An official website of the United States government
Thin film deposition is a fundamental technology for the discovery, optimization, and manufacturing of functional materials. Deposition by molecular beam epitaxy (MBE) typically employs reflection high-energy electron diffraction (RHEED) as a real-time in situ probe of the growing film. However, the state-of-the-art for RHEED analysis during deposition requires human observation. Here, we present an approach using machine learning (ML) methods to monitor, analyze, and interpret RHEED images on-the-fly during thin film deposition. In the analysis workflow, RHEED pattern images are collected at one frame per second and featurized using a pretrained deep convolutional neural network. The feature vectors are then statistically analyzed to identify changepoints; these changepoints can be related to changes in the deposition mode from initial film nucleation to a transition regime, smooth film deposition, and in some cases, an additional transition to a rough, islanded deposition regime. The feature vectors are additionally analyzed via graph analysis and community classification. The graph is quantified as a stabilization plot, and we show that inflection points in the stabilization plot correspond to changes in the growth regime. The full RHEED analysis workflow is termed RHAAPsody and includes data transfer and output to a visual dashboard. We demonstrate the functionality of RHAAPsody by analyzing the precaptured RHEED images from epitaxial depositions of anatase TiO2 on SrTiO3(001) and show that the analysis workflow can be executed in less than 1 s. Our approach shows promise as one component of ML-enabled real-time feedback control of the MBE deposition process.
more »
« less
This content will become publicly available on March 1, 2027
RHEED pattern classification by a convolutional neural network for the growth of chalcogenide thin films and nanostructures
The use of reflection high-energy electron diffraction (RHEED) plays a critical role for in situ characterization in molecular beam epitaxy, pulsed laser deposition, and sputtering. While sensitive to crystal symmetries and morphology, it is used ubiquitously to determine the growth modes of thin films. However, analysis of RHEED patterns depends on skilled experts and is, therefore, difficult to incorporate into the growth strategy in real time. The development of machine learning (ML) processes, specifically convolutional neural networks (CNNs), presents a unique opportunity toward real-time RHEED pattern recognition. In this study, we develop a CNN model that can accurately classify four common and distinct RHEED patterns present in chalcogenide thin film growth. Its accuracy reached 94.9% for single run and 91.2% when averaged over 20 seeds. Our network is able to distinguish the nucleation of three common growth modes encountered in epitaxy, namely, Volmer–Weber (VW), Stransky–Krastanov (SK), and Frank–van der Merwe, potentially enabling future automation of substrate temperature and shutter control informed by RHEED data. The network is material-agnostic and distinguishes the VW process with greater than 98% accuracy but is somewhat more limited in its ability to properly classify roughening and the initiation of SK growth. Our findings show that ML techniques can be successfully implemented even in cases where there is no detailed knowledge of growth chemistry providing an avenue toward real-time incorporation of ML to control nanostructure nucleation and thin film morphology.
