zirconia-coated foam device prepared by spin coating for solar demonstration of thermochemical water-splitting

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/he

Ferrite/zirconia-coated foam device prepared by spin coating for solar demonstration of thermochemical water-splitting Nobuyuki Gokon a,b,c,*, Tatsuya Kodama b,c, Nobuki Imaizumi c, Jun Umeda c, Taebeom Seo d a

Center for Transdisciplinary Research, Niigata University, 8050 Ikarashi 2-nocho, Nishi-ku, Niigata 950-2181, Japan Department of Chemistry and Chemical Engineering, Faculty of Engineering, Niigata University, 8050 Ikarashi 2-nocho, Nishi-ku, Niigata 950-2181, Japan c Graduate School of Science and Technology, Niigata University, 8050 Ikarashi 2-nocho, Nishi-ku, Niigata 950-2181, Japan d Department of Mechanical Engineering, Inha University, #253 Yanghyundong, Namgu, Incheon 402-751, Republic of Korea b

article info

abstract

Article history:

A Ferrite/zirconia foam device in which reticulated ceramic foam was coated with zirconia-

Received 31 August 2010

supported Fe3O4 or NiFe2O4 as a reactive material was prepared by a spin-coating method. The

Received in revised form

spin-coating method can shorten the preparation period and reduce the coating process as

5 November 2010

compared to the previous wash-coating method. The foam devices were examined for

Accepted 9 November 2010

hydrogen productivity and cyclic reactivity in thermochemical two-step water-splitting. The

Available online 21 December 2010

reactivity of these foam devices were studied for the thermal reduction of ferrite on a laboratory scale using a sun simulator to simulate concentrated solar radiation, while the ther-

Keywords:

mally reduced foam devices were reacted with steam in another quartz reactor under

Thermochemical cycle

homogeneous heating in an infrared furnace. The most reactive foam device, NiFe2O4/m-ZrO2/

Two-step water-splitting

MPSZ, was tested for successive two-step water-splitting in a windowed single reactor using

Hydrogen

solar-simulated Xe-beam irradiation with a power input of 0.4e0.7 kWth. The production of

Ferrite

hydrogen continued successfully in the 20 cycles that were demonstrated using the NiFe2O4/

Ceramic foam

m-ZrO2/MPSZ foam device. The NiFe2O4/m-ZrO2/MPSZ foam device produced hydrogen at

Spin coating

a rate of 1.1e4.6 cm3 per gram of device through 20 cycles and reached a maximum ferrite conversion of 60%. ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.

1.

Introduction

Thermochemical water-splitting using concentrated solar energy has attracted interest as a renewable method for producing hydrogen from water [1e5]. The direct thermal dissociation of water requires temperatures above 2500
C to achieve a reasonable degree of dissociation; further, it requires an effective technique for the separation of H2 and O2 to avoid their explosive recombination. In addition, the requisite high temperatures restrict the materials that can be used in the

construction of a reactor, which makes it difficult to design a reactor that can perform the thermal dissociation of water. Thermochemical water-splitting is divided into several steps that allow the operation to proceed at relatively moderate upper temperatures and without the problem of H2/O2 separation. The most interesting approach involves a two-step cycle based on metal-oxide redox reactions. This two-step water-splitting thermochemical cycle using a redox pair of Fe3O4/FeO was first proposed by Nakamura [6] and proceeds as follows:

* Corresponding author. Center for Transdisciplinary Research, Niigata University, 8050 Ikarashi 2-nocho, Nishi-ku, Niigata 950-2181, Japan. Tel./fax: þ81 25 262 6820. E-mail address: [email protected] (N. Gokon). 0360-3199/$ e see front matter ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2010.11.034

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

Fe3 O4 /3FeO þ 1=2O2

1

DH
298K ¼ 319:5 kJ mol

H2 O þ 3FeO/Fe3 O4 þ H2

(1) 1

DH
298K ¼ 33:6 kJ mol

(2)

The first step, the high temperature thermal reduction (T-R) of Fe3O4, is highly endothermic, whereas the second step, the low-temperature water-decomposition (W-D) by FeO, is slightly exothermic. Several oxide pairs of multi-valent metals [7e11] have also been developed and evaluated as redox materials. Mixed solid solutions of (Fe1xMx)3O4/(Fe1xMx)O between the redox system of Fe3O4/FeO and M3O4/MO are expected to decrease the reaction temperatures relative to those used in the Fe3O4/FeO system. The possibility exists of thermodynamically combining good H2 yields in the Fe3O4/ FeO system with the low reduction temperature of an M3O4/ MO (M ¼ Mn, Co, Mg) system [1,7]. One of the problems presented by the (Fe1xMx)3O4/ (Fe1xMx)O redox pair as the reactive working material, however, is the rapid deactivation of the iron oxide particles during the cyclic reaction. This takes place because of the high-temperature melting and sintering of the iron oxide particles, which results in a rapid decrease in the iron oxide surface area. The present authors successfully demonstrated a repeatable two-step water-splitting cycle using Fe3O4 or other metal-doped iron oxides (ferrites) supported on monoclinic ZrO2 (m-ZrO2) [12e16]. The ZrO2 support alleviated the coagulating or sintering of the solid reactant of the iron oxides, so the two-step reaction could be repeated with a relatively strong degree of activity in a temperature range of 1000e1400
C. The m-ZrO2 support is thus a chemically inert material for a repeatable twostep water-splitting cycle using Fe3O4 or ferrites at high temperatures. The present authors also demonstrated that a new redox reaction occurs on Fe3O4 supported on yttrium-stabilized cubic zirconia (Fe3O4/c-YSZ) [17e19]. When c-YSZ particles were used as a support for Fe3O4, the XRD studies indicated that the new cyclic reaction proceeds as follows:

2015

proposed solar chemical reactors are assumed to be combined with developed solar reflective towers, parabolic dishes, or beam-down optics. Thus, solar reactors-receivers are equipped with a transparent quartz window to allow the direct heating of redox working materials by the passage of concentrated solar radiation, which yields high temperatures at a reaction site. The present authors have developed ceramic foam devices whose foam matrix is made of MgO-partially stabilized zirconia (MPSZ); the foam matrix is coated with zirconia and ferrite particles (ferrite/zirconia/MPSZ foam device). The advantage of using a ceramic foam device is that it makes possible the effective absorption of light irradiation due to the large specific surface area. Multicycling of the two-step watersplitting process have been demonstrated using these foam devices [37e40]. When these foam devices were prepared previously using a wash-coating method, however, a great deal of time was required for the coating processes with the zirconia and ferrite particles, including a number of calcinations at a high temperature of 1000
C (for example, 40e50 coating steps with NiFe2O4/m-ZrO2 particles and same number of calcinations [39]). In the present study, the authors address a method of preparation of NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/MPSZ foam devices that can shorten the preparation period and reduce the coating process as compared to the previous preparation method. The water-splitting foam devices are prepared using a new method of spin coating. The activity and reactivity of the water-splitting foam devices are first examined using solar-simulated Xe-light irradiation for the T-R step and then using another quartz reactor and an electric furnace for the subsequent W-D step. Most reactive foam device of NiFe2O4/m-ZrO2/MPSZ was tested in a windowed single reactor for successive two-step water-splitting under Xebeam irradiation to simulate solar radiation. The hydrogen productivity and reactivity of the NiFe2O4/m-ZrO2/MPSZ foam devices were evaluated.

3þ Fe2þ x Yy Zr1y O2y=2þx þ x=2H2 O ¼ Fex Yy Zr1y O2y=2þ3x=2 þ x=2H2

2þ Fe3þ x Yy Zr1y O2y=2þ3x=2 ¼ Fex Yy Zr1y O2y=2þx þ x=4O2

(3)

2.

