# Fileset
[Tansho_Main_Article_revised_20220706_cleaned_2.pdf](https://mdr.nims.go.jp/filesets/f7ae4282-76d6-4f7a-a8f3-68674b485742/download)
## Creator
[Masataka Tansho](https://orcid.org/0000-0001-7986-3199), [Atsushi Goto](https://orcid.org/0000-0002-9472-4098), [Shinobu Ohki](https://orcid.org/0000-0002-7357-3833), [Yuuki Mogami](https://orcid.org/0000-0002-9807-3165), Yuichi Sakuda, Yuta Yasui, Taito Murakami, Kotaro Fujii, Takahiro Iijima, Masatomo Yashima
## Rights
This document is the Accepted Manuscript version of a Published Work that appeared in final form in The Journalof Physical Chemistry C, copyright © 2022 American Chemical Society after peer review and technical editing by the publisher. To access the final edited and published work see https://doi.org/10.1021/acs.jpcc.2c03429[In Copyright](http://rightsstatements.org/vocab/InC/1.0/)
## Other metadata
[Different Local Structures of Mo and Nb Polyhedra in the Oxide-Ion-Conducting Hexagonal Perovskite-Related Oxide Ba3MoNbO8.5 Revealed by 95Mo and 93Nb NMR Measurements](https://mdr.nims.go.jp/datasets/f18cccdb-dc90-4e12-a4e5-4af8f9b9d55a)
## Fulltext
1 Different Local Structures of Mo and Nb Polyhedra in the Oxide-Ion-Conducting Hexagonal Perovskite-Related Oxide Ba3MoNbO8.5 Revealed by 95Mo and 93Nb NMR Measurements Masataka Tansho,∗,† Atsushi Goto,† Shinobu Ohki,† Yuuki Mogami,† Yuichi Sakuda,‡ Yuta Yasui,‡ Taito Murakami,‡∇ Kotaro Fujii,‡ Takahiro Iijima,§ and Masatomo Yashima‡ †NMR Station, National Institute for Materials Science (NIMS), 3-13 Sakura, Tsukuba 305-0003, Japan ‡Department of Chemistry, School of Science, Tokyo Institute of Technology, 2-12-1-W4-17, O-okayama, Meguro-ku, Tokyo, 152-8551, Japan §Institute of Arts and Sciences, Yamagata University, Yamagata 990-8560, Japan 2 ABSTRACT The oxide-ion conductor Ba3MoNbO8.5, the oxide-ion and proton conductor Ba7Nb4MoO20, and their hexagonal perovskite-related oxides are important groups of materials because of their high ionic conductivity. The structure of the ion-conducting layer of these materials has not been clarified because of their complex structure and the difficulty in distinguishing between Mo and Nb. In this study, we separately detected 95Mo and 93Nb by solid-state nuclear magnetic resonance (NMR) measurements to directly observe the Mo/Nb coordination in the high oxide-ion conductor Ba3MoNbO8.5. The results showed that the number of revealed peaks was different for 93Nb and 95Mo. For the two chemical shifts from 93Nb NMR, the more intense peak was attributed to a NbO6 octahedron in the conducting layer, while the less intense peak was ascribed to a NbO4 tetrahedron in the conducting layer or a NbO6 octahedron in the non-conducting layer. Four peaks were observed in the 95Mo NMR of the 95Mo-enriched sample. One peak was attributed to the MoO6 octahedron in the non-conducting layer. The other three peaks attributed to the conducting layer were only interpreted by assigning either one or two of them to the MoO5 polyhedra, which are speculated to play an important role in ionic conduction. Presumably, these are the first results supporting the presence of MoO5 in the ion-conducting layer of oxide-ion conductors, and Mo likely plays an important role in ionic conduction. The analysis of the local structure of Mo and Nb oxide polyhedra by NMR is an important tool for understanding the nature of ionic conduction because it provides element-independent information. It is therefore expected to contribute to the further development of oxide-ion conductors. INTRODUCTION 3 The development of solid oxide-ion (O2-) conductors has led to a range of energy and environmental technologies, including gas sensors, solid oxide fuel cells (SOFCs), and oxygen separation membranes. Research on the high energy efficiency of SOFCs at low temperatures has led to the discovery of new oxide-ion conductors and an improved understanding of the diffusion mechanisms of oxide ions.1–5 Recently, it has been reported that hexagonal perovskite-based materials such as Ba3MoNbO8.5 and the related materials,6–17 which are oxide-ion conductors, as well as Ba7Nb4MoO2018,19 and Ba7Ta3.7Mo1.3O20.15,20 which are mixed conductors of oxide ions and protons, show high ion conductivity. For example, Ba3MoNbO8.5 exhibits oxide-ion conduction over a pO2 range of 10−20–1 atm with a bulk conductivity of 2.2 × 10−3 Scm−1 at 600 °C. The complex structures of these materials have been shown to be related to their high ionic conductivity. In Ba7Ta3.7Mo1.3O20.15, the structural analysis of synchrotron