RELATED APPLICATION

This application claims the benefit of U.S. Provisional Application No. 61/203,073, filed on December 18, 2008. The entire teachings of the above application are incorporated herein by reference.

BACKGROUND OF THE INVENTION

A fuel cell is a device that generates electricity by a chemical reaction. Among various fuel cells, solid oxide fuel cells use a hard, ceramic compound metal (e.g., calcium or zirconium) oxide as an electrolyte. Typically, in solid oxide fuel cells, an oxygen gas, such as O ₂ , is reduced to oxygen ions (O ^2" ) at the cathode, and a fuel gas, such as H ₂ gas, is oxidized with the oxygen ions to form water at the anode. Fuel cells are generally designed as stacks, whereby subassemblies, each including a cathode, an anode and a solid electrolyte between the cathode and the anode, are assembled in series by locating an electrical interconnect between the cathode of one subassembly and the anode of another.

A major materials challenge in solid oxide fuel cell (SOFC) development is the interconnect material, which provides the conductive path for electrical current to pass between the electrodes and to the external circuit. The interconnect also acts as a separator plate, physically separating the fuel in the anode cavity from the air or oxygen in the cathode cavity. Therefore, the interconnect material must have good electrical conductivity to minimize resistive losses and must be stable in both oxidizing and reducing conditions. SOFCs operate at high temperatures, and therefore the coefficient of thermal expansion (CTE) of the interconnect material must be close to those of the other cell components in order to minimize thermal stresses. Other requirements of interconnect materials include adequate mechanical strength, relatively low permeability to oxygen and hydrogen, and reasonable thermal conductivity. Jeffrey W. Fergus, Lanthanum chromite-based materials for s olid oxide fuel cell interconnects, Solid State Ionics, 171 pp. 1-15 (2004); Zhenguo Yang, Recent advances in metallic interconnects for solid oxide fuel cells, International Materials Reviews, 53[1] pp. 39-54 (2008).

There are two approaches, in general, to developing interconnect materials for SOFCs: metallic and ceramic, each of which offers different advantages and challenges. Metallic interconnects have good electrical conductivity, but are not stable when exposed to the oxidizing conditions at the cathode of SOFCs, so that coating of the surface with a conductive oxide (such as spinels) is usually required. Ceramic interconnect materials are oxides, and are thus stable in oxidizing atmospheres, but typically have lower electrical conductivities as compared to metals. High cost of the raw materials and fabrication difficulties in obtaining high density are some other issues related to ceramic interconnects which have hindered their application in SOFCs.

Lanthanum-doped strontium titanates (LST) are promising interconnect materials with good chemical and dimensional stability in reducing environments, low oxygen ion mobility and acceptably low resistance when used as thin layers. They also exhibit good compatibility with both cathode and anode materials without formation of deleterious second phases. However, sinterability typically is a limitation when co-firing with yttria-stabilized zirconia (8mol% YSZ) electrolytes. Currently, LST materials generally need to be sintered at 1,460-1,500 ⁰ C to enable complete densification, while YSZ (such as, for example, TZ-8Y from Tosoh (Tosoh USA, Grove City, OH)) can be sintered to full density at 1,350 ⁰ C. Therefore, LST materials generally need to be modified either physically or chemically to improve their sinterability to enable co-firing with YSZ electrolyte.

SUMMARY OF THE INVENTION

This invention generally is directed to a method of forming an interconnect material, the interconnect material formed, and to a solid oxide fuel cell that includes the interconnect material of the invention. The invention is also directed to a strontium titanate varistor and a strontium titanate capacitor that includes an interconnect material of the invention. In one embodiment, a method of forming an interconnect material includes the steps of combining a lanthanum-doped strontium titanate with an aliovalent transition metal to form a precursor composition and sintering the precursor composition to form the interconnect material. In certain embodiments, the amount of aliovalent transition metal present in the precursor composition is sufficient to cause the precursor composition to sinter at a temperature in a range of between about 1,100 ⁰ C and about 1,500 ⁰ C. The amount of aliovalent transition metal present in the precursor composition can be in a range of between about 1 mol% and about 8 mol%. In some embodiments, the aliovalent transition metal can be an electron-acceptor dopant. In those specific embodiments, the aliovalent transition metal can be at least one member selected from the group consisting of manganese, nickel, cobalt and iron. In other embodiments, the aliovalent transition metal can be an electron-donor dopant. In those specific embodiments, the aliovalent transition metal can be at least one member selected from the group consisting of niobium and tungsten. In certain embodiments, the amount of aliovalent transition metal present in the precursor composition is sufficient to cause the precursor composition to be sintered to full density at a temperature in a range of between about 1,200 ⁰ C and about 1,35O ⁰ C. In another embodiment, the method of forming an interconnect material further includes the step of combining the precursor composition with at least one of a precursor cathode or anode layer of a solid oxide fuel cell prior to sintering the precursor composition.

