Resistive switching of tin dioxide for memristor applications

In the last decades new electrical devices have been investigated in order to overcome problems caused by integrated circuit shrinking. Within these devices there is one that has attracted much attention due to its high scalability and simple structure. This device was proposed theoretically by Leon Chua in 1971 who gave it the name of “Memristor”. In the recent years thin films structures composed of a metal/oxide/metal (MOM) stack have been linked to this new device, also known as “Resistive switching”. There are several applications for the memristor; one of the most important is for resistive random access memory (ReRAM) where there are two states, low (LRS) and high (HRS) resistance. These states are achieved by a set and reset voltage respectably. The purpose of this work is to investigate the resistive switching behavior of SnO2 films as the middle stack material. MOM structures were fabricated on glass substrates, as bottom film Ti was deposited by RF magnetron sputtering. Then after SnO2 was grown by reactive RF magnetron sputtering and finally silver paint was applied with micropipettes and used as top metal. For the microstructure analysis scanning electron microscope (SEM), grazing incidence x-ray diffraction (GIXRD), and energy dispersive spectroscopy EDS techniques were used. An electrical analysis was performed to study the memristor behavior. Repeatable unipolar switching was observed using a compliance current of 10 mA and limiting the reset voltage between 0.8 and 1.2 V. In order to get a repeatable and reliable device an optimal SnO2 thickness of 45-53 nm was found. A top contact area analysis was carried out indicating a filamentary forming mechanism. The forming voltage was reduced using lower oxygen partial pressures, however the HRS was affected. Furthermore, a retention memory analysis was performed indicating an acceptable device behavior through time. And finally a Ohmic conduction process was found in LRS and HRS, however for the HRS this process was observed only at voltages lower than 0.3 V, at higher voltage there was not a well fit with any conduction process studied in this document indicating a possibility of a change in material structure that is not accounted