# Fileset

[Exploration - 2023 - Abe - Exploration of heterogeneous catalyst for molecular hydrogen ortho‐para conversion.pdf](https://mdr.nims.go.jp/filesets/0d67d7ec-b579-4073-8e12-c01f7c860cc6/download)

## Creator

[Hideki Abe](https://orcid.org/0000-0002-8392-7586), [Hiroshi Mizoguchi](https://orcid.org/0000-0002-0992-7449), [Ryuto Eguchi](https://orcid.org/0009-0003-2859-6934), [Hideo Hosono](https://orcid.org/0000-0001-9260-6728)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Exploration of heterogeneous catalyst for molecular hydrogen ortho-para conversion](https://mdr.nims.go.jp/datasets/7421fc3f-d061-4a69-9726-3fac74c74019)

## Fulltext

Exploration of heterogeneous catalyst for molecular hydrogen ortho‐para conversionReceived: 20 April 2023 Accepted: 1 September 2023DOI: 10.1002/EXP.20230040RESEARCH ARTICLEExploration of heterogeneous catalyst for molecular hydrogenortho-para conversionHideki Abe, Hiroshi Mizoguchi Ryuto Eguchi Hideo Hosono,1Center for Green Research on Energy andEnvironmental Materials, National Institute forMaterials Science (NIMS), Tsukuba, Ibaraki, Japan2Graduate School of Science and Technology,Saitama University, Saitama, Japan3Research Center for Materials Nanoarchitectonics(MANA), National Institute for Materials Science(NIMS), Tsukuba, Ibaraki, Japan4Faculty of Pure and Applied Sciences, Universityof Tsukuba, Tsukuba, Ibaraki, Japan5MDX Research Center for Element Strategy,International Research Frontiers Initiative, TokyoInstitute of Technology, Yokohama, JapanCorrespondenceHideki Abe, Center for Green Research on Energyand Environmental Materials, National Institutefor Materials Science (NIMS), 1-1 Namiki,Tsukuba, Ibaraki 305-0044, Japan.Email: abe.Hideki@nims.go.jpHideo Hosono, Research Center for MaterialsNanoarchitectonics (MANA), National Institutefor Materials Science (NIMS),1-1 Namiki, Tsukuba,Ibaraki 305-0044, Japan.Email: hosono@mces.titech.ac.jpFunding informationGrant-in-Aid for Scientific Research, Grant/AwardNumber: 22H02172; Japan Science and TechnologyAgency, Grant/Award Number: JPMJMI18A3AbstractMolecular hydrogen (H2) ortho-para conversion (O/P conversion) proceeds slowly atlow temperatures accompanying a heat release. Thus, catalysts for accelerating this con-version rate are highly demanded in terms of the storage and utilization of liquidH2. Thecatalysts for this purpose are experimentally screened by examining a broad range ofmaterials covering magnetic, non-magnetic, metallic, and nonmetallic oxides. The pri-mary conclusions obtained are summarized below. (1) activematerials are required to benon-metallic and to bear the cations with ionic radii smaller than the bond length of H2.(2) Metallic materials have almost no activity irrespective of with or without magnetism(3) The activity of materials belonging to (1) is largely enhanced when the constitutingcation has a magnetic moment. In addition, there is a class of materials for which theactivity is distinctly enhanced just upon substitution by the foreign ions.KEYWORDScatalyst, hydrogen, hydrogen economy, hydrogen liquefaction, ortho-para conversion INTRODUCTIONThe dominant role of hydrogen in sustainable future energy orhydrogen economy, is widely accepted based on its excellentnature as an energy source for alternative fuel.[1,2] It is a sci-entific consensus that four key technologies are required forthe promotion of the hydrogen economy: hydrogen produc-tion, transportation, storage, and utilization.[3] The coolingand liquefaction of molecular hydrogen (H2) enable the stor-age of large quantities of hydrogen fuel.[4,5] However, there isa technical obstacle to overcome for the industrial use of liq-uefied H2: There are two isomers of molecular H2, ortho H2This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the originalwork is properly cited.