In yet another embodiment, the invention is directed to an interconnect material comprising a lanthanum-doped strontium titanate that is further doped with an aliovalent transition metal. In still another embodiment, the invention is directed to a solid oxide fuel cell comprising a interconnect material, including a lanthanum-doped strontium titanate that is further doped with an aliovalent transition metal. In one embodiment, the interconnect material is adjacent an anode. Alternatively, the interconnect material can be adjacent a cathode. In certain embodiments, the solid oxide fuel cell further includes an interface layer between the interconnect material and either an anode or a cathode. In other embodiments, the invention is directed to a strontium titanate varistor or a strontium titanate capacitor comprising an interconnect material including a lanthanum-doped strontium titanate that is further doped with an aliovalent transition metal. This invention has many advantages, such as enabling co-firing of LST interconnects with YSZ electrolyte layers at a temperature in a range of between about 1,200 ⁰ C and about 1,35O ⁰ C in SOFC stack applications.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing will be apparent from the following more particular description of example embodiments of the invention, as illustrated in the accompanying drawings. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating embodiments of the present invention.

FIG. 1 is a graph of the median particle size (dso), specific surface area, and pH value (of attrition milling solution) of LST28 powder after attrition milling. FIG. 2 is a graph of the shrinkage rate of LST28 after attrition milling as measured by dilatometry using water for mixing with binders.

FIG. 3 is a graph of the peak temperatures and maximum shrinkage rates, (dL/dT) _ma χ, of the attrition milled LST28 powder shrinkage curves shown in FIG. 2.

FIG. 4 is a graph of the shrinkage rate of LST28 after attrition milling and Mn ₂ O ₃ doping as measured by dilatometry.

FIG. 5 is a graph of the shrinkage rate of Mn ₂ O ₃ -doped and as-received LST28 powder as measured by dilatometry.

FIG. 6 is a graph of the peak temperatures and maximum shrinkage rates, (dL/dT) _max , from Mn ₂ O ₃ -doped attrition-milled LST28 and as-received LST28 shrinkage curves shown in FIG. 4 and FIG. 5.

FIG. 7 is a graph of the shrinkage rate of Nb ₂ θ ₅ -doped and as-received LST28 as measured by dilatometry.

FIG. 8 is a graph of the peak temperatures and maximum shrinkage rates, (dL/dT) _max , from Mn ₂ O ₃ -doped LST28 and Nb ₂ Os-doped LST28 shrinkage curves shown in FIG. 1. FIG. 9 is a graph of the shrinkage rate of an A-site deficiency LST355 (LaojoSr _0.55 TiO ₃ ) as measured by dilatometry.

FIG. 10 is a graph of the chemical expansion of Mn ₂ O ₃ -doped LST28 measured by dilatometry. FIG. 1 1 is a graph of the chemical expansion of Nb ₂ O ₅ -doped LST28 and

LST355 measured by dilatometry.

FIG. 12 is a graph of the electrical conductivity of Mn ₂ O ₃ - and Nb ₂ θ ₅ -doped LST28 measured with a four-probe method in air. LST355 data are included in the figure for comparison. All samples were sintered at the same condition of 1,500 ⁰ C for 6 hrs.

FIG. 13 is a graph of the electrical conductivity of Mn ₂ O ₃ - and Nb ₂ O ₅ -doped LST28 measured with a four-probe method in forming gas (96%N ₂ +4%H ₂ ). LST355 data are included in the figure for comparison. All samples were sintered at the same condition of 1,500 ⁰ C for 6 hrs. FIGS. 14A-E are graphs illustrating XRD patterns of powder mixtures: (FIG.

14A) pure LST28 powder; (FIG. 14B) 2mol% Mn ₂ O ₃ + 98 mol% LST28 powder mixture; (FIG. 14C) 4mol% Mn ₂ O ₃ + 96 mol% LST28 powder mixture; (FIG. 14D) 2mol% Nb ₂ O ₅ + 98 mol% LST28 powder mixture; (FIG. 14E) 6mol% Nb ₂ O ₅ + 94 mol% LST28 powder mixture. FIGS. 15A-F are graphs illustrating XRD patterns of undoped and doped

LST28 samples (bar samples which were pressurelessly sintered at the same condition of 1,500 ⁰ C for 5 hrs.): (FIG. 15A) undoped LST28 (LSTMO.O); (FIG. 15B) 2mol% Mn ₂ O ₃ doped LST28 (LSTM2.0); (FIG. 15C) 4mol% Mn ₂ O ₃ doped LST28 (LSTM4.0); (FIG. 15D) 2mol% Nb ₂ O ₅ doped LST28 (LSTN2.0); (FIG. 15E) 6mol% Nb ₂ O ₅ doped LST28 (LSTN6.0). (FIG. 15F) 8mol% Nb ₂ O ₅ doped LST28 (LSTN8.0).

