A Continuous Hydrogen Reduction Process for the Production of Water on the Moon

Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021); doi: 10.15011/jasma.38.380203 1 of 9 . A Continuous Hydrogen Reduction Process for the Production of Water on the Moon Eri KUMAI1, Manabu TANAKA1, Takayuki WATANABE1*, Takeshi HOSHINO2, Satoshi HOSODA2, Hiroshi KANAMORI2 1 Department of Chemical Engineering, Kyushu University, Fukuoka Japan. 2 Japan Aerospace Exploration Agency (JAXA), Sagamihara, Kanagawa Japan.. * Correspondence: watanabe@chem-eng.kyushu-u.ac.jp Department of Chemical Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan . Abstract: A continuous screw reactor for hydrogen reduction of lunar soil simulant was assembled. The water production rates were measured with different reduction conditions. Reduced simulants were analyzed by XRD (X-ray Diffraction), SEM (Scanning Electron Microscope), and EDS (Energy dispersive X-ray spectroscopy). The effect of reduction temperatures in the range of 1173-1373 K in 3 vol% hydrogen was clarified. The highest water production rate at the steady-state is obtained at 1273 K. The water production rate becomes higher with the higher reduction temperature up to 1273 K. It becomes lower at the temperature above 1273 K because Na-rich components in lunar simulant melt and it inhibits the hydrogen diffusion. The impact of hydrogen concentrations between 3-10 vol% was revealed with the fixed reduction temperature of 1273 K. The reaction rate has linear relationship to the hydrogen concentration. The reduced lunar simulant contain α-Fe, and the amount increases with higher hydrogen concentrations. The reduced ilmenite has porous structures due to the vacancy of oxygen. This work suggests the continuous hydrogen reduction system as a promising process to acquire oxygen on the moon.. Keywords: ISRU, Screw Reactor, Continuous Process, Hydrogen Reduction, Lunar Simulant Article History: Received 28 Janualy 2021, accepted 7 April 2021, published 30 April 2021.. 1. Introduction. Several space exploration projects have been planned recently. Especially lunar exploration and development are attracting more and more attention. Considering the manned missions, oxygen and water play essential roles. Oxygen will be used for life support, and also for propellant of rockets in the form of liquid oxygen. Water will be used firstly for life supporting such as drinking and sanitation, secondly for construction materials such as concrete, and also for the various experiments on the moon.. Along with the effective recycling systems of the water and oxygen1-3), in-situ resource utilization (ISRU) technology is necessary for sustainable missions on the moon. ISRU is a strategy to use locally derived materials in order to get life supporting consumables and other necessary materials. This strategy is crucial because the transportation from the earth requires much cost, time and labor.. Over 20 approaches for oxygen production on the moon have been proposed4). Among these processes, oxygen extraction from lunar regolith by hydrogen reduction is one of the simplest and the best methods5). The hydrogen reduction system of lunar soil is described in Fig. 1. Two main steps are involved to obtain. mailto:watanabe@chem-eng.kyushu-u.ac.jp. . Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 2 of 9. oxygen. First, oxides in lunar soil are reduced with hydrogen producing water. The main target of the reduction in lunar regolith is ilmenite6). The reaction is described as following.. FeTiO3 + H2 → Fe + TiO2 + H2O (1). This process requires relatively low temperature compared to other methods7). Useful metals for the lunar infrastructure are also produced in this step. The second step is electrolysis of the produced water. The reaction is written as following.. H2O → H2 + 1 2. O2 (2). The hydrogen produced by electrolysis can be recycled as a reductant of (1). Hydrogen reduction of lunar rock was performed6). All Fe2+ in ilmenite and some olivine and pyroxene. were