Crystalline as-deposited TiO 2 anatase thin films grown from TDMAT and water using thermal atomic layer deposition with in situ layer-by-layer air annealing

Jamie P. Wooding; Kyriaki Kalaitzidou; Mark D. Losego

doi:10.3897/aldj.2.117753

Research Article

Crystalline as-deposited TiO ₂ anatase thin films grown from TDMAT and water using thermal atomic layer deposition with in situ layer-by-layer air annealing

Jamie P. Wooding^‡, Kyriaki Kalaitzidou^‡, Mark D. Losego^‡

‡ Georgia Institute of Technology, Atlanta, United States of America

Corresponding author: Mark D. Losego ( losego@gatech.edu )

Academic editor: J. Ruud Van Ommen

This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Citation: Wooding JP, Kalaitzidou K, Losego MD (2024) Crystalline as-deposited TiO ₂ anatase thin films grown from TDMAT and water using thermal atomic layer deposition with in situ layer-by-layer air annealing. Atomic Layer Deposition 2: 1-18. https://doi.org/10.3897/aldj.2.117753

Abstract

We report a new thermal atomic layer deposition (thermal-ALD) process including an air exposure as a third precursor to deposit crystalline TiO₂ anatase thin films from tetrakis(dimethylamido)titanium(IV) (TDMAT) and water at deposition temperatures as low as 180 °C and film thicknesses as low as 10 nm. This ALD process enables TiO₂-antase crystal growth during the deposition at low temperatures (< 220 °C). This additional oxidant pulse is used to fully oxidize the Ti to a 4+ state in the amorphous film, lowering the barrier to crystalline anatase formation. This new approach is informed by preliminary studies of post-deposition annealing (PDA) of thermal ALD films in both nitrogen and air atmospheres, which demonstrate the importance of having an oxidizing atmosphere to achieve the nucleation of the crystalline anatase phase. This oxidizing atmosphere is subsequently introduced into the ALD cycle as a third precursor and is shown to be more effective and efficient in promoting the crystalline transformation than even by post-deposition annealing. The crystalline anatase phase is verified by Raman spectroscopy and grazing incidence X-ray diffraction (GIXRD). The mechanism for crystallization during the TDMAT/H₂O/air ALD cycle is probed by chemical state analysis via X-ray photoelectron spectroscopy (XPS). We propose that sub-oxidation in TiO₂ thin films deposited by the thermal-ALD process inhibits crystallization during ALD from TDMAT/H₂O chemistry. Scanning electron microscopy (SEM) is used to investigate the microstructure of these TiO₂ thin films as a function of thickness (5 nm to 50 nm) and deposition temperature (180 °C to 220 °C). The reported layer-by-layer air anneal process is found to crystallize entire films in shorter total process times than thermal-ALD with ex situ post deposition annealing at identical temperatures, presumably due to the improved surface diffusion kinetics accessed during the deposition process.

Key words

Atomic layer deposition, TiO₂, anatase, crystallization, thin films

1. Introduction

Titanium dioxide is a versatile metal oxide with phase-dependent applications in photocatalysis [1], corrosion prevention [2], photovoltaics [3], and high-κ dielectrics [4]. Atomic layer deposition (ALD) is a sequential vapor deposition process, defined by self-limiting surface half reactions [5], that is widely employed to deposit conformal titanium dioxide (TiO₂) thin films [6]. A typical ALD cycle consists of delivering vapor precursors to a substrate, having these precursors react with activated surface sites on the substrate, and then removing the excess precursor and the reaction by-products in a purge step. A vapor-phase co-reactant, commonly an oxidant, is then delivered in a similar manner and reacts to restore the initial surface chemistry such that the surface is receptive to repeating the precursor chemistry. Typical ALD-TiO₂ precursors include titanium tetrachloride (TiCl₄), titanium tetraisopropoxide (TTIP), and titanium (IV) tetrakis(dimethylamido) titanium (TDMAT) [6, 7]. Titanium precursor chemistry, oxidant chemistry, and ALD deposition temperature affect the deposited TiO₂ phase—amorphous, anatase, or rutile—and the resultant film properties [6].

Here, we highlight relevant prior work in thermal TiO₂-ALD from titanium (IV) tetrakis(dimethylamido) titanium (TDMAT) and H₂O, as this is the chemistry used in this study. TDMAT can be preferred as a precursor over TiCl₄ due to its lack of chlorine contamination in the deposited film [8, 9], higher deposition rate at lower temperatures, and high reactivity of the Ti-N bond [6]. However, because the ALD process temperature window is limited to below TDMAT’s decomposition temperature of approximately 220 °C, TDMAT/H₂O films are nearly always amorphous as-deposited [9–12]. Films deposited from TDMAT/O₃ chemistry show similar behavior, with the anatase phase appearing in deposition temperatures above 200 °C [13, 14]. There is a notable exception to this trend where a study from Pheamhom et al. using TDMAT/H₂O₂ chemistry observed TiO₂-anatase crystallite formation from XRD at 150 °C [15]. As such, post-deposition annealing (PDA) is usually necessary to achieve crystalline TiO₂ films from TDMAT/H₂O ALD chemistry [16, 17]. In comparison, ALD of TiO₂ thin films from the TiCl₄/H₂O chemistry permits anatase crystallization during deposition at about 150 °C and above [18–20]. And TiO₂ films deposited from TTIP and H₂O chemistry show evidence of the anatase phase at 180 °C with well-developed crystallites observed between 250–300 °C [21–23]. As a result, here we question what is limiting crystallization during ALD in TiO₂ films deposited from TDMAT and H₂O at deposition temperatures below 220 °C.

Additional observations of relevance include differences in electronic properties for amorphous-TiO₂ films deposited from the TDMAT/H₂O chemistry versus the TiCl₄/H₂O chemistry. The TDMAT/H₂O amorphous TiO₂ films are often reported as moderately conductive while the TiCl₄/H₂O deposited amorphous TiO₂ films are generally electrically insulating. For example, in 2019 Nunez et al. reported that amorphous-TiO₂ films deposited from TDMAT/H₂O had higher carrier concentrations than those deposited from TiCl₄/H₂O [24]. Nunez et al. ascribed these differences to a Ti³⁺ defect-mediated transport mode and not due to the presence of external dopants [24]. Indeed, the most common TiO₂ defect is oxygen vacancies which create unpaired electrons or Ti³⁺ centers [25]. In 2021, Babadi et al. sought to identify the cause of increased electronic conductivity in TDMAT-TiO₂ thin films [26]. Babadi et al. found that the reaction by-product from the TDMAT/H₂O ALD process, dimethylamine (DMA), can act as a reducing agent for TiO₂. Increased exposure of TiO₂ thin films to DMA caused an increase in electrical conductivity and an increase in the concentration of Ti in the 3+ oxidation state [26]. Finally, in 2022, Saari et al. reported a comprehensive electron and optical absorption spectroscopy study finding that the concentration of intrinsic Ti³⁺ defects in amorphous TiO₂-ALD films can be controlled by varying deposition temperatures from 100 to 200 °C [27]. Thus, current understanding is that Ti³⁺ states are a common defect in TDMAT/H₂O ALD films and contribute to the enhanced electrical conductivity of these predominantly amorphous TiO₂ films.

