[Introduction]Requirements for high-dose ion-implanted photoresist stripping technology in the field of semiconductor manufacturing are to achieve with no residue, low damage and less chemical consumption. To realize these requirements simultaneously have been extremely difficult for decades. It is said that highly ion-implanted photoresist is highly carbonized and hardened, so it is extremely difficult to strip this. Traditional plasma ashing process tends to introduce a plasma damage, and batch wet process tends to cause much chemical consumption. Therefore, we tackled with this issue by employing single-wafer wet process system with high temperature SPM (Sulfuric acid and hydrogen Peroxide Mixture) process, and obtained interesting results.SPM produces a strong oxidant, peroxymonosulfuric acid, to decompose organic matter [1]. The reactivity of SPM with the photoresist increases at higher temperatures. However, at temperatures above 220°C, oxidant in the SPM becomes deactivated [2] and photoresist-stripping performance is said to be deteriorated, resulting in longer processing times, thereby increasing use of chemicals.Whereas, we have confirmed that a high-dose photoresist with a dose quantity of 1E16atoms/cm2 can be stripped successfully using SPM at a temperature of about 250°C. In light of that, we have evaluated the etching characteristics in the high temperature region and from the results, we have discussed the mechanism of the deactivation process of active species and photoresist strip performance.[Experimental procedure]The sample structure used are shown in Tab.1. We evaluated the etching characteristics considering amorphous-Carbon (a-Carbon) as a completely hardened form of photoresist layer. Tab.2 shows the chemicals that we used, Fig.1 shows outline of the experimental concept and Fig.2 shows the evaluation flow. The SPM was formed by mixing of H2SO4 and H2O2 in advance at room temperature, and then supplied them on a wafer, rapidly heated with a lamp heater, thereby confirming the temperature dependence of the etching rate. The thickness of a-Carbon film was measured before and after treatment with a spectroscopic ellipsometer (Onto Innovation Inc.: Atlas). And then we calculated the etching rate. Photoresist samples were observed with cross sectional SEM images (Hitachi High-Tech Corp.: S-5500) at an accelerating voltage of 10 kV. Accordingly, we checked for the presence of resist residue after treatment.[Results and Discussion]As shown in Fig.3, temperature dependence of the etching rate of the a-Carbon shows that increases from 165 to around 210 °C and then decreases at higher temperature. A concentration of H2SO4 included in SPM is about 81 wt%, and the boiling point of 81 wt% of the H2SO4-aq is about 209 °C [3]. When the boiling point of SPM is assumed same as that of H2SO4-aq, the temperature dependence of the etching rate of a-Carbon has a peak at boiling point. According to the Arrhenius plot of the etching rate of a-Carbon [Fig.4], a mechanism of the etching of a-Carbon is changed at the boiling point. We discuss the reaction mechanism at higher than 210 °C, the original boiling point of SPM, below. SPM is consisted of mixture of H2SO4, H2O2 and H2O. By heating up at 210 °C, first H2O is evaporated away and H2O2 will be start to be evaporated, and then the H2SO4 concentration will increase. It means that the boiling point of SPM will shift to higher temperature. Existence of H2SO4 and H2O2 is essential to form oxidant: peroxymonosulfuric acid that is delivered from SPM, at higher than 210 °C the concentration of the oxidant becomes lower. This will lead a decrease of the a-Carbon etching rate. At higher than 210 °C, a difference of actual data and the extrapolated plot from Arrhenius plot shows a concentration of the oxidant when all the oxidant contribute to the reaction. Temperature dependence of estimated concentration of the oxidant is plotted at Fig. 5. The concentration of the oxidant is estimated about 10 % at 250 °C. On the contrary, a good stripping performance shows at 250 °C; the oxidant concentration becomes very low. It is thought that the stripping performance of the high-dose photoresist is determined not only by the oxidant concentration, but also another factor. We will discuss in detail the mechanism of this at the presentation.[1] N. Hayamizu et al., Toshiba Review, 64, 38 (2009).[2] M. Uchida, Japanese Patent No. 5668914 (2014).[3] C.M. Gable, et al., J. Am. Chem. Soc., 72, 1445, (1950). Figure 1
Read full abstract
Sep 5, 2024
A Yu Sidorenko + 5
Open Access