reduced at 900-1050°C. Hydrogen reduction experiments of lunar simulants, lunar rock, soil, and volcanic glass were conducted8-10). They reported that the oxygen yields are affected by the amount of iron, not the types of the material. A scale-upped hydrogen reduction system for treating 10-15 kg of lunar soil was tested11). The results evaluated the capability of producing oxygen in different conditions.. Many kinds of batch processes for hydrogen reduction such as fixed bed reactors and fluidized bed reactors have been investigated, however much less work has done with continuous processes. Continuous processes are strongly required to produce large amount of water and oxygen. This research suggests a continuous hydrogen reduction system for extracting oxygen from lunar soil. The purpose of this study is to investigate the oxygen production mechanism with the continuous hydrogen reduction system.. Figure 1. Schematic of hydrogen reduction system.. 2. Experiments. 2.1. Experimental configuration and procedure. A continuous screw reactor was assembled in this work. Figure 2 shows the schematic diagram of the experimental apparatus. The set-up mainly consists of a hopper, a screw to carry the samples, a reactor, an electric furnace for heating the reactor, and a measurement line including a moisture meter. The screw is 1 m long and the furnace region is 30 cm in horizontal length.. Experiments are started by flowing argon at 10.0 NL/min for the reactor heating up to pre-determined temperature. Argon is used for preventing reactor from oxidation. After the reactor is heated, argon is quickly turned to the determined flow rate and hydrogen injection is started. At the same time the screw starts rotating in pre-determined rotation speed. The screw transports the sample from the hopper to the reactor. Hydrogen reduction proceeds in the heated region. The pressure is kept at 200 kPa by controlling the pressure valve at the exit. Argon, hydrogen and produced water vapor is transported to a moisture meter to measure the water production rate. Other detailed operating conditions is summarized in Table 1.. . Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 3 of 9. Figure 2. Schematic diagram of the continuous screw reactor for hydrogen reduction system.. Table 1. Experimental conditions for the continuous hydrogen reduction system.. Parameter Condition Effect of reduction temperature Effect of H2 concentration Temperature [K] 1173, 1273, 1373 1273 Gas flow rate (Ar+H2) [NL/min]. 10 10. H2 concentration [vol%] 3 3, 6, 10 Pressure [kPa] 200 200 Reduction time [min] 10 10 Screw rotation speed [rpm] 0.8 0.8. 2.2. Samples. Lunar soil simulant FJS-1 was used for experiments. FJS-1 was produced by Shimizu Corp., Japan12). It is made of basalt in Mt. Fuji, Japan. The components of FJS-1 are similar to that of lunar mare regolith. Specific characteristics of FJS-1 were reported by Kanamori, et al.12) FJS-1 contains iron oxides such as FeO, Fe2O3 and FeTiO3. These oxides are likely to be reduced by hydrogen. The Gibbs free energy of iron oxides are shown in Fig. 3. The Gibbs free energies at the temperature range of our experiments have positive value. The reaction will not be expected to proceed when the Gibbs free energy is positive. However, by removing water vapor from the reaction region by large amount of argon and hydrogen, forced reduction is available in our system.. Figure 3. Gibbs free energy charge of hydrogen reduction with metal oxides.. . Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 4 of 9. 2.3. Analysis methods. Product identification analysis was carried out with X-ray diffraction (XRD, Rigaku Smartlab) equipping with Cu-Kα source. The diffraction data were collected using the continuous scan mode at a speed of 10°/min in the region of 20–50° with a step of 0.01°. The accelerating voltage and applied current were 40 kV and 30 mA, respectively. The morphology of the particles was observed through scanning electron microscope (SEM, Hitachi SU3500) with accelerating voltage of 15 kV. Element mapping of samples was conducted by energy- dispersive X-ray spectroscopy (EDS, Hitachi SU3500). The samples were mounted in epoxy resin and polished to expose the cross section before the SEM and EDX analysis.. 