In this report, we describe the first use of an in situ layer-by-layer air anneal during the ALD cycle for TiO₂-anatase thin film growth during deposition. Functionally, this ALD process operates as a three-precursor deposition, where the precursor pulsing time (s) follows the format t₁/t₂/t₃/t₄/t₅/t₆ compared to the two-precursor deposition format of t₁/t₂/t₃/t₄. Related modification of the deposited film oxidation state during atomic layer deposition (ALD) has been demonstrated in, for instance, the vanadium oxide (VO_x) system. Weimer et al. reported using a second co-reactant (i.e. third precursor) as a means to adjust the oxidation state of VO_x, applying various combinations of H₂O, H₂O₂, O₂, O₃, and H₂ [28]. Here, we similarly modify the TiO₂ ALD growth process, using an air dose as the third precursor and second oxidant, to encourage stoichiometric TiO₂ deposition and subsequent crystalline nucleation. We suggest that the presence of the Ti³⁺ chemical state, indicative of oxygen vacancies, inhibits crystalline anatase formation and that the introduction of an additional oxidizing step in the ALD cycle removes these defects and lowers the barrier to crystal nucleation.

Methods

ALD conditions

ALD was performed in a home-built, hot-walled, flow-tube reactor (~4 cm tube diameter) with custom LabVIEW control software [29]. Purified nitrogen (99.995%+ N₂, On Site Model Pro N-8TGNPSA Nitrogen Generator, Newington, CT) was used as the carrier gas with a chamber pressure of 1.2 Torr. The chamber deposition region was held at 180 °C, 200 °C, or 220 °C, depending on desired deposition temperature. The process temperature was kept at or below 220 °C to prevent TDMAT decomposition [11]. All gas process line temperatures were maintained between 82 °C and 110 °C to prevent precursor condensation prior to reaching the deposition region. Depositions were performed on 10 cm × 1 cm Si substrates (University Wafer, P/Boron<100>, 1–100 Ohm-cm, Test Grade) and the Si wafer was diced into 1 cm × 1 cm coupons for analysis following deposition.

Tetrakis(dimethylamino)titanium(IV) (TDMAT, 99% purity) from Strem Chemicals, Inc. (Newburyport, MA, USA) and deionized water were used as the precursor and co-reactant for TiO₂ ALD. TDMAT was held at 23 °C and was dosed into the reaction chamber in a bubbler configuration with N₂ flow. The TDMAT delivery lines to the reactor were heated to 82 °C. The standard thermal-ALD cycle included 1.0 s TDMAT dose / 5 s N₂ purge / 0.4 s H₂O dose / 85 s N₂ purge for deposition of TiO₂ reference films. This two-precursor ALD process is referred to as follows: TDMAT/H₂O. Note that the TDMAT dose and H₂O dose times remain fixed throughout this report. For the ALD cycle with the third precursor, an air dose (Zero Grade Air, Airgas) was included to increase the chamber process pressure from 1.2 Torr to 4.5 Torr for the specified time duration. This additional cycle step is referred to as the air anneal since its time duration is much longer than a 0.1–1 s precursor dose and it includes active chemistry compared to an inert gas purge. Including the air anneal, the standard three-precursor ALD process sequence was: 1.0 s TDMAT dose / 5 s N₂ purge / 0.4 s H₂O dose / 15 s N₂ purge / 35 s zero grade air / 35 s N₂ purge. As in a typical thermal-ALD process, the N₂ purge following the TDMAT dose is intended to remove the reaction byproducts and any unreacted precursor from the reaction chamber. Similarly, the N₂ purge following the H₂O dose is intended to remove the reaction byproducts and any unreacted oxidant from the reaction chamber. The purpose of the air anneal is to provide an oxidizing atmosphere during the ALD cycle. The purpose of the last N₂ purge in the cycle is to remove all air from the reaction chamber to not adversely affect the self-limiting surface half reactions between TDMAT and H₂O, and to restore process pressure to 1.2 Torr. The three-precursor ALD process is abbreviated as TDMAT/H₂O/air or TDMAT/H₂O/35 s air if the air dose time duration is being highlighted.

Characterization

Spectroscopic ellipsometry (alpha-SE, J.A. Woollam) was used to measure film thickness and the index of refraction, n. A CodyLor model, for films with UV absorption, was used to describe the TiO₂ layer on top of a 1.76 nm native oxide layer on a Si wafer substrate. Raman spectroscopy (Renishaw InVia Qontor Raman Microscope) was performed for TiO₂ phase identification using a 488 nm laser with 10% power, 1200 l/mm grating, and 2 s acquisition time. Raman spectroscopy measurements were made in at least three different lateral locations on the 1 cm × 1 cm TiO₂ film. To support TiO₂ phase identification, GIXRD was conducted on a PANalytical Empyrean system using Cu-Kα radiation, a BBHD source optic, and a PIXcel 2-dimensional detector. 2θ-ω scans (corresponding to θ-2θ geometry) were taken. Scanning electron microscopy (SEM) is used to image the resultant TiO₂ plan-view microstructure as a function of process conditions. SEM was performed on a Hitachi SU8230 field emission SEM (FE-SEM) at 1 kV accelerating voltage and 20 µA emission current. In mixed phase TiO₂ films, amorphous regions appear dark and anatase crystals appear bright. In films appearing fully crystalline, the image contrast is such that different grains are distinguishable. For each condition, at least two TiO₂ films were imaged, with top-down images taken at three different lateral locations on the 1 cm × 1 cm film surface. ImageJ analysis was employed to quantify the crystalline fraction. Additional information about this analysis can be found in our prior publication [30].

To study surface roughness and topography, a Bruker Icon AFM (Bruker, Billerica, MA, USA) was used in standard tapping mode with a n-Si tip (MikroMasch Hq:XSC11/AL), a scan rate of 0.5–1 Hz, and scan area ranging from 1–4 μm².

X-ray photoelectron spectroscopy (XPS) is used to evaluate the chemical state of Ti as a function of ALD process conditions. XPS was conducted on a Thermo Scientific K-Alpha XPS coupled with a monochromated Al Kα (1486.6 eV) X-ray source. This system uses an X-ray incident angle of 60° from sample normal and the photoemission angle is 0° from sample normal. For each TiO₂ thin film analyzed, a survey spectrum was measured and high-resolution elemental scans for C 1s, O 1s, N 1s, and Ti 2p were collected with a 0.100 eV step size. The presented high-resolution scans are averaged from 5+ data collection scans. Binding energy calibration was performed based on the adventitious carbon C1s peak at 284.8 eV. Ti3/2p and Ti1/2p peaks were fitted to discern Ti⁴⁺ from Ti³⁺ states. Applied constraints were selected based on TiO₂ references [31, 32].

