No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Modeling incomplete conformality during atomic layer deposition in high aspect ratio structures
Abstract
Atomic layer deposition allows for precise control over film thickness and conformality. It is a critical enabler of high aspect ratio structures, such as 3D NAND memory, since its self-limiting behavior enables higher conformality than conventional processes. However, as the aspect ratio increases, deviations from complete conformality frequently occur, requiring comprehensive modeling to aid the development of novel technologies. To that end, we present a model for surface coverage during atomic layer deposition where incomplete conformality is present. This model combines existing approaches based on Knudsen diffusion and Langmuir kinetics. Our model expands the state-of-the art by (i) incorporating gas-phase diffusivity through the Bosanquet formula as well as reaction reversibility in the modeling framework first proposed by Yanguas-Gil and Elam, and (ii) being efficiently integrated within level-set topography simulators. The model is manually calibrated to published results of the prototypical atomic layer deposition of Al2O3 from TMA and H2O in lateral high aspect ratio structures. We investigate the temperature dependence of the H2O step, thus extracting an activation energy of 0.178eV which is consistent with recent experiments. In the TMA step, we observe increased accuracy from the Bosanquet formula and we reproduce multiple independent experiments with the same parameter set, highlighting that the model parameters effectively capture the reactor conditions.
... Originally, these models were developed to aid the design of chemical vapor deposition (CVD) reactors [5]. More recently, significant contributions were made in applying them to atomic layer deposition (ALD) [6]- [9]. However, many of the historical misconceptions relating to the applicability of Knudsen diffusion to features with arbitrary cross-sections re-appeared in these recent works. ...
Unparalleled conformality is driving ever new applications for atomic layer deposition (ALD), a thin film growth method based on repeated self-terminating gas-solid reactions. In this work, we re-implemented a diffusion-reaction model from the literature to simulate the propagation of film growth in wide microchannels and used that model to explore trends in both the thickness profile as a function of process parameters and different diffusion regimes. In the model, partial pressure of the ALD reactant was analytically approximated. Simulations were made as a function of kinetic and process parameters such as the temperature, (lumped) sticking coefficient, molar mass of the ALD reactant, reactant's exposure time and pressure, total pressure, density of the grown material, and growth per cycle (GPC) of the ALD process. Increasing the molar mass and the GPC, for example, resulted in a decreasing penetration depth into the microchannel. The influence of the mass and size of the inert gas molecules on the thickness profile depended on the diffusion regime (free molecular flow vs. transition flow). The modelling was compared to a recent slope method to extract the sticking coefficient. The slope method gave systematically somewhat higher sticking coefficient values compared to the input sticking coefficient values; the potential reasons behind the observed differences are discussed.
Atomic layer deposition (ALD) raises global interest through its unparalleled conformality. This work describes new microscopic lateral high-aspect-ratio (LHAR) test structures for conformality analysis of ALD. The LHAR structures are made of silicon and consist of rectangular channels supported by pillars. Extreme aspect ratios even beyond 10 000 : 1 enable investigations where the adsorption front does not penetrate to the end of the channel, thus exposing the saturation profile for detailed analysis. We use the archetypical trimethylaluminum (TMA)-water ALD process to grow alumina as a test vehicle to demonstrate the applicability, repeatability and reproducibility of the saturation profile measurement and to provide a benchmark for future saturation profile studies. Through varying the TMA reaction and purge times, we obtained new information on the surface chemistry characteristics and the chemisorption kinetics of this widely studied ALD process. New saturation profile related classifications and terminology are proposed.
The conformality of a film grown by atomic layer deposition (ALD) is strongly affected by the reactivities of the precursor and coreactant, which can be expressed in terms of their sticking probabilities toward the surface. We show that the leading front of the thickness profile in high-aspect-ratio structures gives direct information on the sticking probabilities of the reactants under most conditions. The slope of the front has been used to determine the sticking probabilities of Al(CH3)3 and H2O during ALD of Al2O3. The determined values are (0.5–2) × 10⁻³ for Al(CH3)3 and (0.8–2) × 10⁻⁴ for H2O at a set-point temperature of 275 °C, corresponding to an estimated substrate temperature of ∼220 °C. Additionally, the thickness profiles reveal soft-saturation behavior during the H2O step, most dominantly at reduced temperatures, which can limit the conformality of Al2O3 grown by ALD. This work thus provides insights regarding quantitative information on sticking probabilities and conformality during ALD, which is valuable for gaining a deeper understanding of ALD kinetics.