Experimental procedure

(4)

2.1.

Synthesis of ferrite/zirconia particles

2þ

where y  0.15. Fe -YSZ is first formed by the high-temperature reaction between the YSZ and the Fe3O4 supported on the YSZ, at a temperature above 1400
C and in an inert atmosphere: x=3Fe3 O4 þ Yy Zr1y O2y=2 ¼ Fe2þ x Yy Zr1y O2y=2þx þ x=6O2

(5)

When Fe3O4/c-YSZ ( y ¼ 0.15) is thermally reduced at 1400
C, the x value as estimated by chemical wet analysis is 0.08 [18,19]. In the subsequent hydrolysis reaction, the formed Fe2þ-YSZ is reacted with steam at 1000
C to generate hydrogen and is oxidized to Fe3þ-YSZ. The repeatability and stoichiometry of hydrogen/oxygen production has been examined using T-R step temperatures of 1450e1600
C [20]. Different concepts regarding solar reactor-receivers that use ferrites or mixed iron oxides have been proposed and demonstrated by several research groups [5,21e40]. These

Zirconia-supported ferrites were prepared as follows. Monoclinic zirconia (m-ZrO2), or cubic yttrium-stabilized zirconia (c-YSZ) doped with 8 mol%-Y2O3 ((ZrO2)1z(Y2O3)z at z ¼ 0.08), were used as supports for Fe3O4 or NiFe2O4. The m-ZrO2 support had a purity of 98%, a particle size smaller than 1 mm, and a BrunauereEmmetteTeller (BET) relative surface area of 13 m2 g1. The c-YSZ support had a purity of 99%, a particle size smaller than 1 mm, and a BET relative surface area of 7  2 m2 g1. The zirconia-supported ferrites were prepared by coating the m-ZrO2 or c-YSZ particles with the ferrite (Fe3O4 or NiFe2O4), using the aerial oxidation method of aqueous suspension of the Fe(II) hydroxide with/without Ni(II) hydroxide [12e20]. The loadings of Fe3O4 or NiFe2O4 in the zirconia-supported ferrites were about 20% on a weight basis. The zirconia-supported ferrites (Fe3O4/m-ZrO2, Fe3O4/c-YSZ, NiFe2O4/m-ZrO2 and NiFe2O4/c-YSZ) were calcined at 900
C in

2016

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

b 4500

a 3000

3500

In ntensity / cps

In ntensity / cps

c-YSZ Ni F e 2 O4

4000

m-ZrO2 Ni Fe 2 O4

2500 2000 1500 1000

3000 2500 2000 1500 1000

500

500

0 20

30

40

50

60

70

2θ CuKα / degrees

80

0 20

30

40

50

60

70

80

2θ CuKα / degrees

Fig. 1 e XRD patterns of the prepared (a) NiFe2O4/m-ZrO2 and (b) NiFe2O4/c-YSZ samples.

an N2 atmosphere before the two-step water-splitting cyclic reactions were employed. The solid samples were characterized using X-ray diffractometry (XRD) with CuKa radiation (MAC Science, MX-Labo) for identification of the formed phases. Fig. 1 shows a typical XRD pattern of the prepared NiFe2O4/m-ZrO2 and NiFe2O4/c-YSZ samples. For both samples, a spinel phase that was due to the ferrite and monoclinic or cubic zirconia phase were observed in the XRD patterns. To determine the Fe3O4 or NiFe2O4 loadings on the zirconia support, the zirconia-supported ferrite was dissolved in an HCl aqueous solution and the dissolved solution containing Fe (and Ni) ions was analyzed using inductively coupled plasma-atomic emission spectrometry (ICP-AES, Seiko Instrument SPSe1500PV).

2.2.

Activity tests of ferrite/zirconia particles

The four zirconia-supported ferrites Fe3O4/m-ZrO2, Fe3O4/cYSZ, NiFe2O4/m-ZrO2 and NiFe2O4/c-YSZ were tested for reactivity for the two-step water-splitting cycle under the same reaction condition. Fig. 2 shows the experimental apparatus for (a) the thermal reduction (T-R) step, and (b) the subsequent water-decomposition (W-D) step of the two-step water-splitting cycle. About 1 g of the zirconia-supported ferrite was mounted in a platinum cup (with a 10 mm diameter and 7 mm length) that was then placed on the ceramic bar in a quartz reaction chamber (ULVAC-RICO, SSA-E45) with an inner diameter of 45 mm (Fig. 2a). The zirconia-supported ferrite was first heated in an infrared furnace (ULVAC-RICO,

RHL-VHT-E44) up to the desired temperature (1400, 1450, and 1500
C) of the thermal reduction step within 10 min, while an N2 gas (99.999% N2) was passed through the reactor at a flow rate of 1.0 N dm3 min1. The temperature was controlled using an R-type thermocouple that was in contact with the platinum cup. After the solid sample was heated at a constant temperature for 30 min to carry out the T-R step, it was cooled to room temperature. After the T-R step, the solid sample was taken from the quartz reaction chamber, and then pulverized using a mortar and pestle. After pulverization, the solid sample was packed in another tubular quartz reactor (Fig. 2b) with an inner diameter of 7 mm in order to carry out the W-D step. A H2O/N2 gas mixture was produced by passing N2 gas through the distilled water at 80
C at an N2 flow rate of 4 N cm3 min1, after which it was introduced into the reactor. The partial pressure of the steam in the H2O/N2 mixture for the system was estimated to be 47% with the steam vapor pressure at 80
C and 1 atm. The reactor was heated to 1000
C in an infrared furnace (ULVAC-RIKO, RHL-E45P) within 10 min. The temperature was controlled using a K-type thermocouple that was in contact with the sample bed located inside the reactor. The W-D step was performed at 1000
C for 60 min. To determine the amounts of hydrogen evolved during the W-D step, the effluent was collected in a bottle by means of water displacement. After the W-D step, the volume of the collected effluent was measured and the gas compositions were determined by gas chromatography using a thermal conductivity detector (TCD) (Shimadzu, GC-8A). The solid powder

Fig. 2 e Experimental setups for reactivity test of zirconia-supported ferrite particles, for (a) the thermal reduction (T-R) step and for (b) the water-decomposition (W-D) step.

2017

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

Fig. 3 e Preparation of NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/MPSZ foam devices by spin coat method.

samples were subjected to X-ray diffractometry (XRD) with CuKa radiation for identification of the formed phases. After the W-D step, the solid sample was again pulverized with a mortar and pestle. It was then mounted in a platinum cup and again placed on the ceramic bar in a quartz reaction chamber, as shown in Fig. 2a. The solid sample was then heated at 1400e1500
C in an N2 gas stream in the same manner as in the first heat treatment mentioned above, in order to perform the T-R step. For repetition testing of the cyclic reactions, the T-R and W-D steps were repeated in alternation 6 times.

2.3.