Results and discussion

TDMAT/H₂O/air ALD Overview

Figure 1 presents SEM micrographs from a) as-deposited TiO₂ thin films deposited from TDMAT/H₂O thermal-ALD with an 85 s purge after the water dose step, and b) TiO₂ thin films deposited from TDMAT/H₂O/air ALD with a 35 s in situ air anneal after the water dose step. To maintain an equivalent thermal budget, the time per cycle for both ALD processes, TDMAT/H₂O and TDMAT/H₂O/air, is kept constant at 91.4 s. TiO₂ thin films deposited from TDMAT/H₂O chemistry at 180 °C, 200 °C, and 220 °C show a uniformly amorphous microstructure except for some white hillocks. A previous post-deposition annealing (PDA) study has demonstrated that these hillocks do not serve as seeds for 2-dimensional crystal growth [30], nor do these hillocks purport a crystalline signature. In Figure 1b, the 50 nm TiO₂ thin films deposited from TDMAT/H₂O/air chemistry show what appears to be a fully crystalline microstructure. This result is very different from the TDMAT/H₂O ALD films and demonstrates that the in situ oxidizing atmosphere facilitates anatase nucleation and growth, achieving an as-deposited, fully crystalline microstructure from the TDMAT/H₂O/air ALD process at low temperature without an activated species.

Figure 1.

SEM micrographs for (a) as-deposited film morphologies at ALD temperatures increasing from 180 °C (bottom row) to 220 °C (top row) with 1120 ALD cycles from TDMAT/H₂O chemistry at 91.4 s per cycle. (b) SEM micrographs for TDMAT/H₂O/air ALD film morphologies at temperatures increasing from 180 °C (bottom row) to 220 °C (top row) with 1120 ALD cycles with 91.4 s per cycle, including a 35 s air dose per cycle.

To further emphasize the importance of annealing atmosphere, we also examined post-deposition annealing behavior in these films. Suppl. material 1: fig. S1 presents SEM micrographs of TDMAT/H₂O ALD TiO₂ films of 50 nm thickness deposited at 180 °C and then annealed in either N₂ or air for 22.5 h. Post deposition annealing (PDA) in N₂ does not enable anatase nucleation and growth on this time scale while PDA in air does, further emphasizing the importance of oxidizing atmospheres to promote TiO₂ crystallization.

Figure 2 presents Raman spectra to identify the TiO₂ phase in films deposited at a) 180 °C, b) 200 °C, and c) 220 °C. For each deposition temperature, spectra are presented for a (i) 50 nm TiO₂ film from TDMAT/H₂O/air ALD, (ii) 25 nm TiO₂ film from TDMAT/H₂O/air ALD, and (iii) 50 nm TiO₂ film from TDMAT/H₂O ALD. For all deposition temperatures, the 50 nm TDMAT/H₂O ALD films do not exhibit any Raman vibrational modes other than the Si substrate signal at 521 cm^-1 (Figure 2a–c, iii), confirming their amorphous structure. In contrast, all 50 nm and 25 nm TiO₂ films from TDMAT/H₂O/air ALD exhibit Raman signatures consistent with the anatase crystalline phase (Figure 2a–c, i), providing further evidence that the in situ layer-by-layer air anneal during the ALD cycle drives crystallization.

For the 50 nm TDMAT/H₂O/air ALD films deposited at 180 °C and 200 °C (Figure 2a, b, i), the Raman spectra show the vibrational modes for anatase, both the E_g (145 cm^-1, 636 cm^-1) and B_1g (394 cm^-1). The E_g peak results from the symmetric stretching of the O—Ti—O bonds and the B_1g results from the symmetric bending of O—Ti—O bonds [33–35]. The A_1g peak for TiO₂-anatase at 514 cm^-1 is not visible due to overlap with the Si substrate vibrational mode. In Figure 2c, (i), the 50 nm TDMAT/H₂O/air ALD film deposited at 220 °C shows a weaker, but still present, anatase signal. In Raman spectroscopy, the signal intensity is proportional to the population of bonds contributing to the vibrational mode. For this reason, we believe the 50 nm TDMAT/H₂O/air ALD films deposited at 220 °C to be more defective than the lower temperature (180 °C and 200 °C) films, but to still have the anatase phase present. This result appears consistent with the finer grain microstructure (i.e., more defective grain boundaries) observed at this deposition temperature in Figure 1.

Figure 2.

Raman spectra for TDMAT/H₂O ALD and TDMAT/H₂O/air ALD TiO₂ films deposited at (a) 180 °C, (b) 200 °C, and (c) 220 °C. For each deposition temperature, (i) 50 nm TDMAT/H₂O/air ALD, (ii) 25 nm TDMAT/H₂O/air ALD, and (iii) 50 nm TDMAT/H₂O ALD spectra are presented. Spectra are normalized to the Si peak at 521 cm^-1. TiO₂ stretching modes are labeled as E_g (*) and B_1g (^#).

Figure 2a–c, (ii) presents Raman spectra for 25 nm TDMAT/H₂O/air ALD films deposited at 180 °C to 220 °C. Uniformly, the E_g and A_1g peaks have decreased. While these spectra still demonstrate the presence of the TiO₂-anatase phase in these thinner films, the intensity is decreased due to the reduction in film volume and reduced crystallinity likely from film thickness effects.

To determine how including air as a third ALD precursor alters the ALD growth per cycle (GPC), Figure 3a reports deposited film thickness as a function of the number of TDMAT/H₂O/air ALD cycles at deposition temperatures of 180 °C, 200 °C, and 220 °C. GPC values for TDMAT/H₂O ALD-TiO₂ films at each of these temperatures are also included in Figure 3a as dotted lines. TDMAT/H₂O/air ALD-TiO₂ films deposited at 180 °C and 200 °C had similar GPCs of 0.046 nm/cycle and 0.047 nm/cycle, respectively. Corresponding TDMAT/H₂O ALD-TiO₂ films deposited at temperatures of 180 °C and 200 °C have GPC values of 0.041 nm/cycle and 0.038 nm/cycle respectively. These values are in good agreement with literature for thermal-ALD of TDMAT/H₂O chemistry at temperatures around 200 °C [6, 36]. As such, the 180 °C and 200 °C TDMAT/H₂O/air ALD-TiO₂ films from this work are about 6 nm and 10 nm thicker than their TDMAT/H₂O counterparts. This behavior, whereby the TDMAT/H₂O/air TiO₂ films have increased growth per cycle compare to two-precursor ALD films, aligns with literature reports describing GPC as higher for crystalline as-deposited films compared to amorphous films and further corroborates formation of the crystalline phase during TDMAT/H₂O/air ALD [37].

Figure 3.

(a) TiO₂ film thickness measured via spectroscopic ellipsometry (SE) for TDMAT/H₂O ALD (hollow, crossed symbols) and TDMAT/H₂O/air ALD (filled symbols) processes for between 114 and 1120 cycles at deposition temperatures: 180 °C, 200 °C, and 220 °C. GPC values are also notated in Angstrom per cycle. (b) Index of refraction (n) at 550 nm for 50 nm TiO₂ thin films deposited at 180 °C, 200 °C, and 220 °C by both TDMAT/H₂O ALD and TDMAT/H₂O/air ALD processes.

