Advanced Interconnect Development at PNNL

Advanced Interconnect and Interconnect/Electrode Interfaces Development at PNNL Z.G. Yang, G.G. Xia, G.D. Maupin, Z.M. Nie, X.S. Li, J. Templeton, J.W. Stevenson, P. Singh Pacific Northwest National Laboratory Richland, WA 99352 8th Annual SECA Workshop and Peer Review San Antonio, TX, August 6-9, 2007 Objectives and Approach Objectives Develop cost-effective, optimized materials and fabrication approaches for SOFC interconnect and interconnect/electrode interface (i.e. contacts) applications Identify and understand degradation processes in interconnects and interconnect/electrode interfaces Approach Materials and process development Cost-effective oxidation resistant alloys Surface modification via coatings Interconnect/electrode contact materials Materials evaluation and degradation study Screening study of alloys and ceramics for interconnect and interface applications, respectively Investigation and understanding of oxidation/corrosion and interfacial reactions and stability under SOFC operating conditions. 2 Accomplishments in FY07 Investigation and development of cost-effective ferritic stainless steels (In collaboration with Allegheny Ludlum Corp.) Systematically investigated 430 Identified and evaluated 439 and 441, two modified versions of 430 Applied protection layers onto candidate alloys and evaluated their performance Development of protection layers and fabrication approaches Completed long-term thermal stability and electrical performance evaluation Initiated optimization of materials and fabrication for further cost-reduction Investigation and development of contact layers between metallic interconnects and electrodes Screening-studied more than a dozen materials systems via different fabrication approaches Identified two promising material groups and three approaches Evaluated electrical performance of selected candidates 3 S3 Investigation and Development of Novel Interconnect Alloys Goal: Identify/develop a novel ferritic stainless steel (FSS) with an optimized alloy chemistry that offers comparable or improved performance relative to the state-of-the-art compositions such as Crofer 22 APU, while being more cost-effective. Approach: To achieve the desired alloy chemistry or control residual alloy elements of Si, C, N, etc., via alloying, instead of extra refining that adds cost. Accomplishments Investigated properties of 430 relevant to interconnect applications Identified potential candidates 441 and 439, two modified versions of 430 Evaluated their properties relative to interconnect requirements Surface-modified the potential candidates with spinel protection layers and investigated their stability and electrical performance 4 Slide 4 S3 Staff, 7/30/2007 Oxidation Kinetics of Bare and Coated 430 Why 430: cost reduction 430 430: 17% Cr, via conventional melting – more cost-effective Crofer 22 APU: 23%Cr, extra refining (e.g. vacuum refining) for cleaning residual elements, Si, C, N, etc. Bare 430 demonstrated a fairly low scale growth rate at early stages Leveling off of the weight gain indicated likely spallation Mn1.5Co1.5O4 (MC) spinel protection layers drastically mitigated the scale growth beneath the coating δω , (g/cm ) 2 2 Time (hours) 0 2.0E-7 300 600 900 1200 1.6E-7 Bare 430 1.2E-7 δω = 6.0x10 t 2 R = 0.9995 8.0E-8 2 -14 800oC, air 2 4.0E-8 δω = 5.0x10 t 2 R = 0.9928 2 -15 MC coated 430 0.0E+0 0.0E+0 1.0E+6 2.0E+6 3.0E+6 4.0E+6 5.0E+6 Time (seconds) 5 Surface Stability of 430 Unlike bare 430, no spallation observed on MC 430 Fe transported through the coating, BUT not Cr No solubility of SiO2 in Cr2O3 Formation of continuous, insulating SiO2 layer b/w scale and Fe-Cr substrate (Mn,Co,Fe)3O4 Bare 430 1,200 hrs, air, 800oC MC coated 430 6 Long-Term Performance of MC Coated430 The formation of a continuous insulating SiO2 layer at the scale/metal interface led to a high ASR. The ASR became unstable after about 4,000 hours, likely due to detachment of scale from the metal substrate. 