PacSurf2024 Session TF-WeP: Thin Films and Surface Modification Poster Session II

The (113) LiNbO₃ layers were grown onto the (111) Pt/SiO₂/Si substrates at 180 – 320 °C by radio-frequency magnetron sputtering. The samples grown at 250 °C displayed the improved crystallinity as well as the smooth surface morphology without any hillocks and pits. The memristive devices, fabricated in the form of the top-to-bottom Au/LiNbO₃/Pt two-terminal device scheme, clearly exhibited the external electric field polarity-dependent asymmetric memristive hysteresis loops in their current-voltage characteristic curves. When repeating the current-voltage sweep at an appropriate program voltage range, the on-state current was gradually increased with increasing sweep number. Through analyzing the transport mechanism in Au/LiNbO₃/Pt, such a behavior was confirmed to be attributable to the Schottky barrier modulation, arising from the ionic migration of oxygen vacancies inside the LiNbO₃ layer. In other words, the electro-migrated oxygen vacancies in LiNbO₃ lead to the Schottky barrier modulation particularly at the LiNbO₃/Pt side; and it eventually gives rise to the switchable diode effect in the Au/LiNbO₃/Pt device. Since the degree of the switchable diode effect relies on the pulse parameters of the applied voltage stresses, the memristive characteristics (e.g., data storage speed, multiple resistance states, data retention, etc.) could be effectively controlled by changing the pulse magnitude and the pulse duration of the program/erase voltages. Using these unique features, various synaptic functions such as a short-term memory, long-term potentiation/depression, and spike-timing dependent plasticity were effectively demonstrated. The results suggest that the LiNbO₃ based memristors hold great promise for the future neuromorphic applications.

[1] J. Wang et al., Adv. Electron. Mater. 9, 2201064 (2023)

[2] J. Wang et al., Adv. Intell. Syst. 5, 2300155 (2023)

The ever-growing problem in modern science is that data is being collected at a rate faster than analysis can be performed by characterization experts. The analysis that is done for many recently published x-ray photoelectron spectroscopy (XPS) and x-ray emission spectroscopy (XES) data, is often incorrect/irreproducible and leads to a cycle of incorrect fits in this spectroscopy data. In this work, a genetic algorithm (GA) is being constructed to potentially minimize this human error. This GA code known as XPS Neo/XES Neo, is based on the Neo package which exists for EXAFS (EXAFS Neo) and Nanoindentation (Nano Neo) data. GAs are based on biological methods and depend on parameters such as populations size, number of generations, genes, crossover, and mutation. The GA takes in a certain population size and constructs individual vectors each with their own unique genes i.e the fitting parameters we are trying to optimize. It then performs crossover and mutation to these individual vectors to progress toward a lower global minimum. This GA allows for a variety of mutation options including, Random Perturbations, Rechenberg, Metropolis mutation, and Self Adapting Differential Evolution. The methods of how a GA works in relation to XPS and XES datasets are discussed. The difficulties in making this work for XPS data arise from complicated backgrounds due to many effects such as plasmon loss, Auger peaks, and satellite peaks. XPS and XES data can also have many peaks that are difficult to distinguish from one another. Currently, the algorithm is only able to fit simplistic XPS spectra such C, O, N, and Si and is being worked on with the hopes of it becoming applicable for more difficult data. The goal is to make the algorithm applicable to all XPS data, with a greater focus given to the actinides, specifically for the use of fitting plutonium data as analysis of this spectra is highly sought after and difficult to fit. Through proper use of an informed GA, and collaboration with the XPS/XES database website XPSOasis.org, theoretically correct fitting of this data is hoped to be achieved.

Molecular clouds are composed of gases and interstellar dust grains. The dust grains are covered with ice mantles predominantly composed of H₂O [1]. When the densities of the clouds increase up to 10⁴ cm⁻³ or more and their temperatures drop below 20 K, CO is condensed on H₂O-rich ice. The ice mantles are exposed to ultraviolet (UV) radiation, which causes a variety of photochemical reactions. CO₂ is one of the abundant molecules in the ice mantles, and the following two formation channels have been discussed [2]: (1) reaction of two CO molecules, one of which is electronically excited by UV light; (2) reaction of CO with the OH radical which is a dissociation product of H₂O. We focused on the fact that these two channels can be distinguished from each other by isotope labeling, i.e., by employing H₂¹⁸O instead of H₂¹⁶O. In the present study, we investigated the UV photolysis of CO-H₂O ice and determined the effective rates of the CO₂ formation channels with use of isotope labeling and infrared spectroscopy [3].

A CO gas and an H₂¹⁸O vapor were mixed in a gas handling system. The mixing ratio was changed in the range of CO/H₂¹⁸O = 1000-0.1. The gaseous mixture was introduced into an ultrahigh vacuum chamber and was condensed on a gold substrate cooled down to 10 K. After the condensation, the CO-H₂¹⁸O ice was irradiated with UV light from a deuterium lamp for 120 min. Reflection-absorption infrared spectra were recorded during the condensation and UV-irradiation.