Films deposited at 220 °C by TDMAT/H₂O/air ALD exhibit a GPC of 0.053 nm/cycle, higher than both the 180 °C and 200 °C TDMAT/H₂O/air ALD films. For the 220 °C TDMAT/H₂O/air ALD film, there is a corresponding decrease in the Raman TiO₂-anatase signal. TDMAT is known to undergo gas-phase unimolecular decomposition above 220 °C, mechanistically driven by the low binding energy of the Ti-N bond and moderated by pressure and surface-to-volume ratio of the ALD chamber [6, 11]. The high TDMAT/H₂O/air ALD GPC at 220 °C likely has a primary contribution from TDMAT decomposition. This decomposition typically would result in CVD-like deposition, which can increase the GPC due to the gas phase, non-surface limited reactions, as consistent with Xia et al. [38]. We propose that during TDMAT/H₂O/air ALD at 220 °C, competing reactions occur between TDMAT decomposition and surface adsorption and reaction, increasing the film growth rate and reducing anatase crystallization.

Figure 3b presents the refractive index (n, 550 nm) for 50 nm TDMAT/H₂O and TDMAT/H₂O/air ALD-TiO₂ films. Amorphous TDMAT/H₂O ALD-TiO₂ films have refractive indices varying from 2.26 to 2.43. These values are in agreement with n = 2.3 reported for amorphous TiO₂ from the TiCl₄/H₂O chemistry [39]. TDMAT/H₂O/air TiO₂ films (anatase) have refractive indices varying from 2.52 to 2.55. These values are close to literature reports for the refractive index of annealed TiO₂-anatase from TDMAT/H₂O chemistry at 2.54 [7] and close to the n = 2.6 value reported for as-grown, high deposition temperature anatase from TiCl₄/H₂O chemistry [39]. Refractive index is documented as an accurate indicator of the crystalline quality of TiO₂ thin films [7]. Thus, these TDMAT/H₂O/air ALD-TiO₂ thin films with refractive index values approaching those measured for bulk anatase indicate increased thin film density and improved anatase crystal quality.

To summarize, including a 35 s in situ air anneal step as a third precursor dose during the ALD cycle enables the growth of crystalline TiO₂ as confirmed with Raman spectroscopy, SEM, and greater refractive index. Ideal, self-limiting ALD behavior appears to occur at deposition temperatures of 200 °C and below, with higher temperatures likely leading to TDMAT decomposition and higher GPC values. Below, we provide further evidence that the air dose acts to fully oxidize the depositing TiO₂, lowering the barrier to crystallite nucleation.

TiO₂-anatase crystallization

To evaluate the hypothesis that sub-oxidized film chemistry inhibits TiO₂ crystallization during the ALD cycle, X-ray photoelectron spectroscopy (XPS) is used to probe the oxidation state of 50 nm TiO₂ thin films (1120 cycles). Figure 4 compares Ti 2p spectra for TDMAT/H₂O ALD-TiO₂ films (Figure 4a, c, e) and TDMAT/H₂O/air ALD-TiO₂ films (Figure 4 b, d, f) deposited at 180 °C, 200 °C, and 220 °C. As expected, the Ti 2p doublets display spin orbital splitting for Ti 2p_3/2 (~458.6 eV) and Ti 2p_1/2 (~464.3 eV). Peak positions are in good agreement with each other and literature reports [27, 31, 32]. TDMAT/H₂O ALD-TiO₂ films display a clear shoulder for the Ti 2p_3/2 peak around 457.5 eV. This shoulder corresponds to the presence of sub-oxidized titanium oxide states, likely Ti³⁺ [27]. For TDMAT/H₂O ALD-TiO₂ films, this Ti³⁺ peak area contribution ranges from 15–17%. In contrast, the sub-oxide shoulder on the Ti 2p_3/2 peak for the TDMAT/H₂O/air ALD-TiO₂ films is not readily visible and contributes only 4–7% of the total area, indicating that Ti³⁺ states are much less prevalent. This sub-oxidation result for as-deposited, TiO₂ ALD films from TDMAT/H₂O chemistry is consistent with the report from Babadi et al. which identifies the ALD reaction by-product, dimethylamine (DMA), as a reducing agent during the ALD process [26]. As such, here we demonstrate that DMA-induced reduction of ALD-TiO₂ can be counteracted with an in situ layer-by-layer air anneal, as evidenced by decreased presence of the Ti³⁺ chemical state in TDMAT/H₂O/air ALD-TiO₂ films.

Figure 4.

XPS Ti2p elemental scan for (a) TDMAT/H₂O ALD at 180 °C, (b) TDMAT/H₂O/air ALD at 180 °C, (c) TDMAT/H₂O ALD at 200 °C, (d) TDMAT/H₂O/air ALD at 200 °C, (e) TDMAT/H₂O ALD at 220 °C, and (f) TDMAT/H₂O/air ALD at 220 °C.

Interestingly, the reduction in Ti³⁺ chemical state in TDMAT/H₂O/air ALD films is accompanied by significant TiO₂-anatase crystallization. In our previous work studying the kinetics of the TiO₂ amorphous to anatase phase transformation during post-deposition annealing (PDA) in air, we found that the energy barrier for the amorphous to anatase phase transformation is approximately ~1.5 eV atom^-1, and was specifically attributed to the nucleation step, not crystal growth [30]. This energy barrier to nucleation appears to be partly associated with overcoming the incomplete oxidation of the as-deposited amorphous film. If the in situ air dose creates an amorphous state with a lower energy barrier to overcome, this would result in faster nucleation of the TiO₂ layers during TDMAT/H₂O/air ALD compared to the two-precursor TDMAT/H₂O ALD. Mechanistically, we suggest that the air dose oxidizes the Ti³⁺ states to Ti⁴⁺ states, which better form TiO₆ octahedra that lower critical nucleus size and the energy barrier to nucleation. To support this hypothesis, we recommend in situ chemical state analysis in future work, such as in situ FTIR or in situ XPS, in order to confirm any surface functionality change after this air anneal.

Probing TDMAT/H₂O/air ALD growth parameters

Next, we examine how various process parameters affect TiO₂ crystallization in the TDMAT/H₂O/air ALD process. Specifically, we look at crystallization behavior with film thickness and the differences between the TDMAT/H₂O/air ALD process versus an ex situ post-deposition anneal (PDA). Influence of cycle times and atmosphere selection are also reported.

First, TDMAT/H₂O/air ALD-TiO₂ films are grown to thicknesses ranging from 5 nm to 50 nm at deposition temperatures of 180 °C, 200 °C, and 220 °C. Plan view SEM images of each of these films are collected to evaluate the percent of crystalline regions versus amorphous regions. Figure 5 presents the SEM micrographs for TiO₂ films deposited TDMAT/H₂O/air ALD at 180 °C for 5 nm (114 cycles) to 51 nm (1120 cycles). At this temperature, the crystalline phase is first observed at 18 nm (409 cycles) and full crystallization is achieved at 35 nm (800 cycles). Figure 6 presents the SEM micrographs for TiO₂ films deposited by the TDMAT/H₂O/air ALD process, now at 200 °C. The behavior is similar to the 180 °C processed film, albeit with onset and full crystallization reached at lower cycle numbers, here at around 10 nm (227 cycles) and 25 nm (568 cycles). From SEM imaging of at least two TiO₂ films and analyzing the crystallinity with ImageJ, the percent crystallinity at 10 nm (Figure 6b) is less than 1% while the percent crystallinity at 20 nm (Figure 6d) is 97.5%. From top-down SEM, the 25 nm (Figure 6e) and 50 nm (Figure 6f) films do not show any amorphous regions. In comparing the 15 nm (Figure 6c) to the 20 nm TiO₂ film (Figure 6d), the 20 nm film has an increased crystal density and increased crystal size, implying both nucleation and grain growth occur with subsequent TDMAT/H₂O/air ALD cycles until the aerial density is fully crystalline.