© 2023 The Authors. Exploration published by Henan University and John Wiley & Sons Australia, Ltd.with nuclear spin (J) of 1 and para H2 with J = 0 -forms. Inthe thermal equilibrium state, ortho H2 and para H2 occupythe rotational ground states of J = 1 and J = 0, respectively,and the O/P ratio is 3 near the ambient temperature, but thisratio becomes 0 (only para H2 exists) at the liquefied temper-ature of H2, 20 K (Figure 1). Since ortho H2 has a non-zeromoment of J = 1, it relaxes to the para-state in the condensedstate where themolecular separation is 0.37 nm at low temper-atures at a time constant of ∼100 h at 20 K (O/P conversion).The rotational energy released from the O/P conversion cor-responding to J = 1→0 is 15 meV, which is larger than theevaporation energy (∼12meV) of liquefiedH2 (Figure 1, inset).Exploration 2024;4:20230040. wileyonlinelibrary.com/journal/exploration  of https://doi.org/10.1002/EXP.20230040https://orcid.org/0000-0002-8392-7586https://orcid.org/0000-0002-0992-7449https://orcid.org/0009-0003-2859-6934https://orcid.org/0000-0001-9260-6728mailto:abe.Hideki@nims.go.jpmailto:hosono@mces.titech.ac.jphttp://creativecommons.org/licenses/by/4.0/http://wileyonlinelibrary.com/journal/explorationhttps://doi.org/10.1002/EXP.20230040http://crossmark.crossref.org/dialog/?doi=10.1002%2FEXP.20230040&domain=pdf&date_stamp=2023-12-14 of F IGURE  Equilibrium ratio of ortho/para H2 concentration. Insetshows the energy diagram for molecular hydrogen.As a result, ∼55% of liquefied H2 is lost through evaporationby the O/P conversion heat when conventional H2 gas withan O/P ratio of 3 is equilibrated at 20 K. Such an H2 evapora-tion loss is called “boil-off.” To avoid the boil-off, catalysis topromote the O/P conversion is needed to store liquid para H2with J = 0.In 1933, E. Wigner reported a theoretical considerationon the O/P conversion and proposed that this conversion isinduced on the adsorbed H2 on the magnetic materials by aninhomogeneous magnetic field arising from the electron spinmagnetic dipole moment.[6–8] The interaction is proportionalto μ2/r6, where μ is the magnetic moment of the magnetic ionand r is the collision distance. Considering Wigner’s theory,various O/P catalysts have been developed so far. Much efforthas been devoted to the development of new catalysts sincethe 1960s.[9–13] Currently, 3d-transition metal oxides such asCr2O3 and hydrated iron(III) oxide (Fe2O3•nH2O) are usedas catalysts for this purpose.[14–18] However, it has continuedto seek more efficient catalysts as well as understanding theeffective conversion mechanism until recently.[19–22]Since then, it has been believed that this conversion isnot induced on the surface of diamagnetic materials. How-ever, in the 1980s, it was found that this conversion occurseven on various nonmagnetic materials, such as amorphousice.[23,24] Recently, the conversion of H2 confined into thenano space of metal-organic framework (MOF) has gainedattention.[25,26] So far, fundamental research on the O/P con-version on solid surfaces has been performed employing cleanmaterials, surfaces and physical techniques.[24] According tothese researches, the O/P conversion is induced by three ori-gins: magnetic fields,[6–8] charge transfer,[27–29] and electricfields.[24] Although each mechanism appears to be valid forthe materials system examined, we think it is pivotal at thepresent stage for rational catalyst design to get a rough butcomprehensive image of the concrete catalytic materials.In this work, we explore the O/P conversion at 77 Kfor a wide range of solid materials covering magnetic/non-magnetic andmetallic (no opened band gap)/semiconductingmaterials and classify these materials into different typesdepending on the conversion activity. As a consequence, theexperimental results may be classified into four types, andeach type is featured by the mechanism except for severalexceptions. The