3. Results and discussions. 3.1. Performance evaluation of the screw reactor. In screw reactors, the powder feed rate and the reduction time are fixed by the screw rotation speed. The reduction time in this study was defined as the time that sample stay in the heated area of the reactor.. The sample is transferred by the distance of the screw pitch while the screw rotates once. Therefore, the relationship between the screw rotation speed 𝑁𝑁 [rpm] and the reduction time 𝑡𝑡 [min] can be written as following equation:. 𝑡𝑡 = 𝐿𝐿 𝑎𝑎𝑁𝑁. (3). where 𝐿𝐿 [m] is the length of the heated region, 𝑎𝑎 [m] represents the screw pitch. The powder feed rate in general screw feeder 𝑄𝑄 [kg/min] is expressed as following correlation13):. 𝑄𝑄 = 𝜂𝜂 𝜋𝜋 4. (𝐷𝐷2 − 𝑑𝑑2)𝑎𝑎𝑁𝑁𝑎𝑎 (4). where 𝜂𝜂 [-] is the packing fraction of the powder in the screw, 𝑎𝑎 [kg/m3] is the bulk density of the powder, and 𝐷𝐷 [m] and 𝑑𝑑 [m] represent the screw diameter and the shaft diameter, respectively. The powder feed rate in general screw feeder is in proportion to the screw rotation speed according to this correlation. The actual sizes of the apparatus are listed in Table 2.. Figure 4 shows the reduction time and the sample feed rate. The green line, corresponding to the right- hand axis, is the calculated relationship between the screw rotation speed and the reduction time. The blue plots show the actual sample feed rates during experiments with different screw rotation speeds, and correspond to the left-hand axis. The feed rate in our system has a linear relationship with the screw rotation speed (𝑅𝑅2 = 0.99,𝑅𝑅2: the coefficient of determination). The packing fraction is calculated to be 𝜂𝜂 = 0.13. The powder feed rate is considered to be stable even though the gas mixture of argon and hydrogen is supplied as a counter-flow.. Table 2. Actual sizes of the apparatus.. Symbol Size. 𝐿𝐿 0.3 m. 𝑎𝑎 38 mm. 𝐷𝐷 31 mm. 𝑑𝑑 7 mm. 𝑎𝑎 1.8×103 kg/m3. Figure 4. Relationship between screw rotation speed and. feed rate of lunar simulant.. . Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 5 of 9. 3.2. Effect of reduction temperature. The effect of reduction temperature on water production rate during continuous hydrogen reduction was investigated. Figure 5 shows the water production rate with different reduction temperatures. The temperatures were set as 1173, 1273, and 1373 K. The time 0 in the graph represents the time when the valve in hydrogen flow line is opened. Water is not detected at the beginning in all conditions. This is because the amount of untreated sample in the reactor is insufficient to produce the water at higher humidity than the detection limit of 1.5%RH, where %RH represents the relative humidity. The water production rate increases slowly during untreated sample is increasing. After the reactor is filled with sample, the water production rate should become constant. However the obtained water production rates in some conditions do not seem to reach steady state because the changes in the temperature and the pressure of the moisture meter affect the measurement. This issue of inconsistent temperature and pressure will be improved by the improvement of the apparatus in future experiments. Though the obtained results are not desirable, they still have enough accuracy to continue qualitative discussions as follows.. The water production rate per unit sample at the steady-state is shown in Fig. 6. Water production rates after 30 minutes were considered to be steady-state and employed in the calculations in this discussion. The water production rate at the steady-state is highest at the reduction temperature of 1273 K. Higher reduction temperature results in a higher reduction rate up to 1273 K. On the other hand, water production rate decreases with higher temperature