Study of Cu Migration-Induced Failure of Inter-Layer Dielectric

Modeling of Interconnect Dielectric Lifetime Under Stress Conditions and New Extrapolation Methodologies for Time-Dependent Dielectric Breakdown

“…Earlier work (including long term constant low-field stress at various temperatures) on thick gate oxides showed substantial evidence in support of the E-model [5] ( ln(t) cc E, where t = lifetime and E = applied field). Other models followed, such as the 1/E model [6,7] ( ln(t)oc ), and recently also a F -model [3,8,9] ( ln(t) x £ ). The physical reasoning behind these functional dependencies can be easily disputed, and there is not enough experimental data to clearly distinguish between them phenomenologically.…”

Electrical Breakdown in Advanced Interconnect Dielectrics

“…For the designs of such structures (sometimes referred to as damascene 'MIM' structures for metal-insulator-metal), the interconnect spacing between the comb finger and serpent or adjacent comb fingers can be varied as well as the lengths and widths of the individual fingers or serpent metal lines. An example is a planar metal-oxide-semiconductor (MOS) capacitor (see examples in Figure 11.4 (d); some refer to these structures as 'MIS' for metalinsulator-semiconductor), where a dielectric is deposited upon a heavily doped Si wafer and then capped by a suitable metal layer [256][257][258][259][260]. Because both CC and CS structures are necessarily terminated with ends or turnabouts (within the serpent meander), there is some concern that they might limit the reliability of a given test structure [27].…”

“…Because both CC and CS structures are necessarily terminated with ends or turnabouts (within the serpent meander), there is some concern that they might limit the reliability of a given test structure [27]. The advantage of such a structure is that the LK/ULK is not damaged by the damascene process flow so that an intrinsic reliability study of bulk LK/ULK is possible, and if Cu is chosen as the metal layer, then a Cu drift-diffusion reliability assessment can be done [256][257][258][259]. The effect of having line terminations and turnabouts may be partially mitigated by ensuring that adequate space exists between the line end and the opposing comb spine or serpent turnabout, although one might expect that making adequate adjustments will be more difficult with newer BEOL integration approaches such as double-patterning [27].…”

“…Because Cu metallization is involved, any free Cu from either corrosion defects or weak barrier confinement will be activated at higher temperatures. Dielectric breakdown is generally expected to have low activation energy (E a ≤ 0.5 eV) [221,[267][268][269] as opposed to Cu drift-diffusion controlled breakdown (E a ≥ 1.0 eV) [259] so that their individual contributions to dielectric breakdown can be determined (rather painfully) by testing over a wide temperature range [270]. Too high a testing temperature cannot be used [265,266], unless one is testing to directly assess the impact of such free Cu.…”