X-ray diffraction data with the support of density functional theory calculations shows that Mo6+ is preferentially present in the conduction layer.20 For Ba3MoNbO8.5, represented by A3B2O8.5 (A = Ba, B = M = Mo/Nb), the structure is based on the 9R polytype A3B3O9 and a cation-anion deficient derivative of palmierite A3B2O8.6 Although it was initially believed that one-third of the vacancies were distributed in the B-site,7 it has recently been demonstrated that up to two cations can occupy each M1–M2–M1 polyhedral stack, and that the M2-site is distributed in two equivalent positions along the c-axis, maintaining the M2O6 octahedral structure (Figure 1).14 Here, the O2 and O3 oxygen atoms coordinated to M1 are statistically distributed with an occupancy of less than 1, whereas the O1 oxygen atom coordinated to M2 is coordinated with an occupancy of 1.7,14 In addition, the M1 sites were initially believed to contain only M1O4 and M1O6 polyhedra, and the oxide-ion conductivity was correlated with the change in the ratio of M1O4 tetrahedra to M1O6 octahedra with increasing temperature using 4 Rietveld refinement.8,10 However, the very large atomic displacement parameter of O2 obtained from single-crystal diffraction results suggests that five-fold coordination is also very likely (Table 1).14,17 Oxide ions were also found to migrate two-dimensionally through the mixed oxygen sites of O2 octahedra and O3 tetrahedra on the O2–O2–O2 faces of the M1O5–Ɛ polyhedra, where Ɛ is the oxygen vacancy concentration.15 The O2/O3 disordering makes the minimum neutron scattering length densities on the O2−O3 path higher, which enhances oxide-ion conductivity, leading to higher activation energies of Ba3WNbO8.5 compared with Ba3MoNbO8.5.12 Recently, modeling using a combination of Bragg, D(r), and F(q) data from neutron scattering experiments suggested a preference for M1O5 polyhedra in Ba3MoNbO8.5.16 In this study, the coordination numbers of the M cations were determined using a cut-off length based on the M–O distance histograms obtained from reverse Monte-Carlo simulations. The variable coordination environment and coordination dynamics of the M1 site in Ba3MoNbO8.5 are believed to be important for creating a low-energy transfer pathway for oxide ions; however, these structures are usually studied by X-ray diffraction, neutron scattering, and electron diffraction, assuming an even distribution of Mo and Nb at each site and analyzed as if there were averaged Mo and Nb atoms, although there were no averaged Mo/Nb atoms.6–17 This is because X-rays, neutrons, and electrons all have a close scattering ability, rendering distinguishing between Nb and Mo difficult. In addition, the reverse Monte-Carlo simulation is very model-dependent. Hence, more reliable and direct evidence of the presence of M1O5 polyhedra in Ba3MoNbO8.5 is necessary. 5 Figure 1. Part of the crystal structure of Ba3MoNbO8.5. (a) Averaged structure is represented as a hybrid consisting of the overlapping 9R polytype and palmierite subunits, where M is Mo/Nb, the space group is R3�m, and M2 sites are C3 symmetric and distributed in two positions along the c-axis. The occupancies of M1 and M2 at room temperature are 0.881 and 0.119, respectively.14 (b) Structural cross-sectional in the c-axis direction. The layer containing partially occupied O2/O3 sites is shown. The arrows in both directions indicate the diffusion directions of oxide ions. In the M1 site, in addition to M1O4 and M1O6, the M1O5 polyhedron is likely an important structure related to oxygen diffusion in the pathway via O2 and O3. Each site is described in the same way as in ref 7. 6 Table 1. Local structures known and suggested to exist in Ba3MoNbO8.5 site a Local structure known to exist Local structure suggested to exist M1 M1O13O31 and M1O13O23 M1O5 M2 M2O16 – aBoth M1 and M2 are for (Nb/Mo) sites, and the notation is the same as in ref. 7. Nuclear magnetic resonance (NMR), which, in principle, directly reflects the electronic state around the nucleus, is advantageous for determining local structures. Specifically, we measured the solid-state NMR of 95Mo and 93Nb in Ba3MoNbO8.5 for two reasons: the ease of distinguishing between Mo and Nb, and the possibility of splitting the signal for each coordination number, which would include the MO5 polyhedron.21–26 For the measurement of Mo, 95Mo was selected from NMR-active nuclei because of its suitability for high-resolution analysis; NMR measurements were also conducted on a sample enriched with 95Mo to improve