FIGS. 16A-B are SEM photographs of (FIG. 16A) LSTM4.0 surface and (FIG. 16B) LSTN8.0 surface after sintered at 1,500 ⁰ C for 6 hrs.

FIG. 17 is a schematic diagram of a fuel cell of the invention in a planar, stacked design.

FIG. 18 is a schematic diagram of a fuel cell of the invention in a tubular design. DETAILED DESCRIPTION OF THE INVENTION

This invention generally is directed to a method of reducing the sintering temperature and modifying the sintering behavior of a lanthanum-doped strontium titanate (LST) interconnect material. Methods employed by the method of the invention to reduce the sintering temperature of LST can include: (1) attrition milling (to reduce the particle size); (2) doping (to add one or two oxides to LST by a conventional ceramic process) such as (a) Mn ₂ O ₃ doping and (b) Nb ₂ O ₅ doping; and (3)using an A-site deficiency LST. An objective of the invention is to improve the sinterability of LST by Mn ₂ O ₃ - and Nb ₂ O ₅ -doping and/or milling — to reduce the T _peak > 100 ⁰ C so as to get dense and crack free interconnects, and evaluate the effects of Mn ₂ O ₃ - and Nb ₂ O ₅ -doping on the electrical conductivity, the coefficient of thermal expansion (CTE), and chemical expansion in both oxidizing and reducing atmospheres. A-site deficient La _{0 3} oSr _{o S5} TiO ₃ perovskite (LST355) was also evaluated as a candidate material for interconnects. Without wishing to be bound by any particular theory, it is proposed that the addition of aliovalent transition metals as dopants that replace the Ti within the LST perovskite can help create vacancies and improve sinterability. For this purpose both acceptor dopants such as Mn, Ni, Co or Fe can be added, or donor dopants such as Nb or W can be used. To enhance the effect on sintering, the dopants can be either incorporated into the LST powder or added to the LST powder prior to sintering. An aliovalent transition metal dopant is a transition metal dopant with a different charge state than the ions of the host crystal. The amount of aliovalent transition metal present in the precursor composition is sufficient to cause the precursor composition to sinter at a temperature in a range of between about 1,100 ⁰ C and 1,500 ⁰ C, more preferably in a range of between about 1100 ⁰ C and about

1 ,400 ⁰ C, and most preferably in a range of between about 1,100 ⁰ C and about 1,250 ⁰ C.

A fuel cell device with the doped LST materials can consist of LST material modified with aliovalent dopants, such as Mn, Ni, Co, Fe, Nb or W, as an interconnect layer or part of interconnect multilayer structure, where one layer is doped LST sandwiched between the cathode and anode layers with or without interface layers between the interconnect layer or layers and the cathode and anode. Four kinds of powders, La ₀₂ Sro sTiO _3- s (LST28) and Lao _3θ Sr _{0 55} TiO _3-S (LST355) from American Elements (Los Angeles, CA), Nb ₂ O ₅ , and Mn ₂ O ₃ from Alfa Aesar (Ward Hill, MA) have been selected as the starting materials. The latter two oxides can be employed as the additives/dopants. Attrition milling, for example, can be employed to refine LST28, Nb ₂ O ₅ , and Mn ₂ O ₃ . The median particle size (d ₅ o) Of Mn ₂ O ₃ can be reduced from as-received 4.0 μm to a d ₅₀ in a range of between about 0.2 μm and about 0.5 μm after attrition milling for a time period in a range of between about 18 hours and about 48 hours, and the specific surface area (SSA) of Mn ₂ O ₃ can be increased from as-received 1.04 m ² /g to a SSA in a range of between about 5 m ² /g and about 32 m ² /g at the same time. In a specific example, the d ₅ o of Mn ₂ O ₃ was reduced to 0.2 μm, and the SSA Of Mn ₂ O ₃ was increased to 31.59 m ² /g by attrition milling for about 24 hours. The dso and SSA OfNb ₂ O ₅ can be changed from as-received 0.47 μm and 5.66 m /g, respectively, to a d ₅ o in a range of between about 0.2 μm and about 0.5 μm and a SSA in a range of between about 5 m /g and about 32 m ² /g by attrition milling for a time period in a range of between about 18 hours and about 48 hours. In a specific example, the d ₅ o OfNb ₂ O ₅ was reduced to 0.33 μm, and the SSA OfNb ₂ O ₅ was increased to 1 1.88 m ² /g by attrition milling for about 24 hours. Both attrition-milled Mn ₂ O ₃ and Nb ₂ O ₅ can be employed as the additives in doping LST28. SEM and laser light scattering can be employed to characterize the as- received materials and those after attrition milling. The particle sizes can be obtained by laser scattering measurements with, for example, a Partica LA-950 laser from Horiba (Horiba Instruments, Inc., Irvine, CA). The LST28 and additive powders can be mixed by ball milling for 24 hrs with the additive up to 8 mol% using deionized (DI) water with 3 wt% of Darvan ^® C (R. T. Vanderbilt, Inc., Norwalk, CT) as the dispersant. A large ultrasonic probe can be employed for about 3 minutes at a power range of between about 40 W and about 65 W to improve the mixing of Mn ₂ O ₃ or Nb ₂ O ₅ and LST28 powders before ball milling. The sintering behaviors of the undoped and doped LST series can be studied by pressureless sintering and dilatometry. Densities of sintered LST samples can be measured by the Archimedes method well known in the art. The electrical conductivities of selected compositions of the doped LST series can be measured in the sintered bar samples by a four-probe technique at 800, 900, and 1,000 ⁰ C in both air and forming gas (96%N ₂ +4%H ₂ ). CTE and chemical expansion can also be measured with dilatometry using forming gas.

