Lesley Holmes

From Grey Lab Page

PROJECT: Solid State NMR (SSNMR) studies of dynamic processes in ionically conducting ceramic oxides.

These materials are used as gas separation membranes, oxygen sensors in catalytic convertors, but are primarily of interest as electrolytes in solid oxide fuel cell (SOFC) applications. For the materials currently in use, primarily yttria-stabilized zirconia, conduction is very limited except at high temperatures (>800°C). Due to the practical difficulties and costs associated with the use of materials that can operate at these temperatures, it is important to identify electrolytes with higher levels of anionic conduction at lower temperatures and to understand the structural factors that promote this high conductivity. While impedance studies have yielded the values for the long-range ionic motion in these materials, they cannot be used to determine the pathway(s) or the local dynamics responsible for the conduction, nor do they provide information concerning which type of oxygen site is responsible for the motion. Also, while insight into the possible mobile species can be extracted from diffraction studies, this technique does not provide a direct measure of the exchange rates between different oxygen sublattices. To provide such information, SSNMR methods have been successfully applied previously to a wide range of ionic conductors.^1-11 With localized information about the conduction pathways, better materials for ionically conducting applications can be identified. Due to the proven success of this technique at probing localized dynamics in these types of materials, we have undertaken to use it in the examination of several conducting oxides of interest.

The first material investigated was bismuth molybdate (Bi₂₆Mo₁₀O₆₉). This material is compositionally related to doped bismuth vanadates (so called BiMeVOx's), which have been found to demonstrate high levels of ionic conductivity. The structures of the BiMeVOx's consist of defect Aurivillius phases composed of perovskite-like sheets comprising corner-sharing VO_6-8^7-2δ polyhedra alternating with Bi₂O₂²⁺ layers. The motion predominantly occurs in the perovskite-derived sheets. The molybdate, however, is structurally very different, being composed of MoO₄^2- tetrahedra interspersed with [Bi₁₂O₁₄]<a>&#8734</a> columns. This structure has been found to be monoclinic at ambient temperature, with a transition to triclinic at elevated temperatures. While the conductivity in this material is significantly lower than that of the BiMeVOx's, the reasons for this were still unclear. Using solid state NMR techniques at a wide variety of temperatures, we were able to ascertain the nature of the conduction pathway in this material.⁶

The second group of materials investigated were barium indium oxide (Ba₂In₂O₅) and its lanthanum ((Ba_1-xLa_x)₂In₂O_5+x/2) and gallium (Ba₂(In_1-xGa_x)₂O₅) doped analogs, as they are materials which have been found to have conductivites significantly exceeding that of yttria-stabilized zirconia¹² at high temperatures.^5,12-20 The room temperature structure of Ba₂In₂O₅ is that of a highly oxygen-deficient perovskite; to charge balance, one sixth of all oxygens in the perovskite structure are removed, resulting in a structure laden with intrinsic vacancies, having three crystallographically distinct oxygen sites and orthorhombic symmetry of space group Ibm2.²¹ The structure contains alternating layers of tetrahedrally (T,T') and octahedrally (O) coordinated In³⁺ ions, with alternate tetrahedral layers offset from one another in a ...OTOT'... pattern. This is referred to as the Brownmillerite phase, named for the Ca₂FeAlO₅ mineral originally determined to have this structure.²¹ Oxygen vacancies in materials having the Brownmillerite structure order alternately in the (010) planes of the tetrahedral layers. Ba₂In₂O₅ undergoes a structural phase change to tetragonal above 925 <a>&#176</a>C and then to a disordered cubic phase above 1075 °C.²² While the large number of oxygen vacancies in the tetrahedral layers would appear to be conducive to good oxygen conduction, trapping of the vacancies in the room temperature orthorhombic phase allows for only limited oxygen motion. As the material enters the cubic phase at high temperature and vacancy distribution becomes disordered, conduction is increased by several orders of magnitude. This being true, however, calorimetry measurements by Prasanna and Navrotsky indicate that the experimentally observed disordering of the oxygen vacancies at the cubic transition is only 4.8% of the calculated potential configurational entropy. They interpret this to mean that conduction is still limited even in this temperature regime by extensive short range ordering.¹⁶ The current study hopes to more fully examine the implications of this localized ordering using solid state nuclear magnetic resonance (NMR) techniques. Previous high temperature ¹⁷O solid-state NMR measurements by Adler et al. show that an increasing number of oxygen ions are involved in the conductivity between 925 and 1075 °C, i.e., in the tetragonal phase, and that all oxygens become mobile above 1075 °C as the material enters the cubic phase.⁵ The current work re-examines the undoped material using solid state NMR spectroscopy at higher field strengths and using additional 2-dimensional experiments such as multiple-quantum magic angle spinning (MQMAS) as well as examining materials doped with lanthanum at the barium site and gallium at the indium site for this first time are examined by using both ¹⁷O and ⁷¹Ga solid state NMR techniques.

