Abstract
The development of modern electronics has promoted the advancement of science and technology as well as human life. The exploration of materials makes electronic devices more stable and reliable to meet the requirements of applications. Unfortunately, these reliable electronic devices create undesired issues such as electronic waste, potentially adverse long-term effect during implantation and physical information insecurity. How to tackle these problems by developing new devices without sacrificing the electronic performance has become a hot research topic. Transient electronics, which can physically disappear at a predefined time, is a promising candidate that can overcome these problems and offer unique advantages for medicine, information security and green electronics. Till now, various transient devices including resistor, inductor, energy harvester, transistor, diode, sensor, battery and memory devices have been reported. Resistive random-access memory (RRAM) is one of the most potential non-volatile memories due to its simple structure and high scalability. However, so far, the reported transient RRAM devices are power inefficient, incompatible to the complementary metal-oxide-semiconductor (CMOS) technology, and have uncontrollable degradation, which greatly hinder practical applications. This research project aims to tackle the abovementioned issues by developing novel transient RRAMs with low power consumption (< 1mW) and controllable device degradation process through material and device engineering. Firstly, a transient resistive memory with ultralow power consumption is demonstrated. Although some transient memories have been reported, the power consumption is still relatively high (~mW). How to reduce power consumption of transient memory remains a big challenge. Here, a reliable W/Ag/MgO/W transient memory was designed without sacrificing the memory performance, dissolution behaviour and non-cytotoxicity. The memory device shows a high ON/OFF ratio and stable switching performance under a compliance current of 5 µA. The total power consumption of W/Ag/MgO/W transient memory is estimated as 0.5 mW, which is record low and reduced by half compared to the reported work previously. The dissolution results in water and PBS solution prove that the insertion of ultrathin Ag layer has no negative effect on the complete dissolution of the whole device. Moreover, no obvious cytotoxicity is observed after the disappearance of the transient memory device, predicting its biocompatibility and biodegradation. Secondly, a transient resistive memory that is compatible with current complementary metal-oxide semiconductor (CMOS) technology with prolong lifetime (up to 3 months at room temperature) via Al2O3 encapsulation layer was reported. The previous research works about transient memory are generally based on materials such as Mg, MgO, silk fibroin, Zn, ZnO. These materials option may cause two possible problems: a) Incompatible to CMOS technology may hinder its mass product and wide application, b) Ultra-sensitivity of the metal electrode to aqueous environment may greatly decrease the lifetime of transient resistive memory and increase the difficulty of encapsulation design. In this work, a new transient resistive memory structure with CMOS compatible materials (W/Cu/SiO2/W) was demonstrated. Stable bipolar memory performance with high uniformity was observed. Increasing the temperature didn’t deteriorate the memory properties, indicating its suitability for back-end-of-line (BEOL) process. In addition, cytotoxicity testing results show that the growth and viability of cells are not adversely affected by the dissolution of the resistive memory device. More importantly,through a 30 nm Al2O3 protection layer, the W/Cu/SiO2/W resistive memory shows an estimated lifetime of ~ 3 months at room temperature, which is 3 orders higher than that of Mg-based transient memory. Furthermore, a hydrogel-oxide bilayer encapsulation strategy was designed to achieve the controllable degradation of the transient memory device which is rarely reported previously. The encapsulation materials reported includes SiNx, SiO2, silk fibroin etc. They can provide a lifetime ranges from minutes to days which is definitely determined by the thickness of the encapsulation layer while lacking tunability. This predefined lifetime has greatly limited the application of transient memory devices and makes the following situation happen: for different scenarios, a different encapsulation layer (thickness, materials) need to be redesigned. Aiming to solve this problem, in this work, a hydrogel-oxide bilayer encapsulation was employed to control the lifetime of the transient resistive memory. The light responsive hydrogel serves as a barrier between the water and oxide. Interestingly, the gel-to-sol transition of hydrogel happens under UV condition, greatly changing the water flow through the oxide layer, which means that the controllability degradation can be reached by UV treatment. Experimental results demonstrate there is a minimum requirement for the thickness of the encapsulation layer in order to realize the degradation tunability of transient memory devices. Finally, a simple and universal photo-triggerable method to control the degradation of transient memory was demonstrated by fullerene-mediated degradation of polymer:PCBM encapsulation layer. Here, the PCBM was used as an additive to two polymer candidates with the phenyl group. Compare transient memory performance under the condition with and without UV trigger, the lifetime becomes much shorter under UV trigger. The possible reason can be ascribed to the polymer photodegradation caused by UV light, the generated defects in polymer allow more water to pass through and react with the dissolvable metals, which significantly accelerate the degradation of the transient memory device. It also has been demonstrated that this strategy can be applied to the flexible substrate. In summary, this research has conducted a comprehensive study of transient resistive memory from materials (performance, fabrication technology), devices to controllable degradation. The demonstration of high performance, CMOS compatible, low power and degradation on-demand transient resistive memories paves a way for practical applications. More importantly, the strategies to optimize performance and fabrication process, as well as the encapsulation design to control degradation, will provide a useful guideline for advancing the development of not only transient memories but also the entire transient electronics.