The goal of this study was to investigate how changing the pulsed frequency affects the deposition process and correlates with AlN film properties. The resulting films were then characterized in terms of their crystallinity, microstructure, and surface roughness to identify any correlations with the pulsed frequency. This approach was used to determine the optimal pulsed conditions for film deposition. Each dataset spans the wavelength range of 190nm to 850nm, comprising 1,900 features. Following data collection, we employed traditional ensemble learning methods (Random Forest), tree-based gradient boosting (Categorical Boosting), and the improved gradient-boosted algorithm (Histogram Gradient Boosting), for predicting the quality of thin films. This analysis aimed to clarify which method excels in handling semiconductor process OES data to obtain an optimal processing pulsed frequency on reactive pulsed DC sputtering of aluminum nitride films.

The optical emission spectroscopy (OES) data provide multi-featured and high-dimension data, which exhibit rich physical phenomena. The novel unsupervised learning uniform manifold approximation and projection (UMAP) and t-distributed stochastic neighborhood embedding (t-SNE) are implemented to deal with OES data and keeping data structure during the aluminum nitride (AlN) film deposition process on Si (100) substrate. After dimensionality reduction, we hand over data to a tree-based ensemble model for training, and the performance of the model through indicators such as recall, precision, and F1-score was also determined. The optimal testing results of this study of compressive and tensile stress classification prediction were presented by UMAP including true positive (TP) 34.32%, false positive (FP) 0.21%, false negative (FN) 0.71%, and true negative (TN) 64.76%. Critical indicators were also evaluated with excellent results such as recall 0.9797, precision 0.9939, and F1-score 0.9867. Therefore, we proved a residual stress prediction with machine learning studies would become workable in the thin film deposition process.

Piezoelectric aluminum nitride (AlN) and scandium nitride (AlScN) thin films using reactive pulsed DC magnetron sputtering have shown the grown films on silicon (100) substrates without arcing during the deposition. Limited in-depth DC pulse studies have been done on the role played by variables like pulse frequency, duty cycle, and reverse voltage in the deposition process, despite the fact that many researchers have now demonstrated that pulsed DC magnetron sputtering can be used frequently to produce fully dense, defect-free films. Operating conditions were routinely altered, and the deposition technique was continuously updated. The goal of this study is to look at how the pulse parameters affected the deposition process and then how the pulse parameters and the film properties are correlated. Fifteen (15) distinct design of experiment (DOE) combinations by Box-Behnken experimental method on pulse parameters were executed for film deposition and characterized the films crystallinity, microstructures, and surface roughness to find out film properties in correlation with pulse parameters. This is the way that used for obtaining optimal pulse conditions is based on the subsequent response surface method (RSM) of DOE model.

In the present study, the sputtered aluminum nitride (AlN) films were processed in a reactive pulsed DC magnetron system. We applied a total of 15 different design of experiments (DOEs) on DC pulsed parameters (reverse voltage, pulse frequency, and duty cycle) with Box-Behnken experimental method and response surface method (RSM) to establish a mathematical model by experimental data for interpreting the relationship between independent and response variables. For the characterization of AlN films on the crystal quality, microstructure, thickness, and surface roughness, X-ray diffraction (XRD), atomic force microscopy (AFM), and field emission-scanning electron microscopy (FE-SEM) were utilized. AlN films have different microstructures and surface roughness under different pulse parameters. In addition, in-situ optical emission spectroscopy (OES) was employed to monitor the plasma in real-time, and its data were analyzed by principal component analysis (PCA) for dimensionality reduction and data preprocessing. Through the CatBoost modeling and analysis, we predicted results from XRD in full width at half maximum (FWHM) and SEM in grain size. This investigation identified the optimal pulse parameters for producing high-quality AlN films as a reverse voltage of 50 V, a pulse frequency of 250 kHz, and a duty cycle of 80.6061%. Additionally, a predictive CatBoost model for obtaining film FWHM and grain size was successfully trained.

