ALD/ALE 2025 Session AF3-TuM: Precursor Chemistry II

Tin dioxide (SnO₂), with an exceptional combination of high transparency, conductivity and stability in harsh chemical environments, has gained a lot of attention as a potential material for various optoelectronic applications including transparent electrodes for solar cells, anode materials for batteries, smart windows and displays, gas sensing and catalysis. Many of these applications require scalable, highly uniform and conformal thin films with superior quality and performance. The atomic layer deposition (ALD) technique offers a unique combination of scalability, uniformity and conformality with angstrom level thickness control.

Most of the previously studied ALD precursors for SnO₂ require highly oxidizing coreactants, such as O₃, H₂O₂, and O₂ or H₂O plasma. These processes are unsuitable for various sensitive applications, including biological systems and polymeric substrates. Among the available precursors, SnCl₄ and tetrakis(dimethylamino)tin(IV) react with water; however, SnCl₄ presents challenges due to the high deposition temperatures (300–600°C) and formation of corrosive by-products. Consequently, development of new precursors and processes are essential to fully explore the potential applications of ALD SnO₂ thin films.

The development of new precursors is essential for expanding atomic layer deposition (ALD) applications in microelectronics and beyond.^[1-2] However, a persistent challenge remains: the limited availability of precursors that combine high volatility, high reactivity and thermal stability, and while meeting industrial scalability requirements. The transition from cutting-edge scientific research to commercially available precursors is particularly difficult, with only a few dedicated R&D projects and companies addressing this gap. This is where VolaChem steps in.

VolaChem, founded by Dr. Dominik Naglav-Hansen and Dr. des. Martin Wilken, is a science-to-market transfer project dedicated to the development of volatile chemicals for technical applications. A key focus is the research, development, and up-scaling of ALD precursors, ensuring their applicability in the processing of advanced materials. By employing a database-driven approach for synthesis R&D, specialized precursor analytics, and consulting services, VolaChem aims to accelerate the transition of novel precursor concepts from academia to industry.

In this presentation, the authors will introduce their business concept and service offerings as well as innovative methodologies, while fostering a scientific dialogue on the needs, challenges, and opportunities related to volatile compounds in both research and industrial settings.

VolaChem is built upon the founders’ extensive academic research and expertise in designing tailored compounds for use as volatile precursors in technical processes.^[3-6] By bridging the gap between laboratory innovation and industrial application, VolaChem aspires to become a driving force in the next generation of ALD chemistry.

Borazine (B₃N₃H₆) has emerged as a promising precursor for synthesizing high-quality hexagonal boron nitride (h-BN) and amorphous boron nitride (a-BN), particularly after the groundbreaking work of Hong et al. (Nature 582, 511–514 (2020)), where they claimed to grow ultralow dielectric amorphous boron nitride using chemical vapor deposition method. Leveraging its intrinsic B₃N₃ ring, borazine enables facile atomic layer deposition (ALD) without the need for additional nitrogen sources. In this work, we utilize density functional theory (DFT) to investigate the reactivity of borazine with diverse surfaces, including those with defects, hydrogen terminations, and varying degrees of hybridization (sp2 and sp3). Our findings emphasize the critical roles of surface defects and hydrogenation in governing the deposition process. Furthermore, we will discuss the self-limiting behavior observed during ALD using borazine, a key characteristic for achieving precise film thickness control.

Vertical channel transistors (VCT) with InGaZnO¹ and InGaZnTinO² channels deposited by ALD have been gaining attention for the 3D integration of devices, especially DRAM. A supercycle ALD with homoleptic alkyl metal precursors is commonly used: multiple metal precursors and reactants are injected step by step. It allows very controlled components, yet the process is long, and the precursors are pyrophoric³. Overcoming a long process, alternative approaches have been suggested such as co-dosing, which supplies multiple precursors simultaneously, and cocktail precursor, which combines precursors in one canister. However, these are challenging because all precursors must have similar volatility and mixing stability, and no precursors satisfy both requirements.

