4 Point defects and their correlation to hydrogen

4.1 Terminology and defect classes

Defect research in crystalline and amorphous SiO trace back to 1956 when Robert A. Weeks of Oak Ridge National Laboratory [79] first used electron paramagnetic resonance (EPR) to study radiation-induced defects in quartz. In 1978, a committee was established by the Electronics Division of the Electrochemical Society and the IEEE-sponsored Semiconductor Interface Specialists Conference in order to clarify the terminology of defects associated with reliability issues in thermally oxidized MOS devices. Following Bruce E. Deal [80, 81], the Chairman of this committee, four different types of oxide charges could be identified, cf. Fig. 4.1.

The defects were entitled in the following way: Fixed Oxide Charge Q, Oxide Trapped Charge Q, Interface Trapped Charge Q and Mobile Ionic Charge Q. At that time the different types of charges could not be ascribed to a microscopic defect, but could at least be classified by their process sensitivity, their ability to exchange carriers with the silicon substrate, and by their response to thermal treatment. (i) Fixed Oxide Charges: oxidation process dependent fixed positive charges which do not exchange carriers with the silicon substrate. (ii) Oxide Trapped Charges: stress induced (avalanche injection or ionizing radiation) positive or negative charges, depending on whether holes or electrons are injected; cannot be neutralized electrically but can be annealed out by low temperature treatment (<500 °C). (iii) Interface Trapped Charges: structural oxidation-induced defects located at the Si/SiO interface, which communicate (exchange carriers) rapidly with the underlying silicon substrate and thus can be either positive or negative, depending on the surface potential; most of the interface trapped charge can be neutralized by low-temperature (450 °C) hydrogen annealing. (iv) Mobile Ionic Charge: Ionic impurities such as Li, Na, K and possibly H, which move across the oxide when a field is applied.

Concerning the degradation of the oxide as a consequence of NBTI, all of those charge types may come into consideration, except for the mobile ionic charges which are rather process contaminations than real defects. In the last 30 years a lot of progress has been made to understand the microscopic nature and degradation dynamics behind the above listed charge species. Against Deal’s request, not to modify the terminology of charges, scientists began to invent new names for the defect precursors and defect states causing the observed oxide and interface charge. Furthermore, due to process innovations like PNOs and high– dielectrics, additional defects and charge types had to be incorporated to the above framework.