Exemplification

Reducing the sintering temperature of LST28

Attrition milling

Attrition milling of LST28 powder was performed from 1 hr to 24 hrs, and the results of powder characterization and dilatometry are shown in Figures 1-3. As expected, the median particle size (dso) decreased significantly from 0.5 μm to 0.20 μm, and SSA increased from 23 m ² /g to 37 m ² /g with attrition milling for 12 hrs. After attrition milling the LST powder up to 24 hrs, however, the median particle size (d ₅ o) was only reduced from 0.20 μm to 0.19 μm, and the SSA was increased to 41 m ² /g. Attrition milling more than 12 hrs under these conditions did not further refine the LST28 particles. Figure 1 showed that the pH value of LST28 solution increased from 8.54 to 1 1.88, indicating that some phases had been changed during attrition milling. Inductively coupled plasma (ICP) results showed that only a 61 μg/ml Sr concentration was detected in the water (the total water was about 1.5 liters) used for attrition milling and washing milling media, but almost no La or Ti were found. The total Sr leached out during milling was about 0.1 g, which is negligible compared to 90 g Sr out of the 200 g LST powder in the batch and may not be able to explain the pH value change. Furthermore, XRD was not able identify any change in the phase structure attrition-milled LST28 powder. Figure 2 shows the first derivative of the shrinkage curves as measured by dilatometry for the LST28 powders milled for various amounts of time. The peak temperatures, T _pea k, and maximum shrinkage rates, (dL/dT) _max , from Figure 2 are summarized in Figure 3. The shrinkage curves shifted to slightly lower temperatures with attrition milling, as shown in Figure 2. The T _peak of LST28 powder decreased steadily in the first 4 hrs attrition milling, from 1 ,510 ⁰ C to 1 ,454 ⁰ C, but did not decrease further by extending the milling time to 8 hrs or 12 hrs. The T _pea k of LST28 powder decreased from 1,510 ⁰ C to 1 ,410 ⁰ C after attrition milling for 24 hrs. Attrition milling 24 hrs also broadened the shrinkage curve of LST28 powder. The (dL/dT) _max decreased after 1 hr attrition milling, then increased steadily with further milling up to 12 hrs. It is reasonable that finer powder started to sinter earlier (lower T _peak ) and sintered faster (higher (dL/dT) _max ), but it would be a challenge to cast a LST28 powder with a high specific surface area (SSA=41 m ² /g). Based on the above attrition milling study, LST28 powder after 4 hrs milling (d ₅₀ = 0.34 μm, SSA= 28 m ² /g) was selected as the baseline for Mn ₂ O ₃ doping.

