Combinatorial Measurement Methods for Inorganic Materials

Our goal is to develop novel, combinatorially-compatible measurement methods and metrologies, as well as comprehensive and consistent data sets, that will enable manufacturers of devices based on inorganic functional materials to select new materials more rapidly and intelligently. For example, the silicon microelectronics industry is currently materials limited; the traditional transistor and capacitor formation materials — silicon, silicon dioxide, and polysilicon — have been pushed to their undamental material limits, and continued scaling will require the introduction of novel materials.

Our approach is to develop a novel metrology for interrogating the property of interest in the combinatorial library film. A good example of the application of this approach is the advanced gate stack of the integrated circuit MOSFET device. The traditional gate stack layers (SiO2 gate dielectric and polycrystalline Si gate electrode) in current Si microelectronic devices must be replaced with a high dielectric constant (high-.) gate dielectric, and a metal gate electrode. We will develop combinatorial methodologies to: (1) fabricate compositionally graded thin film libraries of novel gate metal electrode/high-. gate dielectric/ substrate combinations (“gate stack” structures); and (2) measure the key electronic properties (e.g., work function) and thermal stability of such libraries. Comprehensive data sets of electronic properties as a function of composition will be generated for materials systems identified as high priority by the microelectronics industry. In addition, first-principles modeling techniques will be used to predict the electronic properties of such systems.

Impact and Customers:

• The silicon (Si) microelectronics industry and the consumer electronics and information revolution that it fuels is, at $750 billion, one of the largest sectors of the U.S. (and global) economy. The development of materials with superior properties that will enable device scaling (dimensional shrinkage of integrated circuit device elements according to Moore’s Law) and enhanced device performance is critical to continued innovation in Si microelectronics.

• Currently, there are no rapid measurement techniques to determine the critical properties of novel functional materials, e.g., the work functions of metal gate electrodes for metal-oxide-semiconductor-field-effect-transistor (MOSFET) devices, or the Seebeck coefficients of thermoelectric materials for vehicular waste heat recovery devices. Development of such methods would enable more rapid selection, optimization, and commercialization of systems incorporating these devices.
  

• Key customers include Sematech and major U.S. semiconductor manufacturers such as Intel and Micron, as well as GM and Honda for the use of thermoelectrics for vehicular waste heat recovery applications. We are very active in promoting the use of combinatorial methodologies with our customers.

We commissioned a state-of-the-art combinatorial tool capable of producing thin film libraries by reactive sputtering or pulsed laser deposition (PLD). Both chambers are equipped with multiple targets, allowing for the deposition of ternary films of metals and nitrides (by sputtering) and oxides (by PLD) with monolayer (0.5 nm) thickness control. At present, this tool produces library films for our thermoelectric materials and advanced MOS gate stack materials projects.

Another driver for our combinatorial project is the MOS gate stack (gate dielectric and gate metal electrode), which must be replaced by advanced materials if scaling is to continue. The microelectronics industry has identified HfO₂, among other high-k gate dielectrics, as the leading candidate replacement materials for the SiO₂ gate dielectric. However, dielectrics with dielectric constants even greater than HfO₂ (k = 20), i.e., between the range of 50 and 100, are now needed. We have investigated the HfO₂-TiO₂-Y₂O₃ ternary oxide system, and determined that there are many compositions with dielectric constants between 50 and 100, notably TiO₂-rich oxides containing about 30% Y₂O₃ and 15% HfO₂. 

We have also studied the other gatestack layer, the gate metal electrode. The selection of gate electrode materials to replace polysilcon is not as far along as for the gate dielectric case. Recently, we have focused our efforts on gate electrode

Dielectric constants of HfO₂-TiO₂-Y₂O₃