Corinne Miramond, Dominique Goubier & Michael Chomat, STMicroelectronics, Crolles, France; Yorick Trouiller, LETI-CEA, France; Yves Rody, Philips Semiconductors, France; Olivier Toublan, Mentor Graphics

ABSTRACT

The introduction of strong optical proximity correction (OPC) to process 0.13 μm designs and below has induced a new data-processing flow. This has been implemented at STMicroelectronics Crolles using a Mentor Graphics software suite. To deal more easily with model-based corrections and additional verification on critical layers, a separation of the design database into critical and non-critical layers has been introduced. The resist model and the correction parameters needed during the OPC processing are developed in an iterative way. File sizes and data-processing time are the main issues in mask data preparation. The impact on mask manufacturing is also addressed in this article.

Kay Lederer, Steffen Hornig & Ralf Schuster, Infineon Technologies SC300, Dresden, Germany 

ABSTRACT 

In this paper, we discuss implications currently relevant to 300mm lithography resist processing. The larger wafer size has highlighted a range of resist process issues that were not encountered at the smaller wafer sizes. In particular, resist thickness fluctuations around the edge area of a silicon wafer typically occur if high spin speeds are applied during the coating process. 300mm coating processes are particularly prone to the occurrence of such spin marks as the operating range of applicable spin speeds is lower in comparison to processes applied to smaller sized wafers. We show in our example how such local resist thickness fluctuations impact the product yield. The paper will also show how we investigated a novel resist development process based on a new developer dispense nozzle to combat this problem.

Kiril A. Pandelisev, Single Crystal Technologies, Gilbert, AZ, USA 

ABSTRACT 

The Bridgman Method [1] has undergone many modifications to make it suitable for growth of various crystals for the semiconductor and optical industries. Growth of Calcium Fluoride (CaF₂) for 193nm and 157nm technology has long been contemplated. Many hundreds of millions of dollars have been spent to adapt this method for CaF₂ in order to grow crystals having suitable quality for 193nm and 157nm technology [2]. Nevertheless, the reported crystal yield is still very, very low, in the single digits [3]. This is a very small improvement to the digit fraction (less than 1%) reported by Nestor [4] over ten years ago. At this pace of Bridgman method crystal lens development, one might question the further progress and practicability of this route in the semiconductor industry. Sidestepping CaF₂ by using fused silica may suffice for 193nm technology, but it will not serve for 157nm applications. A new crystal lens fabrication method is required.

Ernst Richter, Michael Sebald, Linda Chen, Günther Schmid & Günther Czech, Infineon Technologies AG, Erlangen, Germany

ABSTRACT

The process of chemical amplification of resist lines (CARL) is outlined in this article. An improvement in depth of focus in lithography operations is achieved leading to improved resolution of features in ICs. This new concept is already being implemented in some production environments. 

Hans C. Pfeiffer & John Hartley, US IBM, Hopewell Junction, NY, USA

ABSTRACT

Feature reduction and pattern shape count have outpaced Moore's Law. Shaped-beam lithography offers the possibility of cost-effective mask-making. A new mask writer that has recently completed pre-integration testing is described. Performance specifications and projected throughput are given. The mask writer has a number of novel features. 

Kiril A. Pandelisev, Single Crystal Technologies, Gilbert, AZ, USA

Materials of poor heat conducting properties and poor starting material quality, such as Calcium Fluoride, have severe constraints on the usable material produced from each crystal grown in a Bridgman-Stockbarger type crystal grower. By using new purification techniques together with crystal growth method for plate or slab form, SCT has obtained ways for production of Calcium Fluoride that have far superior yield and quality. The new methods for material purification and crystal plate growth and their advantages over the Bridgman-Stockbarger method are discussed.

John H. Burnett, Zachary H. Levine & Eric L. Shirley, National Institute of Standards and Technology, Gaithersburg, MD, USA

ABSTRACT

Calcium fluoride and other crystalline fluoride materials are being exploited for latest generation lithography optics, making up a significant component of the optics of 193 nm lithography systems and being potentially the exclusive optical materials for 157 nm systems. Improvements in the crystalline material quality seemed to have brought the most troubling material property, stress-induced birefringence, into specification.

Roger H. French, Jerald Feldman, Fredrick C. Zumsteg, Michael K. Crawford, Andrew E. Feiring, Viacheslav A. Petrov, Frank L. Schadt III & Robert C. Wheland, DuPont Co., iTechnologies and Central Research and Development, Wilmington, DE, USA; Joseph Gordon & Edward Zhang, DuPont Photomasks Inc., Danbury, CT, USA

Substantial progress has been made in developing novel fluoropolymer materials for 157nm lithography. Materials with sufficient transparency at 157 nm to enable both thick single layer resists and high transmission pellicle membranes have been demonstrated. We have shown that tetrafluoroethylene (TFE)-containing 157 nm photoresist binder resins can be made that are sufficiently transparent to be used at film thickness greater than 200nm, have good photosensitivity, exhibit low outgassing upon exposure, are compatible with aqueous base development, and have etch rates comparable to PHOST resins. Optical absorbance on the order of 1.5/µm @ 157 nm should be possible for photoresits.

Geert Vandenberghe, Young-Chang Kim, Christie Delvaux, Kevin Lucas, Sang-Jun Choi, Monique Ercken & Kurt Ronse, IMEC, Leuven, Belgium; Bert Vleeming, ASML, Veldhoven, Holland

As ArF resists mature, lithographers are pushing the imaging limits as far as possible. As shown earlier, ArF lithography is getting ready for the 130nm technology node and currently even the 100nm node printability with ArF is being studied. Since high numerical aperture (NA) ArF scanners are not yet available in volume, strong enhancement techniques will be required to meet these challenging targets at lower NA (0.63NA). In this paper we give an overview of the status of 193nm lithography towards 100nm patterning of memory and logic front-end features, and explore the various enhancement techniques needed.

Jan-Willem Gemmink, Philips Semiconductors, Nijmegen, The Netherlands; Peter Zandbergen, Philips Semiconductors, Kapeldreef, Leuven, Belgium

ABSTRACT

As the ITRS roadmap confirms its accelerated pace of two years per technology node, the status and maturity of the lithography solutions upon introduction of the nodes is influenced. Though on paper the technological solutions on individual domains as mask, resist and tools are ready just in time, integration into full process flowcharts, however, is not thoroughly addressed. In this paper we will address the impact of this shrinking time window on the focus areas going from basic R&D of process step concepts
down to release of the intended solutions in process technology at product sampling. In the domain of SOC manufacturing this also implies modifications in process solutions that have to come after initial introduction of a technology. In addition we indicate how we see that the global programs as executed in various pre-competitive consortia by themselves not necessarily prepare for a smooth integration of technologies such as 157 nm or EUV lithography into manufacturing situations. We give some examples of such less
considered challenges that have to be addressed by device makers to assure robust and proven manufacturing boundaries for the use of these challenging lithography technologies.