Electromigration in Bicrystal Al Lines

Abstract

Wehave developed a new experimental technique to study electromigration in bicrytal Al lines as a function of the type and location of the grain boundary as well as the testing temperature.The failure times of these lines are found to be lognormally distributed.The median time to failure (MTTF) depends more strongly on the boundary orientation than the type of grain boundary.The dependence of lifetimes on the type and orientation of grain boundaries, the location and appearance of the failure sites, and the measured activation energy (E.) of 0.94eV suggest that both interfacial and grain boundary diffusion contribute to failure in bicrystal lines, and likely in bamboo and near-bamboo lines as well.

Irreversible thermodynamics of triple junctions during the intergranular void motion under the electromigration forces

“…9 the MTTF reported by Longworth and Thompson (1992) for R13[100] and (115)/(100) bi-crystals are in agreement with our LGMFT prediction. However, it seems that this agreement is fortuitous because the appearance and location of failure sites observed by Longworth and Thompson (1992) indicate that the failure mechanism is likely to involve accelerated grain boundary grooving induced by electromigration. The experimental observations by Kinsbron (1980) on large blocking grains and by Cho and Thompson (1989) on bamboo structures are also studied and presented in Fig.…”

“…Analysis of numerous experimental data cited in this paper (Black, 1969;Longworth and Thompson, 1992;Kinsbron, 1980;Cho and Thompson, 1989;Seeger and Mehrer, 1969;Ogurtani, 1979) gives very consistent and highly accurate values for the surface diffusion constant D 0 r ¼ 3:0 Â 10 À6 m 2 s À1 and the enthalpy H 1V r ¼ 0:62 eV if one assumes thatZðAlÞ ¼ 4 for voids formed in the interior of the aluminum interconnects avoiding any interface contamination such as oxides. These values are unusually close to those values calculated in this paper by utilizing a simple model based on coordination numbers supplemented by a bond-breaking and forming model as applied to mono-vacancies created at (1 1 1) planes of clean or freshly formed aluminum surfaces.…”

“…The aluminum interconnect test materials chosen use structural parameters deduced from a number of experimental studies. Data is taken from (1) Longworth and Thompson (1992) in bicrystal aluminum lines: 2w 0 = 2.0 lm, L = 1000 lm, ' g = 500 lm, j = 2.5 MA cm À2 and T = 523 K, (2) Kinsbron (1980): 2w 0 = 1.0 lm, L = 250 lm, ' g = 10 lm, j = 2.0 MA cm À2 and T = 473 K, and (3) Cho and Thompson (1989): 2w 0 = 2.2 lm, L = 100 lm, ' g = 6.5 lm, j = 1.2 MA cm À2 and T = 548 K in bamboo structures. In the computer simulations plotted in this figure the void surface diffusion coefficient D r = 3.0 · 10 À6 exp(À0.62 eV/kT) m 2 s À1 and the effective value ofZ ¼ 4 is chosen by reevaluating the experimental MTTF versus temperature data presented by three different authors (Lytle and Oates, 1992;Black, 1969;Schreiber and Grabe, 1981).…”

“…Cathode failure time vs. current intensity graph on a double-logarithmic scale. s: Longworth and Thompson (1992); h: Kinsbron (1980); e: Cho and Thompson (1989). mono-vacancy diffusion in aluminum, namely: D 1V b ðAlÞ ¼ 1:66  10 À6 exp ðÀ0:62 eV=kT Þ m 2 s À1 .…”

Electromigration-induced drift in damascene and plasma-etched Al(Cu). II. Mass transport mechanisms in bamboo interconnects

“…Most reports that use median time to failure ͑de-rived from EM failure statistics on single-level structures͒ suggested Al/AlO x diffusion to be the rate-controlling mass transport mechanism. [7][8][9] In that case, one common failure characteristic appears to be the prevailing mechanistic role of GB grooves. 10 On the other hand, when considering drift data, contributions from EM-induced mass transport along the Al/AlO x interface have been excluded, 11 due to the absence of any width dependence in the Al drift velocity of bamboo RIE via-type test structures.…”