Material optimization via combinatorial deposition and analysis for thermoelectric thin films

This work presents a custom, high-throughput combinatorial approach for the optimization of thermoelectric thin films consisting of materials with complex chemistry and structures (e.g., the layered misfit cobaltite, Ca 3 Fe x Co 4−x O 9 ). Combinatorial thin films with graded compositions are produced on 100 mm Si wafers from multiple target materials using pulsed laser deposition. Film thickness and composition are mapped as a function of wafer location. Crystal structures are determined using x–y mapping XRD analysis with specially designed algorithms for automated peak location and analysis. Thermoelectric properties, specifically the Seebeck coefficient and the electrical resistivity, are screened using a custom designed automated probe system. By combining the rapid synthesis of many compositions and structures simultaneously using combinatorial deposition and automated analytical tools capable of spatial mapping, trends in material performance are shown to be quickly obtained primarily due to the elimination of one-at-a-time synthesis and analysis. The possible approaches for such complex multivalent combinatorial optimization of thin films are identified and discussed. For the Ca 3 Fe x Co 4−x O 9 system presented, variations to the thermoelectric power factor are dominated by changes in the electrical resistivity. Enhancements to the Seebeck coefficient are observed due to the incorporation of Fe into the Ca 3 Fe x Co 4−x O 9 structure; however, this improvement is overshadowed by increases in the electrical resistivity due to variations in film thickness and the presence of secondary phases (Co 3 O 4 and Ca 2 Fe 2 O 5 ) which result from increasing Fe content and off-axis pulsed laser deposition.