The focus of this chapter is on fundamental atomic data; other sorts of fundamental data (related to the erosion of surfaces by plasma ions, for example) may be needed as well. The key aspects of plasma modeling are shown in a block diagram in Fig. 3.1. Accurate fundamental input data (the left block of Fig. 3.1) are crucial to the success of this overall scheme. Toward the end of this chapter, the discussion turns to benchmarking the atomic physics aspects of the plasma model (right block of Fig. 3.1). Other chapters in this volume discuss the other blocks.
The necessary fundamental atomic data sets are typically very large because of the huge number (infinite, actually) of internal quantum states that even a single atom presents. By one estimate, the subset of the atomic data set needed to accurately model an EUVL plasma is on the order of 1 million numbers (including the ionization energies, transition energies, and a large variety of cross sections). Due to the sheer magnitude of the problem, most of these data must be calculated rather than determined experimentally.
Because the EUVL plasmas are so hot that they produce highly charged ions (HCIs), and because few labs are equipped to study the atomic physics of HCIs, the reliability of the huge input data set (what I have called type I data) is largely untested. Therefore, it is important not only to begin testing as many of the input data as is practical, but also to jump ahead and begin benchmark-testing the predicted output spectra (what I have called type II data). These type II data can be used to assess the accuracy of the overall plasma model (including the accuracy and completeness of the type I input data). Both of these types of benchmarks are discussed further below.
In order to assess the availability of relevant data and to evaluate the issues discussed in this chapter, International SEMATECH formed a Fundamental Data Working Group (ISMT FDWG) in 2003. Since that time, this group has held a number of deliberations, and their input helped form the substance of this chapter. Notice of significant omissions would be gladly welcomed by the author. The membership of the ISMT FDWG is listed in Appendix A.
Online access to SPIE eBooks is limited to subscribing institutions.