essential factor for the high catalytic activityis not magnetism but bearing the cation with small ionic radiiin the non-metallic materials, to our surprise. EXPERIMENTAL. Catalyst materialsWe investigated a wide range of solid materials coveringmagnetic/non-magnetic, metallic (no opened band gap) andsemiconducting materials. These powder samples were pur-chased from the manufacturers (Sigma-Aldrich, Alfa Aesar,Kishida Chemical, Kojundo Chemical Laboratory, FruuchiChemical, or FuruyaMetal) ormade by ourselves by gas atom-izing or gas reduction of precursors. The particle size of mostof the catalysts was in the range of 1 to 10 μm, except forsome nanoparticulate catalysts, including SnO, Sn3O4, NiO,CuO, and Pr-doped CeO2, whose particle size was smallerthan 100 nm. See Figures S1 and S2, Supporting Information,for powder X-ray diffraction (pXRD) data for the catalysts.The O/P conversion activity of various catalyst materialswas evaluated at 77 K using a batch reactor equipped with aplunger pump and a gas-tight cell, respectively for gas circu-lation and Raman spectroscopy (Figure 2). H2 gas (99.9%) aspurchased was used without further purification. An aliquotof 100 mg of catalyst particles (particle size < 10 μm) wasloaded in a glass-made sample tube with an inner diameterof 6 mm. The sample tube was attached to the batch reactor,evacuated down to 10 Pa, and backfilled with H2 gas up to80 kPa. The H2 gas was repeatedly passed through the cata-lyst layer and the gas cell in sequence, being monitored with aRaman spectrometer (JASCO RMP-510) of the population ofortho- and para H2.Figure 3 shows a series of Raman spectra acquired at 77 K atdifferent duration times after the circulating H2 gas was sub-jected to an iron oxide (Fe2O3) catalyst. The volume fractionof para H2 was calculated as 29± 1.0% from the intensity ratioof the major Raman peaks at 354.4 and 588.4 cm−1 that cor-respond to the J = 0 to J = 2 and J = 1 to J = 3 transitions(see the inset of Figure 1), respectively.[30] The evaluated frac-tion was close to 25%, which is theoretically expected at roomtemperature from the Boltzmann statistics. The para H2 frac-tion increased to reach 39 ± 1.0%, 4 h after the catalysis wasstarted by immersing the sample tube into liquid nitrogen.By contrast, the O/P conversion was hardly promoted whenaluminum oxide (Al2O3) was used as the catalyst. The paraH2 fraction remained around 25% even 5 h after the catalysiswas started, showing that theO/P conversion activity of Al2O3is negligibly low compared to that of Fe2O3 in this catalysiscondition. RESULTSFigure 4 shows the trend of the O/P conversion catalysis at77 K for different materials, including metal and single metal 27662098, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/EXP.20230040 by National Institute For, Wiley Online Library on [24/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License of F IGURE  Experimental setup for the catalytic O/P conversion measurement.F IGURE  Time evolution of Raman spectra for the O/P conversionover Fe2O3 catalysts.oxides (AOx), which are categorized into four groups. The firstis a group of materials that exhibit no finite activity towardthe O/P conversion (group 1, black curves in Figure 4). Thisgroup involves all of the metallic materials and most of theoxides containing low-valence metal cations. The para H2fraction stayed around the initial value of 25% over hourseven though H2 gas repeatedly passed through the metal cat-alysts, including bismuth (Bi), gold (Au), platinum (Pt), orintermetallic compounds such as ErAl2, GdAl2, or HoB2. Themetal oxides containing hollow or filled d-orbitals such asCu2O or ZnO were as inert as the metal catalysts. None ofthe low-valence metal oxides, such as Mn2+O, Fe2+O, Ni2+O,Sn2+O, Pb2+O, V3+2O3, In3+2O3 or Bi3+2O3 efficiently pro-moted theO/P conversion. The second group consists