What Do We Certainly Know About 1/f Noise in MOSTs?

1/f noise is a fluctuation in the conductance of semiconductors and metals. This noise could be a fluctuation in the number of free electrons or in their mobility. Many experimental studies have proved that the 1/f noise in homogeneous samples is a fluctuation in the mobility. There is a reliable empirical relation for the power density of the mobility noise. Theoretical models for mobility noise have not had much influence in the discussions. The theoretical position for number fluctuations is quite the opposite. The generally accepted McWhorter model is simple; however, very few experimental studies definitely prove number fluctuations and exclude uncorrelated mobility fluctuations. There is an extensive literature on noise in MOSTs. Submicrometer MOSTs are notorious for their high low-frequency noise. Almost all studies start from the McWhorter generation-recombination (GR) model. The discussion is confused by the many complications of the MOSTs, such as mobility degradation, surface and contacts effects, size of the samples, etc. An additional problem is that the real 1/f noise is often mixed up with other types of low-frequency noise, such as burst and GR noises. A critical discussion is given for the Kirton and Uren model. Analytical expressions are presented for the results of their computer simulations. This explains their surprising discovery that the density of McWhorter states may be substantially different from the usual 1/tau distribution.