Atomic layer deposition (ALD) relies on alternated, self-limiting reactions between gaseous reactants and an exposed solid surface to deposit highly conformal coatings with a thickness controlled at the submonolayer level. These advantages have rendered ALD a mainstream technique in microelectronics and have triggered growing interest in ALD for a variety of nanotechnology applications, including energy technologies. Often, the choice for ALD is related to the need for a conformal coating on a 3D nanostructured surface, making the conformality of ALD processes a key factor in actual applications. In this work, we aim to review the current status of knowledge about the conformality of ALD processes. We describe the basic concepts related to the conformality of ALD, including an overview of relevant gas transport regimes, definitions of exposure and sticking probability, and a distinction between different ALD growth types observed in high aspect ratio structures. In addition, aiming for a more standardized and direct comparison of reported results concerning the conformality of ALD processes, we propose a new concept, Equivalent Aspect Ratio (EAR), to describe 3D substrates and introduce standard ways to express thin film conformality. Other than the conventional aspect ratio, the EAR provides a measure for the ease of coatability by referring to a cylindrical hole as the reference structure. The different types of high aspect ratio structures and characterization approaches that have been used for quantifying the conformality of ALD processes are reviewed. The published experimental data on the conformality of thermal, plasma-enhanced, and ozone-based ALD processes are tabulated and discussed. Besides discussing the experimental results of conformality of ALD, we will also give an overview of the reported models for simulating the conformality of ALD. The different classes of models are discussed with special attention for the key assumptions typically used in the different modelling approaches. The influence of certain assumptions on simulated deposition thickness profiles is illustrated and discussed with the aim of shedding light on how deposition thickness profiles can provide insights into factors governing the surface chemistry of ALD processes. We hope that this review can serve as a starting point and reference work for new and expert researchers interested in the conformality of ALD and, at the same time, will trigger new research to further improve our understanding of this famous characteristic of ALD processes.
Semiconductor device dimensions have been decreasing steadily over the past several decades, generating the need to overcome fundamental limitations of both the materials they are made of and the fabrication techniques used to build them. Modern metal gates are no longer a simple polysilicon layer, but rather consist of a stack of several different materials, often requiring multiple processing steps each, to obtain the characteristics needed for stable operation. In order to better understand the underlying mechanics and predict the potential of new methods and materials, technology computer aided design has become increasingly important. This review will discuss the fundamental methods, used to describe expected topology changes, and their respective benefits and limitations. In particular, common techniques used for effective modeling of the transport of molecular entities using numerical particle ray tracing in the feature scale region will be reviewed, taking into account the limitations they impose on chemical modeling. The modeling of surface chemistries and recent advances therein, which have enabled the identification of dominant etch mechanisms and the development of sophisticated chemical models, is further presented. Finally, recent advances in the modeling of gate stack pattering using advanced geometries in the feature scale are discussed, taking note of the underlying methods and their limitations, which still need to be overcome and are actively investigated.
In this paper, the authors present the temperature dependent sticking coefficient (SC) of bis-diethyl aminosilane (BDEAS) and trimethylaluminum (TMA) in atomic layer deposition. SiO2 from BDEAS and ozone at substrate temperatures between 200 and 350 °C as well as Al2O3 from TMA and water deposited at substrate temperatures between 150 and 300 °C was deposited on our likewise in this journal published cavity test structures. The SC of BDEAS shows an Arrhenius dependence while for TMA no temperature dependent SC could be resolved. The activation energy for BDEAS which is extracted from a linear fit to the Arrhenius diagram is compared to the value gained by density functional theory calculations from the literature. Furthermore, the different growth behavior of BDEAS and TMA under substrate temperature considerations is identified with different deposition regimes as proposed in the literature.