ofmetaloxides that contain high-valence cations such as V5+2O5,Mn4+O2, Ta5+2O5, and Sb5+2O5 (group 2, green curves inFigure 4).The group 2materials exhibited finiteO/P conversion activ-ity; the para H2 fraction monotonously increased from 25%,showing a tendency to saturate toward 50%. Most of thegroup 2 oxides comprise non-magnetic metal ions without d-F IGURE  Trends of the O/P conversion at 77 K over differentcatalysts. The black, green, blue, and red curves are assigned to the materialsof groups 1, 2, 3, and 4, respectively. Red-filled circles and open squarescorrespond to Mn3O4 and CoO, respectively. Blue-filled circles, opensquares, and filled triangles correspond to SnO2, Ho2O3, and Sn3O4,respectively. Green-filled circles, open squares, and filled trianglescorrespond to Sb2O5, V2O5, and Ta2O5, respectively. Black-filled circles andopen squares correspond to FeO and metallic Gd5Si3, respectively. The errorbars were calculated as a standard deviation of the background spectrumacquired by filling the sample tube with no catalyst.electrons (V5+, Y3+, Ta5+, Sb5+). The third group involvesmetal oxides comprising high-valence metal ions such asMn3+2O3, Cu2+O, Zr4+O2, Sn4+Sn2+2O4, Sn4+O2, Ho3+2O3and Gd3+2O3 (group 3, blue curves in Figure 4). The group3 materials are similar in ionic and catalytic nature to thegroup 2 materials, yet showed superior catalytic activity. Thelast group involves some of the metal oxides and hydroxides 27662098, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/EXP.20230040 by National Institute For, Wiley Online Library on [24/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License of F IGURE  Rate constants of the O/P conversion over representativecatalyst materials. The rate constants were obtained by numerical fitting tothe experimental data using an exponential function (see Figure S3,Supporting Information, and the caption for details on the numerical fitting).that consist of magnetic metal cations (group 4, red curvesin Figure 4). Mn2+Mn3+2O4[31] and Co2+O showed muchhigher catalytic activity than any other materials belongingto groups 1, 2 or 3. The O/P conversion over Mn3O4 and/orCoO was so fast that the para H2 fraction reached the theo-retical maximum of 50%, within a half hour. Cerium dioxide(CeO2) exhibited higher activity even than Mn3O4 or CoO.Mn3O4 and CoO have been known as semiconductors havinga bandgap of 2.8 and 2.1 eV, respectively.[32,33] Moreover, theinherent activity of CeO2 was significantly improvedwhen thetetravalent Ce4+ was partially substituted with trivalent Gd3+as gadolinium-doped ceria (Gd: CeO2). DISCUSSIONFirst, we discuss the effect of an electric field which worksin nonmagnetic ionic compounds categorized into group 3.H2 molecules are modulated in nuclear spins through aninteraction with the electrons of ions constituting the sur-face when weakly adsorbed on the catalyst surface retainingthe H-H bond (physisorption). It is acknowledged that quan-tum transition in the total spin momentum of H2 nucleifrom J = 1 to J = 0, namely the O/P conversion, can bepromoted especially when the physisorbed H2 molecules aresubjected to localized electric fields with high intensity andspatial anisotropy.[6] The effect has been considered to be keyto realizing the conversion on MOF or amorphous ice.