Prototyping Permanent Data Storage Elements with Electron Beam Lithography Kevin Laughlin 1 , Hao Wang 2 , Barry M. Lunt 3 , Robert C. Davis 1 , Matthew R. Linford 2 1 Department of Physics and Astronomy, Brigham Young University, Provo, UT 84602 2 Department of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84602 3 Department of Information Technology, Brigham Young University, Provo, UT 84602 Introduction Most data today, including pictures, videos, documents, technical data, etc., is stored digitally. Much of this information is of great importance both to individuals, e.g., pictures and video, and to governmental organizations and corporations, e.g., documents and technical data. Unfortunately, almost all of the data storage options available today show high degrees of volatility – they are for the most part ephemeral, lasting in general only a few years to about a decade. 1 Our group has been actively working in this area to develop new materials and data storage options that will offer greater permanence. In particular, we have recently been developing permanent solid-state storage devices that use nanofuses as the basic storage elements. Clearly, to be competitive with current data storage densities, features sizes will need to be around those in current Flash technology. Accordingly, tools are need for prototyping at these dimensions. Here we will present electron beam lithography as an effective patterning/prototyping tool for this kind of work, and then describe its use in the fabrication of contact pads and carbon nanofuses for permanent solid-state storage devices. Electron Beam Lithography Traditional semiconductor manufacturing employs advanced lithographic technology that can now produce ca. 20 nm features. However, the specialized equipment, masks, cleanrooms, and other accessories needed to work at these dimension, and also to produce large numbers of devices, are extremely expensive. In contrast, a university researcher will, in general, be much more interested in ‘proof of concept’ experiments that may only require a handful of different prototypes. In this environment, a premium is placed on flexible equipment that can make many different types of structures. In addition, the university researcher will often be able to tolerate larger feature sizes, less precise placement of those features, and slower write/development speeds. Electron beam lithography (EBL) is an advanced materials patterning tool that fits this role. For example, while it is rather slow, it also requires no mask, which allows different patterns to be directly made. It can also pattern very large areas – on the order a micron – or very small ones – down to tens of nanometers. The physics behind EBL hinges on the de Broglie wavelength, Ȝ, of the electron, which is inversely proportional to its momentum, pμ Ȝ = h/p. (Here, h is Planck’s constant.) Thus, a higher accelerating voltage on the electron gives it greater momentum, which in turn gives it a shorter wavelength. Shorter wavelengths reduce diffraction effects and allow smaller features to be patterned. For example a 30 KeV electron will have a de Broglie wavelength less than 1 nm. Fabrication of Contact Pads Because the nanofuses we are currently investigating as candidates for our permanent solid-state storage devices are much smaller than our micromanipulators, contact pads are required for making electrical contact to them. We make these pads as follows. We begin with ca. 2 cm x 2 cm shards of silicon terminated with ca. 500 ȝm of oxide. Obviously the thick oxide removes any conducting path through the substrate. These pieces of silicon are spin coated with ca. 600 nm of an electron beam resist (ZEP520A) (ZEP). This resist-coated sample is placed on a hot plate at ca. 180 °C for 2 min. The sample is then mounted on a special stand that also has a colloidal gold stigmation sample and a Faraday cup. Scratches are placed around the sample edges for planar alignment correction. The stand with the sample in it is then placed in an FEI XL30 ESEM and pumped down to 5 ∗ ͳͲ −ହ Torr. The colloidal gold sample allows us to focus and correct for stigmation, which would cause the electron beam to become elliptical instead of