In this work, a unique approach in experiments was carried out by a pulsed DC sputtering for the 60/30-min process in one-step and multi-step (two- and four-step) deposition to obtain trends of residual stress and film properties of aluminum nitride (AlN). Film characteristics comparison between one-step and multi-step deposition was used to study the correlation among film crystallite orientation, residual stress, and film features. The thickness, microstructure, residual stress, and crystallite status of AlN films were examined by scanning electron microscopy (SEM), X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, and atomic force microscopy (AFM). The results show that the films have different stress characteristics in different microstructures when processed at the same amount of time in different deposition intervals. The increase in the degree of ion bombardment and induced substrate temperature raised leads to the interpretation of different microstructural changes in films. In this investigation, the film with thickness about 1300 nm for one-step deposition has a surface roughness of 2.67 nm, the grain size of 65.1 nm, and residual stress of 1575 MPa. In the four-step deposition, these values were reduced to 2.46 nm, 51.4 nm, and 841 MPa, respectively. In situ optical emission spectroscopy (OES) data were collected during plasma deposition of selected dominant wavelengths of N 2 (315 and 336 nm) and Al (394 and 396 nm) for large-scale data analysis. Based on a cross-validation test executed from OES data preprocessing, the methodology with the principal component analysis (PCA) and value of microstructure characteristics (VMC) algorithm demonstrated that this unique multi-step deposition technique can provide a good stress gradient control from microstructural analysis and can effectively classify film characteristics in AlN film deposition.

The amount of residual stress generated during the growth process of aluminum nitride (AlN) thin films prepared by pulsed DC reactive magnetron sputtering was investigated. The evaluated process parameters were pulsed DC frequency, DC power and flow gas ratio. AlN film may crack or peel from the substrate due to significant film residual stress. Therefore, the control of residual stress in films is very important for the synthesis of mechanically stable AlN films. A correlation between the residual stress and the crystal orientation of the films was studied. X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), transmission electron microscope (TEM) and scanning electron microscopy (SEM) were used to measure the residual stress, crystal structure and thickness in AlN films. The results show that AlN film has a different structure and stress characteristic under different deposition conditions. The film residual stress correlates well with the film thickness. Besides, in situ optical emission spectroscopy (OES) big data were analyzed using principal component analysis (PCA) in this study. The PC1-DEV (standard deviation in the first principal component direction) was used to calculate the value residual stress (VRS) to accurately predict and classify the stress state of the deposited film, i.e., compression stress or tensile stress. The Box–Behnken experimental design was applied, a mathematical design of experiment (DOE) model was established based on the response surface method (RSM), and the optimum conditions for generating the minimum residual stress were determined.

In this study, the minimum residual stress of aluminum nitride (AlN) thin films was obtained controllably using a multi-step deposition technique in a pulsed DC sputtering system. The link between film crystal orientation and residual stress received from X-ray diffraction (XRD) data analyses on this growing AlN experiment was investigated by the comparison of film stress characteristics between one-step deposition and multi-step deposition. The structure, thickness, crystalline status and residual stress of AlN thin films were measured using scanning electron microscope (SEM), atomic force microscope (AFM) and X-ray diffraction (XRD) respectively. The results reveal that, under the same processing duration, with various intervals of deposition, the AlN film has a distinct structure with various stress properties. It can be concluded from this experiment that the AlN film residual stress can be effectively reduced by multi-step deposition. In addition, we are aiming to examine the spectrum data acquired by optical emission spectroscopy (OES) in order to see whether there is any association with OES data, the crystalline state, and thin-film quality obtained by the measured result in this study. In our findings, OES data in conjunction with XRD analyses, it is convinced that this multi-step deposition approach can be used to determine the minimum residual stress of thin-film characteristics in AlN thin-film deposition.

Plasma enhanced chemical vapor deposition (PECVD) is commonly known to be used in the field of silicon thin-film solar systems for the application of nanocrystalline silicon (nc-Si:H) film. The chemical deposition is a rather lengthy process, and it is difficult to determine the crystallization and crystalline phase of the thin film prior to X-ray diffraction (XRD) measurements. In this study, we are trying to analyze the spectral data collected by optical emission spectroscopy (OES) to find out there is any correlation between OES data and crystalline status. We used machine learning onto an in-situ detection tool to forecast this correlation. The collected large-scale OES spectral data obtained via principal component analysis (PCA) was used for the prediction of the crystalline phase in films without necessary experiments performed afterwards. Therefore, this method can be applicable to the field of thin film deposition for the detection of properties on thin films.