[23,25]Such anisotropic electric fields are scarcely formed over themetal surface, where the local charge is evenly screened andsmoothed by itinerant electrons. The screened, isotropic elec-tric fields can hardly promote the nuclear spin transition,resulting in the very low O/P conversion activity of metallicmaterials (Figure 5).Here, we attempt to compare the size-matching betweenH2 and the catalyst surface. The ionic radius proposed byShannon[34] is utilized as a measure. The ionic radii for Bi3+(96 pm), Y3+ (90 pm), and In3+ (80 pm) are larger than thebond length of H2: 74.1 pm (Figure 6). The ionic radius ofPb4+ in PbO2, 78 pm, is also larger than 74.1 pm. As alreadyaddressed in the experimental section, the low-valence metalF IGURE  Relation between the O/P conversion activity of singlemetal oxides and the radius of the constituent metal ions.F IGURE  Polarization of H2 molecules by anisotropic electric fieldsbetween the oxygen- and metal ions. The steep gradient in the electricpotential (φ) at the border of the metal- and oxygen ions polarizes the H2molecule (± δ). The atomic environment around the small metal (b) ion ismore favorable than the large metal ion (a) to develop electric dipolemoments (μ).oxides such as Bi2O3, Y2O3, MnO, In2O3, and FeO showmuch low O/P conversion activity. The activity of PbO2 is aslow as that of the other low-valence metal oxides. The large-sized metal ions are not favorable to offer sufficiently strongelectric fields to the H2 molecules, which results in low O/Pconversion activity. Unlike the metal oxides comprising largemetal ions, the metal oxides composed of small metal ions(ionic radii < 74.1 pm) exhibit high O/P conversion activity(Figure 6).The H2 molecules physisorbed on the small metal ionsare strongly polarized to interact with the anisotropic electricfield, resulting in promotedO/P conversions. Over the surfaceof the metal oxides belonging to group 1, such as Mn2+O2−,Fe2+ O2−, Sn2+ O2−, and Pb2+ O2−, there are distributedlocal extrema of electric potentials at each of the surface ions.The electric potential abruptly changes the polarity at the ionboundary to develop an electric field with high anisotropy(Figure 7). However, low-valence metal ions are large: theionic radii of Fe2+ (HS: 78 pm), Mn2+ (HS: 83 pm), Sn2+ (122pm), and Pb2+ (119 pm) are larger than the bond length of theH2 molecule, 74.1 pm.[34] TheH2 molecules adsorbed on suchlarge ions are subjected to a weak electric field, which maylead to a sluggish O/P conversion similarly occurring over the 27662098, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/EXP.20230040 by National Institute For, Wiley Online Library on [24/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License of F IGURE  (A) Crystal structure of Mn3O4 and (B) CeO2. Mn3O4adopts a distorted spinel-type crystal structure.[36] The cationic distributionof Mn2+/Mn3+ obeys normal spinel type. The Mn3+O6 octahedron showsJahn-Teller distortion, inducing symmetry lowering from cubic to tetragonalcells. CeO2 takes a fluorite-type crystal structure.metal catalysts. The radii of the high-valence metal ions aregenerally smaller than those of the low-valence ions. Indeed,V5+ (ionic radius: 54 pm), Sb5+ (ionic radius: 60 pm), andTa5+ (ionic radius: 64 pm) are smaller than the correspond-ing low-valence ions, V3+ (ionic radius: 64 pm), Sb3+ (ionicradius: 76 pm) and Ta3+ (ionic radius: 72 pm), respectively.The polarization of H2 molecules on the high-valence metalsurface is large when themolecules are adsorbed, pointing theside to the surface (side-on adsorption), subjected to a strong,anisotropic electric field. High-valence metal oxides such asV5+2O5, Sb5+2O5, Ta5+2O5, and Zr4+O2 exhibit prominentO/P conversion activity due to the promoted nuclear spintransition by the electric field.Second, we discuss the character of group 4 where themag-netic field plays a crucial role. The nuclei of H2 molecules areaffected not only by electric fields but also by magnetic inter-actions with the local moments that are distributed over thecatalyst surface.[6,24] Indeed, the O/P conversion of hydrogennuclei is promoted by a magnetic dipole-dipole coupling withthe electrons of hydrogen molecules in direct contact withelectrically non-polar matter such as oxygen molecules.