Thermally Sensitive Ionic Redox Transistor for Neuromorphic Memory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional neuromorphic computing devices, such as RRAM and PCM, are inefficient due to high voltage and current requirements, unpredictability, nonlinearity, and information loss issues, making them unsuitable for scalable, energy-efficient analog memory devices, especially at micro- and nano-scales.
Innovation Solution
A thermally sensitive ionic redox transistor with a variable-conductance solid channel, a solid electrolyte layer, and a solid reservoir layer, where ionic conductivity is high at elevated temperatures, allowing programming, and low at low temperatures, preventing information loss, using materials like nonstoichiometric transition metal oxides and compatible with CMOS fabrication techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If RRAM or PCM are used for analog memory, then neuromorphic computing functionality is achieved, but large voltages and currents are required and the devices are highly unpredictable and nonlinear
Solution Approach 1:
The patent changes the operating parameters by using ionic transport mechanisms instead of electronic switching, enabling linear conductance modulation through controlled ion migration in the electrolyte layer, which resolves the unpredictability and nonlinearity issues of RRAM and PCM
Solution Approach 2:
The patent replaces the electronic switching mechanism with ionic transport mechanisms, where ion migration in the electrolyte layer controls channel conductance, providing predictable and linear analog memory behavior suitable for neuromorphic computing
2Adaptability or versatility
If RRAM or PCM are used for analog memory, then neuromorphic computing functionality is achieved, but the devices draw micro-amperes of current during read operations, precluding scaling into large arrays
Solution Approach 1:
The patent replaces electronic current flow with ionic transport mechanisms, where ion migration controls channel conductance without requiring continuous high current flow, enabling low-power read operations and scaling to large arrays
3Manufacturing precision
If ion transport from electrochemical gate layer to channel layer is used to vary conductance, then programming linearity is improved, but at micro- and nano-scale the devices are prone to information loss within seconds after programming
Solution Approach 1:
The patent introduces a solid electrolyte layer as an intermediary between the gate and channel, where ion migration in the electrolyte controls channel conductance. The electrolyte acts as a mediator that maintains ionic concentration gradients, preventing charge leakage and information loss while preserving programming linearity
Solution Approach 2:
The patent establishes ionic concentration gradients in the electrolyte layer during programming, creating a preliminary ionic distribution that maintains the programmed state without requiring continuous power, thereby preventing information loss at micro- and nano-scales
4Reliability
If lithium polymers and ionic liquids are used in redox transistors, then ionic conductivity is achieved, but the devices are difficult to fabricate using conventional CMOS fabrication techniques
Solution Approach 1:
The patent changes the material parameters by using solid electrolyte materials that can be deposited using conventional CMOS-compatible techniques such as atomic layer deposition (ALD) or sputtering, maintaining ionic conductivity while enabling standard fabrication processes
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The device retains programmed states for extended periods with minimal degradation, achieving efficient and predictable conductance changes, suitable for neuromorphic computing applications, and is scalable for large arrays with improved energy efficiency.
Implementation Method 1
The ionic conductivity of one or more layers, or at the interfaces between layers, is thermally activated such that the ionic conductivity is high at elevated temperatures (e.g., between about 80° C. and about 300° C.) and ions are mobile between the layers, and at low temperatures (e.g., less than 50° C.) the conductivity is low and ions in the one or more layers are effectively immobile
Implementation Method 2
Electronic conductance of the channel (e.g., conductance of the channel as measured between the source and drain contacts) varies with the concentration of ions in the channel layer
Implementation Method 3
at low temperatures (e.g., less than 50° C.) the conductivity is low and ions in the one or more layers are effectively immobile. Thus, as will be described in greater detail herein, at elevated temperatures, the device can be programmed, and at low temperatures, the device retains state for extended periods (e.g., about 10 days with less than 2% change in state value)
Data Source
AI summary
A thermally sensitive ionic redox transistor comprises a solid channel, a solid reservoir layer, and a solid electrolyte layer disposed between the channel and the reservoir layer. A conductance of the channel is varied by changing the concentration of ions such as oxygen vacancies in the channel layer. Ionic conductivity of the gate, electrolyte, and channel layers increase with increasing temperature. Ion or vacancy transport between the channel and the reservoir layer across the electrolyte layer occurs in response to applying a voltage between the channel and the reservoir layer when the device is heated to an elevated temperature. When the device is cooled below the elevated temperature, the ions are trapped in one or more of the layers because the materials lose their ionic conductivity. A state of the redox transistor can be read by measuring the conductance of the channel.