Atomic layer deposition (ALD) is a technique for depositing thin films of materials with a precise thickness control and uniformity using the self-limitation of the underlying reactions. Usually, it is difficult to predict the result of the ALD process for given external parameters, e.g., the precursor exposure time or the size of the precursor molecules. Therefore, a deeper insight into ALD by modeling the process is needed to improve process control and to achieve more economical coatings. In this paper, a detailed, microscopic approach based on the model developed by Yanguas-Gil and Elam is presented and additionally compared with the experiment. Precursor diffusion and second-order reaction kinetics are combined to identify the influence of the porous substrate's microstructural parameters and the influence of precursor properties on the coating. The thickness of the deposited film is calculated for different depths inside the porous structure in relation to the precursor exposure time, the precursor vapor pressure, and other parameters. Good agreement with experimental results was obtained for ALD zirconiumdioxide (ZrO2) films using the precursors tetrakis(ethylmethylamido)zirconium and O2. The derivation can be adjusted to describe other features of ALD processes, e.g., precursor and reactive site losses, different growth modes, pore size reduction, and surface diffusion.
Thermal atomic layer etching (ALE) was studied in HfO2-based 3D NAND test structures with an aspect ratio of more than 50:1. Etching was performed via ligand exchange with dimethyl-aluminum chloride (DMAC) after surfaces had been fluorinated with hydrogen fluoride (HF). In these 3D NAND structures, we found that the horizontal etch rate of HfO2 as a function of depth (depth loading) depended on the DMAC dosing but was nearly independent of the HF dose. The HF dose and the process pressure were keys to increasing the overall etch amount per cycle. With the highest tested HF dose of 192 Torr s and a total process pressure of 8 Torr, we achieved a uniform etch amount of 0.6 nm per cycle. In addition, we investigated the impact of film quality and film coating conformality in these structures on the depth loading in the succeeding ALE processes. The type of precursor, precursor dosing, deposition rate, and substrate temperature played a fundamental role in controlling the film quality and conformality of the deposited HfO2 layers inside high aspect ratio holes. Fluorination studies on blanket films revealed that fluorination efficiency is improving for pressures in the Torr range compared to previous milliTorr experiments and that only temperatures above 250 °C increased the fluorine concentration in HfO2 significantly, whereas fluorine levels were unchanged between 150 and 250 °C.
We have developed a simulation tool to model self-limited processes such as atomic layer deposition (ALD) and atomic layer etching inside reactors of arbitrary geometry as well the output of in situ quartz crystal microbalance and mass spectrometry. We have applied this model to two standard types of cross-flow reactors: a cylindrical tube reactor and a model 300 mm wafer reactor, and explored both ideal and nonideal self-limited kinetics. The model results are in agreement with experimental results and analytic expressions obtained using a simple plug-flow model for the cylindrical tube reactor. We also extended the simulations to consider two nonideal self-limited processes: soft-saturating processes characterized by a slow reaction pathway and processes where surface by-products can compete with the precursor for the same pool of adsorption sites. Our results show that it is possible to have a self-limited process with saturated yet inhomogeneous growth profiles due to the competition of reactor by-products. This is in agreement with experimental observations for titanium dioxide ALD from titanium tetraisopropoxide and titanium tetrachloride precursors.
Coating porous electrodes of Li-ion batteries with functional materials by atomic layer deposition (ALD) can enhance their capacity stability. Because of the complex microstructure of the porous electrodes, predicting the required exposure time of the precursors for achieving a desired coating and the coating characteristics for a given exposure time remains difficult. Here, a three-dimensional (3D) pore-scale lattice Boltzmann model is developed to investigate the reactive transport processes during the ALD of reconstructed electrodes and to assess the accuracy of one-dimensional (1D) mean-field models. The effects of the hierarchical structure of pores and their connectivity on the coating process of ALD are investigated and the detailed coating characteristics in the electrodes are resolved. Electrodes with smaller pore sizes requires a longer exposure time to achieve full coating. At the same depth within an electrode, the smaller pores are coated more slowly than the wider pores (especially at high Damkohler numbers) and the coating speed of well-connected pores are faster than that of poorly connected pores. Simulations also reveal that the 1D mean-field model can capture the average coating characteristics reasonably well when pores in the electrodes are well connected but may perform poorly if the pores are poorly connected.
