Browse Prior Art Database

Use of Resistance as a monitor for Thermal Cycle Reliability

IP.com Disclosure Number: IPCOM000127305D
Original Publication Date: 2005-Aug-22
Included in the Prior Art Database: 2005-Aug-22
Document File: 2 page(s) / 47K

Publishing Venue

IBM

Abstract

Reliability under thermal cycle conditions is one of the main concerns when integrating BEOL structures with low dielectric constant dielectrics. While the thermal cycle performance of a given process can be evaluated by stressing specifically designed test structures, this can only be done on a limited sampling of parts and only on complete builds of the structure. It is highly desirable to have a none destructive method for in line testing of circuits so that the thermal cycle performance provided by a manufacturing process can be continuously monitored. We have identified line resistance as a parameter that directly correlates with the thermal cycle reliability of via stacks during thermal cycling.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Use of Resistance as a monitor for Thermal Cycle Reliability

We are proposing that a measurement of line resistance be used during manufacturing to continuously monitor thermal cycle performance. This is a simple nondestructive test that can be done each level during the building of BEOL structures. Our data show that such a measurement would provide a sensitive test for when process conditions fall outside a window that assures thermal cycle reliability during the manufacture of BEOL structures.

We have been using a test structure called BOCD, which is a stacked via chain, to verify the stability of SilK/Copper BEOL structures during thermal cycling. SiLK is an organic dielectric that has been considered for use in advanced integrated circuit technologies. We have noted in comparing two different processes, Phase3 and Phase 4, that there was a one to one correspondence between the resistance of that structure as built and the failure rate of the structure during thermal cycle testing. Below is a log-log plot of N50 (the number of cycles to 50% failure) versus Average BOCD Resistance for different lots thermal cycled from -65 C to +150 C. The data is fit to an empirical relation, N50=A(Rc-R)-n, where Rc is a critical resistance above which thermal cycle fails should not occur. The plot shows good agreement between the model, given by the solid line, and the actual data. The idea behind this model is that a higher BOCD resistance corresponds to a thicker liner. In a sense, the liner can be thought of as sharing the stress in the stacked via s...