more »
« less
- Award ID(s):
- 2313441
- PAR ID:
- 10693440
- Publisher / Repository:
- American Institute of Physics
- Date Published:
- Journal Name:
- AIP Advances
- Volume:
- 16
- Issue:
- 3
- ISSN:
- 2158-3226
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Perovskite oxides such as LaFeO3 are a well-studied family of materials that possess a wide range of useful and novel properties. Successfully synthesizing perovskite oxide samples usually requires a significant number of growth attempts and a detailed film characterization on each sample to find the optimal growth window of a material. The most common real-time in situ diagnostic technique available during molecular beam epitaxy (MBE) synthesis is reflection high-energy electron diffraction (RHEED). Conventional use of RHEED allows a highly experienced operator to determine growth rate by monitoring intensity oscillations and make some qualitative observations during growth, such as recognizing the sample has become amorphous or recognizing that large islands have formed on the surface. However, due to a lack of theoretical understanding of the diffraction patterns, finer, more precise levels of observations are challenging. To address these limitations, we implement new data analytics techniques in the growth of three LaFeO3 samples on Nb-doped SrTiO3 by MBE. These techniques improve our ability to perform unsupervised machine learning using principal component analysis (PCA) and k-means clustering by using drift correction to overcome sample or stage motion during growth and intensity transformations that highlight more subtle features in the images such as Kikuchi bands. With this approach, we enable the first demonstration of PCA and k-means across multiple samples, allowing for quantitative comparison of RHEED videos for two LaFeO3 film samples. These capabilities set the stage for real-time processing of RHEED data during growth to enable machine learning-accelerated film synthesis.more » « less
-
Design of an atomic layer deposition system with in situ reflection high energy electron diffractionWe report the design, fabrication, and testing of an atomic layer deposition (ALD) system that is capable of reflection high energy electron diffraction (RHEED) in a single chamber. The details and specifications of the system are described and include capabilities of RHEED at varied accelerating voltages, sample rotation (azimuthal) control, sample height control, sample heating up to set temperatures of 1050 °C, and either single- or dual-differential pumping designs. Thermal and flow simulations were used to justify selected system dimensions as well as carrier gas/precursor mass flow rates. Temperature calibration was conducted to determine actual sample temperatures that are necessary for meaningful analysis of thermally induced transitions in ALD thin films. Several demonstrations of RHEED in the system are described. Calibration of the camera length was conducted using a gold thin film by analyzing RHEED images. Finally, RHEED conducted at a series of increasing temperatures was used to monitor the crystallization of an ALD HfO2 thin film. The crystallization temperature and the ring pattern were consistent with the monoclinic structure as determined by separate x-ray diffraction-based measurements.more » « less
-
We investigate the surface morphologies of two series of homoepitaxial GaSb(100) thin films grown on GaSb(100) substrates by molecular beam epitaxy in a Veeco GENxplor system. The first series was grown at temperatures ranging from 290 to 490°C and serves as a control. The second series was grown using the same growth parameters with bismuth used as a surfactant during growth. We compared the two series to examine the impacts of bismuth over the range of growth temperatures on the GaSb surface morphologies using atomic force microscopy and the film properties using Raman spectroscopy and scanning electron microscopy. High-resolution x-ray diffraction was performed to confirm that bismuth was not incorporated into the films. We found that the morphological evolution of the GaSb series grown without bismuth is consistent with the standard surface nucleation theory and identified the 2D-3D transition temperature as close to 290° C. In contrast, the presence of a Bi surfactant during growth was found to significantly alter the surface morphology and prevent undesired 3D islands at low temperatures. We also observed a preference for hillocks over step morphology at high growth temperatures, antistep bunching effects at intermediate temperatures, and the evolution from step-meandering to mound morphologies at low temperatures. This morphological divergence from the first series indicates that bismuth significantly increases in the 2D Erlich–Schwöebel potential barrier of the atomic terraces, inducing an uphill adatom flux that can smoothen the surface. Our findings demonstrate that bismuth surfactant can improve the surface morphology and film structure of low-temperature grown GaSb. Bismuth surfactant may also improve other homoepitaxial III-V systems grown in nonideal conditions.more » « less
-
In this work, we present the investigation of InN/GaN multiple-quantum-well (MQW) growth by plasma-assisted molecular beam epitaxy using in-situ reflection high-energy electron diffraction (RHEED) to monitorthe growth process. The analysis of the RHEED intensity and pattern transitions identified an indium surface ac-cumulation even with a nominal thickness of InN as small as 0.5 monolayer (ML). This result explicitly showsthat, even at low growth temperaturesof ~550 °C, not all of the supplied indium isincorporated into the quantumwell (QW). Moreover, the residual indium can become incorporated into the GaN matrix on either side of theQW. Both QW thickness and the photoluminescence (PL) emission energy showed a self-regulating behavior.The apparent thickness did not exceed 2 MLs even when the deposited InN thickness is as large as 5 MLs. ThePL emission shows a continuous redshift with the deposited InN from ~370 nm for 0.5 ML until it saturates at~423 nm forN2 ML. Based on the observed growth phenomena, a qualitative growth model was developed to ex-plain the self-limited epitaxial growth of ultrathin In(Ga)N/GaN QWsmore » « less