Mn ₂ O ₃ doped attrition milled and as-received LST28 Efforts in this subtask were devoted to making Mn ₂ O ₃ doped LST28 powders both from attrition milled for 4 hrs and as-received LST28, and to the dilatometry measurement on all the mixtures. The Mn ₂ O ₃ content was 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 8.0 mol% for both LST28 powders (AM and AR) and named as LSTMl .0(AM4hrs), LSTMLO(AR), ..., LSTM8.0(AM4hrs), and LSTM8.0(AR), respectively. Their dilatometry results are shown in Figures 4-6. The results have shown that Mn ₂ O ₃ doping erased the contribution of attrition milling. Both LSTM(AM4hrs) and LSTM(AR) materials with the same Mn ₂ O ₃ content have very similar free sintering behaviors: the T _peak of the shrinkage curve of LST28 could be reduced sharply over 300 ⁰ C after doping 3.0 mol% Mn ₂ O ₃ , from 1,510 ⁰ C to 1,206 ⁰ C and it almost reached a saturation point with further doping. The T _pea k was just reduced 20 ⁰ C more after adding 5.0 mol% more Mn ₂ O ₃ , from 1,206 ⁰ C to 1,186 ⁰ C; however, the maximum shrinkage rate, (dL/dT) _max , increased sharply after doping 3 mol% Mn ₂ O ₃ , from 0.10 to 0.20, which is similar to an A-site deficiency LST28 sintering behavior as discussed below, since doping Mn ₂ O ₃ would cause A-site deficiency in LST28 (Mn would diffuse into B-site in the Ti position during sintering.). Then the (dL/dT) _max decreased to 0.15 with addition Of Mn ₂ O ₃ up to 5 mol% because other peaks in the shrinkage curve became stronger and stronger. But it could not be reduced further with increasing dopant because this effect was compensated by more and more A-site deficiency at the same time; the peak at about 1 ,000 ⁰ C when doping more than 1.0 mol% should be related to the melting point of Mn ₂ O ₃ (1,080 ⁰ C). But the peak at about 1,300 ⁰ C when doping more than 3.0 mol% may be related to the structure change of LST28 because of Mn diffusion or Mn valence change.

Nb ₂ O ₅ doped as-received LST28 and LST355 The dopant selection was extended to Nb ₂ O ₅ in this subtask. Nb ₂ O ₅ has been selected because, like La _{O 4} Sr _{O 6} Ti ₀₄ Mn _{O 6} O ₃ , Nb doped SrTiO ₃ (like Sr _{0 9} Η _{0 8} Nbo ₂ O ₃ ) has lower sintering temperature and higher conductivity. T. Kolodiazhnyi and A. Petric, The Applicability of Sr-deflcient n-type SrTiOs for SOFC, Journal of Electroceramics, 15 pp. 5-11 (2005). Only as-received LST28 was used as the baseline for Nb ₂ O ₅ doping, because Mn ₂ O ₃ doping erased the contribution of attrition milling. Therefore, efforts in this subtask were devoted to mixing Nb ₂ O ₅ -doped LST28 powders and to the dilatometry measurement on the mixtures. The Nb ₂ O ₅ content was 1.0, 2.0, 3.0, 4.0, 5.0, 5.5, 6.0, and 8.0 mol% for doping as received LST28 powder and named as LSTNLO(AR), LSTN2.0(AR), - and LSTN8.0(AR), respectively. Their dilatometry results are shown in Figures 7 and 8. Figure 8 also shows the T _pea k and (dL/dT) _max data Of Mn ₂ O ₃ doped as received LST28 for comparison. In Mn ₂ O ₃ doped LST28, the T _peak of the shrinkage curve of LST28 could be reduced sharply over 300 ⁰ C after doping 3.0 mol% Mn ₂ O ₃ , from 1,510 ⁰ C to 1,206 ⁰ C and it almost reached a saturation point with further doping. The Tp _eak was just reduced 20 ⁰ C more after adding 5.0 mol% more Mn ₂ O ₃ , from 1,206 ⁰ C to 1,186 ⁰ C; however, the maximum shrinkage rate, (dL/dT) _max , increased sharply after doping 2 and 3 mol% Mn ₂ O ₃ , from 0.10 to 0.165 and 0.20, respectively, which were similar to an A-site deficiency La _{0 3} oSro ₅₅ Ti0 ₃ (LST355) sintering behavior as shown in Figure 9, since doping Mn ₂ O ₃ would cause A-site deficiency in LST (Mn would diffuse into B-site in the Ti position during sintering.). Then the (dL/dT) _max decreased to 0.15 if adding Mn ₂ O ₃ up to 5 mol% because other peaks in the shrinkage curve became stronger and stronger. But it could not be reduced further with increasing dopant because this effect was compensated by more and more A-site deficiency at the same time; in Nb ₂ O ₅ doped LST28, the free sintering behaviors were somehow different from the Mn ₂ O ₃ doped LST28. The Tp _eak OfNb ₂ O ₅ doped LST28 was gradually reduced from 1 ,510 ⁰ C of as-received LST28 to 1,328 ⁰ C with 5.5 mol% Nb ₂ O ₅ -doped LST28 (in fact, there were three peaks in this shrinkage curve), then sharply to 1,209 ⁰ C with 6.0 mol% Nb ₂ O ₅ doping. Further doping with 8.0 mol% Nb ₂ O ₅ only reduced the T _pcak to 1,201 ⁰ C but increased the (dL/dT) _max to a large number, 0.26, which is exactly like the sintering behavior of LST355 in Figure 9. The (dL/dT) _max of Nb ₂ O ₅ -doped LST28 is gradually reduced in the first 4 mol% Nb ₂ O ₅ doping and increased gradually with 5-6 mol% Nb ₂ O ₅ doping. But overall, it is much lower than that of the Mn ₂ O ₃ - doped LST28 except for the 8.0 mol% Nb ₂ O ₅ doping. It is in the range of 0.085 to 0.127. All of the shrinkage curves of Nb ₂ O ₅ -doped LST28 are relatively asymmetric. They are relatively broad between the onset and the T _pea k- A second peak and even a third peak have also been found in the curves from 1,000 ⁰ C to 1,300 ⁰ C, depending on the dopant content. It was assumed above that the peak at about 1 ,000 ⁰ C in Mn ₂ O ₃ doped LST should be related to the melting behavior Of Mn ₂ O ₃ (its melting point is 1,080 ⁰ C), and the peak at about 1,300 ⁰ C when doping more than 3.0 mol% might be related to the structure change of LST28 because of Mn diffusion or Mn valence change. There should be other reasons for the presence of the other peaks and the broadening in the shrinkage curves of Nb ₂ O ₅ -doped LST28, because the melting point OfNb ₂ O ₅ is as high as 1,520 ⁰ C, which will be studied by XRD below. The low (dL/dT) _max and broad shrinkage curves on the onset side of M^Os-doped LST28 are ideal for co-densification with the electrolyte in SOFC stacks.