Preparation of ferrite/zirconia/MPSZ foam devices

Fig. 3 shows a schematic diagram of the preparation of NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/MPSZ foam devices using the spin-coating method. The matrix of the ceramic foam devices was made of MPSZ. The white-colored MPSZ foam had a diameter of 30 mm, a thickness of 10 mm, and a cell size of 7 cpi (number of cells per linear inch). The diskshaped MPSZ foam was impregnated with a ferrite-containing slurry aqueous solution consisting of 10 g of NiFe2O4/m-ZrO2 (or Fe3O4/c-YSZ) powder, 12.5e14 cm3 of distilled water, 0.15 g of dispersant (Sodium polyacrylate), and 0.10 g of binder (acrylic resin). The impregnated MPSZ foam was set in a spincoater and then rotated at a rate of 600 rpm in order to remove the excess slurry on the foam. After rotation, the foam was dried at 100
C for 1 h in an air stream. Next, the

temperature was increased stepwise to 1100
C (200
C for 0.5 h, 300
C for 0.5 h, 500
C for 0.5 h, 800
C for 0.5 h, and then 1100
C), and the foam device was calcined for 1 h in an air stream. This process of spin coating was repeated 8e10 times. Next, the NiFe2O4/m-ZrO2- or Fe3O4/c-YSZ-loaded MPSZ foam was calcined at 1400
C for 2 h in an N2 stream. The loading amount of the NiFe2O4/m-ZrO2 (or Fe3O4/c-YSZ) powder was estimated from the difference in the weight before and after loading. The ferrite loading and weight of the prepared foam devices are listed in Table 1. As a result, NiFe2O4 loadings of 5.9 and 6.9 wt% were obtained with respect to the total weight of the materials (NiFe2O4, m-ZrO2 and MPSZ foam), and Fe3O4 loading of 7.9 wt% was obtained with respect to the total weight of the materials (Fe3O4, c-YSZ and MPSZ foam). If a slurry with a high solid content is used for coating a foam matrix in a wash-coating method, openings in the

Table 1 e Loading of ferrite and weight of foam devices prepared. Foam device

Fe3O4/c-YSZ/MPSZ NiFe2O4/m-ZrO2/MPSZ NiFe2O4/m-ZrO2/MPSZ

Ferrite loading/%

Weight of foam device/g

Porosity/cpi

7.9 5.9 6.9

12.8 10.3 11.5

7 7 7

2018

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

foam structure will become clogged with the dense slurry. For this reason, it was necessary to keep the density of this slurry low enough in the wash-coating method [37e40]. The preparation of the foam device required a great deal of time to be spent in the coating processes with zirconia and ferrite particles, and also required a number of calcinations at high temperatures. In the case of the spin-coating method, however, after impregnating the ceramic foam with a slurry containing high solid content, the impregnated foam is spun at a high rate that is sufficient to remove the excess slurry loaded on the foam and to prevent clogging of the pores of the foam structure. Thus, the spin-coating method makes it possible to use slurry with much higher solids content in the coating of ceramic foam than is used in a wash-coating method. Therefore, with a much-reduced coating process and preparation time, the foam device was prepared using dense slurry. In short, the advantages of the spin-coating method in the preparation of water-splitting foam devices are: 1) the preparation period can be shortened, and 2) the coating processes are reduced in comparison to a wash-coating method.

2.4.

Activity tests of ferrite/zirconia/MPSZ foam devices

Fig. 4 shows the laboratory-scale experimental setups of the T-R and W-D steps of two-step water-splitting carried out by using the prepared foam devices. The reactor used for the T-R step was composed of double quartz tubes that contained the foam device (Fig. 4a). The inner diameters of the inner and outer tubes were 31 and 39 mm, respectively. The prepared foam device was placed on a porous quartz plate fixed in the inner quartz tube and was fixed with quartz wool inside the quartz tube so it would not fall off. A high-purity N2 gas stream of 0.3 dm3 min1 was introduced into the inner tube of the reactor and passed through the foam device in the outer tube. The reactor was exposed to a 6-kW Xe lamp (Cinemeccanica, ZX-8000H) that simulated solar radiation, with the central axis of the reactor parallel to the axis of the concentrator of the sun simulator. The surface of the foam device was set on the focal spot of the concentrated light irradiation. The diameter of the focal spot was fixed at 6 cm. The energy flux density of the Xelamp beam spot was measured beforehand using a heat flux

transducer with a sapphire window attachment (Medtherm, 64-100-20/SW-1C150). After the prepared foam device was preheated to 800
C, one side of the disk-like device was irradiated for 30e60 min in order to perform the T-R step. The reaction temperature was measured at the center position of the light-irradiated surface using an R-type thermocouple, and temperatures of 1450e1500
C were recorded. A thermally reduced foam device was then set in a quartz tube reactor (Fig. 4b) with an inner diameter of 32 mm in order to perform the W-D step. As in the T-R step, the foam device was fixed with quartz wool inside the quartz tube to prevent it from falling off. After an H2O/N2 gas mixture was produced by bubbling N2 gas that was passed through distilled water heated to 93
C at a flow rate of 10 cm3 min1, it was then introduced into the reactor. The partial pressure of the steam in the H2O/N2 mixture was estimated to be 75% using a steam vapor pressure of 93
C and 1 bar. The thermally reduced foam device was mounted in the reactor and then heated to 1100e1200
C in an infrared furnace (ULVAC-RIKO, RHLP65CP), and the temperature of the foam device was controlled using a K-type thermocouple placed at the center of the irradiated surface of the device. The W-D step was performed at 1100e1200
C for 80 min. The T-R and W-D steps were repeated in alternation 6 times. To determine the amount of hydrogen evolved during the W-D step, the effluent was collected in a bottle using the water displacement method. After the W-D step, the amount of collected effluent was measured and the gas composition was determined by gas chromatography (GC; Shimadzu, GCe8A) using a TCD.

2.5. Successive two-step water-splitting in a single reactor Fig. 5 shows the experimental setup for successive two-step water splitting in a single reactor. The reactor was composed of double quartz tubes that contained the NiFe2O4/m-ZrO2/ MPSZ foam device. The NiFe2O4/m-ZrO2/MPSZ foam device, which was not fixed in place, was placed on the fixed quartz plate inside the reactor and was alternately subjected to the TR and W-D steps (Fig. 5a). In this series of experiments, twostep water-splitting using the NiFe2O4/m-ZrO2/MPSZ foam

Fig. 4 e Experimental setups of (a) T-R and (b) W-D steps of a two-step water splitting with the NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/MPSZ foam devices. The T-R and W-D steps were respectively performed using different two reactors.

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

2019

Fig. 5 e Experimental setup of (a) a single reactor for performing cyclic two-step water-splitting using a foam device. The T-R and W-D steps were alternatively performed in a same reactor by solar-simulated Xe-light irradiation. The temperature of the foam device was measured by using an R-type thermocouple and at (b) the following three points: the center position of the light-irradiated surface (Tcenter); the edge position of that (Tedge); the center position of the back side (Tback).

device was performed by direct irradiation with concentrated Xe-light without preheating of the foam device. The temperature of the foam device was measured using an R-type thermocouple located at the following three points: the center position of the light-irradiated surface (Tcenter), a position at the edge of this surface (Tedge), and the center position of the back side (Tback) (Fig. 5b). This reactor was placed beneath the 6-kW Xe-arc lamp of the sun simulator (Nihon Koki, UXL6000H) with the central axis of the reactor aligned with the axis of the oval concentrator of the sun simulator. The concentrator of the sun simulator reflected the Xe-lamp beam downwards to the focal spot. The surface of the foam device was set on the focal spot of the concentrated light irradiation. The diameter of the focal spot was fixed at 6 cm. The energy flux density of the Xe-lamp beam spot was measured beforehand using a heat flux transducer with a sapphire window attachment (Medtherm, 64-100-20/SW-1C150). Except for where otherwise indicated, the reactor dimensions and the experimental conditions were the same as those described for the abovementioned activity test of the NiFe2O4/ m-ZrO2/MPSZ foam device. One side of the foam device was irradiated for 30 min in an N2 gas stream in order to perform the T-R step. The temperature of the foam device for the T-R step was controlled at Tcenter ¼ 1480e1500
C using an R-type thermocouple placed at the center of the irradiated surface of

the device. After the T-R step, the input power of the Xe-lamp beam was decreased and the gas stream introduced into the bottom of the reactor was changed from N2 gas to an H2O/N2 gas mixture in order to perform the W-D step. The thermally reduced foam device was subjected to steam for 60 min. The temperature of the foam device for the W-D step was controlled at Tcenter ¼ 1200
C using an R-type thermocouple placed at the center of the irradiated surface of the device. The T-R and W-D steps were repeated in alternation for 20 cycles. After the W-D step in each cycle, the remaining hydrogen and steam in the reactor were exhausted by passing N2 gas through the reactor for 30 min. This passing of N2 gas was continued for another 30 min before the next T-R step was performed.