sensitivity. These NMR measurements revealed the diffusion mechanism of oxide ions by analyzing the individual coordination structures of Nb and Mo, providing new opportunities for the design of oxide-ion conductors with hexagonal perovskite-related structures. EXPERIMENTAL SECTION Ba3MoNbO8.5 and 95Mo-enriched samples were synthesized using a solid-state reaction method. The starting materials were high-purity (>99.9%) powders of BaCO3, MoO3, and Nb2O5 for Ba3MoNbO8.5 and high-purity (>99.9%) powders of BaCO3; 95MoO3 (95Mo = 95.5 atom% of Mo isotopes, Trace Sciences International); and Nb2O5 for 95Mo-enriched samples. The 7 Ba3MoNbO8.5 sample was synthesized as previously reported.15 The 95Mo-enriched sample was synthesized as follows: the starting materials were weighed in a molar ratio of Ba:Mo:Nb = 3:1:1 and then mixed and ground for approximately 1 h in an agate mortar as dried powders and as ethanol slurries. This mixture was calcined in air at 900 °C for 12 h. The calcined sample was ground for approximately 1 h in an agate mortar. Thereafter, the mixture was uniaxially pressed into pellets at approximately 250 MPa and sintered in air at 1100 °C for approximately 48 h. The calcined pellets were crushed in a WC mortar and ground for approximately 1 h in an agate mortar. All NMR measurements were conducted with a fabricated 3.2-mm single resonance magic angle spinning (MAS) probe at 18.79 T without temperature control, where the resonance frequencies for 95Mo and 93Nb were 52.16 and 195.84 MHz, respectively. The natural abundances and spin number, I, were 15.7% and 5/2 for 95Mo and 100% and 9/2 for 93Nb. All samples were rotated at 20 kHz. A 0 ppm 2.0 M Na2MoO4 solution was used as a reference for the chemical shifts of 95Mo, and −1093 ppm22,27 at 18.79 T in NaNbO3 (Sanwa Chemical Industry, 99.9%) was used as a convenient secondary chemical shift reference for the chemical shifts of 93Nb instead of a saturated solution of NbCl5 in acetonitrile at 0 ppm. A JEOL ECA 800 NMR spectrometer was used for the one-dimensional measurement of 95Mo and two-dimensional measurement of 93Nb. A JEOL ECZR 800 NMR spectrometer was used for the one-dimensional measurement of 93Nb. One-dimensional 95Mo spectra were acquired with 13,000–22,000 scans using either a single pulse of 1.2–1.6 μs, corresponding to a liquid standard π/6 pulse, or a spin-echo sequence28 (2.5 and 5.0 μs) with a relaxation delay of 20 s. Due to the severe baseline distortion in single-pulse 93Nb MAS measurements, one-dimensional 93Nb spectra were acquired in a spin-echo sequence28 (2.0 and 4.0 μs) with 1,024 scans, and the 8 relaxation delay was set to 1 s. Here, 2.0 μs corresponds to the π/3 pulse of solid NaNbO3. For the multi-quantum (MQ) MAS NMR measurement of 93Nb, we employed three-quantum (3Q) MASNMR, a type of MQMAS NMR, measurement with a three-pulse (2.0, 0.9, and 15 μs) sequence using a zero-quantum filter, as proposed by Amoureux et al.29 Here, the spectrum was recorded with 264 transients averaged for each of the 1024 t1 points and a relaxation delay of 0.2 s, which was sufficient to obtain an adequate signal-to-noise ratio. The magnitude of the quadrupole coupling constant, CQ, of the 95Mo resonance for 95Mo-enriched Ba3MoNbO8.5 was evaluated using the DMFIT simulation software package.30 In the case of quadrupole nuclei such as 93Nb, the apparent shift measured by one-dimensional MAS NMR, δ MAS, is usually different from the isotropic shifts, δiso. For solid solutions such as Ba3MoNbO8.5, it is difficult to separate the contribution of asymmetric parameters from the PQ values because the line widths are broadened by the distribution of chemical shifts,31 and the analysis of the 3QMAS result was performed as follows:32–35 the positions of the experimental resonances (δF1 and δF2) in the F1 and F2 dimensions of the sheared MQMAS spectrum and quadrupole product, PQ, (the same as the second-order quadrupolar effect (SOQE)) are calculated as δ iso =17δF1 + 10δF227 (1) and 𝑃𝑃Q = �1027‧173�1/2 [4𝐼𝐼(2𝐼𝐼 − 1)] [4𝐼𝐼(𝐼𝐼 + 1) − 3]1/2 𝜈𝜈L‧10−3 (δF1 − δF2)12. (2) 9 Here, δF1 and δF2 are chemical shifts in the triple and single quantum dimensions, respectively, and νL is the Larmor frequency. PQ is related to CQ by PQ = CQ (1 + η2/3)1/2. The asymmetry parameter η takes values between 0 and 1, and PQ becomes equal to CQ when η = 0. In the case of I = 9/2 as 93Nb, PQ values are obtained by 𝑃𝑃Q = �1224‧1027(δ𝐹𝐹1 − δ𝐹𝐹2)�1/2𝜈𝜈L‧10−3. (3) RESULTS AND DISCUSSION 93Nb NMR. Figures 2a and 2b show, respectively, the 93Nb MAS NMR measurements using the spin-echo method28 and 3QMAS NMR spectrum29 for a powder sample of Ba3MoNbO8.5 at room temperature. In Figure 2a, two chemical shifts are observed at approximately −750 and −950 ppm, apart from the ghost signal called the spinning sideband. In general, MQMAS measurements are less sensitive than single-pulse MAS measurements, but as shown in Figure 2b, an adequate spectrum was obtained, and the distribution along the chemical shift axis is clearly shown for the more intense signal. From the peak position (δF1, δF2) in the contour map of the 3QMAS spectrum, the values of δiso and PQ were estimated using Eqs. (1)–(3), and the results are listed in Table 2. 