Summer 2008
Back to Front Page
Previous Article | Next Article

CEA-LETI and Nikon Update on Double Patterning Program

Speaking at the LithoVision technical symposium, Dr. Laurent Pain, Lithography Lab Manager at CEA-LETI-Minatec, provided an overview of the double patterning (DP) joint development program (JDP) between LETI and Nikon, which has been underway since early 2007. This work is being performed using an NSR-S307E (0.85 NA) dry ArF scanner in LETI’s Nanotec 300 research facility in Grenoble, France.

Dr. Pain reported on the study of topography impacts on double patterning lithography, noting that higher topography at the wafer level during the second exposure induces larger CDs and increased CD nonuniformity. He commented that litho-etch-litho-etch (LELE) or litho-freeze-litho-etch (LFLE) DP applications necessitate development of advanced bottom anti-reflective coatings (BARC) and use of the thinnest stack possible to minimize the impact of topography.

Pain also discussed the status of efforts to verify the Nikon DP error models. At LithoVision 2008, Andrew Hazelton from Nikon Corporation had presented the Nikon theoretical models that would be used to establish CD and overlay requirements for DP system design. Pain announced that both the Nikon line and space error models have been validated experimentally at LETI.

Figure 1.  The Nikon line and space error models used to establish CD and overlay requirements for DP system design were validated experimentally at LETI. Click image to enlarge. Figure 1.  The Nikon line and space error models used to establish CD and overlay requirements for DP system design were validated experimentally at LETI. Click image to enlarge.

Figure 1. The Nikon line and space error models used to establish CD and overlay requirements for DP system design were validated experimentally at LETI.

A major portion of the Nikon error model budgets was due to mask related components, therefore reticle and pellicle impacts on CD uniformity/overlay were investigated through the JDP. Pain reported that CD mean-to-target and image placement on the reticle were determined for 45 nm and 32 nm features, with results close to the 32 nm specifications for CD mean-to-target, and the image placement data predominantly below 32 nm ITRS mask specifications. Equivalent measurements were then performed on the wafer, and Pain highlighted that mask CD error was < 14 the wafer CDU after the first litho step, with mask overlay error representing < 10% of the wafer overlay budget.

Figure 2.  Mask CD error was < 1⁄4 the wafer CDU after the first litho step, with mask overlay error representing < 10% of the wafer overlay budget. Click image to enlarge.

Figure 2. Mask CD error was < 14 the wafer CDU after the first litho step, with mask overlay error representing < 10% of the wafer overlay budget.

Pain also reported that mechanical performance of the pellicle may introduce mask distortion resulting in pattern placement error, and showed pellicle distortion data x = 3.2 and y = 2.4 nm (3σ) after isotropic and orthotropic corrections. He cautioned that this pellicle-induced distortion represented 15% additional error in the image placement budget for the 45 nm node.

Concluding his presentation, Pain announced that CEA-LETI-Minatec had developed a DP process using a custom illumination solution and a 0.85 NA dry ArF Nikon scanner that enabled 32 nm patterning. They demonstrated a k1 of 0.14 using double patterning, which showed the potential extendibility for dry ArF tools.

Figure 3.  A 0.85 NA dry scanner was used to demonstrate a k1 of 0.14 with double patterning. Click image to enlarge.

Figure 3. A 0.85 NA dry scanner was used to demonstrate a k1 of 0.14 with double patterning.

View Full Presentation

Mr. Masato Hamatani, General Manager of Nikon’s Stepper Mechanical Design Department, followed Pain’s presentation with an update on the NSR-S620 DP scanner. Mr. Hamatani reported that the S620 will deliver dramatically reduced overlay to satisfy the aggressive requirements for double patterning, coupled with ultra-high throughput to ensure superior cost of ownership. He also described the improved S620 system design, which streamlines installation and provides optimal flexibility. By further improving upon Nikon Local Fill Technology and the Tandem Stage, and incorporating them within a new platform that provides increased precision and higher throughput, the NSR-S620 will fully satisfy customers' mass production double patterning requirements. Hamatani concluded his presentation announcing that the S620 will be available beginning Q4 2009.

Figure 4.  The NSR-S620 ultra-high throughput immersion scanner for double patterning applications will be available beginning Q4 2009.

Figure 4. The NSR-S620 ultra-high throughput immersion scanner for double patterning applications will be available beginning Q4 2009.

Back to Front Page
Previous Article | Next Article

©2009. Nikon Precision, Inc. All rights reserved. Vist www.nikonprecision.com