[35]However, the present experimental results show none of themagnetically metallic materials exhibits finite O/P activity,although some of the metals, such as Fe (Co, Ni) showinherentlymagnetism. Thus,magnetic interactions appear lesspredominant than electric interactions judging as a whole.Prominent O/P conversions are realized only if both theelectric- and magnetic interactions are constructively appliedto the H2 molecules, that is, in the case of group 4 materi-als includingMn3O4. Mn3O4 crystallizes in the normal spinelstructure (Figure 8A).[36] The Mn3+ cation is coordinated bysix oxygen atoms to form a MnO6 octahedron. The octahe-drally coordinatedMn3+ cation is so small in ionic radius thatit is likely able to apply an anisotropic electric field to the H2molecule and provide an opportunity for magnetic exchangesbetween the H2 nuclei and metal d-electrons to accelerate theO/P conversion.The same scenario may be valid for the other mag-netic materials belonging to group 3, where the metal ionsare allowed to strongly interact with the nuclei of H2molecules such as Co(OH)2, Cr(OH)3, Fe2O3, or FeOOH.This effect seen in magnetic insulators obeys Wigner’sF IGURE  Adsorption of ortho- and/or para H2 onto the (A) metalions and (B) oxygen vacancies. Black-open circles and red-closed circlesdenote oxygen (O2−) and metal ions, respectively.theory.[6] Mn2+Mn3+2O4 contains both Mn2+ (3d5 HS elec-tronic configuration) and Mn3+ (3d4 HS) ions. The effectivemoment spin-only value, μeff is given by 2[S(S+1)]01/2, whereS is the total spin of the ion. Theμeff of d5 or d4 configuration is5.92 or 4.90 BM, respectively.Mn2+ ions cannot satisfy the cri-terion of cationic radius described above, despite the large μeff(5.92 BM). On the other hand, the Mn3+ ion having a smallerionic size contributes to the magnetic interaction through amoderately large μeff (4.90 BM), even if it undergoes Jahn-Teller distortion leading to the buildup of a large anisotropicelectric field. Finally, mentioned are some exceptional cases inour classification. According to the observation in Figure 5,there seems to be another group of materials in group 4.This group consists of CeO2 and gadolinium oxide (Gd2O3).Figure 8B shows the crystal structure of CeO2. It adopts afluorite-type crystal structure in which the Ce ion is sur-rounded by eight O ions (and we can see a cavity surroundedby eight O ions). They showed similar O/P conversion trendsas the group 4 materials, such as Mn3O4, realizing 50% ofthe para-H2 fraction within 20 min (Figure 4). However,the ionic radii for Ce4+ and Gd3+ being 87 pm and 94 pm,respectively,[34] are significantly larger than the bond length ofH2, 74.1 pm.TheH2 molecules likely receive sufficiently strongelectric fields from the surface of neither Ce4+ nor Gd3+ ions.It is acknowledged that lanthanide oxides can contain highconcentrations of oxygen vacancies in the bulk and/or on thesurface, in particular when the lanthanide ions are capableof adopting different valences. As-synthesized CeO2 mate-rials often contain trivalent Ce3+ ions, being accompaniedby equivalent oxygen vacancies. The charged oxygen vacancymay adsorbH2 molecules in a similar way as themetal cations(Figure 9). Ortho H2 molecules are efficiently converted topara H2 before leaving the surface due to the anisotropic elec-tric field at the oxygen vacancies. This scenario is supported bythe experimental fact that a solid solution of CeO2 andGd2O3(Ce4+0.8Gd3+0.2O1.9) containing a high concentration of oxy-gen vacancy exhibits prominent O/P conversion activity. Arti-ficially impregnated oxygen vacancies in Ce4+0.8Gd3+0.2O1.9most likely play the role of a catalysis center. SUMMARYWehave screened awide range ofmaterials to explore effectivecatalysts for the O/P conversion of H2 at 77 K. The primaryconclusions are summarized as follows: (1) The O/P conver-sion catalysts are categorized into four groups. (2) The first 27662098, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/EXP.20230040 by National Institute