Recently, Interface Dipole Modulation (IDM) has been observed in amorphous HfO2/1-Monolayer (ML) TiOx/SiO2 stack structures [1]. To fabricate IDM devices, the thickness of the modulation layer, which is at the atomic layer level, must be precisely controlled to produce a uniform film with high-quality. Therefore, Atomic Layer Deposition (ALD) is one of the promising methods. However, it is difficult to form very thin oxide films on thermally oxidized SiO2/Si or GeO2/Ge by ALD because the oxide film surface is chemically inert. In fact, deposition delays, which is called incubation cycles, and island growth have been observed in the 0-20 cycle range for Al2O3 with Trimethylaluminum (TMA) and H2O as precursors [2, 3]. This is thought to depend on the time it takes for -OH bonds to form on the surface, so surface modification with low damage is necessary. Several surface modification methods exist, especially UV-Ozone (UVO) treatment [4], which has been reported to modify the surface of oxide films such as SiO2 and Al2O3 and has the advantage of being environmentally harmless and compatible with CMOS processes. In this study, the effect of surface modification of insulating films by the UVO method on ALD deposition was investigated for the purpose of controlling atomic layer thickness and producing a uniform oxide film.The samples for the measurement were prepared by the following procedure. First, the chemically cleaned Si or Ge substrate was thermally oxidized to form a ~10 nm thick oxide film. Next, after 15 min of UVO treatment in air at room temperature, the oxide films of Al2O3 or TiO2 were deposited using ALD method on two kinds of substrate. The precursor used for Al2O3 deposition and for TiO2 deposition were TMA and Tetrakis(dimethylamido)titanium (TDMAT), respectively. The oxidant is H2O. The samples were evaluated by Atomic Force Microscopy (AFM), Spectroscopic Ellipsometry (SE) and X-ray Photoelectron Spectroscopy (XPS).Al 2p photoelectron spectra from Al2O3 films deposited after UVO treatment were measured by XPS to confirm the composition of Al2O3. As shown in Fig. 1(a), fitting is possible with two peaks of Al 2p 3/2 and 2p 1/2, which can be said to be stoichiometric Al2O3. Next, Fig. 1(b) shows the photoelectron spectrum of the TiO2 film used as the modulation layer in the IDM device. The Ti 2p spectra, as same as the Al 2p spectra, can be fitted with two peaks, and the photoelectron intensity increases almost in proportion to the number of cycles. The thickness of the ALD film was then calculated from the photoelectron intensity ratio from the Al2O3, TiO2 and the lower oxide film. As shown in Fig. 1(c), there is a delay of one-cycle for the deposition of Al2O3 on SiO2, but it can be seen that the film thickness increases in proportion to the increase in the number of cycles. This is thought to have occurred by the following mechanism. First, UVO treatment breaks the Si-O or Ge-O bonds on SiO2 or GeO2 surface. This dangling bond is the starting point for precursor formation during initial layer formation. Therefore, the H2O during the ALD process is thought to have formed a Si-OH or Ge-OH bond and improved reactivity with methyl groups. The growth per cycle (GPC) can be calculated to be about 0.10 nm/cycle, which is in close agreement with the literature value [5]. Furthermore, SE measurements of Al2O3 films deposited on SiO2 at 50 and 200 cycles yielded thicknesses of 4.83 and 19.97 nm, respectively, which were in close agreement with the GPC calculated using XPS. The TiO2 film deposited on SiO2 also experienced a one-cycle delay similar to that of the Al2O3 film, and the GPC could be calculated to be 0.046 nm/cycle. Furthermore, as shown in Fig. 1(d), UVO treatment increased the average film thickness and reduced the surface roughness from the AFM images. These results indicate that UVO treatment is effective for surface modification of SiO2 and GeO2, and it can be concluded that the ALD method can deposit uniform oxide films without delay.This work was partly supported by TOKYO CITY UNIVERSITY Interdisciplinary Research Center for Nano Science and Technology for instrumental analysis.[1] N. Miyata, Sci. Rep., 8, pp. 8486, 2018.[2] H. Fukumizu et al., Jpn. J. Appl. Phys. 59 016504, 2020.[3] E. Shigesawa et al., Semicond. Sci. Technol. 33 124020, 2018.[4] D. Guo et al., Chem. Phys. Lett., 429, 124–128, 2006.[5] A.W. Ott et al., Thin Solid Films, 292, pp. 135-144, 1997. Figure 1