Finally, uptake of protons in humid atmospheres by vacancy-containing oxide ceramics has been documented since the 1950s when it was first reported in ZnO.^23,24 This was initially thought to be problematic for applications involving oxide ion conduction as the formation of hydroxide defects is a competitive process to that of oxide conduction (i.e. protons are dissolved into the matrix at the expense of oxygen vacancies).²⁵ In other materials, however, the hopping of protons unto itself has been determined to have many uses in low temperature applications requiring ionic conduction such as humidity and hydrogen sensors.²⁶ It was originally thought that proton conduction could not have uses in solid oxide fuel cell (SOFC) electrolyte materials because of the requirement for high operating temperatures. Materials such as hydroxides, hydrates and acid salts were found to have high protonic conductivity, but are unstable at high temperatures.^25,26 Because it has been found that some perovskite-based materials retain large quantities of water even in a moderately high temperature (500-800 <a>&#176</a>C) range, these materials would seem to make high temperature proton conduction for intermediate temperature solid oxide fuel cell (IT-SOFC) applications a possibility.^27-31 Barium Indium Oxide and its doped analogs have been found to readily take on water below 300 <a>&#176</a>C where they are then able to act as low temperature proton conducting materials. While this material is less than ideal for practical applications as continued hydration/dehydration has been found to lead to cracking, it is of great theoretical interest due to its unusual ordering as it incorporates hydrogen into the lattice. In the case of undoped samples, a hydrated tetragonal phase Ba₂In₂O₅.H₂O is formed.^{32 17}O spectra of the hydrated form Ba₂In₂O₅<a>&#183</a>H₂O contain a sharp resonance at 220 ppm displaying nutation and relaxation rates similar to liquid water, but with weak satellite transitions, indicating small residual quadrupolar interactions. The resonance is therefore attributed to water within the crystal lattice and not simply to surface water. Proton SSNMR data of this system have indicated the presence of two distinct proton sites, one with very strong hydrogen bonding, and one where the hydrogen bonding is quite weak. The implications of this are undergoing continued investigation.

