Imec, Lam Research improve wafer drying with test method

October 24, 2011 — Imec and Lam Research have developed a quantitative evaluation method for wafer drying techniques. Silicon nano-pillar test structures with high aspect ratios (up to 28) demonstrate the mechanisms of pattern collapse and sticking in the wafer drying step.

The team recorded results with top-down scanning electron microscopy (TD-SEM) inspection.

Shallow trench isolators, gate structures, and stacked capacitors can fall victim to capillary forces during the dry step of wet cleaning processes. Improved drying techniques will reduce the risk of stiction.

Also read: Imec, ASML extend litho work, debut EUV sensors and Imec demos faster HBTs in SiGe for wireless, imaging

Imec and Lam Research focused on an appropriate test vehicle and quantitative metrology to evaluate the effect of structure geometry, materials, and cleaning technique on pattern collapse, and to explore the potential of structure repair.

Nano-pillars were fabricated by first printing 90nm pitch nano-dots in 193nm photoresist. Subsequently, they were transferred into an amorphous carbon patterning layer and Si substrate using plasma etch technology on 300mm wafers. As such, Si nano-pillars were obtained with an aspect ratio up to 28. In contrast to other proposed test structures such as cantilevers and DRAM cylinders, these nano-pillars do not require any wet etch treatment during patterning, making the structures suitable for investigating various cleaning techniques. Inspection of collapse events after cleaning was performed using TD-SEM, followed by dedicated post-inspection image analysis.

Figure 1. Cross-section SEM image of Si nano-pillars after plasma etch and mask strip.

A first observation is a sharp dependence of pattern collapse on aspect ratio. By measuring the percentage of collapse at different aspect ratios, a threshold aspect ratio can be defined below which the structures are stable.  We investigated two cleaning methods and found threshold values of 18 and 13, respectively. Based on these results, we can conclude that the first method is superior to the second in terms of pattern collapse prevention. The experiments further suggest that the collapse events are at random and statistically determined within a pillar field.

Another important result is that once collapse occurred, repair is possible when applying HF chemistry, while de-ionized water rinses using a better drying technique do not restore the pattern. This finding proves that the deformation in pillar bending is elastic. Moreover it suggests a chemical bond formation between pillars under stiction similar as observed for particle ageing on silicon wafers.

Figure 2. Pattern collapse threshold curve for a specific cleaning


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.