Search Results (63)

In Czochralski (CZ) silicon growth, controlling the ratio of crystal growth velocity to axial temperature gradient ($V/G$) is critical for defect management. However, the $V/G$ value is difficult to measure in real-time. Furthermore, it exhibits strong multivariate coupling and extreme non-stationarity under complex thermal fields. While standard deep learning models like LSTM are used for soft sensing, they often misidentify high-frequency hardware noise as true process dynamics, causing severe error amplification in multi-step predictions. To address this, we propose a Dual Information Filtering LSTM (DIF-LSTM). It utilizes an external context-aware mechanism to screen long-term steady-state redundant information and an internal denoising gate coupled with the LSTM input to explicitly block transient high-frequency noise. Furthermore, a confidence evaluation branch and residual decay fusion ensure stable multi-step forecasting. Experimental results an industrial-scale experimental silicon single crystal furnace show DIF-LSTM achieves superior accuracy, obtaining an $R^2$ of 0.9935 and a Mean Squared Error of $1.60\times {10}^{-6}$ at a 3-step horizon. Even at a 9-step horizon, it maintains an $R^2$ of 0.9422, significantly outperforming the baseline IF-LSTM (0.8498). Full article

A series of co-doped LiNbO₃:Mg:Er crystals were grown in a single technological cycle and under the same technological conditions by Czochralski. In each subsequent step of the growth cycle, the content of Mg and Er dopants decreased. The initial concentration of dopants in the melt was [Mg] = 4.0 mol% and [Er] = 0.78 mol%. The melt was obtained from a homogeneously doped batch. The batch included the Nb₂O₅:Mg:Er precursor synthesized by the liquid-phase method. The physicochemical features of crystallization were studied. The optical properties of the crystals were investigated using laser conoscopy and photoinduced light scattering. Macro- and microdefect structures were studied by optical microscopy. Quantitative phase analysis was performed for single-crystal samples. The defect structures of powdered LiNbO₃:Mg:Er samples were determined by refining XRD patterns by Rietveld. The optical quality of doubly doped crystals corresponds to that of singly doped LiNbO₃:Er crystals. Mg significantly reduces the transparency of LiNbO₃:Mg:Er crystals in the ultraviolet and violet spectral ranges. The optimal dopant concentration in the melt was [Er] = 0.63 mol% and [Mg] = 3.0 mol%, and [Er] = 0.47 mol% and [Mg] = 3.07 mol% in crystal. The optical properties of LiNbO₃:Mg:Er crystals make them promising active nonlinear optical materials for generating and converting laser radiation. Full article

This study proposes a data-driven framework with post hoc interpretability analysis for one-step-ahead crystal diameter prediction in the Czochralski (CZ) silicon single-crystal growth process. To address the strong multivariable coupling, nonlinear dynamics, variable-specific delays, and difficulty of online measurement in CZ growth, the maximal information coefficient (MIC) was first used to screen key auxiliary variables from industrial process data. The Grey Wolf Optimizer (GWO) was then employed for multi-variable delay estimation and feature alignment, and a hybrid temporal convolutional network (TCN)–long short-term memory (LSTM) model was constructed to combine local temporal feature extraction with long-term dependency learning. Four input configurations were designed according to whether lag alignment and diameter history were included, and the proposed TCN-LSTM was systematically compared with standalone TCN and LSTM models. The results show that both diameter history and delay alignment improve prediction performance. Under the current single-run evaluation protocol, the TCN-LSTM configurations yielded lower prediction errors than the corresponding TCN and LSTM models under the same input settings. Under the withlag-withY configuration, the TCN-LSTM model achieved MSE = 0.00259, RMSE = 0.05087, MAE = 0.03949, and R² = 0.96982. After GWO-based hyperparameter optimization, the best TCN-LSTM configuration further improved to MSE = 0.00239, RMSE = 0.04894, MAE = 0.03651, and R² = 0.97207. SHAP-based analysis was further used to provide a post hoc interpretation of the relative contributions of key process variables to diameter variation. Overall, the proposed framework provides a data-driven prediction approach and may support subsequent process analysis and optimization in industrial CZ growth. Full article

The V/G criterion is a critical indicator for monitoring dynamic changes during Czochralski silicon single crystal (Cz-SSC) growth. However, the inability to measure it in real time forces reliance on offline feedback for process regulation, leading to imprecise control and compromised crystal quality. To overcome this limitation, this paper proposes a novel soft sensor modeling framework that integrates both mechanism-based knowledge and data-driven learning for the real-time prediction of the crystal quality parameter, specifically the V/G value (the ratio of growth rate to axial temperature gradient). The proposed approach constructs a hybrid prediction model by combining a data-driven sub-model with a physics-informed mechanism sub-model. The data-driven component is developed using an attention-based dynamic stacked enhanced autoencoder (AD-SEAE) network, where the SEAE structure introduces layer-wise reconstruction operations to mitigate information loss during hierarchical feature extraction. Furthermore, an attention mechanism is incorporated to dynamically weigh historical and current samples, thereby enhancing the temporal representation of process dynamics. In addition, a robust ensemble approach is achieved by fusing the outputs of two subsidiary models using an adaptive weighting strategy based on prediction accuracy, thereby enabling more reliable V/G predictions under varying operational conditions. Experimental validation using actual industrial Cz-SSC production data demonstrates that the proposed method achieves high-prediction accuracy and effectively supports real-time process optimization and quality monitoring. Full article