3.

Results and discussion

3.1.

Activity tests of ferrite/zirconia particles

Fig. 6 shows the XRD patterns of the original NiFe2O4/m-ZrO2, and of the solid materials obtained after the first T-R step at 1500
C and the subsequent W-D step. After the T-R step, the NiFe2O4 was partially reduced to Ni-containing wustite phase or NiyFe1-yO, as corroborated by XRD analysis. After the

2020

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

m-ZrO2

Original After T-R step After W-D step

400

Intensity / cps

NiyFe1-yO NiFe2O4

300

200

100

0 40

41

42

43

44

45

2θ CuKα / degrees

Fig. 6 e XRD patterns of the NiFe2O4/m-ZrO2 samples original, after the 1st T-R step at 1500
C and after the subsequent W-D step at 1000
C.

subsequent W-D step, the wustite peaks disappeared, indicating that the reduced phase was completely oxidized back to NiFe2O4 on the inert m-ZrO2 support. In the cases in which the T-R step was performed at 1400 and 1450
C, similar changes in the XRD pattern were observed following the T-R and W-D steps. Fig. 7 shows the hydrogen production profiles for the W-D step of the first cycle for the solid samples conducted after thermal reduction at 1400, 1450, and 1500
C. The hydrogen production rates (N cm3 min1) per weight of NiFe2O4/m-ZrO2 are plotted with respect to the reaction time. For the first cycle of two-step water splitting in each profile, a hydrogen production peak was observed at a temperature of around 1000
C, after which the rate of hydrogen production decreased rapidly at 10e25 min. In addition, the maximum rate of hydrogen production was significantly enhanced by higher thermal reaction temperatures. Similar profiles were obtained in the W-D steps of other cycles. The amount of evolved hydrogen per gram of NiFe2O4/mZrO2 in each cycle was plotted against the cycle number, as shown in Fig. 8. The other powder samples of Fe3O4/m-ZrO2, Fe3O4/c-YSZ, and NiFe2O4/c-YSZ were examined with regard to hydrogen productivity and cyclic reactivity under the same reaction conditions as in the two-step water-splitting cycle, and the results were also plotted in Fig. 8. Each powder sample could cyclically produce hydrogen in the two-step water splitting, and the level of evolution was nearly the same in each of the six cycles. At T-R temperatures of 1400e1500
C, NiFe2O4/m-ZrO2 showed the highest degree of activity among the ferrite/zirconia samples tested. The average amount of evolved hydrogen through the six cycles was plotted against the thermal reduction temperature, as shown in Fig. 9. When the T-R step was performed at a temperature of 1500
C, the hydrogen productivity for Fe3O4/m-ZrO2 increased greatly and was close to the level of productivity observed for NiFe2O4/mZrO2. The greater activity of Fe3O4/m-ZrO2 at a T-R temperature of 1500
C is due to the melting of Fe3O4 and the reduced phase of wustite. According to the thermodynamic equilibrium calculation of Allendorf et al. [41], Fe3O4 is completely melted at 1730 K (1457
C), and the reduced solid phase of wustite (FeO) is formed up to a T-R temperature of 1705 K

(1432
C). When the temperature exceeds 1457
C, all metaloxide phases are melted; i.e., the literature indicates that Fe3O4 and the wustite phase in Fe3O4/m-ZrO2 is completely melted during the T-R step at 1500
C, but the melted wustite phase and the remaining Fe3O4 solid phase coexist at 1450
C. The complete melting of Fe3O4 on the surface of m-ZrO2 particles enhances the oxygen release during the T-R step due to the higher ionic diffusivity of liquid phase as compared to solid phase. From a practical point of view, however, the complete melting of Fe3O4 and wustite phase during the T-R step is undesirable for thermochemical two-step water splitting, as it makes it more difficult to react with steam for hydrogen production in the subsequent W-D step. On the contrary, for NiFe2O4, thermodynamic calculations [41] indicate that the reduced phase of Ni-doped wustite is not melted at a T-R temperature less than 1800 K (1527
C), and NiFe2O4 is completely reduced at temperatures greater than 1700 K (1427
C). From the calculation results, it is considered for NiFe2O4/m-ZrO2 that NiFe2O4 and the reduced phase remain solid phases without any melting phases during the T-R step at 1500
C. For a repeated two-step reaction in a single solar chemical reactor, this is an advantage of NiFe2O4/m-ZrO2 over Fe3O4/m-ZrO2. The ferrite conversions to wustite in the T-R steps were estimated from the amounts of hydrogen that were produced, assuming that the wustite phase formed in the T-R step was completely reoxidized to ferrite in the subsequent W-D step, as corroborated by XRD analysis (Fig. 6). The average NiFe2O4 conversions for one cycle were 59%, 66%, and 70% for T-R step temperatures of 1400, 1450 and 1500
C, respectively. As can be seen in Fig. 9, it is quite interesting that the hydrogen productivity of NiFe2O4/c-YSZ was much lower than that of NiFe2O4/m-ZrO2. A similar phenomenon can be observed for Fe3O4/m-ZrO2 and Fe3O4/c-YSZ when the T-R step is performed at 1500
C; the reactivity of Fe3O4/c-YSZ was much lower than that of Fe3O4/m-ZrO2. This would be due to the fact that NiFe2O4 or Fe3O4 could chemically react with the c-YSZ support to form Fe(II)- and Ni(II)-containing YSZ during the first T-R step. In the case of NiFe2O4/c-YSZ, if Fe(II)- and Ni (II)-containing YSZ is formed during the T-R step, the remaining Ni-ferrite will be depleted of nickel, leading to the formation of Ni-deficient Ni-ferrite (Ni1xFe2þxO4; 0 < x < 1). According to a previous paper [15], in comparison to the stoichiometric composition of NiFe2O4, the hydrogen productivity of Ni-deficient Ni-ferrites was greatly decreased. Detailed knowledge of the reaction mechanism of NiFe2O4/cYSZ, however, will require further investigation, and will thus be reported elsewhere.

3.2.

Activity tests of ferrite/zirconia/MPSZ foam devices

The energy flux density of the incident solar-simulated Xebeam on the irradiated surface of the foam device was measured prior to the T-R step. The central peak of flux density reached 998 kW m2. The average flux density on the exposed surface of the foam device was 470 kW m2. From the result of the energy flux density, the input power QHeat of the incident Xe-beam was estimated at 0.3 kW. In the T-R step, NiFe2O4/m-ZrO2/MPSZ (5.9 wt% of ferrite loading) and Fe3O4/c-YSZ/MPSZ (7.9 wt% of ferrite loading)

2021

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

1200

Average hydrogen production amount / Ncm3-g-1-material

16

1000

0.8

0.6

1400

TR

1450

TR

1500

TR

800 600

0.4 400 0.2

Temperature / ºC

Productio rate of hydro Production hydrogen/ Ncm3min-1-g-1-material Nc

1

200

0

14 12 10 8 6

2 0 1350

0 0

10

20

30

40

50

Fe3O4/m-ZrO2 Fe3O4/c-YSZ NiFe2O4/m-ZrO2 NiFe2O4/c-YSZ

4

60

1400

Time me / min

1450

1500

1550

T-R temperature / ºC

Fig. 7 e Hydrogen production profiles in the W-D step of the 1st cycle by the NiFe2O4/m-ZrO2 particles when conducting the T-R step at 1400, 1450 and 1500
C.