After the UV irradiation of any sample, we detected infrared absorption bands of C¹⁶O₂ (2346 cm^-1) and C¹⁸O¹⁶O (2328 cm^-1), which were generated through the CO–CO and CO–H₂O reactions, respectively. The absorption band of C¹⁸O₂ was also detected at 2308 cm^-1 when the mixing ratio was in the range of CO/H₂¹⁸O = 100-0.1. This indicates that the photodissociation and regeneration of CO₂ took place in ice. We analyzed the irradiation-time dependence of the C¹⁶O₂ and C¹⁸O¹⁶O column densities to determine the effective cross sections of the CO₂ formation through the CO–CO and CO–H₂O reactions simultaneously.

References

[1]E. L. Gibb et al., Astrophys. J.Suppl. Ser. 151, 35 (2004).

[2]N. Watanabe and A. Kouchi, Astrophys. J. 567, 651 (2002).

[3]A. Hirayama, I. Arakawa, and K. Yamakawa, Astrophys. J. 951, 132 (2023).

Two-dimensional transition metal dichalcogenides (TMDs), particularly MoS₂ and WS₂, are regarded as potential substitutes for graphene in electronic devices due to their semiconducting properties, exhibiting both indirect and direct band gaps depending on the layer. Various methods have been employed in the synthesis of MoS₂ and WS₂ thin films. In chemical vapor deposition (CVD) and atomic layer deposition (ALD) processes, metal precursors are typically used in conjunction with a sulfur source, such as hydrogen sulfide (H₂S), sulfur powder, and so forth. In the majority of cases, the synthesis of crystalline MoS₂ has been achieved at elevated deposition temperatures, with the subsequent treatment at high temperatures being necessary for the attainment of crystallinity. In this study, precursors suitable for the deposition of MoS₂ and WS₂ were synthesized, characterized, and even MoS₂ deposition studies with molybdenum precursors were conducted at relatively low temperatures without any post-treatment process. Precursors (Mo(NtBu)₂(StBu)₂ (1), W(NtBu)₂(StBu)₂ (2)) suitable for the deposition of MoS₂ and WS₂ were synthesized and characterized. The molecular structures of 1 and 2 exhibit a tetrahedral geometry according to single-crystal x-ray crystallography. Thermogravimetric analyses of 1 and 2 showed two-step weight loss. The residues from each step of 1 were MoS₃ and MoS₂, and these results were consistent with the subsequent deposition results of 1. We successfully established a PEALD-MoS₂ process using 1 and H₂S plasma as the precursor and reactant, respectively, at relatively low temperatures of 150–300 ^oC without any post-sulfurization process. A temperature-dependent selective deposition of MoS_x phases was observed with the growth of amorphous MoS3 films (150–200 ^oC), and crystalline MoS₂ films (250–350 ^oC).

Excessive burning of fossil fuels and changes in land use have led to rising atmospheric concentrations of carbon dioxide (CO₂), resulting in dramatic and potentially irreversible changes in the world's climate. Nanoscale metal-organic framework (MOF) MIL-101(Cr) with high porosity and large specific surface area is an excellent porous nanomaterial for catalytic reduction of CO₂. However, its applications have been typically restricted by nanoscale size. Herein, we assembled MIL-101(Cr) nanocrystals on TiO₂-coated polyimide (PI) substrates via atomic layer deposition (ALD) for high-efficient photocatalytic CO₂ conversion. The high CO₂ capturing capability of MIL-101(Cr) is combined with the unique property of inorganic semiconductor nanoparticles (TiO₂) to generate photoexcited electrons. Carbon monoxide (CO), methanol (CH₄), and hydrogen (H₂) were the main products of CO₂ photoreduction. The prepared MIL-TiO₂-PI felts with 1000 cycles of TiO₂ deposition exhibited the best photocatalytic performance, with high CO, CH₄, and H₂ yields of 158.6, 83.1, and 22.9 μmol g⁻¹, respectively. This work opens new routes for CO₂ reduction using MOF-based membrane hybrids prepared by ALD technology.

Yttrium and lanthanide metal oxides are known for their high permittivity and wide band gaps, making them ideal materials for metal-oxide semiconductor field-effect transistors (MOSFETs) and dynamic random-access memory (DRAM). Although various thin film deposition techniques are used to make these metal oxide films, atomic layer deposition (ALD) is a useful method because it produces high quality films, allows self-limited growth, and allows precise control of film thickness.

Achieving the desired crystal structure and crystallinity during ALD requires forming crystallization seeds during the deposition process. To improve the crystal structure and crystallinity of the thin films, deposition must occur at high temperatures. Therefore, developing new precursors with excellent thermal stability, which can withstand high temperatures without decomposing, is critical for the ALD process.

In this study, we developed new yttrium and lanthanide precursors and evaluated their suitability as high-temperature ALD precursors. Various analytical techniques, including NMR, EA, TGA, and vapor pressure measurements, were used to confirm the characteristics of the newly synthesized precursors. Additionally, we proposed an ALD process for La₂O₃ using the new lanthanum precursor and compared it to the commonly used La(iPrCp)₃ precursor. By processing at high temperatures with the new lanthanum precursor, we successfully obtained La₂O₃ thin films with excellent crystallinity, higher density, and improved leakage current characteristics. These findings suggest that the new precursors are promising candidates for high-temperature growth of yttrium and lanthanide-based thin films via ALD.