCM with the perovskite structure was slightly shifted toward a higher diffraction angle. For the second and third runs of the cycle, similar behavior of peak shift involving redox reaction in the oxygen release uptake mechanism was observed. Finally, there is a clear benefit of using LSCM powders in order to favor the oxidation (hydrogen production) kinetics which are the limiting factor for the round-trip efficiency of the materials evaluated in this work. In fact, it is shown in the results how the oxidation kinetics are in overall slower than reduction kinetics. Therefore, LSCM powders can favor the overall efficiency of the perovskite oxide-based thermochemical cycles. Challenges for the future are to further improve a hydrogen productivity and oxygen kinetics by optimization of reaction conditions and chemical composition of the LSCM powders. 4. SUMMARY Perovskite oxide (LaSr(Al, Cr)MnO-δ) powder was studied for a thermochemical two-step water-splitting cycle to produce hydrogen from water. The perovskite oxide was synthesized by using a modified pechini method, and calcined in air atmosphere during 6 hours at temperature of 5-4 C. The perovskite oxides of some series of LSAM and LSCM powders were successfully synthesized as a single phase. The reactivity and reproducibility of oxygen release and hydrogen production was evaluated through the cyclic reaction. The results for a series of LSAM powders indicate the amount of oxygen release was enhanced, but those of hydrogen production was low than the stoichiometory. In addition, in the case of high Mn contents in the LSAM powder, it was tend to become melting and sintering of the material at high-temperature during the TR step of the thermochemical cycle. By increasing La contents in the LSAM powders, the melting could be alleviated during the TR step. In order to improve the hydrogen productivity for the LSAM powders, new perovskite oxides, LSCM powders, which Al was substituted into Cr in the perovskite structure was synthesized and tested to evaluate oxygen and hydrogen productivities and cyclicity for the thermochemical two-step water splitting cycle. The hydrogen production for LSCM powders was enhanced in comparison to LSAM powder with the same chemical composition. ACKNOWLEDGMENT This research was partially supported by the Ministry of Education, Science, Sports, and Culture, Grant-in-Aid Challenging Exploratory Research, JSPS KAKENHI Grant Number REFERENCES [] MA Rosen. Advances in hydrogen production by thermochemical water decomposition: a review. Energy ; 5(): [] JA Trainham, J Newman, CA Bonino, PG Hoertz, N Akunuri. Whither solar fuels Current Option in Chemical Engineering ; : 4. [] T Kodama and N Gokon. Thermochemical cycles for high-temperature solar hydrogen production. Chemical Reviews 7; 7: [4] A Steinfeld. Solar thermochemical production of hydrogen a review. Solar Energy5 ; 78 (5): [5] C Agrafiotis, M Roeb, C Sattler. A review on solar thermal syngas production via redox pair-based water/carbon dioxide splitting thermochemical cycles. Renewable and Sustainable Energy Reviews 5; 4: [6] A. H. McDaniel, E. C. Miller, D. Arifin, A. Ambrosini, E. N. Coker, R. O'Hayre, W. C. Chueh and J. Tong.Sr- and Mn-doped LaAlO δ for solar thermochemical H and CO production. Energy Environ. Sci.;, 6, [7] J. R. Scheffe, D. Weibel and A. Steinfeld, Lanthanum Strontium Manganese Perovskites as Redox Materials for Solar Thermochemical Splitting of HO and CO,Energy Fuels ;7, [8]Antoine Demont, Stephane Abanades, and Eric Beche.Investigation of Perovskite Structures as Oxygen-Exchange Redox Materials for Hydrogen Production from Thermochemical Two-Step Water-Splitting CyclesJ.Phys.Chem. C 4 ; 8, [9]A.H. Bork, M. Kubicek, M. Struzik and J. L. M.Rupp. Perovskite La.6Sr.4Cr-xCoxO-s solid solutions for solar-thermochemical fuel production: strategies to lower the operation temperature. J. Mater. Chem A 5 ;, [] A. M. Deml, V. Stevanovi c, A. M. Holder, M. Sanders, R. O'Hayre and C. B. Musgrave.Tunable Oxygen Vacancy Formation Energetics in the Complex Perovskite Oxide SrxLa xmnyal yo, Chem. Mater 4 ; 6,