In this study, we submit a complex set of in-situ data collected by optical emission spectroscopy (OES) during the process of aluminum nitride (AlN) thin film. Changing the sputtering power and nitrogen(N2) flow rate, AlN film was deposited on Si substrate using a superior sputtering with a pulsed direct current (DC) method. The correlation between OES data and deposited film residual stress (tensile vs. compressive) associated with crystalline status by X-ray diffraction spectroscopy (XRD), scanning electron microscope (SEM), and transmission electron microscope (TEM) measurements were investigated and established throughout the machine learning exercise. An important answer to know is whether the stress of the processing film is compressive or tensile. To answer this question, we can access as many optical spectra data as we need, record the data to generate a library, and exploit principal component analysis (PCA) to reduce complexity from complex data. After preprocessing through PCA, we demonstrated that we could apply standard artificial neural networks (ANNs), and we could obtain a machine learning classification method to distinguish the stress types of the AlN thin films obtained by analyzing XRD results and correlating with TEM microstructures. Combining PCA with ANNs, an accurate method for in-situ stress prediction and classification was created to solve the semiconductor process problems related to film property on deposited films more efficiently. Therefore, methods for machine learning-assisted classification can be further extended and applied to other semiconductors or related research of interest in the future.

In this study, aluminum nitride (AlN) thin films were deposited on Si (100) and investigated on the minimum film residual stress with varying two critical deposition conditions in N2 gas flow and power. The pulsed DC reactive sputtering of aluminum targets was carried out in gas ratio of nitrogen (N2) and argon (Ar) plasma with N2:Ar ratios from 15:15 to 60:15 and power from 400 to 1000 W. According to the large-scale data of in situ optical emission spectroscopy (OES) study in connection with film microstructure information by analyzing X-ray diffraction (XRD) patterns, scanning electron microscope (SEM) micrographs, transmission electron microscope (TEM) analyses, Fourier-transform infrared spectroscopy (FTIR) spectra, and alpha-step profilers, the deposited film stress states can be highly affected by the critical processing parameters of N2 flow rate and power. In addition, the proposed value of residual stress (VRS) can be calculated by PC1-DEV (the standard deviation in the first principal component direction) method for accurate prediction of the stress states of deposited films, i.e., compressive stress or tensile stress, which will provide valuable information on residual stress characterization as the in situ monitoring tool for the AlN thin film deposition process. In summary, a methodology based on large data of OES by which principal component analysis (PCA) was utilized to reduce its dimension can be used to determine residual stress characterization from a simple measurement in situ plasma monitoring tool is suggested.

In this study, pulsed dc reactive sputtering of aluminum nitride (AlN) thin films was investigated. The aluminum nitride thin films were deposited on Si (100) using a reactive direct current (DC) unbalanced magnetron sputtering system. The DC reactive sputtering was used in sputtering the aluminum targets in a mixture of argon (Ar) and nitrogen (N 2 ) plasma. Processes of aluminum target sputtering were carried out in an atmosphere of a mixture of Ar and N 2 . However, pulsed DC reactive sputtering of aluminum targets was carried out at total pressures with N 2 :Ar ratios from 7:30 to 45:15. In-situ optical emission spectrometry (OES) was applied to obtain the optimal deposition rate and the highest sputtering yield from the effects of flow nitrogen/argon (N2:Ar) ratio and pulse frequency on OES intensity. Thus, we have compared Fourier-transform infrared spectroscopy (FTIR) spectra and X-ray diffraction (XRD) patterns of AlN films deposited on Si (100) by DC reactive sputtering with an Al target in the mixture of Ar and N 2 . FTIR and XRD investigated the quality of the films and the preferred orientation.