Figure 5.

SEM micrographs for films deposited at 180 °C by TDMAT/H₂O/air ALD from (a) 114 cycles, (b) 409 cycles, (c) 444 cycles, (d) 568 cycles, (e) 800 cycles, and (f) 1120 cycles. This corresponds to approximately 5 nm, 18 nm, 19 nm, 25 nm, 35 nm, and 51 nm film thickness respectively.

Figure 6.

SEM micrographs for films deposited at 200 °C by TDMAT/H₂O/air ALD from (a) 114 cycles, (b) 227 cycles, (c) 341 cycles, (d) 444 cycles, (e) 568 cycles, and (f) 1120 cycles. These number of cycles correspond to approximately 5 nm, 10 nm, 15 nm, 20 nm, 25 nm, and 52 nm film thickness respectively.

In comparing the SEM micrographs in Figure 5 and Figure 6, increasing deposition temperature increases crystal nucleation at equivalent film thicknesses and process times. This is consistent with literature reports for thermal ALD with H₂O as the oxidant [20, 40]. Suppl. material 1: fig. S2 includes the SEM micrograph series for TDMAT/H₂O/air ALD process with 220 °C growth temperature and reports complete aerial crystallinity TiO₂ anatase thin films as thin as 10 nm. Figure 7 summarizes the calculated percent crystallinity with increasing film thickness for TDMAT/H₂O/air ALD films deposited at 180 °C, 200 °C, and 220 °C as determined from top-SEM and analysis via ImageJ. For the three deposition temperatures studied, increasing deposition temperature causes TiO₂ anatase nucleation onset at a lower number of TDMAT/H₂O/air ALD cycles, and therefore at thinner film thicknesses.

The data series presented in Figure 7 exhibit a sigmoidal shape similar to that expected for a crystallization kinetics curve. Thus, we attempt to fit this data to the Johnson-Mehl-Avrami-Kolmogorov (JMAK) equation [41–43]. The general form of the JMAK equation (Equation 1) is applied to model the fraction transformed to the crystalline phase X (t) over ALD process time t (seconds):

X (t)= 1 − exp(−ktⁿ) Equation 1

The JMAK equation as presented in Equation 1 is linearized to the following form:

$\ln [\ln \frac{1}{1 - X (t)}] = n \ln k + n \ln t$ Equation 2

Figure 8a plots $\ln [\ln \frac{1}{1 - X (t)}]$ vs. ln t for each ALD temperature. Here k is the reaction rate for TiO₂ anatase nucleation and crystal growth and n is the Avrami exponent. An Arrhenius relationship is applied to the reaction rates for a given ALD process temperature to describe:

$k = k_{0} * \exp (\frac{- E_{n u c - g r o w t h}}{k_{B} T_{A L D}})$ Equation 3

In Equation 3, E_nuc−growth is the activation energy for crystal nucleation and growth, k_B is the Boltzmann constant, T_ALD is the ALD process temperature, and k₀ is a material-dependent frequency factor. The linearized form of Equation 3 is:

$\ln (k) = (- \frac{E_{n u c - g r o w t h}}{k_{B}}) \frac{1}{T_{A L D}} + \ln (k_{0})$ Equation 4

Figure 8b plots ln k vs. 1/T_ALD to determine the activation energy for anatase nucleation and growth in this system. The activation energy for anatase phase nucleation and growth from TDMAT/H₂O/air ALD is calculated to be 0.09 eV atom^-1. In our previous work, we had reported that the activation energy for anatase phase nucleation and growth to be ~1.5 eV atom^-1 for annealing an amorphous TiO₂ thin film after deposition, with most of that attributed to the nucleation step [30]. Interestingly, the activation energy computed here is an order of magnitude lower than this prior value, suggesting that including a second oxidant has reduced the barrier to crystallization.

Figure 7.

Percent crystallinity determined from image analysis of top-view SEM micrographs versus ALD process time for TDMAT/H₂O/air ALD films deposited 180 °C, 200 °C, and 220 °C.

Figure 8.

(a) Linearized JMAK Equation X(t) vs. time relationships for 180 °C, 200 °C, and 220 °C TDMAT/H₂O/air ALD temperature. (b) ln k plotted at each ALD temperature to extract the activation energy for anatase nucleation and growth.

To further support this claim, we want to evaluate how including this additional oxidant dose within each cycle accelerates the crystallization kinetics compared to a post-deposition anneal. Figure 9a shows the microstructure for a 15 nm TDMAT/H₂O/air ALD-TiO₂ thin film deposited at 200 °C with 29% crystallinity. This is a TDMAT/H₂O/35 s air process with 341 cycles resulting in a total process time of 8.66 h. To evaluate the effect of ex situ PDA, this TiO₂ film is removed from the ALD chamber and placed in an oven at 200 °C and annealed in air for 0.47 h. As seen in Figure 9b, after annealing there is a slight increase in percent crystallinity with crystals appearing larger in size and potentially additional nucleation occurring. To evaluate if the TDMAT/H₂O/air ALD process is more effective at accelerating the crystallization kinetics than the post-deposition annealing process, the in situ, layer-by-layer air anneal time is increased from 35 s to 40 s. This increase in TDMAT/H₂O/40 s air ALD cycle time increases the total process time from 8.66 h to 9.13 h, equivalent to the total process time for the TDMAT/H₂O/35 s air ALD + PDA reported in Figure 9b. Figure 9c shows the resultant microstructure for this 15 nm TDMAT/H₂O/40 s air ALD-TiO₂ film deposited at 200 °C. Unlike the other processes, the TDMAT/H₂O/40 s air ALD process yields a nearly fully crystalline microstructure (> 99%). Thus, extending the in situ oxidation process significantly accelerates the crystallization kinetics compared to the post-deposition anneal, presumably because the atomic motion during the in situ anneal process are able to follow surface-diffusion pathways rather than bulk diffusion pathways necessary for the PDA process.

Finally, we investigate whether extending other purge steps can also affect crystal growth presuming sufficient oxidation is accomplished in the air exposure step. To do this, we examine what happens if we extend the overall cycle time by adding 5 s to the purge steps of the post-H₂O dose N₂ purge, the air anneal, and the post-air N₂ purge. Figure 10a shows the resultant microstructure for the standard TDMAT/H₂O/35 s air process deposited from 341 cycles at 200 °C (8.66 h total). With 341 cycles and 200 °C remaining constant, Figure 10b shows the resultant microstructure for the TDMAT/H₂O/air ALD process when 5 s is added to the post-H₂O N₂ purge: 1/5/0.4/20/35/35 s dose scheme, 9.13 h total. Figure 10c shows the resultant microstructure when 5 s is added to the air dose: 1/5/0.4/15/40/35 s dose scheme, 9.13 h total. Figure 10d shows the microstructure when 5 s is added to the post-air N₂ purge: 1/5/0.4/15/35/40 s dose scheme, 9.13 h total. Interestingly, all of these treatments lead to a significant increase in crystallization. This result suggests that the base condition, dose scheme 1/5/0.4/15/35/35 s with 8.66 h total process duration, provides significant oxidation to fully oxidize the amorphous film, and only additional time at temperature is needed to drive the surface diffusion processes for crystallization.