In this study, we developed new heteroleptic In, Ga, Zn, and Sn precursors: In(Me)₂[^sBuNC(Me)C(H)C(Me)N^sBu] (DKI-6), Ga(Me)₂[^sBuNC(Me)C(H)C(Me)N^sBu] (DKG-6), Zn(Et)[^sBuNC(Me)C(H)C(Me)N^sBu] (DKZ-6), and Sn(NⁱPr₂)[^tBuNC(Me)N^tBu] (HTP-7) (Fig. 1). All precursors showed high volatility and non-pyrophoricity. A key feature of these materials is that they exhibit a vapor pressure difference of less than 10 °C at one torr (Fig. 2). Additionally, these materials remain stable and non-reactive when all mixed in equal moles. TGA of the mixture showed a smooth curve and no residue: 99.9% volatility (Fig. 3). ¹H NMR of the mixture had no byproduct peaks: the peaks came from each precursor (Fig. 4). Depositions of DKI-6 and DKZ-6 were done individually with H₂O on SiO₂ substrates at 150 to 300 °C. Both precursors showed ALD behavior and around 200 °C will be promising for a co-dosing and cocktail process (Fig. 5). DKI-6 and DKZ-6 at 200 °C growth rates were 0.1 and 0.2 Å/cycle respectively. X-ray photoelectron spectroscopy (XPS) indicated that impurity-free In₂O₃ and ZnO films were obtained successfully (Fig. 6). In contrast to In₂O₃, ZnO film did not have stoichiometry. This is believed to be caused by the difference in the sputtering ratio between Zn and O atoms. Moreover, DKI-6 and DKZ-6 were also reacted with O₃, and higher growth rates and impurity-free films were detected. This further investigation will be discussed at the conference.

These novel materials are promising for co-dosing and cocktail methods, offering significant process advantages like reducing flammable concerns and time consumption. The deposition of multi-component films is under investigation.

References:

1. Shosuke Fujii, Tseng Fu Lu, et al., IEDM2024, 6-1
2. Gan Liu, et al., VLSI2024, T16-1
3. Junghwan Kim, Jin-Seong Park, et al., ACS Appl. Mater. Interfaces 2019, 11, 43, 40300

This research focuses on the development of new precursors and processes for thermal atomic layer deposition (ALD) of elemental copper and gold thin films. Most previously reported copper thermal ALD processes require deposition temperatures in the range of 100-500 °C, which can result in agglomeration and discontinuous ultra-thin films.¹ Meanwhile, there have only been three reported processes² for Au thermal ALD – two of these use ozone (O₃) as a co-reactant, which limits substrate compatibility, while the minimum deposition temperature for the third process using [AuCl(PEt₃)] with 1,4-bis(trimethylgermyl)-1,4-dihydropyrazine is limited by the volatility of the precursor and co-reactant. Thus, precursors which offer increased reactivity, thermal stability and volatility compared to those currently available, as well as the development of new reaction chemistries which might facilitate deposition at temperatures lower than previously reported, are of particular significance.

Herein we present a new family of fluorinated alkoxides, [{M(OR^F)(L)}_n] (M = Cu, Au; OR^F = fluorinated alkoxide; L = PR₃, CNR), as potential precursors for thermal ALD of Cu and Au. Complexes were synthesized by straightforward and scalable methodologies, and were crystallographically and spectroscopically characterized. The thermal properties of these complexes were evaluated, displaying a wide range of thermal stability and volatility, with some precursors possessing the required characteristics for use in ALD. Solution-state reactions of these precursors with various ALD co-reactants support the thermodynamic feasibility of potential ALD reaction chemistry, yielding the target metal and volatile by-products. Select precursors were chosen for preliminary studies on a custom-built ALD reactor, leading to deposition of metallic thin films.

References

1. Hagen, D. J.; Connolly, J.; Povey, I. M.; Rushworth, S.; Pemble, M. E. Adv. Mater. Interfaces 2017, 4, 1700274.

2. (a) Mäkelä, M.; Hatanpää, T.; Mizohata, K.; Räisänen, J.; Ritala, M.; Leskelä, M. Chem. Mater. 2017, 29 (14), 6130-6136. (b) Hashemi, F. S. M.; Grillo, F.; Ravikumar, V. R.; Benz, D.; Shekhar, A.; Griffiths, M. B. E.; Barry, S. T.; van Ommen, J. R. Nanoscale 2020, 12 (16), 9005-9013. (c) Vihervaara, A.; Hatanpää, T.; Nieminen, H.-E.; Mizohata, K.; Chundak, M.; Ritala, M. ACS Mater. Au 2023, 3 (3), 206-214.