References
(1) Chaudhuri, S.; Wang, F.; Grey, C. P. Journal of the American Chemical Society 2002, 124, 11746-11757.
(2) Kim, N. Doctor of Philosophy, State University of New York at Stony Brook, 2004.
(3) Kim, N.; Grey, C. P. Science 2002, 297, 1317-1320.
(4) Wang, F.; Grey, C. P. Chemistry of Materials 1997, 9.
(5) Adler, S. B.; Reimer, J. A.; Baltisberger, J.; Werner, U. Journal of the American Chemical Society 1994, 116, 675-681.
(6) Holmes, L.; Peng, L.; Heinmaa, I.; x; Dell, L. A.; Smith, M. E.; Vannier, R.-N.; Grey, C. P. Chem. Mater. 2008, 20, 3638-3648.
(7) Fu, R.; Ma, Z.; Zheng, J. P. J. Phys. Chem. B 2002, 106, 3592-3596.
(8) Maekawa, H.; Ukei, Y.; Morota, K.; Kashii, N.; Kawamura, J.; Yamamura, T. Solid State Communications 2004, 130, 73-77.
(9) Jayaraman, V.; Magrez, A.; Caldes, M.; Joubert, O.; Taulelle, F.; Rodriguez-Carvajal, J.; Piffard, Y.; Brohan, L. Solid State Ionics 2004, 170, 25-32.
(10) Tambelli, C. E.; Donoso, J. P.; Magon, C. J.; Ângelo, A. C. D.; Florentino, A. O.; Saeki, M. J. Solid State Ionics 2000, 136-137, 243-247.
(11) Maekawa, H.; Kashii, N.; Kawamura, J.-I.; Hinatsu, Y.; Yamamura, T. Solid State Ionics 1999, 122, 231-236.
(12) Goodenough, J. B.; Ruiz-Diaz, J. E.; Zhen, Y. S. Solid State Ionics 1990, 44, 21-31.
(13) Islam, M. S.; Davies, R. A.; Fisher, C. A. J.; Chadwick, A. V. Solid State Ionics 2001, 145, 333-338.
(14) Kharton, V. V.; Marques, F. M. B.; Atkinson, A. Solid State Ionics 2004, 174, 14.
(15) Kendall, K. R.; Navas, C.; Thomas, J. K.; Loye, H.-C. z. Solid State Ionics 1995, 82, 215-223.
(16) Prasanna, T. R. S.; Navrotsky, A. Journal of The Materials Research Society 1993, 8, 1484-1486.
(17) Mohn, C. E.; Allan, N. L.; Freeman, C. L.; Ravindran, P.; St<a>&#248</a>len, S. Journal of Solid State Chemistry 2005, 178, 346-355.
(18) Yoshinaga, M.; Fumoto, T.; Hashimoto, T. Journal of The Electrochemical Society 2005, 152, A1221-A1225.
(19) Hashimoto, T.; Ueda, Y.; Yoshinaga, M.; Komazaki, K.; Asaoka, K.; Wang, S. Journal of The Electrochemical Society 2002, 149, A1381-A1384.
(20) Rolle, A.; Vannier, R. N.; Giridharan, N. V.; Abraham, F. Solid State Ionics 2005, 176, 2095-2103.
(21) Colville, A. A.; Geller, S. Acta Crystallographica Section B 1971, 27, 2311-2315.
(22) Steele, B. C. H. In High Conductivity Solid Ion Conductors; Takaharhi, T., Ed.; World Scientific Publishing Co. Inc.: Teaneck, NJ, 1989.
(23) Mollwo, E. Zeitschrift fuer Physik 1954, 138, 478-88.
(24) Thomas, D. G.; Lander, J. J. Journal of Chemical Physics 1956, 25, 1136-42.
(25) Norby, T. In HYPOTHESIS II Symposium; Saetre, T. O., Ed.; Kluwer: Grimstad, Norway, 1998, p 603-608.
(26) Zhang, G. B.; Smyth, D. M. Solid State Ionics 1995, 82, 153-160.
(27) Sutija, D. P.; Norby, T.; BjÖrnbom, P. Solid State Ionics 1995, 77, 167-174.
(28) Azad, A. K.; Irvine, J. T. S. Solid State Ionics 2008, 179, 678-682.
(29) Norby, T.; Larring, Y. Current Opinion in Solid State & Materials Science 1997, 2, 593-599.
(30) Iwahara, H. Solid State Ionics 1996, 86-88, 9-15.
(31) Kreuer, K. D. Annual Review of Materials Research 2003, 33, 333-359.
(32) Fisher, C. A. J.; Islam, M. S. Solid State Ionics 1999, 1999, 355-363.

Structure of Bi₂₆Mo₁₀O₆₉ showing, in yellow, the preferential oxygen pathways proposed by a) Vannier et al.¹ and b) Galy et al.² in the ac and bc planes. Red spheres represent bismuth, green spheres represent oxygen and blue tetrahedra represent MoO₄^2- units. The O(19) interstitial sites proposed by Vannier are indicated. Coordinates were taken from ref. 1, bond lengths up to 2.36Å were used for the Bi-O connectivity (click to enlarge).

[1] Vannier, R.-N.; Danzé, S.; Nowogrocki, G.; Huvé, M.; Mairesse, G. Solid State Ionics 2000, 136-137, 51-59.

[2] Galy, J.; Salles, P.; Rozier, P.; Castro, A. Solid State Ionics 2006, 177, 6.