400 MC protection layer||scale||Fe-Cr 300 2 Area specific resistance (ASR) measurement: IC//contact//cathode/LSM pellete/cathode//contact//IC o 800 C, air Cr Mn Co Si Fe ASR, mhm.cm 200 Cathode: La0.8Sr0.2MnO3 IC: Mn1.5Co1.5O4 coated 430 Contact: La0.8Sr0.2Co0.5Mn0.5O4 100 430 20μm 0 0 1000 2000 Time, h 7 3000 4000 5000 Metallurgy of 441 and 439 Designation T-441 439 HP AL 430 Crofer 22 APU Cr 17.8 17.5 17.0 23.0 Mn 0.33 0.44 ≤1.0 0.4-0.8 Ni 0.20 0.20 ≤0.75 C 0.010 0.012 ≤0.12 0.030 ≤0.50 Al 0.045 0.040 Si 0.47 0.73 ≤1.0 ≤.50 0.020 0.050 ≤0.2 0.04-0.20 P 0.024 0.016 S 0.001 0.0004 Ti 0.18 0.41 Nb 0.46 Re Fractional % of Ti and Ti/Nb were added into Fe-17%Cr substrate for 439 and 441, respectively Nb leads to laves phase (Fe2Nb) precipitation along grain boundaries that significantly improves high temperature strength and creep resistance of the Fe-Cr substrate (double yield strength at 800ºC) As strong carbide/nitride formation elements, Ti and Nd lower interstitial elements C and N in the substrate Can Nb (or Ti) tie up Si to prevent SiO2 layer? 441 d 8 Electrical Evaluation of 441 and 439 500 mA.cm2 FSS IC 45 40 35 ASR (mOhm-cm2) 30 25 20 15 10 5 0 0 La0.8Sr0.2MnO3 cathode and contact, air, Chipping 800oC LSM contact LSM cathode LSM MC layer FSS IC 500 mA.cm2 439 Crofer 22 APU (LSCM contact) 439 HP Scale 441 100 200 300 400 500 Time (hours) 9 Effects of Minor Alloying Elements in 441 There was Si buildup or silica layer formation between scale/metal interface, in spite of about 0.5% residual Si in the metal substrate Nb tied up Si, preventing formation SiO2 layer at the scale/metal interface Atomic%: 21.9Nb, 8.4Si, 2.8Ti, 10.0Cr, 56.9Fe Nb, Si rich 10 Scale Structure and Compositions of 441 Scale grown on 441 is mainly comprised of (Mn,Cr)3O4 and Cr2O3, similar to that of Crofer 22 APU Negligible Fe or iron oxides in the scale, different from that of 430 S M Crofer 22 APU S C C S C S C M: Fe-Cr substrtae C: Cr2O3 S: (Mn,Cr)3O4 spinel 300 hours, 800oC, air S C S S C M M C S C T-441 C S S C S M C C S S C S C 28 32 36 40 44 48 2θ 52 56 60 64 68 11 Kinetics of Scale Growth on 441 Time (hours) 0 2.0E-7 200 400 600 800 1000 1.6E-7 MC 441 Bare 441 Modified MC 441 Linear (Bare 441) Linear (MC 441) 800oC, air Scale growth rate comparable to Crofer 22 APU, but with inferior scale adherence (Local spallation found occasionally after extensive oxidation) δw2 = 5.0x10-14t 2 R = 0.9941 δw , g cm 1.2E-7 2~3 times lower for MC coated specimens; no spallation 2. -4 Ba re 44 1 2 8.0E-8 δw2 = 2.0x10-14t 2 R = 0.9726 4.0E-8 Mn 1 O Co 1.5 .5 coa 4 ted 0.0E+0 0.0E+0 8.0E+5 1.6E+6 2.4E+6 3.2E+6 4.0E+6 Time (seconds) After 900 hours 12 Electrical Performance of Surface-Modified 441 Mn1.5Co1.5O4 spinel protection layers minimized area specific electrical resistance (ASR) ASR of coated sample increased little, if any, over the course of the test 25 Cathode: LSM Bare T-441 20 Contact: LSM IC: bare 441 or MC coated 441 800oC, air ASR (mOhm.cm ) 2 15 Current: 500 mA.cm-2 10 T-441 with Mn1.5Co1.5O4 protection l 5 0 0 200 400 600 800 1000 Time (hours) 13 SEM Cross-Sections of ASR Samples Improved surface stability: no spallation or detachment observed No penetration of Cr through the protection layer, though there appeared Fe migration into the coating (similar to 430). MC coated 441 BS image 14 Summary 441 exhibited promising alloy chemistry: addition of a small amount of Nb helps avoid formation of a continuous silica layer and promote desirable second phase precipitation, thus leading to a lower scale resistance and higher mechanical strength. The alloying approach eliminates the costly refining process that is currently employed for making Crofer 22 APU and other super-grade ferritic stainless steels for IC applications . Protection layers are required to further improve alloy surface stability and electrical performance, and seal off Cr. Future Work Evaluate long-term thermal stability and electrical performance of bare and surface modified 441 Further understand the alloy chemistry via advanced diagnostic study Investigate and optimize bulk alloy chemistry and surface modification for satisfactory long-term stability and performance. (In collaboration with Allegheny Ludlum Corp.) 15 Protection Layer Development and Investigation Goal: develop cost-effective, optimized protection layers that are effective barriers to both oxygen inward and chromium outward diffusion, while being stable over lifetime of SOFC operation. Previous work: work Developed spinel protection layers with a nominal composition Mn1.5Co1.5O4 Systematically studied (Mn,Co)3O4 spinel materials Developed slurry-based approaches for fabrication of the spinel protection layers Evaluated kinetics of scale growth, stability under thermal cycling, electrical and electrochemical performance, chromia volatility, etc., of coated Crofer 22 APU Completed one year thermal stability evaluation of coated Crofer 