Random nanosized metal grains and interface-trap fluctuations in emerging CMOS technologies

This article estimates the influences of work-function fluctuation (WKF) and interface-trap fluctuation (ITF) on emerging high-κ/metal gate complementary metal-oxide-semiconductor (CMOS) devices using an experimentally validated three-dimensional (3D) device simulation. Randomness of nanosized metal grains on a device’s gate is simulated using localized WKF (LWKF) technqiue and 2D random ITs at the HfO₂/silicon interface are considered in contrast with 1D ITF simulation. Proportional to minimal grain size, fluctuation of threshold voltage (σV_{th, WKs}) induced by random WKs is 36.7 and 42.5 mV for 16-nm-gate (width: 16 nm) N- (with TiN gate) and P-MOSFETs (with TiN+Al gate), respectively. Random WKs perturb local potential barrier and result in rather different V_th even when a device has the same number of metal grains owing to random position effect which is beyond the averaged WK method. For devices with random IT’s density D_it varying from 1.51×10¹¹ to 6.32×10¹² eV^-1 cm^-2, the random ITs-induced V_th fluctuations (σV_{th, ITs}) is up to 39 mV and it is reduced to 25 mV around for devices with a tenth of D_it. Statistical sum of WKF and ITF: (σ²V_{th, WKs}+σ²V_{th, ITs})^0.5=53.7 mV overestimates σV_{th,“WKs+ITs”}=46.8 mV (>14% overestimation) of combined random WKs and ITs because assumption of identical independent distribution may not hold owing to their interaction of surface potentials. Fluctuation resulting from the combined random WKs and ITs could be comparable to the random dopant fluctuation.