(305c) Synthesis of Perovskite Materials for Use as Sulfur Tolerant Anodes in SOFCs

As the largest user of energy within the DoD, the US Air Force consumed 2.8 billion gallons of fuel in FY04 in support of both domestic and deployed operations (1,2).  The transport and delivery of fuels into theater can also result in considerable risks to DoD personnel and places a significant burden on the logistical infrastructure.  While fuel is critical to these operations, pressure can be reduced through judicious energy conservation strategies.  One such strategy includes increasing the efficiency of power generation units in these forward locations.  Fuel cell technology has the potential to double current fuel efficiency.  However, current fuel cell technology is not compatible with military logistic fuels because of their complex chemistry and high sulfur content.  Furthermore, the power density of state of the art fuel cells systems are well below that of conventional military power generation approaches limiting their potential utility.  Ongoing efforts at the Air Force Research Laboratory are focused on development of high performance solid oxide fuel cells (SOFCs) to specifically address military requirements. 

The objective of this effort is to explore the use of novel materials to improve the sulfur tolerance of conventional SOFCs.  Military fuels can present a significant challenge in this regard because the sulfur concentration in fuels can be as high as 1 weight percent.  There are two main reasons for performance losses in SOFC due to sulfur poisoning.  The first reason is the physical absorption of hydrogen sulfide (H₂S) at surface active sites results in an effective loss in available reaction sites for H₂/CO oxidation.  A second deactivation pathway likely proceeds though adsorption of sulfur onto a metal surface site.  If the physical bound sulfur has sufficient time, sulfidation of the metal occurs.  These metal sulfides often will have a far lower melting temperature which factors into a breakdown and rearrangement of the microstructure.  Current anode materials of Ni-YSZ cermets are poisoned due to the reactions of Ni and H₂S (3).   Poisoning of Ni anodes has been shown to increase with decreasing temperatures and eventually become irreversible at low temperatures due to the stable adsorption state of sulfur onto the Ni surface.

Perovskite materials are of interest as a potential material capable of withstanding sulfur and carbon depositions.  The perovskite structure is stable with hydrogen sulfide when compared to Ni/YSZ anodes.  A perovskite is a material consisting of the perovskite structure ABO₃.  ABO₃ type perovskites have demonstrated superior sulfur tolerance but lack the high conductivity and catalytic activity of Ni/YSZ cermets. (4)  In order to influence the electrical or ionic conductivity, perovskite lattices can be modified to induce either n or p type conductivity.  With a perovskite structure such as A_1-a A'_aBO₃, oxygen vacancy can come from the substitution of A³⁺ for A²⁺.  Oxygen vacancies can also come from the thermal decomposition or partial reduction of B³⁺ to B²⁺ or B⁴⁺ to B³⁺.  The basis of oxide ion conductivity is having vacancies allowing for oxygen mobility.  There are two components required to calculate oxygen vacancies (x=a/2+d).  The first is defined by the A content by substitution of A³⁺ for A²⁺ and is represented in the equation by a/2.  The second component is d, which is dependent on the partial pressure of oxygen and changes with the reduction/oxidation state of the B site.  By increasing the oxygen vacancies the chemical stability in a reducing atmosphere will decrease due to the oxygen having increased mobility.  The electronic conductivity is dependent on the oxygen partial pressure, causing the B site substitution to have a larger impact on the electronic conductivity as compared to the A site substitution.  The ionic conductivity tends to show the A site substitution playing a larger role due to the a/2 term being larger than d (5).

Experimental

Single and double doped perovskites were synthesized using solid state reactions and characterized using X-Ray diffraction to ensure consistency with the materials crystal structure.  The particle size was examined for each processing step to examine how it would affect obtaining a pure phase.  The conductivity of each material was tested on discs measuring 25.4mm in diameter and 1.0mm in thickness using electrochemical impedance spectroscopy (EIS). The sulfur tolerance of the materials was tested by creating an ink for the anode materials and pasting the anode onto a La_0.8Sr_0.2Ga_0.83Mg_0.17O_2.815 (LSGM) electrolyte disc.  The cells created were exposed to 140 ppm of H₂S at 800°C and the voltage was recorded to show any degradation in performance as shown below in Figure 1. 