10 Figure 2. (a) 93Nb MAS NMR and (b) 93Nb 3QMAS NMR spectra of Ba3MoNbO8.5. The diagonal line in the 2D diagram is called the chemical shift axis; a spreading of the signal along the diagonal line indicates a distribution in chemical shift. Asterisks (*) denote spinning sidebands. Arrows indicate the position of each peak. 11 Table 2. δF1, δF2, δiso, and PQ values obtained from 93Nb 3QMAS measurements on Ba3MoNbO8.5, with attributed sites and attributed coordination numbers δF1 / ppm δF2 / ppm δiso / ppm PQ / MHz Attributed sites a Attributed coordination number −753 −775 −761 20 M1 (or M2) 4 (M1) (or 6 (M2)) −937 −949 −941 14 M1 6 aBoth M1 and M2 are for (Nb/Mo) sites, and the notation is the same as in ref 7. The more intense signal of −937 ppm is attributed to the following: the M1 and M2 sites are 88.1% and 11.9% of the total M sites, respectively.14 Even with all Mo biased toward the M1 site notwithstanding,20 Nb is 38.1% and 11.9% at the M1 and M2 sites, respectively. For the half-integer quadrupolar nuclei, the signal intensity ratio does not often reflect the quantity ratio due to the amplitude of the quadrupole coupling. However, as there is no significant difference in the PQ values of the two peaks (Table 2), the ratio rather reflects the quantity ratio even in the echo measurement. Therefore, the signal at the M2 site in the non-conducting layer cannot be the strongest, so the more intense peak at −937 ppm is identified as the M1 site in the conducting layer. We considered three possible M1 site assignments: M1O4, M1O5, and M1O6. The M1O5 polyhedra has an intrinsically large PQ value owing to its low symmetry, but its PQ value of 14 MHz is much smaller than the value of a possible NbO5 polyhedron of approximately 80 MHz; hence, the M1O5 structure is excluded from consideration.21,22 Of the M1O4 and M1O6 polyhedra, the PQ values of NbO6 and NbO4 are reported to be 0–50 MHz and approximately 80 MHz, respectively, with NbO6 tending to be smaller.21,22 Moreover, in most niobium oxide 12 compounds, sites with higher coordination numbers tend to have more negative chemical shifts.21–23 For example, in Ba4Nb2O9, it has been reported that the δiso of NbO4, NbO5, and NbO6 are −714 and −722 ppm; −822 and −871 ppm; and −818, −832, −865, and −931 ppm, respectively.23 Therefore, of the two possibilities for the M1O4 and M1O6 polyhedra, the −937 ppm of δiso is attributed to the signal from the Nb1O6 octahedron. In addition, the signal spread in the direction of the chemical shift axis in the 2D map is considered to be due to the M1 site being located within the conducting layer with a mixture of different local Mo/Nb structures. For the less intense signal at −753 ppm of δiso, the evidence supporting attribution to M1O4 and M2O6 polyhedra is considerable. The M1O5 polyhedron, however, cannot have a PQ value as small as 20 MHz. For M1O4, the relatively small PQ value can be explained by the possibility that either the electric field gradient around the nucleus is averaged to some extent owing to thermal motions such as lattice vibrations, or the local structure is essentially symmetrical. M2O6 was shifted exceptionally in the positive direction, compared to what is typically observed for NbO6 octahedra.21–23 Therefore, the signal is more likely due to M1O4. For the M1O5 polyhedron, the lack of a 3QMAS signal suggests that either there are few Nb1O5 polyhedra or the line broadening due to large quadrupolar interactions prevented them from being observed.22 Here, in reference 23, a case with the small CQ values is reported for 5-coordinated structures, but we believe that these values should not be used as a reference for the following two reasons. The first reason is the exceptionally small CQ values for similar line widths in this literature compared to our data and other literature. As shown in Equation (2), different PQ or CQ values are obtained for the same line width depending on the magnitude of I. This may be because the formula for I = 3/2, which is used in many nuclei, was incorrectly used 13 in the analysis for 93Nb, where I = 9/2. The second reason is that the 5-coordination studied in Ba3MoNbO8.5 does not depend on hydration, and the 5-coordination by hydration as described in the literature may cause line width sharpening due to accelerated molecular motion. 