Figure 9.

SEM micrographs to compare in situ layer-by-layer anneal (TDMAT/H₂O/air ALD) vs. ex situ PDA air anneal. (a) 15 nm TDMAT/H₂O/35 s air ALD-TiO₂ deposited at 200 °C with 8.66 h total process time. (b) 15 nm TDMAT/H₂O/35 s air ALD-TiO₂ deposited at 200 °C with 8.66 h process time plus 0.47 h post-deposition anneal (PDA) in an oven in air at 200 °C, resulting in a total time of 9.13 h at 200 °C. (c) 15 nm TDMAT/H₂O/40 s air ALD-TiO₂ deposited at 200 °C with 9.13 h total process time.

Figure 10.

SEM micrographs for TDMAT/H₂O/air ALD-TiO₂ 15 nm thin films deposited at 200 °C from 341 cycles: (a) baseline 35 s air anneal process for 91.4 s total cycle time, (b) post-H₂O N₂ purge increased from 15 s to 20 s for 96.4 s total cycle time, (c) 40 s air dose for 96.4 s total cycle time, and (d) post-air N₂ time increased from 35 s to 40 s for 96.4 s total cycle time. In the image titles, x refers to a 1 s TDMAT/5 s N₂/0.4 s H₂O dosing scheme, which is held constant for the different conditions.

Preliminary investigation of TDMAT/H₂O/air ALD process optimization

Finally, some preliminary investigations of TDMAT/H₂O/air ALD templating and interleaving are reported. The purpose of these experiments is to test how two-precursor and three-precursor ALD processes can be combined in one, faster process, such that in situ air anneals are not included during every process cycle. First, a crystalline TDMAT/H₂O/air ALD sub-film is deposited on silicon, followed by a TDMAT/H₂O ALD-TiO₂ film to test if the TDMAT/H₂O/air ALD film can successfully serve as a template for continued anatase growth with subsequent ALD cycles. Second, to explore if the second oxidant air dose is required for each ALD cycle, TDMAT/H₂O ALD cycles are interleaved with TDMAT/H₂O/air ALD cycles.

If templating is effective at maintaining anatase crystal growth, it would only be necessary to use TDMAT/H₂O/air ALD for some of the initial cycles, with subsequent cycles being faster TDMAT/H₂O ALD cycles. Figure 11a presents the TiO₂ film grown from 568 cycles of TDMAT/H₂O/air ALD at 180 °C. Figure 11b presents the TiO₂ film grown from 568 cycles of TDMAT/H₂O/air ALD at 180 °C with an additional 552 cycles of TDMAT/H₂O ALD for a total of 1120 cycles. Figure 11c presents the TiO₂ film grown from 1120 cycles of TDMAT/H₂O/air ALD at 180 °C. Encouragingly, the templated film (Figure 10b) appears to have a similar amount of amorphous regions as the 568 cycles TDMAT/H₂O/air ALD film (Figure 10a), although the crystallization does appear to be receding, with grain boundary regions appearing larger. As such, templating appears to be effective to some extent, implying that TDMAT/H₂O ALD can sustain the deposited anatase structure, although it does not appear to be able to coarsen the existing grains.

To explore if the second oxidant air dose is required for each ALD cycle, TDMAT/H₂O ALD cycles are interleaved with TDMAT/H₂O/air ALD cycles. As previously discussed, ALD literature has reported that crystallization can be thickness dependent in that enough monolayers need to be deposited before crystallization can occur. Here we probe if the TDMAT/H₂O/ 35 s air ALD cycle with the in situ air anneal can promote crystallization of a few ALD layers. In the notation for this recipe, 1:1 TDMAT/H₂O ALD to TDMAT/H₂O/air ALD denotes one TDMAT/H₂O ALD cycle followed by one complete TDMAT/H₂O/35 s air ALD cycle. This cycle ratio is then super-cycled to achieve the desired total number of ALD cycles. The following interleaves have been tested: 1:1, 2:1, and 3:1 TDMAT/H₂O to TDMAT/H₂O/air ALD at ~1120 total cycles. SEM micrographs for these ~1120 cycle films deposited at 180 °C are presented in Figure 12 and grazing incidence X-ray diffraction results for these films are presented in Figure 13 to verify the presence of TiO₂ anatase. The TDMAT/H₂O/air ALD film had the strongest anatase signal, with 2θ = 25.3°, 37.8°, and 48.0° all visible. The quantity of identifiable peaks decreases for the interleaved 1:1 and 3:1 films; this could be indicative of decreasing detectable crystallinity. However, plan-view SEM micrographs do not reveal a significant difference in microstructure. While subtle differences may exist in the microstructure, additional metrology will be required to determine if interleaving is significantly degrading crystalline quality. However, at least to a first order, these preliminary experiments suggest that interleaving may be possible for simplifying process schemes.

Figure 11.

SEM micrographs for films deposited at 180 °C: (a) 568 cycles TDMAT/H₂O/air ALD, (b) 568 cycles TDMAT/H₂O/air followed by 552 cycles TDMAT/H₂O ALD, and (c) 1120 cycles TDMAT/H₂O/air ALD.

Figure 12.

SEM micrographs of TiO₂ films grown from ~1120 total cycles (a) TDMAT/H₂O/air ALD, (b) 1:1 TDMAT/H₂O to TDMAT/H₂O/air ALD, (c) 2:1 TDMAT/H₂O to TDMAT/H₂O/air ALD, (d) 3:1 TDMAT/H₂O to TDMAT/H₂O/air ALD deposited at 180 °C.

Figure 13.

GIXRD scans for 1120 total cycles TiO₂ thin films deposited at 180 °C. Bottom to top: Si wafer substrate, TDMAT/H₂O ALD, 3:1 TDMAT/H₂O ALD to TDMAT/H₂O/air ALD, 1:1 TDMAT/H₂O ALD to TDMAT/H₂O/air ALD, and TDMAT/H₂O/air ALD. The three primary TiO₂-anatase peaks are labeled where present: 2θ = 25.3°, 37.8°, 48.0°.

Conclusion

We present the use of an air dose as a third ALD precursor during the TDMAT/H₂O ALD cycle to accelerate the deposition of crystalline anatase TiO₂ thin films at deposition temperatures and film thicknesses not previously reported with H₂O or O₃ chemistry. Depositing at 180 °C, we report one of the lowest deposition temperatures to achieve TiO₂ crystallinity by TDMAT and H₂O chemistry both with or without an activated species. Consistent with previous reports, TiO₂ films produced from thermal-ALD with the TDMAT/H₂O chemistry exhibit significant Ti³⁺ states in XPS. However, upon the use of a layer-by-layer air anneal via TDMAT/H₂O/air ALD, it is possible to achieve TiO₂ films that are almost fully Ti⁴⁺. This full oxidation is believed to lower the activation barrier to crystallization, enabling nucleation at lower temperatures and film thicknesses than previously reported. For the TDMAT/H₂O/air ALD process, 180 °C and 200 °C deposition temperatures still satisfy ALD self-limiting behavior and increasing air dose time and ALD cycle time in general increases crystallinity. Finally, this TDMAT/H₂O/air ALD is found to be more effective at achieving anatase crystallization in shorter time periods and at lower temperatures than post-deposition annealing (PDA).

