Variability and self-heating analysis of spin transfer torque magnetic random access memory devices

Abstract

Spin Transfer Torque Magnetic Random-Access Memories (STT-MRAMs) are promising candidates for next-generation data storage due to their non-volatility, fast access times, scalability, low power consumption, and compatibility with conventional CMOS technologies. The primary building block of an STT MRAM is the Magnetic Tunnel Junction (MTJ), which consists of two ferromagnetic layers separated by an insulating layer. However, owing to the small dimensionalities, STT-MRAMs are significantly prone to device-to-device and cycle-to-cycle variations. Moreover, the high current density required to program the MTJ invariably leads to self-heating, which significantly influences the device’s performance and may even cause the breakdown of the insulating layer, thereby significantly limiting their reliability. In this thesis, a simulation framework for studying the device-to-device variability along with the self-heating analysis capability of the STT MRAM device has been developed. The proposed framework has been validated against reported experimental results in the literature. Within this framework, device-to-device variability of CoFeB/MgO-based STT-MRAMs is studied, considering the influence of interface quality, temperature variation, and device dimensionality. Metal-induced gap states resulting from electron transfer at the ferromagnet-tunnel barrier interface significantly influence these devices’ effective energy barrier height, irrespective of their diameters. Switching voltage and parallel-antiparallel resistance values vary by as much as 43% and 30%, respectively, for about 13% variation of the energy barrier, whereas the tunneling magnetoresistance remains typically unaffected. WRITE cycles of highly scaled STT-MRAMs are, therefore, more susceptible to device-to-device variations resulting from microscopic variations in the interface quality rather than the READ cycles. Such variations are observed to be independent of temperature as well as the spatial distribution of the defects. Moreover, in this thesis, our proposed simulation framework has been extended to present a theoretical study of self-heating, taking into account the magnetic switching dynamics and three-dimensional heat transfer characteristics of STT MRAM devices. The impact of self-heating has been explored for different dimensions, geometrical aspects, and bias conditions of the device. The temperature rise of the MTJ stack is observed to be strongly dependent on the choice of the encapsulating material and top metal layer thickness. Because of the asymmetry of the stack, device-to-device self-heating variability is expected to be more prominent in undercut structures than in overcut ones. From transient analysis, it is observed that during both unipolar and bipolar pulsing conditions, MRAM switching is accompanied by an abrupt temperature change of around 25-30 K. The results of this work suggest that consideration of magnetic switching dynamics is essential to accurately estimate self-heating during transient and steady-state operations of STT-based MTJ device