95Mo NMR. Figures 3a, b, and c show the results of 95Mo MAS NMR measurements of the powder samples of Ba3MoNbO8.5 at room temperature by the single pulse method, 95Mo-enriched Ba3MoNbO8.5 by the single pulse method, and 95Mo-enriched Ba3MoNbO8.5 by the spin-echo method,28 respectively. In Figure 3a, two signals were observed at approximately −40 ppm to −50 ppm. One is a sharp doublet-like signal typical of half-integer spin quadrupole nuclei in axially-symmetric sites observed at approximately −40 ppm with η ~ 0, as simulated in Figure 3d,30 and the other is a broad signal with a clear tail to low frequency at approximately −50 ppm. The CQ and δiso values obtained from the DMFIT simulation software for the narrow peak at approximately −40 ppm (Figure 3b) and the peak position of the broad signal are listed in Table 3. According to published reports, the 95Mo chemical shifts of the MoO4 tetrahedron in salts cover a wide range of values, from 131 ppm in CdMoO4 to −122 ppm in CsLiMoO4.24,25 The 95Mo chemical shifts of the MoO6 octahedron in salts range from 205 to 29 ppm for three sites of [(NH4)6Mo7O24]‧4H2O.26 Therefore, it is difficult to discuss the coordination number only in terms of the magnitude of the shift. Notably, the CQ values of 95Mo are less than those of 93Nb because those values are dependent on the properties of each nucleus. The signal at −40 ppm has a CQ of 1.8 MHz, within the normal tetrahedral and octahedral values of 0–3 MHz;24–26 hence, we can exclude M1O5, which cannot have spherical symmetry and therefore should have a larger CQ. Furthermore, M1O4 is also excluded, as its asymmetry parameter cannot be zero because O3 oxygen does not have axial C3 symmetry. The M2 site can be split into two equivalent sites,14 but the M2 sites have axial C3 symmetry; hence, it is reasonable to simulate 14 the signal with it when the asymmetry parameter is zero, as shown in Figure 3d. As for M1 site, it also may have axial C3 symmetry when it only adopts the M1O6 structure; hence, it is possible that this signal is derived from M1O6. However, the M1 site is considered to be affected by structural disorder due to the mixing of different local Mo/Nb structures in the conducting layer, even in the M1O6 structure, and the axial-symmetrical signal is attributed to the M2 site in the non-conducting layer. Meanwhile, the broad signal with the tail at approximately −50 ppm is considered suitable to be attributed to one of the M1 sites, regardless of the coordination number, M1O4, M1O5, or M1O6 in the conducting layer (Table 3). Here, for this peak, the distribution of chemical shifts due to the structural disorder renders obtaining the reliable CQ value by line shape analysis difficult. 15 Figure 3. 95Mo MAS NMR spectra of (a) Ba3MoNbO8.5 measured by the single pulse method, (b) 95Mo-enriched Ba3NbMoO8.5 measured by the single pulse method, (c) 95Mo-enriched Ba3MoNbO8.5 measured by the Oldfield echo method and (d) the DMFIT simulation30 result for the sharp peak around −40 ppm. The asterisk (*) indicates a spinning sideband. Arrows indicate the position of each peak. 16 Table 3. 95Mo peak positions, isotropic chemical shifts, quadrupolar parameters, attributed sites, and attributed coordination numbers for Ba3MoNbO8.5 δ MAS / ppm δ iso / ppm CQ / MHz η Attributed sites a Attributed coordination number 160 – – – M1 4, 5, or 6 10 – – – M1 4, 5, or 6 −50 – – – M1 4, 5, or 6 −33 1.8 0b M2 6 aBoth M1 and M2 are for (Nb/Mo) sites, and the notation is the same as in ref. 7. bIn the present simulation, this value was fixed at 0 for axial C3 symmetry. The peak at approximately +350 ppm in Figure 3a is a spinning sideband; however, another signal that appears at approximately +160 ppm has not been assigned. We will now discuss the possible assignment of this peak. When we measured a sample enriched with 95Mo, the signal-to-noise ratio was greatly improved, as shown in Figure 3b, but the width of the signal appears to increase between +200 ppm and 0 ppm. Because the single-pulse MAS measurements shown in Figures 3a and 3b tend to distort the baseline due to the phase shift, the echo method was used to measure the 95Mo-enriched sample, and a signal was observed at +10 ppm, in addition to +160 ppm, as listed in Table 3. The echo method has the disadvantage of a low signal-to-noise ratio and a deviation between intensity and volume ratios related to the amplitude of the quadrupolar coupling; however, it is known for its low baseline distortion. 