Atomic layer deposition (ALD) is a method to fabricate uniform and conformal thin films based on surface chemistry. Al2O3 (alumina) is the most widely and thoroughly studied material in the field of oxide ALD. Most previous mechanistic studies on ALD of Al2O3 have been focused on the surface reaction step of Al-containing precursor: for the popular trimethylaluminum, it is known that the surface becomes terminated with Al-CH3 groups. However, reaction mechanisms for oxygen sources other than H2O has been largely unexplored. In this study, the surface reaction mechanisms for the CH3-terminated alumina surface with H2O, H2O2, and O3 oxidants were investigated at molecular level. Density functional theory (DFT) calculations were used to obtain the energy change along reaction coordinates. It is confirmed that all three oxidants considered in this study would show facile oxidation of surface CH3 into OH under ALD conditions. Still, the exact mechanisms and number of oxidant molecules required for a full reaction to generate hydroxyl-terminated surface vary. The molecular reactivity toward ALD of Al2O3 at low deposition temperatures for the three oxidizing agents are expected to be H2O < H2O2 < O3.
The conformality of thin films grown by atomic layer deposition (ALD) is studied using all-silicon test structures with long narrow lateral channels. A diffusion model, developed in this work, is used for studying the propagation of ALD growth in narrow channels. The diffusion model takes into account the gas transportation at low pressures, the dynamic Langmuir adsorption model for the film growth and the effect of channel narrowing due to film growth. The film growth is calculated by solving the diffusion equation with surface reactions. An efficient analytic approximate solution of the diffusion equation is developed for fitting the model to the measured thickness profile. The fitting gives the equilibrium constant of adsorption and the sticking coefficient. This model and Gordon's plug flow model are compared. The simulations predict the experimental measurement results quite well for Al2O3 and TiO2 ALD processes.
Atomic layer deposition (ALD) is renowned for its step coverage in porous substrates. Several
emerging applications require a combination of this high step coverage with high throughput
ALD, like spatial ALD. Often, high throughput ALD is performed at atmospheric pressure, and
therefore, the effect of reactor pressure on the saturation dose is investigated. ALD inside porous
substrates is governed by three key parameters: the reaction probability, the pore aspect ratio, and
the precursor diffusion coefficient, of which the latter one contains the reactor pressure
dependency. The effect of these parameters on the saturation dose is validated using Monte Carlo
modeling, where the reactor pressure dependency is included through the mean free path. A
reaction-limited and a diffusion-limited regime can be identified, and it is shown that for many
realistic experimental conditions, even at low reactor pressures, the saturation dose is in the
diffusion-limited regime. An expression for the pressure dependent saturation dose in the
diffusion-limited regime is derived. For small pore diameters, the saturation dose is pressure independent,
but for larger pores, higher saturation doses are required for atmospheric reactor pressures
than for low reactor pressures. However, as high reactor pressures enable much higher
precursor partial pressures than low reactor pressures, the resulting saturation times can be much
shorter at atmospheric pressure than low pressure. Often, high surface area porous substrates will
lead to supply limited conditions, and increased saturation times have to be taken into account.
These results show that the atmospheric pressure ALD can be used for high throughput ALD
inside porous substrates, as long as high precursor partial pressures and molar flows can be
applied. This is experimentally demonstrated by a near 100% step coverage obtained by atmospheric
spatial ALD of alumina in high aspect ratio pores.
Atomic layer deposition (ALD), also referred to historically as atomic layer epitaxy, is a vapor-phase deposition technique for preparing ultra-thin films with precise growth control. ALD is currently rapidly evolving, mostly driven by the continuous trend in the miniaturization of electronic devices. In addition, many other innovative technologies are increasingly benefitting from the high-quality thin films. This chapter describes-on an elementary level-the key features of ALD. A standard ALD process scheme is used to discuss the relevant concepts of the technique. Materials that can be deposited by ALD are discussed, including typical precursors and co-reactants that can be used. Several example chemistries, specifically for ALD of Al2O3, HfO2, TiN, and Pt, are presented to illustrate the variety in surface chemistry. ALD reactor types are described and, finally, some cases are addressed to illustrate the virtues and practicalities of ALD that are important to advancing present day and emerging thin-film applications.