than 1273 K. This is due to the melting of some components in the sample. This will be discussed with the sample morphology at next section. Previous studies on hydrogen reduction of FJS-1 with fixed bed and fluidized bed reactors reported that the maximum water production rates were obtained at 1173~1273 K14,15). The same trend has been observed in the present study with the screw reactor. . The cross-sections of the samples were observed to investigate the morphological change of the samples during hydrogen reduction. The SEM images of initial and reduced FJS-1 are shown in Fig. 7. The FJS-1 particles before the reduction have several cracks containing small particles as shown in Fig. 7(a). By contrast,. (a) Raw FJS-1 (b) 1373 K, 10min. Figure 7. SEM image of (a) raw FJS-1 and (b) products reduced for 10 minutes in 3% hydrogen at 1373 K.. Figure 5. Water production rate with different reduction temperatures in 3 vol% hydrogen.. Figure 6. Water production rate at the steady-state. with different reduction temperatures.. . Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 6 of 9. the cracks and the small particles are not observed in the samples reduced at 1373 K. These results demonstrate that the fine particles melted and filled up the cracks and halls of the particles. Na-rich minerals have relatively low melting points at around 1373 K16,17). Then the molten components inhibited the hydrogen diffusion inside particles. That resulted in smaller water production rate at high temperature.. 3.3. Effect of hydrogen concentration. Last section reveals the effect of reduction temperature on hydrogen reduction of FJS-1. The effect of hydrogen concentration is investigated in this section.. Figure 8 shows the water production rate with different hydrogen concentrations. The hydrogen concentrations were fixed at 3 vol%, 6 vol%, and 10 vol%. The maximum hydrogen concentration was set as 10vol% in this work with the consideration of the safety. The water production rate increases with higher hydrogen concentration. As mentioned in the previous section, the obtained water production rates are accurate enough for the following discussions though they are not very stable in some cases. The water production rates at the steady-state per unit sample are shown in Fig. 9. Water production rates after 30 minutes were considered to be steady-state in this discussion. The water production rate at the steady state is in proportion to the hydrogen concentration. Previous work revealed by TG analysis that the hydrogen reduction of FJS-1 is proportion to the hydrogen concentration in the small scale18). The results of our study are consistent with the previous report.. The XRD spectra of raw FJS-1 and reduced products with different hydrogen concentrations are demonstrated in Fig. 10. Diffraction peak associated with α-Fe phase at 2θ = 44.68 is observed only in the reduced samples. This indicates the appearance of pure Fe due to the reduction of iron oxides such as FeO, Fe2O3, and FeTiO3 in FJS-1. The relative intensity of α-Fe becomes larger with higher hydrogen concentration. It represents more reduction yield with higher hydrogen concentrations. This trend is consistent with the results of water production rate.. Raw FJS-1 sample and reduced sample were investigated by SEM and EDS. The sample treated by 10 vol% hydrogen was selected as an example of reduced samples. Figure 11 shows the results of cross-section observation. Several brightnesses are found in the SEM results of both initial and reduced samples. Higher atomic number components generate stronger intensity of back scattered electrons, therefore the regions with larger atomic number show a lighter-colored image19). The brightest regions in FJS-1 contain Fe and Ti according to the EDS element mapping. These regions are recognized as ilmenite. The ilmenite regions have innumerable holes after hydrogen reduction as seen in Fig. 11(b). The porous regions are found even on inside of the particles. These results suggest that