17 The two chemical shifts at +160 and +10 ppm in Figure 3c may originate from different sites or from a single site. If they originate from different sites, the chemical shift at approximately +10 ppm is related to a different local structure than the peak at +160 ppm, and it is reasonable to attribute each of the three peaks at +160, +10, and −50 ppm to one of the three polyhedra, M1O4, M1O5, or M1O6 (Table 3). If both peaks originate from a single site, this implies that a broad spectrum ranging from +160 to +10 ppm has a CQ of 8 MHz and η of almost 0 (refer to the Supporting Information). This CQ value is much larger than the values up to approximately 3 MHz that are typical for other MoO4 and MoO6 polyhedra.24–26 Furthermore, because the PQ results of 93Nb are smaller than usual, it is natural that the CQ values of 95Mo at the same site are also smaller than usual; further, a larger-than-normal CQ of 8 MHz is not possible for either the MoO4 or MoO6 polyhedra. To the best of our knowledge, no 95Mo NMR spectra of MoO5 polyhedra have been reported. However, large CQ values are only possible for MoO5 polyhedra, which are relatively difficult to form in a highly symmetric structure. Here, it is possible that η may be close to zero only for structures with near-axial symmetry, such as a square pyramid or a triangular bipyramid.14,17 In both cases, therefore, the experimental data are mostly likely explained as MoO5 polyhedra in the conducting layer, although the attribution of the broad chemical shifts observed is still unclear (Table 3). CONCLUSIONS The initial analysis of Ba3MoNbO8.5, using Rietveld refinement, suggested that the conducting layer contains only M1O4 and M1O6 polyhedra, whereas single-crystal analysis suggested the possibility of the presence of M1O5 polyhedra. The presence of a five-coordinate (Mo/Nb)O5 has 18 been suggested by reverse Monte-Carlo total neutron scattering; however, the results were not highly reliable. In this study, the local structure of Ba3MoNbO8.5 was investigated by individual 93Nb and 95Mo solid-state NMR measurements, which directly reflect the electronic state around 95Mo/93Nb nuclei. 93Nb NMR data showed the presence of Nb1O6 octahedra and either Nb1O4 tetrahedra or Nb2O6 octahedra. The NMR spectra of 95Mo-enriched Ba3MoNbO8.5 showed four peaks; one was attributed to MoO6 octahedra in the non-conducting layer, and the other three peaks could not be attributed, assuming there are only MoO4 and MoO6 polyhedra in the conducting layer. Therefore, it is concluded that a significant proportion of Mo1O5 polyhedra is present in the oxide-ion conducting layer, and that Mo6+ plays a significant role in the enhanced conductivity of Ba3MoNbO8.5. The direct observation of the local structure by solid-state NMR is important for understanding the nature of ionic conduction and is an important tool in the development of solid electrolytes such as hexagonal perovskite-related oxide-ion conductors. ASSOCIATED CONTENT Supporting Information. Results of the DMFIT simulation in the event that both the +160 and +10 ppm chemical shifts appear from the same local structure in the 95Mo MAS NMR spectrum of 95Mo-enriched Ba3MoNbO8.5 in Figure 3c (blue line). (PDF) AUTHOR INFORMATION Corresponding Author 19 *TANSHO.Masataka@nims.go.jp Present Addresses∇Department of Electronic Engineering, Graduate School of Engineering, Tohoku University, Sendai 987-8579, Japan. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding Sources Any funds used to support the research of the manuscript should be placed here (per journal style). Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS A part of this work was supported by NIMS microstructural characterization platform as a program of "Nanotechnology Platform" of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan, Grant Number JPMXP09A19NM0110. We express our gratitude to Dr. T. Shimizu and Mr. K. Deguchi of NIMS for their help with NMR measurements. We also express our gratitude to Mr. Y. Shimoikeda of Jeol Resonance Inc. for his help with MQMAS analysis. ABBREVIATIONS 20 NMR, nuclear magnetic resonance; SOFCs, solid oxide fuel cells; MAS, magic angle spinning; MQ, multi-quantum; 3Q, three-quantum; SOQE, second-order quadrupolar effect. REFERENCES (1) Skinner, S. J. Recent Advances in Perovskite-Type Materials for Solid Oxide Fuel Cell Cathodes. Int. J. Inorg. Mater. 