Fig. 9 e Variations in the average amount of hydrogen evolved during the repeated water-splitting reaction as a function of the T-R step temperature.

foam devices were uniformly preheated at 800
C in an electric furnace and one side of the foam device was exposed to the Xebeam irradiation of 0.3 kW. Thus, such locations as the edges and back sides of the foam devices were maintained at

a temperature of at least 800
C during the T-R step. At the center of the light-irradiated surface, the temperatures of the foam devices reached Tcenter ¼ 1450e1500
C during the T-R step.

Amount of the evolved hydrogen / Ncm3g-1-material

a

1400˚C

b 14

16

16

14

14

12

12

10

10

8

8

0˚C

6

6 Fe3O4/m-ZrO2 Fe3O4/c-YSZ NiFe2O4/m-ZrO2 NiFe2O4/c-YSZ

4 2 0 1

2

3

4

5

Fe3O4/m-ZrO2 Fe3O4/c-YSZ NiFe2O4/m-ZrO2 NiFe2O4/c-YSZ

4 2 0 1

6

2

4

5

6

7

Cycle number

Cycle number

c

3

1 00˚C

Amount of the evolved hydrogen / Ncm3g-1-material

16 14 12 10 8 6 Fe3O4/m-ZrO2 Fe3O4/c-YSZ NiFe2O4/m-ZrO2 NiFe2O4/c-YSZ

4 2 0 1

2

3

4

5

6

Cycle number

Fig. 8 e Amount of the evolved hydrogen per gram of ferrite/zirconia particles in each cycle. After ferrite/zirconia particles was thermally reduced at a) 1400
C, b) 1450
C and c) 1500
C, the reduced sample was subjected to the W-D step at 1000
C.

2022

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

0.1

1400 1200

0.08

2nd

0.1

1400

1200

0.08

Production rate of H2 / Ncm3 min-1 g-1device

0.04

NiFe2O4/m-ZrO2/ MPSZ foam device

800

0.06

800

600

0.04

600

400

0.02

200

MPSZ foam

0

0 0

20

40

60

0.08

200

0

0 0

1200

0.1

1000

0.08

800 0.06

20

40

60

5th

0

0 0

20

40

60

80

800

0.04

600 400

0.02

200

0

1200

0.1

1000

0.08

0 20

40

60

80

6th

1200 1000

800 0.06

800 0.06

600

600 0.04 400

400

400 200

0.06

0

0.04

0.02

1000

80

600

0. 04

1200

0.08

400 0.02

80

4th

0.1

1400

1000

1000 0.06

3rd

0.1

Temperature / ºC

1st

0.02

200 0

0 0

20

40

60

80

0.02

200

0

0 0

20

40

60

80

Time / min Fig. 10 e Time variations of the hydrogen production rate per gram of foam device materials (NiFe2O4, m-ZrO2, and MPSZ foam matrix) during the W-D steps of the 1ste6th cycles. The solid curve shows the hydrogen production rate for NiFe2O4/ m-ZrO2/MPSZ, and the dotted line curve exhibits that for the original non-coated MPSZ foam as a reference. The W-D step was performed at temperature of 1200
C (cycles 1e3) and 1100
C (cycles 4e6).

In the W-D step, the thermally reduced foam devices were uniformly heated at 1200
C (cycles 1e3) and 1100
C (cycles 4e6) using an infrared furnace. Profiles of the hydrogen production from the NiFe2O4/m-ZrO2/MPSZ foam device during the W-D step are shown in Fig. 10. The rates of hydrogen production in cycles 1e6 are plotted against the reaction time. As a reference, the non-coated MPSZ foam was also tested under the same reaction conditions, and a small amount of hydrogen (0.3e0.9 N cm3 per gram of device) was found to be produced in the W-D step, as shown in cycle 1 in Fig. 10. As can be seen in cycles 1e3 in Fig. 10, the rate of hydrogen production during the W-D step at 1200
C reached a maximum value of 0.08 N cm3 min1 per gram of device at 20e30 min and subsequently decreased to 0.02 N cm3 min1 g1. The profile of hydrogen production through cycles 1e3 was measured in a reproducible fashion. As can be seen in cycles 4e6 in Fig. 10, when the temperature in the W-D step decreased to 1100
C, the production rate decreased to 0.04e0.05 N cm3 min1 per gram of device at 30 min. However, the hydrogen production proceeded intermittently with slight fluctuations during cycles 4e6. The profiles of cycles 4e6 were also measured in a reproducible fashion. Fig. 11 shows the profiles of hydrogen production from the Fe3O4/c-YSZ/MPSZ foam device during the W-D step.

In the first cycle, the hydrogen production proceeded with strong fluctuations, but with slight fluctuations in cycles 2e3. Subsequently, hydrogen production could be continued at nearly the same levels in cycles 4e6. The amounts of hydrogen produced per gram of device during the repeated W-D step using the NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/MPSZ foam devices are plotted against the cycle numbers in Fig. 12a. For the NiFe2O4/m-ZrO2/MPSZ foam device, the hydrogen productivity was 2.3e3.3 N cm3 per gram of device during the W-D step at a temperature of 1200
C (cycles 1e3). By lowering the W-D step temperature to 1100
C, the hydrogen productivity in cycles 4e6 was decreased to 2.2e2.4 N cm3 per gram of device, though some small fluctuation was observed. In the case of the Fe3O4/c-YSZ/MPSZ foam device, the hydrogen productivity was 2.4e2.9 N cm3 per gram of device in cycles 1e3 and 1.8e2.2 N cm3 per gram of device in cycles 4e6. The hydrogen productivity at the W-D step temperature of 1200
C (in cycles 1e3) was slightly greater than it was at 1100
C (cycles 4e6). The values for ferrite conversion for both foam devices are shown in Fig. 12b. The ferrite conversion was estimated from the amount of hydrogen produced in the subsequent W-D step. It is assumed for the NiFe2O4/m-ZrO2/MPSZ foam device that the reduced phase formed in the T-R step was completely oxidized back to

2023

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

0.14

1400

0.14

1400

0.12

1200

0.12

1200

0.12

1200

1000

0.1

1000

0.1

1000

0.08

800

0.08

800

0.08

800

0.06

600

0.06

600

0.06

600

0.04

400

0.04

400

0.04

400

200

0.02

200

0.02

200

Fe3O4/c-YSZ/ MPSZ foam device

0.1

Production rate of H2 / Ncm3 min-1 g-1device

3rd

1400

MPSZ foam

0.02 0

0 0

20

40

60

80

4th

0.14

0.12 0.1

0

1200

0.14

1000

0.12

0.08

20

40

60

5th

1200

0.14

1000

0.12

20

40

60

1200 1000 800

0

80

20

40

60

200

0.02

0 0

400

0.04

200

0.02

0 0

6th

600

400

0.04

0

80

0.06

400 200

60

600 0.06

0.02

40

0.08

600

0.04

20

0.1

800

0.08

0.06

0 0

80

0.1

800

0

0

0

Temperature / ºC

2nd

1st 0.14

0

80

0 0

20

40

60

80

Time / min

4

50

a

b NiFe2O4/m-ZrO2

3

2

Ferrite conversion n/%

Ncm3 g-1-device Evolved amount of H2 / N

Fig. 11 e Time variations of the hydrogen production rate per gram of foam device materials (Fe3O4, c-YSZ, and MPSZ foam matrix) during the W-D steps of the 1ste6th cycles. The solid curve shows the hydrogen production rate for Fe3O4/c-YSZ/ MPSZ, and the dotted line curve exhibits that for the original non-coated MPSZ foam as a reference. The W-D step was performed at temperature of 1200
C (cycles 1e3) and 1100
C (cycles 4e6).