For, Wiley Online Library on [24/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License of is a group of materials that exhibit no finite activity towardthe O/P conversion (group 1). This group involves all of themetallic materials and most of the oxides containing low-valencemetal cations. (3)Group 2 consists ofmetal oxides thatcontain high-valence cations, such as V5+2O5. The materialsexhibited finite O/P conversion activity: the para H2 frac-tion monotonously increased from 25% showing a tendencyto saturate toward 50%. Most of the group 2 oxides com-prise non-magnetic metal ions without d-electrons (V5+, Y3+,Ta5+, Sb5+). (4) The group 3 materials, such as Mn2O3 orCuO are similar in the ionic and catalytic nature to the group2 materials yet show superior catalytic activity. (5) Group 4involves some of the metal oxides and hydroxides that con-sist ofmagneticmetal cations.Mn3O4 andCoO showedmuchhigher catalytic activity among the four groups. (6) Althoughthere are various factors to influence the catalytic activity,the surface electric field in ionic compounds seems to be themost important one, which induces the electric polarizationof adsorbed H2.We can easily estimate this effect considering the cationicsize in metal oxides: the oxides containing smaller cations(ionic radius < 75 pm) tend to show the activity empirically.This effect appears to work well for MOF materials but notfor metallic materials with itinerant electrons. (7) Magnetisminduced by open-shell transition metal cations in the oxidecatalysts enhances the catalytic activity through the theoryproposed by Wigner. The activity of ionic oxide semicon-ductors, including Mn3O4 and CoO originates from bothelectric fields and magnetic interactions. (8) CeO2 also showshigh activity without magnetic ions. It contains a large Ce4+ion (size > 75 pm) and does not satisfy the criterion aboutcationic size. While we cannot find the primary factor forthis oxide, the oxygen vacancy created as a result of the for-mation of Ce3+ would give rise to a large and anisotropicelectric field on the surface. (9) There is a class of materials forwhich the activity change is distinct, just doping the foreignions, but cannot be understood at this stage. While furtherefforts are needed to solidify the scientific base, these find-ings obtained through material exploration would be usefulto design the optimal catalyst. It was unexpected for us thatthe electric field effect was more dominant than the magneticinteraction. This finding is quite consistent with the inert-ness of metallic materials. The electric field gradient over H2adsorbed on the material surface appears to be a critical fac-tor for effective O/P conversion catalysis. The results of thepresent exploratory research imply the conversion would becontrolled by enhanced spin-orbital interaction[23] or nuclearquadrupole interaction.ACKNOWLEDGEMENTSThis work was supported by Grant-in-Aid for ScientificResearch (No. 22H02172) from the Japan Society for thePromotion of Science (JSPS), JST MIRAI Program (No.JPMJMI18A3).CONFL ICT OF INTERESTS STATEMENTThe authors declare no conflicts of interest.DATA AVAILAB IL ITY STATEMENTThe data that supports the findings of this study are availablein the supplementary material of this article.ORCIDHidekiAbe https://orcid.org/0000-0002-8392-7586HiroshiMizoguchi https://orcid.org/0000-0002-0992-7449RyutoEguchi https://orcid.org/0009-0003-2859-6934HideoHosono https://orcid.org/0000-0001-9260-6728REFERENCES[1] L. Barreto, A. Makihira, K. Riahi, Int. J. Hydrogen Energy , , 267.[2] W. McDowall, M. Eames, Energy Policy , , 1236.[3] F. Zhang, P. Zhao, M. Niu, J. Maddy, Int. J. Hydrogen Energy , ,14535.[4] G Nazir, A. Rehman, S. Hussain, S. Aftab, K. Heo, M. Ikram, S. Patil, M.A. U. Din, Adv. Sustainable Syst. , , 2200276.