Statistical analysis of early failures in electromigration

The detection of early failures in electromigration (EM) and the complicate d statistical nature of this important reliability phenomenon have been dif ficult issues to treat in the past. A satisfactory experimental approach fo r the detection and the statistical analysis of early failures has not yet been established. This is mainly due to the rare occurrence of early failur es and difficulties in testing of large sample populations. Furthermore, ex perimental data on the EM behavior as a function of varying number of failu re links are scarce. In this study, a technique utilizing large interconnec t arrays in conjunction with the well-known Wheatstone Bridge is presented. Three types of structures with a varying number of Ti/TiN/Al(Cu)/TiN-based interconnects were used, starting from a small unit of five lines in paral lel. A serial arrangement of this unit enabled testing of interconnect arra ys encompassing 480 possible failure links. In addition, a Wheatstone Bridg e-type wiring using four large arrays in each device enabled simultaneous t esting of 1920 interconnects. In conjunction with a statistical deconvoluti on to the single interconnect level, the results indicate that the electrom igration failure mechanism studied here follows perfect lognormal behavior down to the four sigma level. The statistical deconvolution procedure is de scribed in detail. Over a temperature range from 155 to 200 degreesC, a tot al of more than 75 000 interconnects were tested. None of the samples have shown an indication of early, or alternate, failure mechanisms. The activat ion energy of the EM mechanism studied here, namely the Cu incubation time, was determined to be Q=1.08 +/-0.05 eV. We surmise that interface diffusio n of Cu along the Al(Cu) sidewalls and along the top and bottom refractory layers, coupled with grain boundary diffusion within the interconnects, con stitutes the Cu incubation mechanism. (C) 2001 American Institute of Physic s.