Characterization and Modeling of Out-Diffusion of Manganese and Zinc Impurities from Deep Ultraviolet Photoresist
Journal of the Electrochemical Society
The increasing complexity and miniaturization of modern integrated circuits (ICs) demand a higher device yield and hence lower defect density in the active region of silicon devices. 1 For a deepsubmicrometer device, a single metal precipitate could cause a distortion of electrical properties and result in a faulty IC.     For example, metals can degrade the dielectric properties of gate oxide, causing premature breakdown. They can also diffuse into the bulk of the silicon
... e silicon substrate, resulting in increased junction leakage as well as reduced minority carrier lifetime. Therefore, a better knowledge of the diffusion route and behavior of impurities introduced into the silicon substrate during device fabrication could help to control contamination and thus promote circuit yield. The control of fabrication processes involved in device manufacturing becomes more and more crucial due to increasing complexity of the materials and tools. Among them, lithography plays a very important role because it is applied repeatedly to the wafer surface during device manufacturing. Many lithographic models have been developed and applied for the photoresist profile simulation, 6-8 however, to the best of our knowledge, the effects of impurity segregation and diffusion of Mn and Zn impurities in the deep ultraviolet (DUV) photoresist have not yet been reported. The behavior of metallic contaminants in DUV photoresist and their migration into the underlying substrate is of eminent importance, because proper control of elemental impurity levels in the "high-purity" DUV photoresist would be very crucial for fast proliferation of DUV photoresist in the IC industry in the coming years. 9 Although it is always safer to demand an unnecessarily high degree of purity for DUV photoresist (i.e., 50-100 ppb for i-line and 10-30 ppb for DUV photoresist), the resulting cost would be extremely high and unnecessary. As this would require tedious sample pretreatment procedures, including matrix decomposition by hot plate or microwave oven, separation, preconcentration, and control of sample handling and analytical environment. 10-12 Therefore, a systematic study of the impurity diffusion behavior from DUV photoresist into the underlying substrate is technologically important and urgent from the viewpoint of setting the proper standard level of tolerable impurities in DUV photoresist for ultralarge-scale integrated (ULSI) industrial application. To evaluate the ratios of metallic impurities diffusing from DUV photoresist into the underlying substrate during lithographic processing, e.g., baking, various analytical methods are required for determining the impurities in the resist layer and in the underlying sub-strate. 13 Methods that have been developed for detecting Mn and Zn impurities include graphite furnace atomic absorption spectrometry (GFAAS), inductively coupled plasma mass spectrometry (ICP-MS), total reflection X-ray fluorescence (TXRF) spectrometry, and secondary ion mass spectrometry (SIMS). 14-17 However, these methods suffer from the shortcomings 11,18,19 of the need to develop the decomposition method for the photoresist. They also require high sensitivity for the instrument and are cumbersome in determining spectroscopic interference. In addition, the required blank control can seriously affect the analytical reliability during measurement. 20, 21 The radioactive tracer technique 22-24 has proved very suitable for studying element migration in materials, environment, and biochemistry. In semiconductor fabrication, the well-known RCA cleaning method, 25 universally applied in wafer processing for removing impurities from silicon wafers, was originally developed based on the radioactive tracer method by Kern and Puotinen in 1970. It is still in wide use today and forms the basis of many so-called modified RCA cleaning methods for wafer cleaning. The effectiveness of the RCA cleaning method was first established by deliberately contaminating the wafer surface with radioactive nuclides of 64 Cu and 198 Au. Gamma radioactivity was then recorded before and after RCA cleaning to monitor cleaning efficiency. The advantages of the radioactive tracer technique are well recognized; they include high throughput, easy operation, interference-free from stable isotope, and reliability. Despite the versatility of the radioactive tracer technique, it has not yet been applied to study the impurity behavior during lithographic and subsequent cleaning processes. In this work, the radioactive tracer method is first introduced to investigate the migration ratio of Mn and Zn impurities from DUV photoresist into the underlying substrate. The effects of various baking temperatures are evaluated. Furthermore, a diffusion model, together with a newly proposed parameter, K, is suggested to describe the impurity diffusion in the photoresist layer and its subsequent diffusion into the underlying substrate. In addition, by using the same radioactive tracer method, the impurity cleaning efficiency of various wet chemical cleaning recipes is also evaluated in this work. Possible mechanisms relevant to impurity diffusion associated with various underlying substrates are discussed. Experimental Materials.-p-Type <100> wafers with 15 cm diam were passivated with various films (i.e., polysilicon, silicon dioxide, silicon nitride, and nonpassivated or bare silicon control). They were cut into 2 ϫ 2 cm pieces to serve as test samples. The samples were then The radioactive tracer technique was applied to investigate the out-diffusion of manganese and zinc impurities from deep-ultraviolet (DUV) photoresist. Two important process parameters, viz., baking temperature and the type of substrate (i.e., bare silicon, polysilicon, oxide, or nitride), were evaluated. Our results indicated that diffusion ratios were all below 6%, irrespective of the substrate type and baking temperature. The substrate type did not appear to strongly affect the metallic impurity out-diffusion from DUV photoresist. However, solvent evaporation was found to have a significant effect on impurity diffusion. A new model, together with a new parameter, was proposed to describe the out-diffusion behavior of impurities from DUV photoresist. This model could explain the diffusion ratio of metallic impurities in photoresist layers under various baking conditions. The effectiveness of various wet cleaning recipes in removing metallic impurities such as manganese and zinc was also studied. It was found that (i) bath life due to temperature change can considerably affect the cleaning efficiency, and (ii) hot water immersion can effectively dissolve the impurities from the wafer surface.