Use of atomic layer deposition (ALD) in microelectromechanical systems (MEMS) has increased as ALD enables conformal growth on 3-dimensional structures at relatively low temperatures. For MEMS device design and fabrication, the understanding of stress and mechanical properties such as elastic modulus, hardness and adhesion of thin film is crucial. In this work a comprehensive characterization of the stress, elastic modulus, hardness and adhesion of ALD aluminum oxide (Al2O3) films grown at 110–300 °C from trimethylaluminum and water is presented. Film stress was analyzed by wafer curvature measurements, elastic modulus by nanoindentation and surface-acoustic wave measurements, hardness by nanoindentation and adhesion by microscratch test and scanning nanowear. The films were also analyzed by ellipsometry, optical reflectometry, X-ray reflectivity and time-of-flight elastic recoil detection for refractive index, thickness, density and impurities. The ALD Al2O3 films were under tensile stress in the scale of hundreds of MPa. The magnitude of the stress decreased strongly with increasing ALD temperature. The stress was stable during storage in air. Elastic modulus and hardness of ALD Al2O3 saturated to a fairly constant value for growth at 150 to 300 °C, while ALD at 110 °C gave softer films with lower modulus. ALD Al2O3 films adhered strongly on cleaned silicon with SiOx termination.
We present a general model based on a time-dependent, reaction–diffusion equation to determine the dosing times and coverage profiles in structured substrates during atomic layer deposition (ALD). We first derive expressions comprising a non-linear diffusion–reaction equation coupled to a surface kinetic equation. In their non-dimensional forms, these equations show that coverage dynamics during ALD in nanostructured substrates depend only on two non-dimensional parameters, the Damkoler and precursor excess (number of molecules per surface site in the nanostructure) numbers. Using the assumptions of molecular flow in a circular pore, we derive a general, analytic equation to predict saturation exposure times. To demonstrate the utility of our model, we derive additional expressions incorporating a precursor loss term relevant to predicting exposure times during ozone-based ALD. Because our model makes no assumptions about the diffusion coefficient or sample geometry, it can easily be adapted to describe a broad range of ALD systems such as trenches and vias, anodized alumina, or aerogels under almost any conditions including molecular, viscous, and transition flow regimes
Aluminum oxide was deposited on arsenic doped silicon, using atomic layer deposition (ALD) with either a silicon oxide or a silicon nitride interface. The physical properties of these films were investigated by elastic-recoil-detection, X-ray-photoelectron-spectroscopy and transmission electron microscopy. Special focus was given to contamination of the film and the interface, crystallization and temperature effect on diffusion. The films remained stoichiometric and did not have Al–Al clusters, even post annealing steps. Evidence of diffusion of silicon and arsenic into the dielectric and of aluminum from the film was found. Carbon and hydrogen were seen in the film and at the interface as well, whereas hydrogen diffused out of the film to some extent due to anneal. Carbon content in the layer was reduced by using O3 as oxidant. Grain size of crystalline Al2O3 films was in the order of film thickness.
A calculation is made of the rate of diffusion of "tagged" molecules in a pure gas at uniform pressure in a long capillary tube of half-length L and radius a. At pressures for which the mean free path lambda>>a, the result in the limit L-->∞ reduces to that already obtained by M. Knudsen, the diffusion coefficient D being given by 2av¯3, where v¯ is the mean molecular speed. For a capillary of finite length the diffusion coefficient is, to first order in aL, smaller than this by a factor 1-3a4L. In the opposite limit of high pressures, for which lambdaL. Thus the initial slope of D as a function of pressure is given by dDd(alambda)≅-12v¯a lnLa. It is shown that these results account for the anomalous low pressure minima observed by several investigators who have measured the specific flow GDeltap through long capillary tubes as a function of mean pressure p¯. The failure to observe such minima with porous media, for which effectively L~a in each pore, is also explained by these results. The formulae obtained here represent a rigorous solution to the long capillary diffusion problem, valid at all pressures and subject only to the limitations of the mean free path type of treatment.
