Jonathan M. McKenna

New global insight in ultrathin oxide reliability using accurate experimental methodology and comprehensive database

In this paper, we critically examine several important experimental aspects concerning ultrathin oxide reliability. The statistical nature of breakdown measurements and the impact on data interpretation is discussed. Thickness dependence of Weibull slopes and its impact on reliability projection is reviewed. We also investigate the voltage-dependent voltage acceleration using two independent experimental methods over a wide range of oxide thickness values. Within the framework of a general defect generation model, we explore the possibility of a voltage-dependent defect generation rate to account for the increase in voltage acceleration with decreasing voltages. Using direct experimental results, we clarify that strong temperature dependence found on ultrathin oxides is a voltage effect, not a thickness effect as previously suggested, In the context of voltage-dependent voltage acceleration, we experimentally resolve various seemingly contradicting and confusing observations such as temperature-independent voltage acceleration and non-Arrhenius temperature dependence found on ultrathin oxides. Finally, we provide a global picture for time-to-breakdown in voltage and temperature domain constructed from two important empirical principles based on comprehensive experimental database.

Interplay of voltage and temperature acceleration of oxide breakdown for ultra-thin oxides

In this work, we resolved several seemingly conflicting experimental observations regarding temperature dependence of oxide breakdown in the context of change of voltage acceleration factors with reducing voltages. It is found that voltage acceleration factor is temperature dependent at a fixed voltage while voltage acceleration factors are temperature independent at a fixed TBD. We unequivocally demonstrated that strong temperature dependence of time(charge)to-breakdown, TBDðQBDÞ, observed on ultra-thin gate oxides (<5 nm) is not a thickness effect as previously suggested. It is a consequence of two experimental facts: (1) voltage-dependent voltage acceleration and (2) temperature-independent voltage acceleration at a fixed TBD window. For the first time, time-to-breakdown at low temperature of � 50 Ci s reported. It is found that Weibull slopes are insensitive to temperature variations using accurate area-scaling method. The stress-induced leakage current (SILC) was used as a measure of defect-generation rate and critical defect density to investigate its correlation with the directly measured breakdown data, QBDðTBDÞ. The comprehensive and statistical measurements of SILC at breakdown as a function of temperature are presented in detail for the first time. Based on these results, we conclude that SILC-based measurements cannot adequately explain the temperature dependence of oxide breakdown. Finally, we provide a global picture for time-to-breakdown in voltage and temperature domains constructed from two important empirical relations based on comprehensive experimental database. 2002 Published by Elsevier Science Ltd.

Tunneling current characteristics and oxide breakdown in P+ poly gate PFET capacitors

In this work, we investigate both tunneling current and oxide breakdown characteristics for lightly and heavily doped p+ polysilicon gates of PFET capacitors in inversion mode. It was found that tunneling currents show significantly different magnitude for the two doping conditions over the same applied gate voltages. We present experimental evidence that strongly supports electron energy, as set by the gate voltage, and electron fluence, measured as charge-to-breakdown, Q/sub BD/, as being the physical parameters that control the breakdown process, rather than oxide field and time-to-breakdown, T/sub BD/, as suggested by the thermo-chemical model.

Weibull slopes, critical defect density, and the validity of stress-induced-leakage current (SILC) measurements

Voltage, temperature, and polarity dependence of Weibull slopes are carefully measured using area scaling method over a wide range of voltages and temperatures for several oxide thickness (T/sub OX/) values in comparison with direct method. We investigate the validity of stress-induced-leakage-current (SILC), /spl Delta/J/J/sub 0|BD/, as a measure for the critical defect density, N/sub BD/. Our finding clearly shows that the /spl Delta/J/J/sub 0|BD/ cannot be used as a reliable measure of N/sub BD/. This work suggests that a re-evaluation of the breakdown models constructed from the SILC-based measurements is required in gate oxide reliability community, in particular, their validity in comparison with the statistically accurate breakdown data.

Key measurements of ultrathin gate dielectric reliability and in-line monitoring

High-performance CMOS products depend upon the reliability of ultrathin gate dielectrics. In this paper a methodology for measuring thin gate dielectric reliability is discussed in which the focus is upon the elements of those test structures used in the evaluation, the design of the reliability stress matrix, and the generation of engineering design models. Experimental results are presented which demonstrate the reliability of ultrathin gate dielectrics measured on a wide variety of test structures with dielectric thicknesses ranging from 7 to 3.5 nm. An overview is provided for thin gate oxide reliability that was measured on integrated functional chips-high-performance microprocessors and static random-access memory (SRAM) chips. The data from these measurements spanned the period from early process and device development to full production. Manufacturing in-line monitoring for thin gate dielectric yield and reliability is also discussed, with several case histories presented which show the effectiveness of monitors in detecting process-induced dielectric failures. Finally, causes of oxide fails are discussed, leading to the process actions necessary for controlling thin gate dielectric defects.