22 APU Recent Accomplishments: Accomplishments Completed half year electrical evaluation of MC coated Crofer 22 APU and 430 (with LSM cathode & contact paste) Investigated suitability and performance of the spinel protection layers on 430 and 441 (see previous slides) Started developing alternative fabrication approaches, e.g. electrochemical deposition 16 Summary and Future Work Spinel protection layers with a nominal composition Mn1.5Co1.5O4 and fabricated with slurry coating approaches are an effective oxygen inward and chromium outward diffusion barrier, mitigating scale growth and sealing off chromium Interconnect FSS, e.g. Crofer 22 APU, with the spinel protection layers demonstrated excellent long-term stability and electrical performance Developing cost-effective approaches compatible with mass production and practical shapes of interconnect Electroplating, electrophoresis, etc., in addition to spray process. 17 Contact Layer Investigation and Development Goal: develop cost-effective, optimized contact layers between metallic interconnects and electrodes. Interconnect Protective layer Contact layer Seals Cathode Functions Promote electrical contact Facilitate stack assembling Act as a potential buffer zone to prevent unwanted reactions and transport, such as Cr volatility Challenges Electrolyte A metallurgical bond can be built between a metallic interconnect and Ni-YSZ anode, providing a low resistance path for electrons. Oxide-metal interfaces are present between metallic interconnects and cathodes, increasing electrical resistance and thus causing power loss. 18 Previous Work and Current Strategy Spinel protection layer Previous work and accomplishments: accomplishments Evaluated metals and varied conductive oxides, including LSM, LSCo, LSCM, MC, etc. Investigated interfacial interactions Initiated enhanced sintering approaches for LSM and MC LSM contact 441 Challenges of current materials Precious metals demonstrate suitable properties, but too expensive (Ag, a possible exception). Conductive oxides of high sintering activity, e.g. superconductors, usually too reactive, negatively affecting the stack and interface stability Conductive oxides, e.g. LSM, that are typically used as cathode compositions demonstrated good compatibility, but need improvement in sintering activity at 800-900oC and thus better electrical contact Approach: improve sintering activity via reaction sintering, addition of sintering agents Approach or chemical modification 19 Fabrication of Contact Layers via Reaction Sintering Paste of metals and/or oxides mixture During first stack heating or sealing Conductive oxide contact layer Reactions assisted sintering Conductive oxides 5 μm Without sintering After reaction sintering 20 Contac Layers via Adding Sintering Agents Among studied, CuO and Bi2O3 more effective for LSM To be effective needs 4~5% 5 950oC 0 -5 Shrinkage (%) -10 LSM-6 (LSM only) LSM-5 (5%Sn) LSM-4 (2%Sn) LSM-3 (2%Bi2O3) LSM-2 (2%CuO) LSM-1 (5% CuO, attritioned) LSM-1 Spex (5%CuO, Spexed) -15 -20 -25 0 200 400 600 800 o 1000 1200 1400 Temperature ( C) 21 Electrical Performance and Stability Evaluation 500 mA.cm2 FSS IC Interfacial resistance evaluation unit (IRU) 25.0 Tempearture 800 20.0 700 LSM+1%CuO 15.0 500 LSM+2%CuO 10.0 300 LSM+4%Bi2O3 200 100 0.0 0 100 200 300 400 Time (hours) 0 500 400 600 Tempearture ( C) o contact cathode 900 LSM MC layer FSS IC 500 mA.cm2 ASR, mOhm.cm 2 5.0 22 Enhanced Sintering via Chemical Modifications 5 LSM 0 -5 % Shrinkage -10 Modified LSM -15 -20 -25 0 200 400 600 800 1000 1200 Temperature, °C 23 Summary Reaction sintering appears to be a promising approach to fabricate contact layers between perovskite cathodes and metallic interconnects. Addition of sintering aids and chemistry modifications also help improve sintering activity of conductive oxides. Future work Continue to search and optimize contact materials and processing approaches Systematically evaluate candidate systems: dilatometry, IRU (ASR), SEM, XRD. Evaluate long term electrical performance and interface stability under isothermal and thermal cycling 24 Acknowledgements The work summarized in this paper was funded under the U.S. Department of Energy’s Solid-State Energy Conversion Alliance (SECA) Core Technology Program. The authors wish to thank the SECA management team at NETL for their helpful discussions regarding this work. Metallographic preparation and SEM: Jim Coleman, Shelley Carlson, Nat Saenz 25