Fe3O4/c-YSZ

1

0

NiFe2O4/m-ZrO2

40

30

20

Fe3O4/c-YSZ 10

0

1

2 3 4 Cycle number

5

6

1

2

3

4

5

6

Cycle number

Fig. 12 e (a) Amount of hydrogen produced per gram of foam device materials during the W-D step of the repeated watersplitting cycle. In addition, (b) ferrite conversion during repeated two-step water-splitting using the foam device is plotted against cycle number. The black and white symbols represent temperature of 1200
C (cycles 1e3) and 1100
C (cycles 4e6) during the W-D step, respectively.

2024

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

Fig. 13 e Flux density distribution of the Xe-beam irradiation for (a) T-R step and (b) W-D step. The input energy of the incident Xe-beam into the foam device was estimated by integrating the distribution. The maximum energy flux density is the power density of the incident beam at the central position of the spot. The unit of flux density is kWmL2.

ferrite phase via water decomposition in the subsequent W-D step. For the Fe3O4/c-YSZ/MPSZ foam device, it is assumed that the Fe2þ ions formed by the reduction in the T-R step were all oxidized back to Fe3þ via water-decomposition in the subsequent W-D step. The ferrite conversion in each cycle was over 1.4 times greater for the NiFe2O4/m-ZrO2/MPSZ foam device throughout cycles 2e6 than it was for the Fe3O4/c-YSZ/ MPSZ foam device. Damage to the NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/ MPSZ foam devices was observed after 6 cycles. It appeared that both foam devices were damaged to a similar extent. Morphological differences in the damage to the m-ZrO2 and cYSZ supports could not be observed after repetition testing. We consider the reason for the damage to be as follows. In the present study, the foam devices underwent the T-R step using concentrated Xe-light radiation and the subsequent W-D step

was performed with another quartz reactor using an infrared furnace; the foam device was fixed inside a quartz reactor tube during each step. Therefore, the foam device was subjected to a mechanical stress owing to its detachment from the reactor after each step and underwent thermal stress at high temperatures during each step. It is believed that the cyclic mechanical and thermal stress through the repetition testing of the two-step reaction induced the damage to the foam device.

3.3. Successive two-step water-splitting using foam devices Fig. 13 shows the energy flux profile of the incident solarsimulated Xe-beam on the irradiated surface of the foam device in the T-R step and W-D step. As seen in Fig. 13a, for the

0.6 0.5

Tedge

0.4

1000

0.3 Tback

0.2

500

Device e temperature / ºC

H2 production rate / Ncm3min-1 g-1-device

1500 Tcenter

0.1 0

0 0

500

1000

1500

2000

2500

3000

3500

Time / min

Fig. 14 e Profiles of hydrogen production during the W-D step of cyclic two-step water-splitting using the NiFe2O4/m-ZrO2/ MPSZ foam device. In addition, the temperature variations of the foam device (Tcenter, Tedge and Tback) during the two-step reaction are also shown in the figure.

29 48 20 10 10 44 38 34 0 1 6 27 20 24 18 30 30 29 35

59

2.5 (3.8) 3.8 (5.9) 1.8 (2.8) 1.1 (1.8) 1.1 (1.7) 3.5 (5.4) 3.1 (4.8) 2.8 (4.4) 0.3 (0.5) 0.5 (0.7) 0.8 (1.2) 2.3 (3.8) 1.8 (2.9) 2.1 (3.4) 1.7 (2.7) 2.5 (4.1) 2.5 (4.1) 4.6 (7.5) 2.5 (4.0) 2.8 (4.6)

19th 18th 17th 16th 15th 14th 13th 12th 11th 10th 9th 8th 7th 6th 5th 4th 3rd 2nd 1st

H2 production/Ncm g -foam (Ncm3 cm3-foam) Ferrite conversion/%

3 1

Cycle number

Table 2 e Amount of hydrogen produced per gram of device during the repeated W-D step and ferrite conversions of the NiFe2O4/m-ZrO2/MPSZ foam device.

20th

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

2025

T-R step the central peak of flux density reached 2900 kW m2. The average flux density on the exposed surface of the foam device was 1010 kW m2. The input power QHeat of the incident Xe-beam into the foam device was 0.71 kW, as estimated from the integration of the profile. For the W-D step, the central peak, the average of flux density, and the input power were 1700 kW m2, 590 kW m2 and 0.42 kW, respectively (Fig. 13b). In the T-R and the W-D steps, the foam device was heated only by solar-simulated Xe-beam irradiation, without preheating in an electric furnace. The profiles of hydrogen production during the W-D step of cyclic two-step water-splitting using the NiFe2O4/m-ZrO2/ MPSZ foam device are shown in Fig. 14. In addition, the temperature variations of the foam device (Tcenter, Tedge and Tback) during the two-step reaction are also given in the figure. Two-step water-splitting using the NiFe2O4/m-ZrO2/MPSZ foam device was successfully demonstrated for up to 20 cycles, and in each cycle, hydrogen production was observed immediately following the injection of steam. For cycles 1e2, as shown in Fig. 14, when the temperature of the foam device during the T-R step was set to Tcenter ¼ 1480e1500
C, and Tedge and Tback were in the range of 710e740
C and 810e830
C, respectively. Thus, the thermal reduction of NiFe2O4 proceeds preferentially at the center of the irradiated surface of the foam device, but at the positions of the edge and back side. During the subsequent W-D step in cycles 1e2, the temperature of the foam device was set to Tcenter ¼ 1200
C while Tedge and Tback were in the range of 580e640
C and 650e660
C, respectively. The rate of hydrogen production in cycles 1 and 2 reached a maximum value of 0.05e0.1 N cm3 min1 per gram of device. The behavior of the hydrogen production was similar to that which took place at 1200
C using an electric furnace (Fig. 5). After the W-D step of cycle 2, the foam device mounted in the reactor was left to cool to room temperature in an N2 gas stream. Subsequently, the device was detached from the reactor and was again set at the focal point of the reactor. This process was performed in order to further enhance the temperatures of the foam device at the edge position (Tedge) and the back side (Tback). For cycle 3, shown in Fig. 14, though the T-R temperatures at the center (Tcenter) and back side (Tback) were nearly the same in cycles 1 and 2, the Tedge values during the T-R step increased greatly to 1170e1210
C. Thus, the NiFe2O4 loaded on the irradiated surface of the foam device is thermally reduced over a relatively widespread region in comparison to cycles 1 and 2. Immediately following steam injection, a high rate of hydrogen production was measured in cycle 3. Tedge and Tback during the W-D step in cycle 3 were 930
C and 710
C, respectively. The back side of the foam device, however, will not react with steam due to insufficient temperature for the T-R step. After cycle 3, hydrogen was produced continuously through cycles 4e20, though the profiles of hydrogen production fluctuated for unknown reasons. One likely explanation is that the foam device was not fixed in place on the fixed porous quartz plate inside the reactor and could thus have been slightly displaced by the gas flow of the N2 and H2O/N2 mixture through the porous plate. Table 2 lists the amount of hydrogen produced per gram of device during the repeated W-D step and the ferrite