[5] S.Ghafri, S.Munro,U.Cardella, T. Funke,W.Notardonato, J.M.Trusler,J. Leachman, R. Span, S. Kamiya,G. Pearce, A. Swanger, E.D. Rodriguez,P. Bajada, F. Jiao, K. Peng, A. Siahvashi, M. Johns, E. May, EnergyEnviron. Sci. , , 2690.[6] E. Wigner, Z. Phys. Chem. , B, 28.[7] Y. Ishi, S. Sugano, Surf. Sci. , , 21.[8] E. Ilisca, Prog. Surf. Sci. , , 217.[9] L. Farkas, Usp. Fiz. Nauk, , , 347.[10] A. Zhuzhgov, O. Krivoruchko, L. Isupova, O. Mart’yanov, V. Parmon,Catal. Ind. , , 9.[11] D. Ashmead, D. Eley, R. Rudham, J. Catal. , , 280.[12] P. W. Selwood, J. Am. Chem. Soc. , , 2676.[13] C. Ng, P. W. Selwood, J. Catal. , , 252.[14] Ionex Type OP Catalyst, https://www.molecularproducts.com/products/ionex-type-op-catalyst (accessed: 12/5/2023).[15] D. H. Weitzel, W. Loebenstein, J. Draper, O. Park, J. Res. Nat. Bur. Std., , 221.[16] R. Svadlenak, A. Scott, J. Am. Chem. Soc. , , 5385.[17] R. A. Buyanov, Kinet. i Kataliz , , 306.[18] R. A. Buyanov, Kinet. i Kataliz , , 418.[19] M. Matsumoto, J. Espenson, J. Am. Chem. Soc. , , 11447.[20] T. Das, S. Kweon, J. Choi, S. Y. Kim, I.-H. Oh, Int. J. Hydrogen Energy, , 383.[21] J. H. Kim, S. W. Kang, I. W. Nah, I.-H. Oh, Int. J. Hydrogen Energy ,, 15520.[22] O. Boeva, A. Odintzov, R. Solovov, E. Abkhalimov, K. Zhavoronkova,B. Ershov, Int. J. Hydrogen Energy , , 22897.[23] T. Sugimoto, K. Fukutani, Nat. Phys. , , 307.[24] K. Fukutani, T. Sugimoto, Prog. Surf. Sci. , , 279.[25] T. Kosone, A. Hori, E. Nishibori, Y. Kubota, A. Mishima, M. Ohba, H.Tanaka, K. Kato, J. Kim, J. Real, S. Kitagawa, M. Takata, Roy. Soc. OpenSci. , , 150006.[26] D. Polyukhov, N. Kudriavykh, S. Gromilov, A. Kiryutin, A. Poryvaev,M. Fedin, ACS Energy Lett. , , 4336.[27] T. Sugimoto, K. Fukutani, Phys. Rev. Lett. , , 146101.[28] E. Ilisca, Phys. Rev. Lett. , , 667.[29] E. Ilisca, Europhys. Lett. , , 18001.[30] B. Stoicheff, Can. J. Phys. , , 730.[31] H. Abe, C. Nishimura, Y. Nohara, N. Okura, C. Fukuhara, R. Watanabe,M. Akaishi, K. Toyoshiba, WO Patent PCT/JP2022/040546, .[32] S. Khan, A. Hussain, K. He, B. Liu, Z. Imran, J. Ambreen, S.Hassan, M. Ahmad, S. S. Batool, C. Li, J. Environ. Manage. , ,112854.[33] K. Baraik, A. Bhakar, V. Srihari, I. Bhaumik, C.Mukherjee,M. Gupta, A.Yadav, P. Tiwari, D. Phase, S. Jha, S. Singh, T. Ganguli, RSC Adv. ,, 43497.[34] R. D. Shannon, Acta Crystallogr. , , 751.[35] X. Zhang, T. Karman, G. Groenenboom, A. Avoird, Nat. Sci. , ,10002. 27662098, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/EXP.20230040 by National Institute For, Wiley Online Library on [24/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://orcid.org/0000-0002-8392-7586https://orcid.org/0000-0002-8392-7586https://orcid.org/0000-0002-0992-7449https://orcid.org/0000-0002-0992-7449https://orcid.org/0009-0003-2859-6934https://orcid.org/0009-0003-2859-6934https://orcid.org/0000-0001-9260-6728https://orcid.org/0000-0001-9260-6728https://www.molecularproducts.com/products/ionex-type-op-catalysthttps://www.molecularproducts.com/products/ionex-type-op-catalyst of [36] M. Bekheet, I. Svoboda, N. Liu, L. Bayarjargal, E. Irran, C. Dietz, R.Stark, R. Riedal, J. Solid State Chem. , , 54.SUPPORTING INFORMATIONAdditional supporting information can be found online in theSupporting Information section at the end of this article.How to cite this article: H. Abe, H. Mizoguchi, R.Eguchi, H. Hosono, Exploration , , 20230040.https://doi.org/10.1002/EXP.20230040 27662098, 2024, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/EXP.20230040 by National Institute For, Wiley Online Library on [24/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons Licensehttps://doi.org/10.1002/EXP.20230040 Exploration of heterogeneous catalyst for molecular hydrogen ortho-para conversion Abstract 1 | INTRODUCTION 2 | EXPERIMENTAL 2.1 | Catalyst materials 3 | RESULTS 4 | DISCUSSION 5 | SUMMARY ACKNOWLEDGEMENTS CONFLICT OF INTERESTS STATEMENT DATA AVAILABILITY STATEMENT ORCID REFERENCES SUPPORTING INFORMATION