Search

In this work, static and low frequency noise parameter extraction are carried out on surface- and buried-mode 0.1 μm PMOSFETs. The two architectures are based either on a surface mode of operation (SC) or on a buried one (BC) featuring P+ and N+ polygate, respectively. The impact of the gate architecture i.e. P+ and N+ polysilicon, on the static and noise parameters is analyzed. The 1/f noise, which can be interpreted in terms of carrier number fluctuations, is found to be about one decade lower for buried-mode devices. The statistical sample-to-sample dispersion of the noise level has also been studied and found to be as large as 4–5 decades for the nominal area devices. Standard RTS analysis has also been performed on representative PMOS devices.

In this paper, the short channel effects in buried and surface channel pMOSFETs are studied. The effective channel length, threshold voltage, source–drain series resistance, carrier mobility in various temperature ranges and the drain-induced-barrier-lowering effect are investigated. Hot carrier degradation is performed in order to extrapolate the device lifetimes and the maximum drain bias that can be applied. Low frequency noise measurements are also carried out in order to evaluate the impact of the gate doping type on the device noise performance.

The aim of this paper is to study the electrical properties of advanced nMOSFETs realized with ultra-thin gate oxide in various temperature ranges. The driving and leakage currents, carrier velocity and short channel effects are studied. Hot carrier degradation at low temperature is also addressed in order to evaluate the device lifetimes and the maximum drain biases that can be applied.

This work presents the low frequency noise and the electric performances in terms of output/transfer characteristics, threshold voltage, and short channel effect in both NMOS and PMOS transistors for 0.1 μm technologies. For the last one there are two architectures based either on a surface mode of operation (surface channel) or on a buried one (buried channel) featuring either a P+ or a N+ polysilicon gate material. The impact of the channel length on the noise characteristics as well as on the output/transfer characteristics is studied. We find that the 1/f noise can be interpreted in terms of carrier number fluctuations for both N and P channel MOSFETs for surface and buried mode of operation. The oxide trap density Nt is therefore evaluated, demonstrating an overall good oxide quality.

A detailed characterization of the gate and drain current characteristics has been carried out on NMOS transistors featuring ultra-thin gate oxides (1.2–2.3 nm). A quasi-analytical model of both gate and drain currents has been worked out and validated in ohmic and saturation regions. The impact of gate current on MOSFET operation has been shown to be critical for large area devices but of less importance for nominal devices, indicating that, down to 1.2 nm, the gate oxide thinning is not detrimental to the device functionality.

This paper reports the Capacitance–Voltage (C–V) characterization of gate oxide in the 20 A thickness range, dealing especially with quantum and polysilicon effects. In the first part, the basis of Poisson–Schrodinger modeling is summarized. The extraction procedure of doping level, flat band voltage, threshold voltage, and oxide thickness is presented. A simple method to extract the polysilicon doping level and to correct for polysilicon effect on C(V) curves is proposed. These parameter extraction procedures are illustrated by experimental data obtained on both NMOS and PMOS transistors.

The characterization of LF noise and RTS fluctuations in 0.18μm CMOS devices is presented. It is found that the 1 f noise results from carrier number fluctuations for both N and P MOS devices. The slow oxide trap density deduced from the noise data lies around 1017/eV/cm3 in agreement with state-of-the-art gate oxides. Drain current RTS amplitude as large as 5–10% have been observed, being somewhat larger than for 0.35μm CMOS technology but comparable to those obtained in 0.25μm CMOS devices.

The low frequency (LF) noise and random telegraph signal (RTS) fluctuations in 0.25 μm complementary metal-oxide-semiconductor (CMOS) devices are investigated. We find that the 1/f noise stems from fluctuations in carrier number for both n and p channel MOS devices. The slow oxide trap concentration deduced from the noise data is ≈ 10 17 eV/cm 3 in agreement with state-of-the-art gate oxides. The study of some particular RTSs is performed and provide the gate voltage dependence of the capture/emission times as well as of the drain current RTS amplitude. Drain current RTS amplitude as large as 10% have been observed, being somewhat larger than for 0.35 μm CMOS technology.

A detailed study of the low frequency noise in Si MOS devices from a 0.35 μ m CMOS technology is presented. The normalized drain current noise WLS I d /I d 2 has been systematically investigated at a fixed frequency and constant normalized drain current. In large area devices (>5–7 μ m 2 ), the noise is typically 1/f-like, whereas, in small area devices (≤1 μ m 2 ), the noise stems from the contribution of one or several RTS components. In large area devices, the 1/f noise can easily be interpreted by a classical carrier number fluctuation model. A multi-RTS component scheme has to be employed in the case of small area devices.

A detailed investigation of the low frequency noise in Si MOS devices issued from a 0.35 μm CMOS technology is conducted. The normalized drain current noise WLSid/Id2 has been systematically measured at a fixed frequency (10 Hz) and constant normalized drain current. It is found that, for large area devices (≥7–10 μm2), the sample‐to‐sample noise level variation lies around a factor of 2–3, while, for the smallest devices (0.1–0.3 μm2), the sample‐to‐sample noise level variation can exceed 3 decades. In large area devices, the 1/f noise can easily be described by a classical carrier number fluctuation model using the concept of dynamic flat band voltage. In the case of intermediate or small areas, a multi‐RTS component scheme has to be employed.