Noise Generation
Theoretical frameworks for low-frequency noise in semiconductor devices describe the fluctuation of electrical current over time. The McWhorter model attributes this noise to the random trapping and emission of charge carriers by defects near the semiconductor-dielectric interface. It explains the characteristic inverse-frequency dependence of flicker noise as a superposition of many individual trapping processes with a wide distribution of time constants.
The model is a fundamental tool for analyzing noise in field-effect transistors.
Trapping Mechanism
Charge carriers from the conduction channel tunnel into defects located in the gate oxide and remain trapped there for a duration that depends on the distance from the interface. When these carriers eventually escape back to the channel, their movement creates fluctuations in the local threshold voltage and drain current. The rate of this tunneling process decreases exponentially with the distance of the trap from the semiconductor boundary.
Flicker Characterization
Mathematical integration over a uniform distribution of trap distances results in a power spectral density that varies inversely with frequency. In many silicon devices, the McWhorter model provides an accurate prediction of the noise slope across several decades of frequency.
Interface Quality
Manufacturing processes that minimize the density of interface defects directly reduce the noise level predicted by this framework. In modern transistor fabrication, high-temperature annealing in a hydrogen ambient is used to passivate dangling bonds at the silicon boundary, which lowers the number of active trapping sites. Characterizing the resulting noise spectrum allows process engineers to monitor the effectiveness of these passivation steps.