Following a brief historical introduction an overview is given relating the most recent studies of rarefied gas flow to the early work of Knudsen. The first paper submitted in October 1908 (published in 1909) initiated a period of intense activity by Knudsen, Smoluchowski (1910) and, a little later, by Gaede (1913) and Langmuir (1912). This also covered the transition to the already well established hydrodynamic flow expressed in terms of the ratio of mean free path to critical apparatus dimension: which is now referred to as the Knudsen number. The desorption, evaporation and scattering of molecules from surfaces was described in terms of the Knudsen cosine law of scattering. The Knudsen effusion method for determining vapour pressure, also introduced in 1909, has become the main tool for studies of the related problem of the dissociation, chemical bonding and the vaporisation process itself. Clausing (1926) developed, as an alternative to conductances, the concept of transmission probability, still referred to as the Clausing factor, and provided a procedure for their more accurate evaluation in long and short tubes. A number of misconceptions of these early efforts have found their way into the literature and current books on vacuum science and technology. However, detailed studies have clarified the problem of gas-surface interactions; the gas flow in tubes has been tackled with Clausing-type integral equations and by statistical computation techniques based on Monte Carlo analysis procedures adaptable to more complex systems. Results have been confirmed experimentally using molecular-impact pressure probe measuring techniques.
A feature scale simulator for atomic layer deposition (ALD) is presented that combines a Boltzmann equation transport model with chemistry models. A simple but instructive chemistry is considered; one reactant species adsorbs onto the surface, and a second reactant reacts with it from the gas phase (Eley–Rideal). This work includes potential desorption of the adsorbed species during purge steps, which may or may not play a role in any given ALD system. Three sets (cases) of rate parameters are chosen to compare chemical rates with transport rates. The duration of the ALD pulses and the geometry of the representative feature are the same for each case. Simulation results are presented for all four steps in one ALD cycle, adsorption, post-adsorption purge, reaction, and post-reaction purge. The results are extended to multiple ALD cycles, and the monolayers per cycle are estimated. We highlight the potential trade-off between pulse durations and deposition rate (wafer throughput); e.g. the time penalty required to increase the amount adsorbed during the adsorption step. The simulation methodology we present can be used to determine the pulse durations that maximize throughput for a given chemistry and chemical rate parameters. One overall observation is that transport is fast relative to chemical reactions, for reasonable kinetic parameters. 2002 Elsevier Science B.V. All rights reserved.
Highly conformal coatings can be deposited inside narrow features, such as holes, using complementary self-limiting reactions of two vapors with a surface. A simple theory is developed for the conditions needed to deposit coatings with uniform thickness in narrow holes of arbitrary cross section. Two conditions are found to be necessary: 1) certain minimum amounts of vapor must be supplied, and 2) a certain minimum product of vapor pressure and its exposure time at the entrance to the hole. The theory is in good agreement with data from the reaction of hafnium dimethylamide vapor and water vapor to form hafnium dioxide films that are completely conformal in holes with an aspect ratio of 43. The theory provides simple scaling laws to predict the conditions needed to coat holes with other aspect ratios.
A study of the application of atomic layer deposition (ALD) technique in relation to surface chemistry of trimethylaluminum/water process, is presented. The study also provides an insight to the of surface chemistry of atomic layer deposition (ALD) technique. The study also discusses the mechanism of two-reactant ALD process and includes descriptions about the physicochemical requirements of self-terminating reactions, reaction kinetics, typical chemisorption reactions, and effect of temperature on number of cycles on growth per cycles (GPC). The issues hampering a physicochemical process are also discussed. The results of the study were compared to the growth experiments on flat substrates and reaction chemistry of high-surface-area materials.
Available online: www.silvaco.com/tcad/victory-process-3d
- Silvaco