2001, 3 (2), 113–121. https://doi.org/10.1016/S1466-6049(01)00004-6. (2) Ishihara, T.; Matsuda, H.; Takita, Y. Doped LaGaO3 Perovskite Type Oxide as a New Oxide Ionic Conductor. J. Am. Chem. Soc. 1994, 116 (9), 3801–3803. https://doi.org/10.1021/ja00088a016. (3) Nakayama, S.; T., K.; Sadaoka, Y. Ionic Conductivity of Lanthanoid Silicates. La10(SiO4)6O2 (Ln = La, Nd, Sm, Gd, Dy, Y, Ho, Er and Yb). J. Mater. Chem. 1995, 5 (5), 1801–1805. (4) Lacorre, P.; Goutenoire, F.; Bohnke, O.; Retoux, R.; Lallgant, Y. Designing Fast Oxide-Ion Conductors Based on La2Mo2O9. Nature 2000, 404 (6780), 856–858. https://doi.org/10.1038/35009069. (5) Li, M.; Pietrowski, M. J.; De Souza, R. A.; Zhang, H.; Reaney, I. M.; Cook, S. N.; Kilner, J. A.; Sinclair, D. C. A Family of Oxide Ion Conductors Based on the Ferroelectric Perovskite Na0.5Bi0.5TiO3. Nat. Mater. 2014, 13 (1), 31–35. https://doi.org/10.1038/nmat3782. 21 (6) García-González, E.; Parras, M.; González-Calbet, J. M. Electron Microscopy Study of a New Cation Deficient Perovskite-like Oxide: Ba3MoNbO8.5. Chem. Mater. 1998, 10 (6), 1576–1581. https://doi.org/10.1021/cm9708022. (7) Fop, S.; Skakle, J. M. S.; McLaughlin, A. C.; Connor, P. A.; Irvine, J. T. S.; Smith, R. I.; Wildman, E. J. Oxide Ion Conductivity in the Hexagonal Perovskite Derivative Ba3MoNbO8.5. J. Am. Chem. Soc. 2016, 138 (51), 16764–16769. https://doi.org/10.1021/jacs.6b10730. (8) Coduri, M.; Bernasconi, A.; Fischer, H. E.; Malavasi, L. The Ba3Mo1-xWxNbO8.5 ion Conductors: Insights into Local Coordination from X-Ray and Neutron Total Scattering. J. Mater. Chem. A 2020, 8 (40), 21227–21240. https://doi.org/10.1039/d0ta07073d. (9) Auckett, J. E.; Gutmann, M. J.; Evans, I. R. Understanding the Correlation between Oxide Ion Mobility and Site Distributions in Ba3NbWO8.5. Inorg. Chem. 2020, 59 (19), 14245–14250. https://doi.org/10.1021/acs.inorgchem.0c02035. (10) Fop, S.; Wildman, E. J.; Irvine, J. T. S.; Connor, P. A.; Skakle, J. M. S.; Ritter, C.; McLaughlin, A. C. Investigation of the Relationship between the Structure and Conductivity of the Novel Oxide Ionic Conductor Ba3MoNbO8.5. Chem. Mater. 2017, 29 (9), 4146–4152. https://doi.org/10.1021/acs.chemmater.7b01298. (11) Fop, S.; Wildman, E. J.; Skakle, J. M. S.; Ritter, C.; McLaughlin, A. C. Electrical and Structural Characterization of Ba3Mo1-xNb1+xO8.5-x/2: The Relationship between Mixed Coordination, Polyhedral Distortion and the Ionic Conductivity of Ba3MoNbO8.5. Inorg. Chem. 2017, 56 (17), 10505–10512. https://doi.org/10.1021/acs.inorgchem.7b01488. 22 (12) Yasui, Y.; Tsujiguchi, T.; Sakuda, Y.; Hester, J. R.; Yashima, M. Oxide-Ion Occupational Disorder, Diffusion Path, and Conductivity in Hexagonal Perovskite Derivatives Ba3WNbO8.5 and Ba3MoNbO8.5. J. Phys. Chem. C 2022, 126 (5), 2383– 2393. https://pubs.acs.org/doi/10.1021/acs.jpcc.1c09416... (13) Bernasconi, A.; Tealdi, C.; Malavasi, L. High-Temperature Structural Evolution in the Ba3Mo(1- x)WxNbO8.5 System and Correlation with Ionic Transport Properties. Inorg. Chem. 2018, 57 (11), 6746–6752. https://doi.org/10.1021/acs.inorgchem.8b01093. (14) Auckett, J. E.; Milton, K. L.; Evans, I. R. Cation Distributions and Anion Disorder in Ba3 NbMO8.5 (M = Mo, W) Materials: Implications for Oxide Ion Conductivity. Chem. Mater. 2019, 31 (5), 1715–1719. https://doi.org/10.1021/acs.chemmater.8b05179. (15) Yashima, M.; Tsujiguchi, T.; Fujii, K.; Niwa, E.; Nishioka, S.; Hester, J. R.; Maeda, K. Direct Evidence for Two-Dimensional Oxide-Ion Diffusion in the Hexagonal Perovskite-Related Oxide Ba3MoNbO8.5-δ. J. Mater. Chem. A 2019, 7 (23), 13910–13916. https://doi.org/10.1039/c9ta03588e. (16) Chambers, M. S.; Mccombie, K. S.; Auckett, J. E.; Mclaughlin, A. C.; Irvine, J. T. S.; Chater, P. A.; Evans, J. S. O.; Evans, I. R. Hexagonal Perovskite Related Oxide Ion Conductor Ba3NbMoO8.5: Phase Transition, Temperature Evolution of the Local Structure and Properties. J. Mater. Chem. A 2019, 7 (44), 25503–25510. https://doi.org/10.1039/c9ta08378b. 23 (17) Fop, S.; McCombie, K.; Smith, R. I.; McLaughlin, A. C. Enhanced Oxygen Ion Conductivity and Mechanistic Understanding in Ba3Nb1-xVxMoO8.5. Chem. Mater. 2020, 32 (11), 4724–4733. https://doi.org/10.1021/acs.chemmater.0c01322. (18) Fop, S.; McCombie, K. S.; Wildman, E. J.; Skakle, J. M. S.; Irvine, J. T. S.; Connor, P. A.; Savaniu, C.; Ritter, C.; Mclaughlin, A. C. High Oxide Ion and Proton Conductivity in a Disordered Hexagonal Perovskite. Nat. Mater. 