AVS 71 Session TF1-MoM: Fundamentals of Thin Films I

Atomic layer deposition (ALD) is a useful technique that enables atomic layer-by-layer growth of conformal films, but the technique is inherently wasteful due to the conventional viscous flow of chemical precursor injection, where the flow stream results in faster film growth at the stream-sample interaction point, reducing coating conformality. This work presents a new, “hold-step” ALD reactor that greatly increases efficiency of conformal film growth while reducing chemical use for improved sustainability. This work specifically focuses on iridium and iridium oxide thin films. By delving deeper into the understanding of the iridium precursor physical properties at temperature, the ALD recipe is tuned to best promote efficient iridium film growth kinetics. By introducing a “hold-step”, the reaction zone is isolated from the vacuum pump during dosing and the pressure is held constant for a set amount of time, resulting in improved film growth as the dosing gases are permitted to diffusively permeate and chemisorb onto embedded surfaces more efficiently. Additionally, characterizing the iridium precursors using techniques such as gas chromatography and thermogravimetric analysis, the ALD reaction can be fine-tuned for effective film growth. By changing from viscous to static flow and better understanding of the precursor kinetics, the total amount of gas required per deposition cycle is substantially reduced. A hold- step reactor design will be presented and compared to traditional ALD reactor designs. For comparison, precursors quantities and the resulting film qualities will be compared. The innovative yet simple design of a hold-step reactor not only enhances film quality but also promotes sustainability by reducing waste gas usage.

Microchannel plate (MCP) electron amplifiers are important components in large-area photodetectors, particularly for security-related applications such as night vision, radiation monitoring, and surveillance. Enhancing their performance and gain depends on advanced thin film coatings with high secondary electron yield (SEY). Barium-containing materials, recognized for their high SEY properties, are highly promising candidates for emissive coatings in MCPs. Atomic layer deposition (ALD) has emerged as a leading fabrication technique for such films, offering advantages in uniformity, pinhole-free morphology, and atomic-scale thickness control at low processing temperatures. However, despite ALD’s potential, research on barium-thin films via this method remains under-explored. Notably, no prior work has investigated ALD-synthesized barium-based coatings in MCPs, presenting a significant opportunity to bridge this gap.

In this study, we demonstrate the thermal atomic layer deposition (ALD) of barium-containing thin films using bis(tri(isopropyl)cyclopentadienyl)barium (Ba(iPr₃cp)₂) as the barium precursor and water (H₂O) as the co-reactant. We further investigate the structural and functional impact of incorporating alumina via super-cycle deposition within the Ba(iPr₃cp)₂/H₂O process. The Ba(iPr₃cp)₂ precursor was vaporized at 175°C, and depositions were performed in a hot-wall reactor at 250°C under a pressure of ~1.2 Torr. Systematic saturation studies were conducted to optimize precursor temperature, dose time, purge duration, and co-reactant exposure. Real-time thin film growth was monitored using in-situ ellipsometry, which enabled rapid saturation analysis while providing valuable insights into surface reactions during each ALD cycle.

The barium-containing films were characterized using X-ray photoelectron spectroscopy (XPS) for chemical composition, X-ray diffraction (XRD) for crystalline structure, X-ray reflectivity (XRR) for density, and atomic force microscopy (AFM) for surface topography. Moving forward, we aim to integrate ALD-grown barium layers onto microchannel plate (MCP) substrates and systematically evaluate their resistance, gain, and temporal stability. These metrics will directly assess secondary electron yield (SEY) performance, validating the material’s potential as a high-SEY coating to enhance MCP efficiency.

For a number of ALD processes, it is desirable to employ alternative co-reactants to achieve a variety of objectives, which include modifying the temperature window, optimizing the stoichiometry of the thin film, and eliminating undesirable side reactions. Concerning the latter, we have demonstrated that using t-BuOH in lieu of H₂O as a co-reactant in ALD with trimethyl aluminum (TMA) results in deposition of a thin film of Al₂O₃ that does not oxidize the underlying Co substrate, while use of H₂O does [1]. Here we build upon this previous work using a combination of experiments and theory to examine systematically a series of alcohols—primary, secondary and tertiary—as co-reactants with TMA for the ALD of Al₂O₃. We compare these results to the benchmark TMA|H₂O process and investigate the role of temperature. We have employed a quartz-crystal microbalance to monitor ALD in situ and in real-time, and the deposited thin films have been characterized using XPS. In parallel, we utilized density functional theory (DFT) calculations to identify key reaction intermediates and quantify the kinetics of surface reactions. At a temperature of T = 120 °C we find that none of the 8 alcohols examined result in steady growth of a thin film of Al₂O₃. At a temperature of T = 285 °C the situation is quite different, as steady growth is observed, but only by employing tertiary alcohols as co-reactants. Steady growth does not occur with the 6 primary and secondary alcohols examined. For example, concerning structural isomers of C₄H₉OH and C₅H₁₁OH, t-BuOH and 2-methyl-2-butanol result in steady growth, while 2-butanol and 3-methyl-2-butanol do not. Our calculations using DFT verify the essential role played by the tertiary -OH groups in facilitating the reaction with the chemisorbed species formed in the TMA half-cycle. We find that the important reaction intermediate involves an interaction between an adsorbed alkoxy species with another alcohol, producing -OH(a) species. A final issue we addressed concerned the effect of intentionally introducing a small amount of H₂O into the alcohol co-reactants. We find that a mixture of t-BuOH and a small amount of H₂O results in steady growth at T = 120 °C, whereas pure t-BuOH did not. Similarly, a mixture of i-PrOH and a small amount of H₂O results in steady growth at T = 285 °C, whereas pure i-PrOH did not. Overall, our study highlights the critical roles played by alcohol order, process temperature, and the influence of small amounts of H₂O impurity on the efficacy of using alcohols as co-reactants in ALD.

[1]J. V. Swarup, H.-R. Chuang, A. L. You, and J. R. Engstrom, ACS Appl. Mater. Interfaces 16, 16983–16995 (2024).