2026

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

Fig. 15 e Photographs of a 5-kWt Inha University’s dish concentrator in Korea: Left photograph; front side of dish concentrator; right photograph: solar reactor installed.

conversions of the NiFe2O4/m-ZrO2/MPSZ foam device. The hydrogen productivity and the ferrite conversion were a maximum value of 4.6 N cm3 per gram of device and 59% in cycle 3, respectively. The average of the hydrogen productivity and the ferrite conversion were 2.8 N cm3 per gram of device and 34%, respectively. The average of the hydrogen productivity was comparable to that of the hydrogen productivity of the foam device when homogeneously heated at 1200
C in an infrared furnace (Fig. 12). No damage to the NiFe2O4/m-ZrO2/MPSZ foam device was observed through the 20 cycles of the repetition test in a single reactor under concentrated Xe-light irradiation, although it was observably damaged in the repetition test in which two different reactors were used. This is attributed to the fact that during repetition testing, the foam device was not fixed in place on the quartz plate in the single reactor. The thermal endurance of the foam device against cyclic thermal shock was demonstrated using a 1 kW sun simulator on a laboratory scale. The present authors are presently conducting a solar demonstration project of thermochemical two-step water splitting at Inha University in Korea, using a larger NiFe2O4/mZrO2-coated MPSZ foam device with a 5-kWt dish concentrator. Based on the spin-coating method described above, an MPSZ foam disk with a diameter of 8 cm and a thickness of 1e2 cm was coated with NiFe2O4/m-ZrO2 particles. The NiFe2O4-loading of the foam disk reached 10e33 wt%. The solar demonstration of the two-step water-splitting cycle with this foam device was begun in late summer 2009. A new water-splitting reactor with a NiFe2O4/m-ZrO2-coated MPSZ foam device was designed at Niigata and Inha Universities, and then fabricated at Inha University. The fabricated reactor with the foam device has been installed and tested in Inha University’s 5-kWt dish concentrator (Fig. 15) [42].

4.

Summary

NiFe2O4/m-ZrO2/MPSZ and Fe3O4/c-YSZ/MPSZ foam devices were prepared for a two-step water-splitting thermochemical cycle to produce hydrogen from water. The foam device was prepared using a spin-coating method. A spin-coating method offers the advantages of a shortened preparation period and

a reduced coating process, as compared to the preparation method that we previously reported. Two-step water-splitting using these foam devices was repeated up to 6 cycles. For the NiFe2O4/m-ZrO2/MPSZ foam device, the hydrogen productivity was 2.3e3.3 N cm3 per gram of device during the W-D step at a temperature of 1200
C (cycles 1e3). By lowering the W-D step temperature to 1100
C, the hydrogen productivity in cycles 4e6 was decreased to 2.2e2.4 N cm3 per gram of device, though small fluctuations were observed. In the case of the Fe3O4/c-YSZ/MPSZ foam device, the hydrogen productivity was 2.4e2.9 N cm3 per gram of device for cycles 1e3 and 1.8e2.2 N cm3 per gram of device for cycles 4e6. The most reactive foam device of NiFe2O4/m-ZrO2/MPSZ was tested in a windowed single reactor for cyclic hydrogen production using solar-simulated Xe-beam irradiation with a power input of 0.4e0.7 kWth. Hydrogen continued to be produced successfully in the 20 cycles repeated with the NiFe2O4/m-ZrO2-coated zirconia foam. The NiFe2O4/m-ZrO2/ MPSZ foam device produced hydrogen at 1.1e4.6 cm3 per gram of device through 20 cycles and achieved a maximum ferrite conversion of 60%. The present authors are conducting a solar demonstration project of thermochemical two-step water splitting at Inha University in Korea, using a larger NiFe2O4/m-ZrO2-coated MPSZ foam device with a 5-kWt dish concentrator. Based on the spin-coating method described above, the foam device is prepared. The results of a solar demonstration will be reported elsewhere.

references

[1] Kodama T. High-temperature solar chemistry for converting solar heat to chemical fuels. Progress in Energy and Combustion Science 2003;29:567e97. [2] Steinfeld A. Solar thermochemical production of hydrogen e a review. Solar Energy 2005;78:603e15. [3] Ihara S. On the study of hydrogen production from water using solar thermal energy. International Journal of Hydrogen Energy 1980;5:527e34. [4] Kogan A. Direct solar thermal splitting of water and on-site separation of the products-II. Experimental feasibility

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

[5]

[6] [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

study. International Journal of Hydrogen Energy 1998;23(2): 89e98. Kodama T, Gokon N. Thermochemical cycles for hightemperature solar hydrogen production. Chemical Review 2007;107(10):4048e77. Nakamura T. Hydrogen production from water utilizing solar heat at high temperatures. Solar Energy 1977;19:467e75. Lundberg M. Model calculations on some feasible two-step water splitting processes. International Journal of Hydrogen Energy 1993;18(5):369e76. Bilgen E, Ducarroir M, Foex M, Sibieude F, Trombe F. Use of solar energy for direct and two-step water decomposition cycles. International Journal of Hydrogen Energy 1977;2:251e7. Abanades S, Charvin P, Flamant G, Neveu P. Screening of water-splitting thermochemical cycles potentially attractive for hydrogen production by concentrated solar energy. Energy 2006;31:2805e22. Sibieude F, Ducarroir M, Tofighi A, Ambriz J. High temperature experiments with a solar furnace: the decomposition of Fe3O4, Mn3O4, CdO. International Journal of Hydrogen Energy 1982;7(1):79e88. Steinfeld A, Kuhn P, Reller A, Palumbo R, Murray J, Tamaura Y. Solar-processed metals as clean energy carriers and water-splitters. International Journal of Hydrogen Energy 1998;23(9):767e74. Kodama T, Kondoh Y, Kiyama A, Shimizu K-I. Hydrogen production by solar thermochemical water-splitting/ methane-reforming process. In: Thornbloom MD, Jones SA, editors. Proceedings of international solar energy conference (ISEC) 2003 (Hawaii, 2003); 2003. ISEC2003e44037, [CD publication]. Gokon N, Mizuno T, Takahashi S, Kodama T. A two-step water splitting with ferrite particles and its new reactor concept using an internally circulating fluidized-bed. In: Morehouse JH, Krarti M, editors. Proceeding of ASME 2006 international solar energy conference (ISEC 2006) (Denver, 2006). New York: ASME; 2006. ISEC2006e99063, [CD-ROM publication]. Kodama T, Kondoh Y, Yamamoto R, Andou H, Satoh N. Thermochemical hydrogen production by a redox system of ZrO2-supported Co(II)-ferrite. Solar Energy 2005;78: 623e31. Kodama T, Gokon N, Yamamoto R. Thermochemical twostep water splitting by ZrO2-supported Ni xFe3xO4 for solar hydrogen production. Solar Energy 2008;82:73e9. Gokon N, Murayama H, Umeda J, Hatamachi T, Kodama T. Monoclinic zirconia-supported Fe3O4 for the two-step watersplitting thermochemical cycle at high thermal reduction temperatures of 1400e1600
C. International Journal of Hydrogen Energy 2009;34:1208e17. Kodama T, Nakamuro Y, Mizuno T. A two-step thermochemical water splitting by iron-oxide on stabilized zirconia. ASME Journal of Solar Energy Engineering 2006;128 (1):3e7. Gokon N, Mizuno T, Nakamuro Y, Kodama T. Iron-containing YSZ (Yttrium-Stabilized Zirconia) system for a two-step thermochemical water splitting. In: Romero M, Martinez D, Ruiz V, Silva M, Brown M, editors. Proceedings of SolarPACES2006 (Seville, 2006). CD-ROM Publication, ISBN 847834-519-1; 2006. B2-S12. Gokon N, Mizuno T, Nakamuro Y, Kodama T. Iron-containing yttria-stabilized zirconia system for two-step thermochemical water splitting. ASME Journal of Solar Energy Engineering 2008;130(1). 011018e11021-6. Kodama T, Gokon N, Hiraiwa E. Reactivity of iron-containing YSZ for a two-step thermochemical water splitting using thermal reduction temperatures of 1400-1500
C. In: Proceedings of the ASME Solar Energy Division, Energy