the ilmenite in FJS-1 is reduced by hydrogen even on the inside of particles. Moreover, the absence of O atoms in the porous regions is observed by EDS results. The extraction of oxygen as water produces the pores.. Figure 8. Water production rate with different. hydrogen concentrations at 1273 K.. Figure 9. Water production rate at the steady-state with different hydrogen concentrations.. Int. J. Microgravity Sci. Appl. 38(2) 380200 (2021); doi: 10.15011/jasma.38.380200 7 of 9 . Figure 10. Normalized XRD spectrums of raw material and products in different hydrogen concentrations.. (a) . (b) . Figure11. EDS results of (a) Raw FJS-1 and (b) products reduced for 10 minutes in 10% hydrogen at 1273 K. . 4. Conclusion. A continuous hydrogen reduction system with a screw reactor for the extraction of oxygen from the lunar regolith was established. The main performances of the screw reactor were investigated. Lunar soil simulant was successfully reduced by hydrogen in continuous process. The effects of reduction temperature and hydrogen concentration on the hydrogen reduction were revealed.. Controlling temperature is necessary to improve the effect of oxygen extraction. The reaction yield is highest at the reduction temperature of 1273 K. The water production rate becomes higher with the higher reduction temperature up to 1273 K. On the other hand, it becomes lower at the temperature above 1273 K. This is because some components of lunar simulant melt and hydrogen diffusion is limited by the molten components. The morphological change due to the melting of components was confirmed by cross-section observation with SEM.. Higher hydrogen concentration is recommended for the water production from the lunar soil. The reaction rate has a linear relationship to the hydrogen concentration in our system. The XRD analysis indicates the increasing amount of α-Fe due to the hydrogen reduction with higher hydrogen concentrations. Porous structures with the absence of oxygen in ilmenite regions in lunar simulants are observed by SEM and EDS.. Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 8 of 9. The results obtained in this study are consistent with the previous results with smaller-scaled systems. This indicates that the system is well scale-upped. This work suggests the continuous hydrogen reduction system with a screw reactor as a promising process to acquire oxygen on the moon.. Acknowledgments. This research was carried out as a part of the JAXA Space Exploration Innovation Hub Center (TANSAX) based on the JST Innovation Hub Construction Support Project.. The XRD, SEM, and EDS analysis were performed using Rigaku Smartlab and Hitachi SU3500 at the Center of Advanced Instrumental analysis, Kyushu University.. Nomenclature. 𝑡𝑡 reduction time (min) 𝐿𝐿 heated length (m) 𝑎𝑎 screw pitch (m) 𝑁𝑁 screw rotation speed (rpm) 𝑄𝑄 sample feed rate (kg/min) 𝜂𝜂 packing fraction of particle (-) 𝐷𝐷 screw diameter (m) 𝑑𝑑 screw shaft diameter (m) 𝑎𝑎 bulk density (kg/m3). References. 1) A. Calvo-Lopez, E. Arasa-Puig, M. Puyol, J.M. Casalta and J. A-Chamarro: Biparametric potentiometric analytical microsystem for nitrate and potassium monitoring in water recycling processes for manned space missions, Anal. Chim. Acta, 804 (2013) 190, DOI: 10.1016/j.aca.2013.10.013.. 2) D. Christenson, R. Sevanthi, A. Morse and A. Jackson: Assessment of Membrane-Aerated Biological Reactors (MABRs) for Integration into Space-Based Water Recycling System Architectures, Gravit. Space. Res., 6 (2018) 12, DOI: 10.2478/gsr-2018-0007.. 3) A. Yokohama, S. Suzuki, M. Sakurai, A. Shima and M. Ohnishi: Effect of Heater Preset Temperature on Efficiency and Power Consumption of the CO2 Removal Device for use Onboard the International Space Station, Int. J. Microgravity Sci. Appl., 36 (2019) 360108, DOI: 10.15011//jasma.36.360108.. 4) L.A. Taylor and W.D. Carrier Ⅲ: Production of Oxygen on the Moon: Which Process Are Best and Why, AIAA Journal, 30 (1992) 2860, DOI: 10.2514/3.48974.. 