Coefficient of thermal expansions (CTE) and chemical expansions of undoped and Mn ₂ O ₃ -, Nb ₂ O ₅ -doped LST28

Efforts in this subtask were devoted to preparing dense bar samples of LST355 (La _{0 3θ} Sro ₅₅ Ti0 ₃ ), Mn ₂ O ₃ -, Nb ₂ O ₅ -doped LST with selected compositions, and CTE and chemical expansion measurements. CTE and chemical expansion were measured by dilatometry under air, argon, and a forming gas (96%N ₂ +4%H ₂ ), which will generate different oxygen partial pressures: 0.21 atm in air, and 4.0 x 10 ^"19 atm in the forming gas. All the doped LST powders were die-pressed into 6mm x 6mm x 30mm bar samples in green form, and then sintered pressurelessly in air for one hour at different temperatures. The sintering temperature for each doped LST material was selected to be 50 ⁰ C higher than its T _peak from the dilatometry curve to ensure to get fully dense bar samples after sintering. The Archimedes method revealed that the relative densities of all the sintered bar samples are >99%. The CTE and chemical expansion of selected Mn ₂ O ₃ -, Nb ₂ O ₅ -doped LST samples are summarized in Table 1 below and shown in Figures 10 and 11. The CTE of LST28 is almost stable with Mn ₂ θ ₃ -doping up to 3 mol%, but decreases from 11.53 to 11.30 x 10 ^"6 /K with 4.0mol% Mn ₂ O ₃ doping. However, Nb ₂ O ₅ doping has a more significant effect on reducing the CTE of LST28 than Mn ₂ O ₃ doping. The CTE of LST28 gradually decreases with Nb ₂ O ₅ doping, from 11.53 to 1 1.0, then 10.6 x 10 ^"6 /K with 2 mol% and 6.0 mol% Nb ₂ O ₅ doping, respectively. Based on the thermal model, lowering the CTE of the ceramic interconnect (IC) will be favorable in reducing the thermal stress in the anode functional layer (AFL) which has been realized in Nb ₂ O ₅ doping. The thermal stress in AFL will cause AFL cracking, and then IC cracking. Table 1 also shows that the CTE at 1,200 ⁰ C is higher than at 900 ⁰ C for both undoped LST28 and doped LST28 based on the incomplete data.

Figure 10 and Figure 11 show the chemical expansion of Mn ₂ O ₃ -doped LST28, LST355 and Nb ₂ O ₅ -doped LST in air for one hour, in argon for one hour, and in forming gas for three hours at 900 ⁰ C. All these doped LST28 materials were very stable under the forming gas. The maximum expansion (ΔL/Lo) that has been found in doped LST28 is 0.005% for LSTM4.0 (4.0mol%Mn ₂ O ₃ doped LST28) under the forming gas for three hours, which was much lower than LST28 (-0.01%) and lanthanum chromites (LSC and LCC) interconnects (0.15-0.35% literature data). It can be concluded that Mn ₂ O ₃ -, Nb ₂ Os-doped LST28 materials are stable under forming gas for a limited time at 900 ⁰ C.