2020, 19 (7), 752–757. https://doi.org/10.1038/s41563-020-0629-4. (19) Yashima, M.; Tsujiguchi, T.; Sakuda, Y.; Yasui, Y.; Zhou, Y.; Fujii, K.; Torii, S.; Kamiyama, T.; Skinner, S. J. High Oxide-Ion Conductivity through the Interstitial Oxygen Site in Ba7Nb4MoO20-Based Hexagonal Perovskite Related Oxides. Nat. Commun. 2021, 12 (1), 556. https://doi.org/10.1038/s41467-020-20859-w. (20) Murakami, T.; Shibata, T.; Yasui, Y.; Fujii, K.; Hester, J. R.; Yashima, M. High Oxide‐Ion Conductivity in a Hexagonal Perovskite‐Related Oxide Ba7Ta3.7Mo1.3O20.15 with Cation Site Preference and Interstitial Oxide Ions. Small 2021, 2106785, 2106785. https://doi.org/10.1002/smll.202106785. (21) Hanna, J. V.; Pike, K. J.; Charpentier, T.; Kemp, T. F.; Smith, M. E.; Lucier, B. E. G.; Schurko, R. W.; Cahill, L. S. A 93Nb Solid-State NMR and Density Functional Theory Study of Four- and Six-Coordinate Niobate Systems. Chem. - A Eur. J. 2010, 16 (10), 3222–3239. https://doi.org/10.1002/chem.200901581. 24 (22) Papulovskiy, E.; Shubin, A. A.; Terskikh, V. V.; Pickard, C. J.; Lapina, O. B. Theoretical and Experimental Insights into Applicability of Solid-State 93Nb NMR in Catalysis. Phys. Chem. Chem. Phys. 2013, 15 (14), 5115–5131. https://doi.org/10.1039/c3cp44016h. (23) Dunstan, M. T.; Blanc, F.; Avdeev, M.; McIntyre, G. J.; Grey, C. P.; Ling, C. D. Long-Range-Ordered Coexistence of 4-, 5-, and 6-Coordinate Niobium in the Mixed Ionic-Electronic Conductor γ-Ba4Nb2O9. Chem. Mater. 2013, 25 (15), 3154–3161. https://doi.org/10.1021/cm4014656. (24) Mastikhin, V. M.; Lapina, O. B.; Maximovskaya, R. I. 95Mo Solid-State NMR Spectra of Molybdates. Chem. Physic Lett. 1988, 148 (5), 413–416. (25) Forgeron, M. A. M.; Wasylishen, R. E. Molybdenum Magnetic Shielding and Quadrupolar Tensors for a Series of Molybdate Salts: A Solid-State 95Mo NMR Study. Phys. Chem. Chem. Phys. 2008, 10 (4), 574–581. https://doi.org/10.1039/b713276j. (26) De Lacaillerie, J. B. D. E.; Barberon, F.; Romanenko, K. V.; Lapina, O. B.; Pollès, L. Le; Gautier, R.; Gan, Z. 95Mo Magic Angle Spinning NMR at High Field: Improved Measurements and Structural Analysis of the Quadrupole Interaction in Monomolybdates and Isopolymolybdates. J. Phys. Chem. B 2005, 109 (29), 14033–14042. https://doi.org/10.1021/jp0519621. (27) Johnston, K. E.; Griffin, J. M.; Walton, R. I.; Dawson, D. M.; Lightfoot, P.; Ashbrook, S. E. 93Nb NMR and DFT Investigation of the Polymorphs of NaNbO3. Phys. Chem. Chem. Phys. 2011, 13 (16), 7565–7576. https://doi.org/10.1039/c1cp20258h. 25 (28) Kunwar, A. C.; Turner, G. L.; Oldfield, E. Solid-State Spin-Echo Fourier Transform NMR of 39K and 67Zn Salts at High Field. J. Magn. Reson. 1986, 69 (1), 124–127. https://doi.org/10.1016/0022-2364(86)90224-6. (29) Amoureux, J. P.; Fernandez, C.; Steuernagel, S. Z Filtering in MQMAS NMR. Journal of Magnetic Resonance - Series A. 1996, 123 (1), 116–118. https://doi.org/10.1006/jmra.1996.0221. (30) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calvé, S.; Alonso, B.; Durand, J. O.; Bujoli, B.; Gan, Z.; Hoatson, G. Modelling One- and Two-Dimensional Solid-State NMR Spectra. Magn. Reson. Chem. 2002, 40 (1), 70–76. https://doi.org/10.1002/mrc.984. (31) Iijima, T.; Kato, S.; Ikeda, R.; Ohki, S.; Kido, G.; Tansho, M.; Shimizu, T. Structure of Duplex Oxide Layer in Porous Alumina Studied by 27Al MAS and MQMAS NMR. Chem. Lett. 2005, 34 (9), 1286–1287. https://doi.org/10.1246/cl.2005.1286. (32) Engelhardt, G.; Kentgens, A. P. M.; Koller, H.; Samoson, A. Strategies for Extracting NMR Parameters from 23Na MAS, DOR and MQMAS Spectra. A Case Study for Na4P2O7. Solid State Nucl. Magn. Reson. 1999, 15 (3), 171–180. https://doi.org/10.1016/S0926-2040(99)00054-5. (33) Amoureux, J. P.; Fernandez, C. Triple, Quintuple and Higher Order Multiple Quantum MAS NMR of Quadrupolar Nuclei. Solid State Nucl. Magn. Reson. 1998, 10 (4), 211–223. https://doi.org/10.1016/S0926-2040(97)00027-1. 26 (34) Anupõld, T.; Reinhold, A.; Sarv, P.; Samoson, A. A Comparison of Double Rotation and Multi-Quantum Magic Angle Spinning Spectra. Solid State Nucl. Magn. Reson. 1998, 13 (1–2), 87–91. https://doi.org/10.1016/S0926-2040(98)00068-X. (35) Tansho, M.; Suehiro, T.; Shimizu, T. Constancy of the Quadrupolar Interaction Product in Nanocrystalline Gallium Nitride Revealed by 71Ga MAS NMR Shift Distribution. Solid State Nucl. Magn. Reson. 2019, 97, 25–30. https://doi.org/10.1016/j.ssnmr.2018.08.006. TOC Graphic