Introduction Inspired by Okuyama et al. [1], fundamental studies on the wet-etching behavior of SiO2 films in single-nanometer scale gaps between silicon walls play an extremely significant role in optimizing the manufacturing of advanced semiconductors with complex 3D nanostructures. A well-known explanation for SiO2 etching behavior in narrow spaces is the change in the reactant concentration. Various attempts have been made to supply reactants to narrow spaces [2-4]. However, the wet-etching phenomenon in narrow spaces remains unclear, and further elucidation of the underlying mechanism is required. In this study, the etching behavior of SiO2 films confined between two silicon walls was investigated using a reactant consumption model. We attempted to suppress the reactions on the wall surface using the deuterium isotope effect, which is often used to control reactions [5]. Experimental Methods and Discussion First, we investigated the differences between hydrogen fluoride and deuterium fluoride in reactions on silicon surfaces. Silicon prisms machined from silicon wafers were prepared. The proportion of Si–H bonds on the prism was measured by infrared spectroscopy in multiple internal reflection geometry (MIR-IRAS) after immersing the silicon prisms in a mixture of dilute hydrogen fluoride, D2O, and deionized water for 2 min. The Si–H bond ratio on the prism surface increased with the increase in the proportion of hydrogen in the mixture, observed as the upward concave profile shown in Fig. 1. This suggests that deuterium fluoride reacts more slowly with the silicon surface than hydrogen fluoride, i.e., dilute deuterium fluoride reduces the reactant consumption on the silicon surface compared to dilute hydrogen fluoride. In dilute deuterium fluoride, the hydrogen fluoride is diluted with sufficient D2O such that hydrogen becomes negligible.Second, to examine the influence of the reactant consumption by the silicon walls, we measured the etching rates of SiO2 films with different film thicknesses sandwiched between two silicon layers. We performed etching tests on coupon samples in beakers with dilute hydrogen fluoride or dilute deuterium fluoride (w/o additive) at room temperature. The etching amounts of the confined SiO2 film and blanket SiO2 film as a control were measured by cross-sectional scanning electron microscopy (X-SEM), transmission electron microscopy with a focused ion beam (FIB-TEM), and spectroscopic film measurements. The etching rate was calculated by dividing the etched amount by the etching time. The etching rate in the narrow space was normalized to the reference etching rate. It was found that the etching rate in dilute deuterium fluoride is higher than that in hydrogen fluoride without deuterium fluoride (Fig 2(a)). Moreover, we performed an etching test on coupons with dilute hydrogen fluoride or dilute deuterium fluoride with an acidic additive (w/acid) and discovered that the etching rate in the narrow space was increased upon acid addition. Summary In this study, we verified the effect of deuterium fluoride on the etching behavior in nanometer-sized narrow spaces. We discovered that deuterium fluoride increased the etching rate in narrow spaces compared with hydrogen fluoride. We attribute these advantages of deuterium fluoride in the SiO2 etching rate in the narrow space to the consumption control of the reactant. At the PRiME 2024 conference, we would like to discuss our study and the effects of additives on the etching effectiveness of deuterium fluoride.REFERENCE:[1] A. Okuyama, S. Saito, Y. Hagimoto, K. Nishi, A. Suzuki, T. Toshima and H. Iwamoto, Solid State Phenomena, 2019 (2015) 115-118.[2] D. Ueda, Y. Hanawa H. Kitagawa, N. Fujiwara, M. Otsuji, H. Takahashi and K. Fukami, Solid State Phenomena, 314 (2021) 155-160.[3] S. Kumari, Shan. Hu and P. D’elia, Solid State Phenomena, 346 (2023) 149-154.[4] G. Vereecke, H. Debruyn, Q. D. Keyser, R. Vos, A. Dutta and F. Holsteys, Solid State Phenomena, 282 (2018) 182-189.[5] Kenneth B. Wiberg, Chemical Reviews, 55 (4) (1955) 713-743. Figure 1
Read full abstract