ALBERT

Notes: We have identified several features of the 1/f noise and radiation response of metal-oxide-semiconductor (MOS) devices that are difficult to explain with standard defect models. To address this issue, and in response to ambiguities in the literature, we have developed a revised nomenclature for defects in MOS devices that clearly distinguishes the language used to describe the physical location of defects from that used to describe their electrical response. In this nomenclature, "oxide traps'' are simply defects in the SiO2 layer of the MOS structure, and "interface traps'' are defects at the Si/SiO2 interface. Nothing is presumed about how either type of defect communicates with the underlying Si. Electrically, "fixed states'' are defined as trap levels that do not communicate with the Si on the time scale of the measurements, but "switching states'' can exchange charge with the Si. Fixed states presumably are oxide traps in most types of measurements, but switching states can either be interface traps or near-interfacial oxide traps that can communicate with the Si, i.e., "border traps'' [D. M. Fleetwood, IEEE Trans. Nucl. Sci. NS-39, 269 (1992)]. The effective density of border traps depends on the time scale and bias conditions of the measurements. We show the revised nomenclature can provide focus to discussions of the buildup and annealing of radiation-induced charge in non-radiation-hardened MOS transistors, and to changes in the 1/f noise of MOS devices through irradiation and elevated-temperature annealing.Border-trap densities of ∼1010–1011 cm−2 are inferred from changes in switching-state density during postirradiation annealing, and from a simple trapping model of the 1/f noise in MOS devices. We also present a detailed study of charge buildup and annealing in MOS capacitors with radiation-hardened oxides through steady-state and switched-bias postirradiation annealing. Trapped-hole, trapped-electron, and switching-state densities are inferred via thermally stimulated current and capacitance-voltage measurements. A lower bound of ∼3×1011 cm−2 is estimated for the effective density of border traps that contribute to the electrical response of the irradiated devices. This is roughly 20% of the observed switching-state density for these devices and irradiation conditions. To our knowledge, this represents the first quantitative separation of measured switching-state densities into border-trap and interface-trap components. Possible physical models of border traps are discussed. E' centers in SiO2 (trivalent Si centers associated with oxygen vacancies) may serve as border traps in many irradiated MOS devices.

Notes: A general methodology is developed to experimentally characterize the spatial distribution of occupied traps in dielectric films on a semiconductor. The effects of parasitics such as leakage, charge transport through more than one interface, and interface trap charge are quantitatively addressed. Charge transport with contributions from multiple charge species is rigorously treated. The methodology is independent of the charge transport mechanism(s), and is directly applicable to multilayer dielectric structures. The centroid capacitance, rather than the centroid itself, is introduced as the fundamental quantity that permits the generic analysis of multilayer structures. In particular, the form of many equations describing stacked dielectric structures becomes independent of the number of layers comprising the stack if they are expressed in terms of the centroid capacitance and/or the flatband voltage. The experimental methodology is illustrated with an application using thermally stimulated current (TSC) measurements. The centroid of changes (via thermal emission) in the amount of trapped charge was determined for two different samples of a triple-layer dielectric structure. A direct consequence of the TSC analyses is the rigorous proof that changes in interface trap charge can contribute, though typically not significantly, to thermally stimulated current.

Notes: Thermally stimulated current (TSC) and capacitance–voltage measurements are combined via a newly developed analysis technique to estimate positive and negative oxide-trap charge densities for metal–oxide–semiconductor (MOS) capacitors exposed to ionizing radiation or subjected to high-field stress. Significantly greater hole trapping than electron trapping is observed in 3% borosilicate glass (BSG) insulators. Two prominent TSC peaks are observed in these BSG films. A high-temperature peak near 250 °C is attributed to the Eγ′ defect, which is a trivalent Si center in SiO2 associated with an O vacancy. A lower temperature positive charge center near 100 °C in these films is likely to be impurity related. The higher temperature Eγ′ peak is also observed in 10, 17, and 98 nm thermal oxides. A much weaker secondary peak is observed near ∼60 °C in some devices, which likely is due to metastably trapped holes in the bulk of the SiO2. Negative charge densities in these thermal oxides are primarily associated with electrons in border traps, which do not contribute to TSC, as opposed to bulk electron traps, which can contribute to TSC. Ratios of electron to hole trap densities in the thermal oxides range from ∼30% for radiation exposure to greater than 80% for high-field stress. It is suggested that the large densities of border traps associated with trapped holes in these devices may be due to high space-charge induced electric fields near the Si/SiO2 interface. In some instances, border traps can reduce near-interfacial electric fields by local compensation of trapped positive charge. This may provide a natural explanation for the large densities of border traps often observed in irradiated or electrically stressed MOS capacitors. © 1998 American Institute of Physics.

Notes: Thermally stimulated current- and capacitance-voltage techniques are combined to provide the first quantitative estimates of the contributions of trapped-hole annealing and electron trapping to oxide-trap charge neutralization in metal-oxide-semiconductor devices. For 350-nm nonradiation-hardened oxides, trapped electrons compensate ∼15% of the radiation-induced trapped positive charge after x-ray irradiation (evidently forming dipolar defects), and ∼65% of the trapped positive charge remaining after positive-bias annealing at 80 °C. For 45-nm radiation-hardened oxides, trapped electrons compensate ∼45% of the trapped positive charge after irradiation, and ∼70% after annealing. Implications for models of oxide-trap-charge buildup and annealing are discussed.