Acknowledgements

This material is based upon work supported by the National Science Foundation Graduate Research Fellowship under Grant No. DGE-2039655. A significant portion of this work was performed at the Georgia Tech Institute for Electronics and Nanotechnology, a member of the National Nanotechnology Coordinated Infrastructure (NNCI), which is supported by the National Science Foundation (ECCS-2025462). Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. J.P.W. thanks D. Tavakoli, Rathi, Todd for assistance with metrology.

References

1. Badovinac IJ, Peter R, Omerzu A, Salamon K, Šarić I, Samaržija A, Perčić M, Piltaver IK, Ambrožić G, Petravić M (2020) Grain size effect on photocatalytic activity of TiO₂ thin films grown by atomic layer deposition. Thin Solid Films 709: 138215. https://doi.org/10.1016/j.tsf.2020.138215

2. Hu S, Shaner MR, Beardslee JA, Lichterman M, Brunschwig BS, Lewis NS (2014) Amorphous TiO₂ coatings stabilize Si, GaAs, and GaP photoanodes for efficient water oxidation. Science 344: 1005–1009. https://doi.org/10.1126/science.1251428

3. Seo S, Shin S, Kim E, Jeong S, Park N-G, Shin H (2021) Amorphous TiO₂ coatings stabilize perovskite solar cells. ACS Energy Letters 6: 3332–3341. https://doi.org/10.1021/acsenergylett.1c01446

4. Kim SK, Kim W-D, Kim K-M, Hwang CS, Jeong J (2004) High dielectric constant TiO₂ thin films on a Ru electrode grown at 250 C by atomic-layer deposition. Applied physics letters 85: 4112–4114. https://doi.org/10.1063/1.1812832

5. George SM (2010) Atomic layer deposition: an overview. Chemical Reviews 110: 111–131. https://doi.org/10.1021/cr900056b

6. Niemelä J-P, Marin G, Karppinen M (2017) Titanium dioxide thin films by atomic layer deposition: a review. Semiconductor Science and Technology 32: 093005. https://doi.org/10.1088/1361-6641/aa78ce

7. Dufond ME, Diouf MW, Badie C, Laffon C, Parent P, Ferry D, Grosso D, Kools JC, Elliott SD, Santinacci L (2020) Quantifying the extent of ligand incorporation and the effect on properties of TiO₂ thin films grown by atomic layer deposition using an alkoxide or an alkylamide. Chemistry of Materials 32: 1393–1407. https://doi.org/10.1021/acs.chemmater.9b03621

8. Xie Q, Jiang Y-L, Detavernier C, Deduytsche D, Van Meirhaeghe RL, Ru G-P, Li B-Z, Qu X-P (2007) Atomic layer deposition of TiO2 from tetrakis-dimethyl-amido titanium or Ti isopropoxide precursors and H₂O. Journal of applied physics 102: 083521. https://doi.org/10.1063/1.2798384

9. Lim GT, Kim D-H (2006) Characteristics of TiOx films prepared by chemical vapor deposition using tetrakis-dimethyl-amido-titanium and water. Thin Solid Films 498: 254–258. https://doi.org/10.1016/j.tsf.2005.07.121

10. Xie Q, Musschoot J, Deduytsche D, Van Meirhaeghe RL, Detavernier C, Van den Berghe S, Jiang Y-L, Ru G-P, Li B-Z, Qu X-P (2008) Growth kinetics and crystallization behavior of TiO₂ films prepared by plasma enhanced atomic layer deposition. Journal of The Electrochemical Society 155: H688. https://doi.org/10.1149/1.2955724

11. Norton ET, Amato-Wierda C (2001) Kinetic and mechanistic studies of the thermal decomposition of Ti(N(CH₃)₂)₄ during chemical vapor deposition by in situ molecular beam mass spectrometry. Chemistry of Materials 13: 4655–4660. https://doi.org/10.1021/cm0104708

12. Saari J, Ali-Löytty H, Lahtonen K, Hannula M, Palmolahti L, Tukiainen A, Valden M (2022b) low-temperature route to direct amorphous to rutile crystallization of TiO₂ thin films grown by atomic layer deposition. The Journal of Physical Chemistry C 126: 15357–15366. https://doi.org/10.1021/acs.jpcc.2c04905

13. Jin C, Liu B, Lei Z, Sun J (2015) Structure and photoluminescence of the TiO₂ films grown by atomic layer deposition using tetrakis-dimethylamino titanium and ozone. Nanoscale Research Letters 10: 1–9. https://doi.org/10.1186/s11671-015-0790-x

14. Kim Y-W, Kim D-H (2012) Atomic layer deposition of TiO₂ from tetrakis-dimethylamido-titanium and ozone. Korean Journal of Chemical Engineering 29: 969–973. https://doi.org/10.1007/s11814-012-0072-6

15. Pheamhom R, Sunwoo C, Kim D-H (2006) Characteristics of atomic layer deposited TiO₂ films and their photocatalytic activity. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 24: 1535–1539. https://doi.org/10.1116/1.2172941

16. Huang Y, Pandraud G, Sarro PM (2013) Characterization of low temperature deposited atomic layer deposition TiO₂ for MEMS applications. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 31: 01A148. https://doi.org/10.1116/1.4772664

17. Nabatame T, Ohi A, Chikyo T, Kimura M, Yamada H, Ohishi T (2014) Electrical properties of anatase TiO₂ films by atomic layer deposition and low annealing temperature. Journal of Vacuum Science & Technology B, Nanotechnology and Microelectronics: Materials, Processing, Measurement, and Phenomena 32: 03D121. https://doi.org/10.1116/1.4869059

18. Aarik J, Aidla A, Uustare T, Sammelselg V (1995) Morphology and structure of TiO2 thin films grown by atomic layer deposition. Journal of Crystal Growth 148: 268–275. https://doi.org/10.1016/0022-0248(94)00874-4

19. Jõgi I, Pärs M, Aarik J, Aidla A, Laan M, Sundqvist J, Oberbeck L, Heitmann J, Kukli K (2008) Conformity and structure of titanium oxide films grown by atomic layer deposition on silicon substrates. Thin Solid Films 516: 4855–4862. https://doi.org/10.1016/j.tsf.2007.09.008

20. Lee W-J, Hon M-H (2010) space-limited crystal growth mechanism of TiO₂ films by atomic layer deposition. The Journal of Physical Chemistry C 114: 6917–6921. https://doi.org/10.1021/jp911210q

21. Aarik J, Aidla A, Uustare T, Ritala M, Leskelä M (2000) Titanium isopropoxide as a precursor for atomic layer deposition: characterization of titanium dioxide growth process. Applied Surface Science 161: 385–395. https://doi.org/10.1016/S0169-4332(00)00274-9

22. Qiu S, Starr TL (2007) Zirconium doping in titanium oxide photocatalytic films prepared by atomic layer deposition. Journal of The Electrochemical Society 154: H472. https://doi.org/10.1149/1.2718475

