Combinatorial memristive materials for sensor applications

A memristor, or memory resistor will remember the value of its resistance in time. In other words, the memristor may have a low or high electrical resistance and switching between these states can be done easily by applying a defined voltage to the device. Hence, the memristor will remember its ohmic value until a new switching action occurs. This is primarily used for development of artificial synapses, resistive random access memories and sensors. When used for sensors, the memristor switches its resistive state due to extra charges generated on its surface as a result of interaction with the detected chemical. Very recent reports emphasize the relevance of using semiconducting or insulating oxides of Ti, W and Nb or Hf and Ta. As the most researched memristive material, Ti oxide serves as a reference/guideline material. Also, combined with W and Nb oxides it describes well the semiconductors used in memristive applications. In spite of their poor electrical conductivity, Hf and Ta oxides have the best memristive properties. All these oxides are easily produced by anodization (by applying a positive voltage to the metal while submerged in a water-based electrolyte) which tremendously decreases the production cost of sensing devices. Memristor fabrication by anodization is a poorly researched topic with a great potential for industrial implementation. A combinatorial route is proposed for obtaining new memristive materials. First, atomic mixture deposition in a vacuum chamber allows a metallic deposit with a changing composition to form across a substrate (i.e. Si wafer). The resulting thin film compositional spread will not be described by a single metallic alloy, but by an entire range of alloys formed at the same time. This library of metals will be anodized so that different oxides will form and mix on the surface depending on the composition of their parent metal alloys. Enhanced memristive properties are reported when materials contain regions with different oxides. The proposed anodic oxidation route directly results in a vertical compositional gradient, no additional processing step being necessary. Metal-insulator-metal structures (identical with the structure of a simple capacitor) obtained by top electrodes patterning will be systematically screened as a function of their composition. Memristive electrical testing using lab automation robots for high throughput will reveal the best mixture of oxides for improved devices. Atomic level imaging and analysis will help elucidating the switching mechanism in mixed oxides. The best memristive oxides will be used for carbon monoxide and glucose sensors proof of concept. The project defines a single PhD work for 3 years. The combinatorial route and the high throughput experimentation combined with the very actual search for better memristive materials provide the grounds for a high impact work in a modern scientific world.

A memristor, or "memory resistor" will remember the value of its resistance in time. In other words, the memristor may have a low or high electrical resistance and switching between these states can be done easily by applying a defined voltage to the device. Hence, the memristor will remember its ohmic value until a new switching action occurs. This is primarily used for development of artificial synapses, resistive random access memories and sensors. When used for sensors, the memristor switches its resistive state due to extra charges generated on its surface as a result of interaction with the detected chemical. Very recent reports emphasize the relevance of using semiconducting or insulating oxides of Ti, W and Nb or Hf and Ta. As the most researched memristive material, Ti oxide serves as a reference/guideline material. Also, combined with W and Nb oxides it describes well the semiconductors used in memristive applications. In spite of their poor electrical conductivity, Hf and Ta oxides have the best memristive properties. All these oxides are easily produced by anodization (by applying a positive voltage to the metal while submerged in a water-based electrolyte) which tremendously decreases the production cost of sensing devices. Memristor fabrication by anodization is a poorly researched topic with a great potential for industrial implementation. A combinatorial route is proposed for obtaining new memristive materials. First, atomic mixture deposition in a vacuum chamber allows a metallic deposit with a changing composition to form across a substrate (i.e. Si wafer). The resulting thin film compositional spread will not be described by a single metallic alloy, but by an entire range of alloys formed at the same time. This library of metals will be anodized so that different oxides will form and mix on the surface depending on the composition of their parent metal alloys. Enhanced memristive properties are reported when materials contain regions with different oxides. The proposed anodic oxidation route directly results in a vertical compositional gradient, no additional processing step being necessary. Metal-insulator-metal structures (identical with the structure of a simple capacitor) obtained by top electrodes patterning will be systematically screened as a function of their composition. Memristive electrical testing using lab automation robots for high throughput will reveal the best mixture of oxides for improved devices. Atomic level imaging and analysis will help elucidating the switching mechanism in mixed oxides. The best memristive oxides will be used for carbon monoxide and glucose sensors proof of concept. The project defines a single PhD work for 3 years. The combinatorial route and the high throughput experimentation combined with the very actual search for better memristive materials provide the grounds for a high impact work in a modern scientific world.