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

2027

Sustainability (ES2008). Jacksonville, Florida, USA: CD-ROM Publication; 2008. ES2008-54151. Agrafiotis C, Roeb M, Konstandopoulos AG, Nalbandian L, Zaspalis VT, Sattler C, et al. Solar water splitting for hydrogen production with monolithic reactors. Solar Energy 2005;79:409e21. Roeb M, Sattler C, Klu¨ser R, Monnerie N, de Oliveira L, Konstandopoulos AG, et al. Solar hydrogen production by a two-step cycle based on mixed iron oxides. ASME Journal of Solar Energy Engineering 2006;128:125e33. Agrafiotis C, Lorentzou S, Pagkoura C, Kostoglou M, Konstandopoulos AG. Advanced monolithic reactors for hydrogen generation from solar water splitting. In: Romero M, Martinez D, Ruiz V, Silva M, Brown M, editors. Proceedings of SolarPACES2006 (Seville, 2006). CD-ROM Publication, ISBN 84-7834-519-1; 2006. B2-S11. Roeb M, Sa¨ck J-P, Rietbrock P, Neises M, Ebert M, Reinalter W, et al. Development and verification of a two-step thermochemical process for solar hydrogen production from water. In: Proceedings of 2008 14th Biennial CSP SolarPACES symposium. Las Vegas, Nevada, USA: CD-ROM Publication; 2008. Roeb M, Neises M, Sa¨ck JP, Rietbrock P, Monnerie N, Dersch J, et al. Operational strategy of a two-step thermochemical process for solar hydrogen production. International Journal of Hydrogen Energy 2009;34(10):4537e45. Kaneko H, Fuse A, Miura T, Ishihara H, Tamaura Y. Two-step water splitting with concentrated solar heat using rotarytype solar furnace. In: Romero M, Martı´nez D, Ruiz V, Silva M, Brown M, editors. Proceedings of 13th international symposium on concentrated solar power and chemical energy technologies. Seville, Spain: CD-ROM Publication, ISBN 84-7834-519-1; 2006. B2-S10. Kaneko H, Miura T, Fuse A, Ishihara H, Taku S, Fukuzumi H, et al. Rotary-type solar reactor for solar hydrogen production with two-step water splitting process. Energy and Fuels 2007; 21:2287e93. Tamaura Y, Kaneko H, Taku S, Naganuma Y, Ouchi K, Hasegawa N. Simultaneous production of H2 and O2 with rotary-type solar reactor (TOKYO TECH MODEL) for solar hybrid fuel. In: Proceedings of the ASME solar energy division, energy sustainability (ES2008). Jacksonville, Florida, USA: CD-ROM Publication; 2008. ES2008-54282. Gokon N, Mizuno T, Takahashi S, Kodama T. A two-step water splitting with ferrite particles and its new reactor concept using an internally circulating fluidized-bed. In: Proceedings of the ASME solar energy division, international solar energy conference (ISEC). Denver, Colorado, USA: CD-ROM Publication, ISBN 0-89553-178-X; 2006. ISEC2006-99063. Gokon N, Takahashi S, Yamamoto H, Kodama T. Thermochemical two-step water-splitting reactor with internally circulating fluidized bed for thermal reduction of ferrite particles. International Journal of Hydrogen Energy 2008;33:2189e99. Gokon N, Takahashi S, Yamamoto H, Kodama T. New solar water-splitting reactor with ferrite particles in an internally circulating fluidized bed. In: Proceedings of the ASME solar energy division, energy sustainability (ES2007). Long Beach, California, USA: CD-ROM Publication, ISBN 0-7918-3798-X; 2007. ES2007-36063. Gokon N, Takahashi S, Yamamoto H, Kodama T. New solar water-splitting reactor with ferrite particles in an internally circulating fluidized bed. ASME Journal of Solar Energy Engineering 2009;131. 011007e1-9. Diver RB, Miller JE, Allendorf MD, Siegel NP, Hogan RE. Solar thermochemical water-splitting ferrite-cycle heat engines. In: Proceedings of the ASME solar energy division,

2028

[34]

[35]

[36]

[37]

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 6 ( 2 0 1 1 ) 2 0 1 4 e2 0 2 8

international solar energy conference (ISEC). Denver, Colorado, USA: CD-ROM Publication, ISBN 0-89553-178-X; 2006. ISEC2006-99147. Miller JE, Evans LR, Stuecker JN, Allendorf MD, Siegel NP, Diver RB. Materials development for the CR5 solar thermochemical heat engine. In: Proceedings of the ASME solar energy division, international solar energy conference (ISEC). Denver, Colorado, USA: CD-ROM Publication, ISBN 089553-178-X; 2006. ISEC2006-99152. James DL, Siegel NP, Diver RB, Boughton BD, Hogan RE. Numerical modeling of solar thermo-chemical watersplitting reactor. In: Proceedings of the ASME solar energy division, international solar energy conference (ISEC). Denver, Colorado, USA: CD-ROM Publication, ISBN 0-89553178-X; 2006. ISEC2006-99141. Diver RB, Siegel NP, Miller JE, Moss TA, Stuecker JN, James DL. Development of a CR5 solar thermochemical heat engine prototype. In: Proceedings of 2008 14th Biennial CSP SolarPACES symposium. Las Vegas, Nevada, USA: CD-ROM Publication; 2008. CD-550-42709. Kodama T, Hasegawa T, Nagasaki A, Gokon N. Reactive FeYSZ coated foam devices for solar two-step water splitting.

[38]

[39]

[40]

[41]

[42]

In: Proceedings of the ASME solar energy division, energy sustainability (ES2007). Long Beach, California, USA: CD-ROM Publication, ISBN 0-7918-3798-X; 2007. ES2007-36060. Gokon N, Hasegawa T, Takahashi S, Kodama T. Thermochemical two-step water-splitting for hydrogen production using Fe-YSZ particles and a ceramic foam device. Energy 2008;33(9):1407e16. Gokon N, Murayama H, Nagasaki A, Kodama T. Thermochemical two-step water splitting cycles by monoclinic ZrO2-supported NiFe2O4 and Fe3O4 powders and ceramic foam devices. Solar Energy 2009;83:527e37. Kodama T, Hasegawa T, Nagasaki A, Gokon N. A reactive FeYSZ coated foam device for solar two-step water splitting. Journal of Solar Energy Engineering 2009;131. 021008e1-7. Allendorf MD, Diver RB, Siegel NP, Miller JE. Two-step water splitting using mixed-metal ferrites: thermodynamic analysis and characterization of synthesized materials. Energy and Fuels 2008;22:4115e24. Seo JH, Ma DS, Kim Y, Seo T, Han GY, Lee SN, Kang YH. In: Proceedings of 2008 14th Biennial CSP SolarPACES Symposium, March 4e7, 2008. Las Vegas, NV, USA: CD-ROM Publication; 2008. CD-550-42709.