23. Kim WD, Hwang GW, Kwon OS, Kim SK, Cho M, Jeong DS, Lee SW, Seo MH, Hwang CS, Min Y-S (2005) Growth characteristics of atomic layer deposited TiO₂ thin films on Ru and Si electrodes for memory capacitor applications. Journal of The Electrochemical Society 152: C552. https://doi.org/10.1149/1.1943589

24. Nunez P, Richter MH, Piercy BD, Roske CW, Cabán-Acevedo M, Losego MD, Konezny SJ, Fermin DJ, Hu S, Brunschwig BS (2019) Characterization of electronic transport through amorphous TiO₂ produced by atomic layer deposition. The Journal of Physical Chemistry C 123: 20116–20129. https://doi.org/10.1021/acs.jpcc.9b04434

25. Pan X, Yang M-Q, Fu X, Zhang N, Xu Y-J (2013) Defective TiO₂ with oxygen vacancies: synthesis, properties and photocatalytic applications. Nanoscale 5: 3601–3614. https://doi.org/10.1039/c3nr00476g

26. Babadi AS, Tang-Kong R, McIntyre PC (2021) Link between gas phase reaction chemistry and the electronic conductivity of atomic layer deposited titanium oxide thin films. The Journal of Physical Chemistry Letters 12: 3625–3632. https://doi.org/10.1021/acs.jpclett.1c00115

27. Saari J, Ali-Löytty H, Kauppinen MM, Hannula M, Khan R, Lahtonen K, Palmolahti L, Tukiainen A, Grönbeck H, Tkachenko NV, Valden M (2022a) Tunable Ti3+-mediated charge carrier dynamics of atomic layer deposition-grown amorphous TiO₂. The Journal of Physical Chemistry C 126: 4542–4554. https://doi.org/10.1021/acs.jpcc.1c10919

28. Weimer MS, Kim IS, Guo P, Schaller RD, Martinson AB, Hock AS (2017) Oxidation state discrimination in the atomic layer deposition of vanadium oxides. Chemistry of Materials 29: 6238–6244. https://doi.org/10.1021/acs.chemmater.7b01130

29. Piercy BD, Losego MD (2015) Tree-based control software for multilevel sequencing in thin film deposition applications. Journal of Vacuum Science & Technology B 33: 043201. https://doi.org/10.1116/1.4926676

30. Wooding JP, Gregory SA, Atassi A, Freychet G, Kalaitzidou K, Losego MD (2023a) Transformation kinetics for low temperature post-deposition crystallization of TiO₂ thin films prepared via atomic layer deposition from TDMAT and water. Atomic Layer Deposition 1: 1–18. https://doi.org/10.3897/aldj.1.101276

31. Biesinger MC, Lau LW, Gerson AR, Smart RSC (2010) Resolving surface chemical states in XPS analysis of first row transition metals, oxides and hydroxides: Sc, Ti, V, Cu and Zn. Applied Surface Science 257: 887–898. https://doi.org/10.1016/j.apsusc.2010.07.086

32. Naumkin AV, Kraut-Vass A, Gaarenstroom SW, Powell CJ (2012) NIST X-ray Photoelectron Spectroscopy Database 20, Version 4.1.

33. Hearne G, Zhao J, Dawe A, Pischedda V, Maaza M, Nieuwoudt M, Kibasomba P, Nemraoui O, Comins J, Witcomb M (2004) Effect of grain size on structural transitions in anatase TiO₂: A Raman spectroscopy study at high pressure. Physical Review B 70: 134102. https://doi.org/10.1103/PhysRevB.70.134102

34. Tian F, Zhang Y, Zhang J, Pan C (2012) Raman spectroscopy: a new approach to measure the percentage of anatase TiO2 exposed (001) facets. The Journal of Physical Chemistry C 116: 7515–7519. https://doi.org/10.1021/jp301256h

35. Martínez H, Cáceres MF, Martínez F, Páez-Mozo EA, Valange S, Castellanos NJ, Molina D, Barrault J, Arzoumanian H (2016) Photo-epoxidation of cyclohexene, cyclooctene and 1-octene with molecular oxygen catalyzed by dichloro dioxo-(4, 4′-dicarboxylato-2, 2′-bipyridine) molybdenum (VI) grafted on mesoporous TiO₂. Journal of Molecular Catalysis A: Chemical 423: 248–255. https://doi.org/10.1016/j.molcata.2016.07.006

36. Abendroth B, Moebus T, Rentrop S, Strohmeyer R, Vinnichenko M, Weling T, Stöcker H, Meyer DC (2013) Atomic layer deposition of TiO₂ from tetrakis(dimethylamino)titanium and H₂O. Thin Solid Films 545: 176–182. https://doi.org/10.1016/j.tsf.2013.07.076

37. Luka G, Witkowski BS, Wachnicki L, Andrzejczuk M, Lewandowska M, Godlewski M (2013) Kinetics of anatase phase formation in TiO₂ films during atomic layer deposition and post-deposition annealing. CrystEngComm 15: 9949–9954. https://doi.org/10.1039/c3ce40893k

38. Xia B, Ganem J-J, Briand E, Steydli S, Baron-Wiecheć AW, Vickridge I (2023) A systematical investigation of layer growth rate, impurity level and morphology evolution in TiO₂ thin films grown by ALD between 100 and 300 °C. Vacuum 216: 112408. https://doi.org/10.1016/j.vacuum.2023.112408

39. Saha D, Ajimsha R, Rajiv K, Mukherjee C, Gupta M, Misra P, Kukreja L (2014) Spectroscopic ellipsometry characterization of amorphous and crystalline TiO₂ thin films grown by atomic layer deposition at different temperatures. Applied Surface Science 315: 116–123. https://doi.org/10.1016/j.apsusc.2014.07.098

40. Miikkulainen V, Leskelä M, Ritala M, Puurunen RL (2013) Crystallinity of inorganic films grown by atomic layer deposition: Overview and general trends. Journal of Applied Physics 113: 021301. https://doi.org/10.1063/1.4757907

41. Avrami M (1939) Kinetics of phase change. I General theory. The Journal of Chemical Physics 7: 1103–1112. https://doi.org/10.1063/1.1750380

42. Avrami M (1940) Kinetics of phase change. II transformation‐time relations for random distribution of nuclei. The Journal of Chemical Physics 8: 212–224. https://doi.org/10.1063/1.1750631

43. Avrami M (1941) Granulation, phase change, and microstructure kinetics of phase change. III. The Journal of Chemical Physics 9: 177–184. https://doi.org/10.1063/1.1750872

E-mail and ORCID

Jamie P. Wooding (jwooding3@gatech.edu)
Kyriaki Kalaitzidou (kyriaki.kalaitzidou@me.gatech.edu)
Mark D. Losego (Corresponding author, losego@gatech.edu), ORCID: https://orcid.org/0000-0002-9810-9834

Supplementary material

Supplementary material 1

Supplementary images

Jamie P. Wooding, Kyriaki Kalaitzidou, Mark D. Losego

Data type: docx

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.

https://doi.org/10.3897/aldj.2.117753.suppl1

Download file (1.18 MB)