Table 1. CTE of undoped and Mn ₂ O ₃ -, Nb ₂ O ₅ -doped LST28

Electrical conductivities of undoped and Mn ₂ O ₃ -, Nb ₂ O ₅ -doped LST28

The electrical conductivities of undoped and doped LST28 samples should be compared with caution if the bar samples were sintered at different temperatures. The grain boundary in LST28 works as a high receptivity barrier so that LST28 samples with larger grain size when sintered at a higher temperature and a longer time will have higher values in conductivity. Also, the interstitial oxygen in LST28 might be easily removed under reducing conditions and therefore the reduction of Ti ⁴⁺ to Ti ³⁺ is favored. As a consequence, there is more Ti ³⁺ when tested under the forming gas, which explains the higher conductivities of all LST28 samples in forming gas than in air. There could also be more Ti ³⁺ in the samples when sintered at a higher temperature and longer time, which may explain why the conductivity of LST28 sample (sintered at 1,500 ⁰ C for 6 hrs in air and reduced in 96%N ₂ +4%H ₂ ) is much lower than the literature data (sintered at 1,600 ⁰ C for 48 hrs, then reduced in 5%H ₂ /Ar at 1,000 ⁰ C for 48 hrs, 30 S/cm at 900 ⁰ C in wet 5%H ₂ /Ar). See J. Canales-Vazquez, S. W. Tao, and J. T. S. Irvine, Electrical properties in La ₂ Sr ₄ TigOig^s: a potential anode for high temperature fuel cells, Solid State Ionics 159 pp. 159-165 (2003). Therefore, new bar samples of the undoped LST and doped LST were sintered at the same condition (1 ,500 ⁰ C for 6 hrs) and their conductivities were measured, which eliminated the effects of sintering temperature. Figures 12 and 13 are the electrical conductivities of Mn ₂ O ₃ -doped LST, LST355, Nb ₂ O ₅ -doped LST and LST 28 measured in air and in forming gas at 800, 900, and 1,000 ⁰ C, respectively. The relative densities of the undoped LST (LST28 and LST355) and doped

LST (LSTM2.0, LSTM4.0, LSTN2.0, and LSTN6.0) samples after sintered at 1,500 0C for 6 hrs are summarized in Table 2 below. They were at least over 95% of the theoretical density. Both LSTM4.0 and LSTN6.0 had lower relative densities because of their good sintering ability (lower T _pea k in the dilatometry curve), which meant that sintering at 1 ,500 ⁰ C for 6 hrs resulted in over-sintering, which yielded lower relative densities for these samples.

Table 2. Relative density of undoped and doped LST after sintered at 1,500 ⁰ C for 6 hrs

Figures 12 and 13 show the electrical conductivities of these samples in air and in forming gas, respectively, at 800, 900, and 1,000 ⁰ C. All the data were the average of two samples. Mn ₂ θ ₃ -doping reduced the conductivity of LST28 in both air and forming gas while Nb ₂ θ ₅ -doping had an opposite effect on the conductivity of LST28. LSTN6.0 had the highest conductivity while LSTM4.0 had the lowest conductivity in both air and forming gas. The conductivity increased with temperature for all samples in air and in forming gas except that the conductivity of LSTN6.0 in air decreased with temperature, which is the behavior of a metallic material. But in forming gas, the conductivity of LSTN 6.0 decreased slightly with temperature from 800 ⁰ C to 900 ⁰ C, then increased with temperature from 900 ⁰ C to 1 ,000 ⁰ C. The A-site deficiency material, LST355, had higher conductivity in forming gas, but lower in air than LST28. Usually, doping a acceptor like Mn ³⁺ in the B-site (Ti-site) will decrease the conductivity of SrTiO ₃ while doping a donor like Nb ⁵⁺ in the B-site (Ti-site) will increase the conductivity of SrTiO ₃ .

