Workshop 4

* If you encounter menus do not work upon clicking, delete your browser's cache.

Materials Introductions - A Path forward for All Devices

Organizer: David Thompson (Applied Materials)

Whether it’s moving to the next node of single digit logic, enabling a 3D DRAM embodiment, or bringing disruptive memristors, cryo or quantum based devices into mainstream applications, all paths forward ultimately rely on new material introductions. This workshop is intended to bring together leaders from the industry helping to tackle the myriad of challenges that are behind the inflections enabling the subsequent chipsets that will be the backbone of the next generation of compute. Whether the materials are enabling patterning, novel device physics or not so simply providing the right stress and strain mechanics to keep it all together this panel of speakers will give you perspectives on the effort required across multiple unit processes, modules, companies and industries for new materials introductions.

1. The Power and Limitations of Materials Solutions, John Smythe, Micron

Planning a new silicon technology node or architecture includes some form of evaluating which process modules or individual unit processes can be used again. The motivation is steeped in the desire to extend return on invested research and development expense as well as invested capital equipment. Alternative memory systems have the potential to disrupt established equipment segregation protocols. A robust business process is needed to provide a framework in which scientists and engineers can work to reduce the likelihood of the effects of trace elements, molecules or ligands on device and structural performance. Design, device, and structural requirements of future nodes are pushing the limits of persistent solutions. Examples will be used to illustrate the overhead in analytics, logistics and data management required to be successful within the limitations of available sample sizes in research and development.

2. Withdrawn
3. Materials and Mechanism for High NA EUV Litho, James Blackwell, Intel

Successful screening, optimization and eventual implementation of new high performance materials requires intimate chemical understanding of stability and mechanism of action. Using novel EUV resists as an example, this presentation will walk through how lab experiments and new analytical techniques can be efficiently and creatively used to understand and design better EUV resists for 300mm development.

4. Capital Equipment as the Bridge from Lab to Fab: High k Materials as an Object Lesson, Robert Clark, TEL Technology Center, America, LLC

Capital Equipment as the Bridge from Lab to Fab:
High k Materials as an Object Lesson
Robert Clark, Dina Triyoso, Steve Consiglio, David O’Meara, Kandabara Tapily, Cory Wajda and Gert Leusink
TEL Technology Center, America, LLC
More than 25 years ago device makers anticipated that silicon oxide and oxynitride gate dielectrics would reach their scaling limit in the foreseeable future and began looking for a replacement in earnest. The industry quickly focused on replacing silicon oxide with a polarizable metal oxide film because the dielectric constant (the k value) could be much higher in metal oxides than what could be obtained in the silicon oxide system. Eventually, the industry coalesced around hafnia and zirconia as likely candidates. This development effort encompassed academic laboratories, consortia, chemical vendors and equipment makers, all working to bring practical high k dielectrics into the industry. Initially hafnium oxide was incorporated into DRAM capacitors, providing a benefit in terms of leakage versus electrical thickness. At the 45nm node Hf-based high k dielectrics were implemented in high volume manufacturing for advanced logic devices as well. Since that time, the use of high k dielectrics has proliferated for advanced logic and memory production. As device structures changed, in some cases new precursors were adopted enabling higher aspect ratios with higher throughput and lower cost. And as these once exotic materials gained acceptance, new applications have emerged and been implemented. This talk provides a brief history of high k dielectric adoption to provide perspective on how new materials and chemistry can successfully make the transition from the lab to the fab, and then how new, unforeseen applications for those materials can arise and gain traction. In fact, the industry has gone back to the lab many times to alter, improve or adapt high k materials to new applications, and in each case capital equipment served as the bridge between the lab and the fab by providing the necessary control and processing capability for advanced high volume manufacturing.

5. New Devices from New Materials: Why We Need Them and How to Get Them, Greg Yeric, Cerfe Labs

The semiconductor industry has driven silicon MOS devices (logic transistors, SRAM, DRAM, NAND Flash) on an amazing decades-long period of advancement. Unfortunately, progress in all of these critical technologies is slowing at the very moment that the digital economy they enabled is beginning to consume non-trivial amounts of the global energy supply. Therefore, it is a global imperative that we accelerate the development of more efficient computing devices, which will need to be enabled by something other than “shrinking the MOS”. This moment puts novel materials development and integration at a point of critical importance. There is a great deal of promising R&D-level devices based on new materials, as this conference highlights. But is the development path from R&D labs to manufacturing fabs as efficient as it could be? How might answering this question fit into the renewed political emphasis on semiconductor technology? Drawing from specific experience in this area, Greg will offer his opinions.

6. Integrating New Materials for Device Fabrication, Ronald M. Pearlstein, EMD Electronics

The diversity of new electronic devices is being driven by a wide variety of end-use applications. As electronic device scaling continues to reach fundamental physical limits, manufacturers are turning to many more novel device structures – ushering in the age of materials. This paradigm shift now accentuates the central position of materials suppliers in the semiconductor ecosystem. The suppliers of deposition materials, cleaning and etching chemicals and the like must support this acceleration of innovation required by our industry. It can do so by generating and consolidating knowledge, working closely with Universities and other research consortia, and then formulating innovative solutions which must be developed in conjunction with device processing equipment manufacturers as well as the device makers. Several examples of how this is being practiced related to atomic scale processing and advanced metallization will be reviewed.

7. The Role of Academia in Identifying Compelling Devices and Materials, Sayeef Salahuddin, Univ. of California, Berkeley

Academy provides a flexible and ‘low-stake’ environment to explore new physics phenomena, materials and devices. I shall discuss this within the context of ferroelectricity that has been observed in ultra-thin films of binary oxides such as HfO2. We are witnessing an exciting time when fundamental understanding of the material physics is running in parallel with device level investigations. As appetite for functional devices increases in the industry, opportunities for new materials and devices are likely to increase. Academy will have to rethink as to how to train the next generation of semiconductor engineers that may need a wide and diverse set of skills. The industry can help by providing appropriate encouragement and excitement.

8. Theory and Modelling Guide to Materials Identification, Michael Haverty, Applied Materials

Through the 1980’s the semiconductor industry relied on just a handful of elements. Expansion across the periodic table started in the 90’s and accelerated into the 2000’s until we now cover nearly every corner of the periodic table. Starting in the late 90’s and early 2000’s materials and chemistry modeling started providing meaningful insights to accelerate the pre-screening and down-selection process of experimental trial and error. We broadly discuss Kinetic Monte Carlo, Molecular Dynamics, and Density Functional Theory modeling formalisms strengths and weaknesses. Beyond the model techniques’ theory however, a critical aspect to their effectiveness is the manner of application. The complex fabrication and integration of multiple layers of metals, semiconductors, and insulators results in materials interacting with many chemistries throughout processing. This necessitates modeling’s application focus less on discovering novel materials, and more on evaluating a range of candidate materials and their risks of interaction with other chemistries. Once the R&D funnel has been narrowed modeling’s focus often transitions into a design phase optimizing the atomic structure and composition to further shorten the iterative process. In this manner, the goal of modeling should not be Eureka! moments sidestepping experimentalists, but rather tight collaboration with experimental partners to shrink the width and length of the R&D funnel required to introduce a new material.