Al₂O₃ is a versatile material system and due to its unique properties, thin films of Al₂O₃ find a range of applications in electronics, aerospace, automotive industry, anti-reflective coatings^[1] packaging, gas barrier layers (GBLs)^[2] etc. Especially for packaging and encapsulation of degradable goods, which are typically encapsulated by polymers such as polypropylene (PP) and polyethylene terephthalate (PET), low temperature processing of dense and amorphous Al₂O₃ thin films is required.^[2] These prerequisites render atomic layer deposition (ALD) as a favorable thin film fabrication method. Trimethylaluminum (TMA) is the most commonly precursor for ALD of Al₂O₃.^[3] However, the high reactivity towards moisture requires expensive safety measures and the possible formation of TMA dimers leads to difficulties in predicting and controlling the process characteristics. Combined with the lack of a defined ALD window in TMA/H₂O ALD processes, alternative Al precursors that are non-pyrophoric, but still volatile and reactive to enable low temperature ALD processes with various co-reactants including H₂O and O₂ plasma are of great interest.

We have explored [3-(dimethylamino)propyl]-dimethyl aluminum (DMAD)for PEALD of Al₂O₃ thin films on polymer substrates. The influence of the two precursors on Al₂O₃ thin film growth and film properties was investigated and compared. Thin layers of Al₂O₃ and dyads of Al₂O₃ and SiO₂ were deposited (Figure 1a) with the aim of using them as GBLs on PP substrates. Interestingly, the dyads on PP containing SiO₂ and Al₂O₃ significantly outperformed the Al₂O₃ grown with TMA in terms of barrier properties (Figure 1 b).^[4]

Literature:

[1]: Z. Lin, C. Song, T. Liu, J. Shao, M. Zhu, ACS Appl. Mater. Interfaces, 2024, 16, 31756.

[2]: M.-H. Tseng, H.-H. Yu, K.-Y. Chou, J.-H. Jou, K.-L. Lin, C.-C. Wang, F.-Y. Tsai, Nanotechnology 2016, 27, 295706.

[3]: R. L. Puurunen, J. Appl. Phys., 2005, 97.

[4]: M. Gebhard, L. Mai, L. Banko, F. Mitschker, C. Hoppe, M. Jaritz, D. Kirchheim, C. Zekorn, T. de los Arcos, D. Grochla, R. Dahlmann, G. Grundmeier, P. Awakowicz, A. Ludwig, A. Devi, ACS Appl. Mater. Interfaces, 2018, 10, 7422.

Niobium Oxides (Nb₂O₅) have been extensively utilized in various fields of technology. Traditionally these oxides have been applied as resistive films used as high-k materials for insulating layers. For instance, a thin layer of Nb₂O₅ between two ZrO₂ dielectric layers is expected to help significantly reduce leakage current and stabilize the cubic/tetragonal phases of the ZrO₂, affording higher k values in the current MIM capacitor of a DRAM. Furthermore, the application range of Nb₂O₅ expanded as HZO (HfZrO), which has ferroelectric properties, began to be applied to dielectric materials for DRAM capacitors. It is recently being studied as an interlayer between electrodes and HZO to enhance ferroelectricity of HZO.

Today, there is a growing need for a highly volatile, thermally stable, ideally low-viscosity liquid niobium precursor that allows to deposit films at high temperature in the Atomic Layer Deposition (ALD) process with a low film growth rate for precise thickness control. However, in general, thermal stability has a trade-off relationship with vapor pressure and viscosity, making it challenging to develop an ideal precursor with excellent properties across all these aspects.

To address these requirements, Air Liquide developed three Nb precursors (Nb1, Nb2, Nb3). Among them, Nb3 exhibited the most promising physical properties including not only high vapor pressure (1Torr at 105℃) and low viscosity (6.5cP at 24℃) but also high thermal stability (decomposition onset at 340℃ by DSC). It demonstrated a moderately low film growth rate of 0.4 Å/cycle, enabling precise control of film thickness. Furthermore, Nb3 showed exceptional ALD performance, achieving 100% step coverage at 350℃ and 375℃ (A/R: 1:30) and an ALD window extending up to 350℃, which is superior to the commercially available precursor in the industry.