XRD study on undoped and Mn ₂ O ₃ -, Nb ₂ O ₅ -doped LST28

Efforts in this subtask were devoted to XRD study on the undoped and doped LST28 (Lao _.2 oSro _.8 oTi0 ₃ , the baseline material for the doping study discussed above) samples before and after sintering. As mentioned above, there were other peaks appearing in the dilatomery curves from 1 ,000 ⁰ C to 1 ,300 ⁰ C when both Mn ₂ O ₃ and Nb ₂ O ₅ doping were more than 3.0 mol%. The dilatometry curves were fairly broad between the onset and the T _peak in Nb ₂ O ₅ -doped LST28 samples. Therefore, room temperature XRD was conducted on the following selected samples before and after sintering to study the possible structure change during sintering: pure LST28 powder; 2mol% Mn ₂ O ₃ + 98 mol% LST28 powder mixture; 4mol% Mn ₂ O ₃ + 96 mol% LST28 powder mixture; 2mol% Nb ₂ O ₅ + 98 mol% LST28 powder mixture, and 6 mol% Nb ₂ O ₅ + 94 mol% LST28 powder mixture, and their sintered bar samples: LSTMO.O, LSTM2.0, LSTM4.0, LSTN2.0, LSTN6.0, and LSTN8.0. All these samples were sintered in the same condition of 1 ,500 ⁰ C for 5 hrs. The XRD results are shown in Figure 14 and Figure 15 for the powder samples and sintered bar samples, respectively. As shown in Figure 14, neither Mn ₂ O ₃ nor Nb ₂ O ₅ was detected when the additive was not more than 4 mol% in the powder mixture. Only LST28 (cubic phase) patterns were apparent in Figure 14(a), pure LST28 powder; Figure 14(b), 2 mol% Mn ₂ O ₃ + 98 mol% LST28 powder mixture; Figure 14(c), 4 mol% Mn ₂ O ₃ + 96 mol% LST28 powder mixture; and Figure 14(d), 2 mol% Nb ₂ O ₅ + 98 mol% LST28 powder mixture. But when the additive, Nb ₂ O ₅ , was increased to 6 mol%, Nb ₂ O ₅ (orthorhombic phase), then patterns could be found in the XRD profiles of the powder mixture along with LST28 patterns in Figure 14(e). After pressurelessly sintering at 1,500 ⁰ C for 5 hrs in air, LSTMO.O, LSTM2.0, LSTM4.0, and LSTN 2.0 remained single phase (LST28, cubic) patterns as shown by their XRD profiles in Figures 15(a), (b), (c) and (d), which were similar to their powder samples as shown in Figure 14. The XRD patterns OfNb ₂ O ₅ also disappeared in the sintered LSTN6.0 sample (Figure 15(e)), and even in the sintered LSTN8.0 sample (Figure 15(f)), which indicated that the solubility OfNb ₂ O ₅ in LST28 was at least 8 mol% when sintering at 1,500 ⁰ C. The SEM results shown in Figure 16 also indicated that there was no second phase in the surfaces of LSTM4.0 and LSTN8.0 samples after sintering at 1,500 ⁰ C for 6 hrs. However, there were other peaks appearing in the dilatomery curves from 1,000 ⁰ C to 1,300 ⁰ C when both Mn ₂ O ₃ and Nb ₂ O ₅ doping were more than 3.0 mol% and the dilatometry curves were fairly broad between the onset and the T _Pea i _< in Nb ₂ O ₅ -doped LST28 samples.

Conclusions

(1) The sintering temperature of LST28 could be significantly reduced by Mn ₂ O ₃ doping or Nb ₂ O ₅ doping. The T _peak could be reduced by over 300 ⁰ C by doping with both Mn ₂ O ₃ and Nb ₂ O ₅ .

(2) Mn ₂ O ₃ -doping had more effect on reducing T _peak and increasing (dL/dT) _max than Nb ₂ O ₅ -doping in the range of 1-5 mol% dopant contents; but at high dopant content, both dopants had a similar effect.

(3) Doping erased the contribution of attrition milling in reducing the sintering temperature of LST28.

(4) A-site deficiency LST355 had a lower sintering temperature. (5) There was no phase change in LST28 for both dopings with content up to 8 mol% after sintering.

(6) Mn ₂ O ₃ -, Nb2θ5-doped LST materials were stable under forming gas at 900 0C. (7) Nb ₂ θ ₅ -doping increased the conductivity of LST28 but Mn ₂ O ₃ -doping had the opposite effect. Nb ₂ O ₅ -doped LST28 had a higher conductivity than that of Mn ₂ O ₃ -doped LST28. (8) Three materials, LSTN6.0 (6.0 mol%Nb ₂ O ₅ doped LST28), LSTM2.0 (2.0 mol% Mn ₂ O ₃ doped LST28, and LST355, have been selected as the interconnect material.

The invention can be used as interconnects in fuel cells, SrTiO ₃ varistors, and SrTiO ₃ capacitors. A fuel cell of the invention can be a planar stacked fuel cell, as shown in FIG. 17. Alternatively, as shown in FIG. 18, a fuel cell of the invention can be a tubular fuel cell. Typically, in the planar design, as shown in FIG. 17, the components are assembled in flat stacks, with air and fuel flowing through channels built into the interconnect. Typically, in the tubular design, as shown in FIG. 18, the components are assembled in the form of a hollow tube, with the cell constructed in layers around a tubular cathode; air flows through the inside of the tube and fuel flows around the exterior.

The teachings of all patents, published applications and references cited herein are incorporated by reference in their entirety.

EQUIVALENTS

While this invention has been particularly shown and described with references to example embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.