As the key enabling technology for semiconductor fabrication process, photolithography are behind almost every critical turn of the semiconductor industry. Process development and photoresist material development, as two pillars of photolithography technology, have shared equal importance since the industry inception. Process development such as immersion 193nm photolithography, wafer size increase from 4” to the current 12” process, fan-out wafer level packaging, etc. have undoubtedly changed the industry from many aspects. New ongoing process development such as EUV photolithography, 450mm wafer process, panel-level packaging, etc. will certainly continue driving the industry on the road to future. However, process development/innovation is not the focus of this blog. Here I would like to discuss photoresist material development from historic perspective.
As we walk into the history, I will share some stories and interesting events that led to a number of important photoresists. These photoresists are truly the “invisible hand” that drives the Moore’s Law. They have led our society to the current mobile computing age and the digital revolution. They have changes every aspect of our lives. For those who haven't followed photolithography technology closely, we have experienced four major generations of photoresists up to date. I expect to take some time to write up on this subject so I will split it into the next few blogs: Part 1: Preface; Part 2: The First Generation of KTFR Material; Part 3: The Age of Novolak/DNQ; Part 4: DUV 248nm Photoresists; Part 5: 193nm Photoresists; Part 6: Summary and Outlook.
Firstly, I will share with you how photoresists are categorized. Photoresists are classified by different methods. First, according to the imaging/development mechanism, all photoresists can be classified as negative- and positive-tone. Some photoresists can be both negative- and positive tone when the development condition is changed; Second, according to exposure light wavelength, photoresists can be classified as g-line, I-line, DUV@248nm (KrF laser), DUV@193nm (ArF laser), 157nm (skipped for commercialization) , EUV 13.5 nm (currently under active development), and electron-beam resists. Many photoresists could work at a broad band of wavelength. But most of them are used in a defined narrow wavelength for the purpose of better patterning capability; Third, photoresist are also defined as chemically amplified and non-chemically amplified. In early days, all photoresists were non-chemically amplified. With the application of shorter wavelength of exposure light source in 1980’s, the light power (number of photons available for reaction) is much less than the previous generations of I-line/G-line light sources. New photoresists based on chemical amplification were developed in the context; Fourth, according to the developer used, all photoresists can be classified as solvent-developable and aqueous-developable photoresists. The line between the two is obscure sometimes because some photoresists could be developed in both types of developer solutions.
Secondly, for any practical photoresist, it has to show excellence in certain material properties to meet various technical requirements. Photoresist materials need to undergo solubility change upon exposure of radiation (with aid of post-exposure bake for the case of chemically amplified photoresists), adhere to substrate, possess etch resistance, and striping capability. Of course, as photo-imaging materials, photoresists also need to show merits in the following aspects including resolution (CD), sensitivity, process latitude, viscosity, formulation stability/shelf life, material swelling, heat resistance, stress, etc. All of these have showed how complex it could be to develop a new commercial photoresist. There were many examples that some initial wonder materials failed as practical photoresists due to lacking of merit in one or two critical properties.
In the next few blogs, we will uncover how the material development has contributed to the success of semiconductor industry for the past 60 years. Hopefully, reexamining these significant material development events and milestones would provide us some insights and guidance for future new material development.