5) R.A. Briggs and A. Sacco. Jr.: Hydrogen reduction mechanism of ilmenite between 823 and 1353 K, J. Mater. Res., 6 (1991) 574, DOI: 10.1557/JMR.1991.0574.. 6) M.A. Gibson, C.W. Knudsen, D.J. Brueneman, C.C. Allen, H. Kanamori, D.S. McKay: Reduction of lunar basalt 70035: Oxygen yield and reaction product analysis, J. Geophys. Res., 99 (1994) 10887, DOI: 10.1029/94JE00787.. 7) C. Schwandt, J.A. Hamilton, D.J. Fray and I.A. Crawford: The production of oxygen and metal from lunar regolith, Planet. Space Sci., 74 (2012) 49, DOI: 10.1016/j.pss.2012.06.011.. 8) C.C. Allen, M.A. Gibson, C.W. Knudsen, H. Kanamori, R.V. Morris, L.P. Keller and D.S. McKay: Lunar Resources- Oxygen from Rocks and Soil., LPI Technical Report, 92 (1992) 1.. 9) C.C. Allen, R.V. Morris and D.S. McKay: Experimental reduction of lunar mare soil and volcanic glass, J. Geophys. Res., 99 (1994) 23173, DOI: 10.1029/94JE02321.. 10) D.S. McKay and C.C. Allen: Hydrogen reduction of lunar materials for oxygen extraction on the moon, Proc. of 34th Aerospace Sciences Meeting and Exhibit (1996) 0488, DOI: 10.2514/6.1996-488.. 11) D.L. Clark, B.W. Keller and J.A. Kirkland: Field Test Results of the PILOT Hydrogen Reduction Reactor, Proc. of AIAA SPACE 2009 Conference & Exposition (2009) 6475, DOI: 10.2514/6.2009-6475.. 12) H. Kanamori, K. Matsui, A. Miyahara and S. Aoki: Development of New Lunar Soil Simulants in Japan, Proc. of Space Resources Roundtable VIII (2006) p.35.. 13) A.W. Roberts: The influence of granular vortex motion on the volumetric performance of enclosed screw conveyors, Powder Technol., 104 (1999) 56, DOI: 10.1016/S0032-5910(99)00039-X.. 14) T. Watanabe, H. Kanamori and S. Aoki: Kinetic Investigation of Water Production from Lunar Soil Simulant by Hydrogen Reduction, Proc. of Space Resources Roundtable VIII (2006) p.66.. 15) K. Sueyoshi, T. Watanabe and Y. Nakano: Reaction Mechanism of Various Types of Lunar Soil Simulants by Hydrogen Reduction, Proc. of Earth & Space 2008 Conference (2008) p. 1, DOI: 10.1061/40988(323)134.. https://doi.org/10.1016/j.aca.2013.10.013 https://doi.org/10.2478/gsr-2018-0007 https://doi.org/10.2514/3.48974 https://doi.org/10.1557/JMR.1991.0574 https://doi.org/10.1029/94JE00787 https://doi.org/10.1016/j.pss.2012.06.011 https://doi.org/10.1029/94JE02321 https://doi.org/10.2514/6.1996-488 https://doi.org/10.2514/6.2009-6475 https://doi.org/10.1016/S0032-5910(99)00039-X https://doi.org/10.1061/40988(323)134 https://doi.org/10.15011//jasma.36.360108. . Int. J. Microgravity Sci. Appl. 38(2) 380203 (2021) 9 of 9. 16) A. Meurisse, J.C. Beltzung, M. Kolbe, A. Cowley and M. Sperl: Influence of Mineral Composition on Sintering Lunar Regolith, J. Aerosp. Eng., 30 (2017) 04017014-2, DOI: 10.1061/(ASCE)AS.1943-5525.0000721. 17) W.A. Deer, R.A. Howie and J. Zussman: Framework Silicates, In An Introduction to the Rock Forming Minerals, 16th Impression, John Wiley & Sons, Inc., New York (1989).. 18) K. Sueyoshi, T. Watanabe, Y. Nakano, H. Kanamori, S. Aoki, A. Miyahara and K. Matsui: Kinetic Investigation on Water Production from Lunar Soil Simulant by Hydrogen Reduction, Proc. of SCEJ 39th Autumn Meeting (2007) S202.. 19) S.J. Higgnis, L.A. Taylor, J.G. Chambers, A. Patchen and D.S. McKay: X-ray Digital-Imaging Petrography: Technique Development for Lunar Mare Soils, Meteorit. Planet. Sci., 31 (1996) 356.. © 2021 by the authors. Submitted for possible open access publication under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).. https://doi.org/10.1061/(ASCE)AS.1943-5525.0000721. 1. Introduction Figure 1. Schematic of hydrogen reduction system. 2. Experiments 2.1. Experimental configuration and procedure Figure 2. Schematic diagram of the continuous screw reactor for hydrogen reduction system. Table 1. Experimental conditions for the continuous hydrogen reduction system. 2.2. Samples 2.3. Analysis methods 3. Results and discussions 3.1. Performance evaluation of the screw reactor 3.2. Effect of reduction temperature 3.3. Effect of hydrogen concentration (a) (b) 4. Conclusion Acknowledgments Nomenclature References