Results

SrTiO₃ has been used due to its chemical stability and has been shown to be a mixed ionic and electronic conductor.  SrTiO₃ has demonstrated n-type semi-conducting behavior when it is doped with a donor such as Y³⁺ or La³⁺ (6).  In our study, SrTiO₃ was synthesized using solid state reactions and doped at the A-site with 0.04 mol % Y,0.08 mol % Y,  0.4 mol % La, 0.3 mol % La and 0.2 mol % La.  The tolerance factors were calculated for each composition to examine the relationship between conductivity and tolerance factors.  The following plot shows the various doped single perovskite and double perovskites.  Double perovskites showed a marked improvement over single perovskites in conductivity.

As the dopant level of lanthanum and yttrium was increased for the strontium titanate cells the conductivity was increased while the tolerance factor was decreased.  By decreasing the tolerance factor the oxygen mobility should increase leading to higher conductivity.  The lanthanum doped strontium titanate had the best conductivity performance of the single perovskites.  The double perovskite materials show Sr₂MgMoO₆ (SMMO) with a higher conductivity then Ba₂MgMoO₆.  SMMO also has the best performance out of all the perovskite materials tested.  

SMMO is a mixed oxide-ion conductor that is stable in the anodic atmosphere.  The main reasons for investigating the double perovskite as anode material are the structure is beneficial because it is oxygen deficient and it will give good oxide-ion conduction (7).  SMMO has a mixed valent cation (Mo) from the 5d block, which will provide good electronic conduction even if the ions occupy only one subarray of the double perovskite structure.  To balance the charge, a partner cation Mg (II) is suitable for use with Mo (VI).  With a sixfold coordinate Mo (VI) is able to accept an electron while losing an oxide ligand.  With a double perovskite structure it is important for the two octahedral site cation Mg (II) to be stable when it is in less than a sixfold coordinate system. This allows for the perovskite structure to maintain stability with the partial removal of oxygen (7).  Ba₂MgMoO₆ was also studied to explore the impact of A site dopants on conductivity and effect from changing the tolerance factor.

As observed previously, Ni/YSZ demonstrates degradation in performance when sulfur is added but can be partially regenerated when pure hydrogen is supplied again.  The reversible degradation after cycling was attributed to the physical absorption of hydrogen sulfide (H₂S) on surface active sites effectively reducing the surface area and decreasing oxidation rates.  For the SMMO cell with an LDC buffer layer there is no sulfidation or physical absorption of sulfur.  XRD work is done to verify if sulfur is in the anode materials and to see if any changes in compositions occur.

This technical presentation will describe efforts thus far in the preparations of single and double perovskites prepared through solid state reactions.  Characterization of these materials and their performance as SOFC anodes will be presented. References

1)      DESC FY04 Fact Book. available on-line.

2)      DOE Annual Energy Review. available online, August 2005.

3)      Gong, M, X Liu, J Trembly, and C. Johnson. "Sulfur-tolerant anode materials for solid oxide fuel cell application." Journal of Power Sources, 2007: 289-298.

4)   Sun, C, and U Stimming. "Recent anode advances in solid oxide fuel cells." Journal of Power Sources, 2007: 247-260.

5)   H. Ullman, N. Trofimenko, F. Tietz, D. Stover, A. Ahmad-Khanlou, ?Correlation between thermal expansion and oxide ion transport in mixed conducting perovskite-type oxides for SOFC cathodes? Solid State Ionics 138 (2000) 79-90

6)   O.A. Marina, L.R Pederson,  Fifth European Solid Oxide Fuel Cell Forum, European Fuel Cell Forum, Oberrohfdorf, Switzerland, (2002) 481-489.

7)      Huang, Y.H, R Dass, J Denyszyn, and J Goodenough. "Synthesis and Characterization of Sr₂MgMoO₆." (Journal of the Electrochemical Society) 153 (2006).