|
Inverse technology for 30% better depth of focus |
|
|
|
Feb 28, 2007 at 02:27 PM |
 Luminescent Technologies unveiled the results of its most aggressive deployment of Inverse Lithography Technology (ILT) to date on complex 45nm and 32nm critical layers. One customer has already installed Luminescent's full-chip product, the Luminizer, for its 32nm development efforts. Another Luminizer order was received this month for a customer's 45nm development initiatives.
The latest data emerges from the evaluation of ILT by a selection of semiconductor manufacturers. The line-up includes a mix of foundries and integrated device (memory and logic) manufacturers on multiple continents using both dry and wet immersion lithography techniques; all are exploring ILT's potential to optimize their 45nm and 32nm IC designs for manufacturability.
Luminescent's ILT is an automated resolution enhancement technology (RET). It starts directly with the desired IC pattern on the wafer, explores the entire available optical lithography space by mathematical inversion, and aims to deliver a manufacturable mask pattern that generates the maximum design fidelity with the broadest possible process window.
Luminescent claims that its ILT approach can rapidly solve for the optimal manufacturable photomask design, with turn-around times comparable or superior to other mask synthesis solutions such as OPC that use simple heuristic rules to improve critical features (e.g. scattering bars). The ILT approach poses a functional analysis minimization problem and searches a subset of the possible masks for an optimal solution using iterative techniques and fast parallel processors.
ILT is designed to fit seamlessly into existing tape-out flows for current-generation 193nm lithography equipment to pattern 45nm and 32nm IC designs.
The ILT engine drives the company's Luminizer product family, an integrated hardware and software platform. The Luminizer family includes full-chip correction at nodes of 90nm down to 32nm (Luminizer FX) and small-area cell optimization (Luminizer LE).
According to Luminescent, ILT has demonstrated the ability to solve line-end shortening problems in poly and diffusion layers at 45nm. At this node, manufacturers are exploring the possibility of extending their dry steppers and delaying the move to immersion. With conventional OPC, printing critical layers such as poly and active has been extremely difficult due to the line-end shortening problem. The poly layer is especially complex since the line-end space is tight, leaving no room for OPC to correct the problem. ILT solves the line-end shortening problem, easily printing 45nm poly and diffusion layers with adequate process window, says the company.
Another reported improvement is an increase in depth of focus (DoF) by at least 100nm, which represents an improvement of more than 30% over conventional solutions, the company claims. Using conventional OPC, the best DoF tends to fall below 300nm, which leaves a very small process margin - i.e. everything has to be set just right. The improved DoF generated by ILT expands the process margin, thereby allowing manufacturers to proceed to production with confidence using reticles with currently available critical dimension (CD) controls.
Luminescent's ILT also considers mask-writing and inspection rules during the inversion calculation so the IC design is optimized for manufacturability. Manufacturers using ILT to tape out full masks are reporting mask-write time that is equal to if not better than conventional OPC and the full masks can be comprehensively inspected by current-generation mask-inspection equipment, the company reports.
Cost-of-ownership (CoO) benefits are reported in the form of accelerated turnaround time - with ILT, recipes are developed in days instead of weeks.
Luminescent CEO David Fried, comments: "Beyond the technical superiority of this new-generation RET solution, manufacturers are most excited by ILT's potential to enable the use of more-economical 193nm dry lithography instead of immersion for printing certain 45nm critical layers. They're also excited by the possibility of delaying the need for more advanced immersion steppers for 32nm designs."
By Dr Mike Cooke
|
News 