High-precision prediction of the crystal diameter during the growth of electronic-grade silicon single crystals is a critical step for the fabrication of high-quality single crystals. However, the process features high-temperature operation, strong nonlinearities, significant time-delay dynamics, and external disturbances, which limit the accuracy of conventional mechanism-based models. In this study, mechanism-based models denote physics-informed heat-transfer and geometric models that relate heater power and pulling rate to diameter evolution. To address this challenge, this paper proposes a hybrid deep learning model combining a convolutional neural network (CNN), a bidirectional long short-term memory network (BiLSTM), and self-attention to improve diameter prediction during the shoulder-formation and constant-diameter stages. The proposed model leverages the CNN to extract localized spatial features from multi-source sensor data, employs the BiLSTM to capture temporal dependencies inherent to the crystal growth process, and utilizes the self-attention mechanism to dynamically highlight critical feature information, thereby substantially enhancing the model’s capacity to represent complex industrial operating conditions. Experiments on operational production data collected from an industrial Czochralski (Cz) furnace, model TDR-180, demonstrate improved prediction accuracy and robustness over mechanism-based and single data-driven baselines, supporting practical process control and production optimization. Full article

The Czochralski method is the dominant technique for producing power-electronics-grade silicon crystals. At the beginning of the seeding stage, an excessively high (or low) temperature at the solid–liquid interface can cause the time required for the seed to reach the specified length to be too long (or too short). However, the time taken for the seed to reach a specified length is strictly controlled in semiconductor crystal growth to ensure that the initial temperature is appropriate. An inappropriate initial temperature can adversely affect crystal quality and production yield. Accurately evaluating whether the current temperature is appropriate for seeding is therefore essential. However, the temperature at the solid–liquid interface cannot be directly measured, and the current manual evaluation method mainly relies on a visual inspection of the meniscus. Previous methods for detecting this temperature classified image features, lacking a quantitative assessment of the temperature. To address this challenge, this study proposed using the duration of the seeding stage as the target variable for evaluating the temperature and developed an improved multimodal fusion regression network. Temperature signals collected from a central pyrometer and an auxiliary pyrometer were transformed into time–frequency representations via wavelet transform. Features extracted from the time–frequency diagrams, together with meniscus features, were fused through a two-level mechanism with multimodal feature fusion (MFF) and channel attention (CA), followed by masking using spatial attention (SA). The fused features were then input into a random vector functional link network (RVFLN) to predict the seeding duration, thereby establishing an indirect relationship between multi-sensor data and the seeding temperature achieving a quantification of the temperature that could not be directly measured. Transfer comparison experiments conducted on our dataset verified the effectiveness of the feature extraction strategy and demonstrated the superior detection performance of the proposed model. Full article

The inevitable introduction of oxygen into Czochralski-method-grown single crystal silicon, facilitated by the use of quartz crucibles, can result in the failure of chips and devices. Both the size and position of the heater exert a significant influence on the oxygen concentration within the Czochralski-method-grown silicon. In this study, a novel short-type heater was designed and evaluated for its effect on melt temperature and oxygen diffusion during crystal growth. The silicon melt temperatures and oxygen diffusion coefficients in an MCZ furnace for several heater settings were simulated, and the results were implemented in experiments. From the examination of the growth process through computation, the heater and its positional adjustments were determined to be effective modulators of oxygen concentration during crystal growth, which was consequently reduced to below 4 ppma (ASTM F121-83). Finally, the simulations were validated experimentally, limitations in production were discussed, and possible improvements were outlined. Full article

We demonstrate that selective electrochemical etching is a reliable method for detecting and observing the uneven concentration distribution of impurities in indium phosphide crystals, which accompanies the growth of highly doped crystals using the Czochralski method. Even though selective electrochemical etching, as a method of detecting defects in the crystal lattice, has been discussed many times in the literature, it has not yet been described for indium phosphide. In this work, we investigated etching in compositions of various selective electrolytes for InP of n- and p-type conductivity with different surface orientations. We present in detail the features of detecting the striped inhomogeneity of impurity distribution. The mechanisms and peculiarities of the formation of oxide crystallites on the surface of InP during electrochemical processing are presented, including structures like flower-like and parquet crystallites. The formation of porous surfaces, terraces, tracks, and crystallites is explained from the perspective of the defect-dislocation mechanism. Full article

This paper presents the preparation of Z-cut near-stoichiometric lithium niobate (NSLN) wafers using a combined process of the lithium-rich Czochralski growth and diffusion methods. The fabricated Z-cut NSLN wafers exhibited outstanding comprehensive performance, including a high Curie temperature of up to 1200 °C, a refractive index gradient in the diameter direction below 1.5 × 10⁻⁴ cm⁻¹, and a UV absorption edge shifted 14 nm toward the ultraviolet region compared to congruent lithium niobate crystals, with a coercive field of 1268 V/mm. Additionally, the wafers demonstrated excellent processing characteristics, with the bow of 4-inch wafers controlled within 55 μm, surpassing the machining standards of traditional lithium niobate wafers of the same size. These results indicated the highly uniform chemical stoichiometry and crystallization quality of the wafers. Leveraging the high uniformity and low coercive field of the wafers, periodic triangular domain structure arrays were successfully fabricated, laying the foundation for domain engineering design in electro-optic deflectors and switching devices. This study not only achieves the scalable preparation of NSLN wafers but also provides a reliable technical solution for their practical applications in high-performance electro-optic devices. Full article

In this paper, we report on the growth and characterization of high doping concentration (2.1 at%) ytterbium (Yb) doped lithium niobate (Yb:LiNbO₃) crystal. By using a slightly modified Czochralski method, we have successfully grown a usable size (2 mm × 2 mm × 30 mm) Yb:LiNbO₃ single crystal. We also conducted the energy-dispersive X-ray spectroscopy (EDS) and the X-ray diffraction (XRD) analyses, which experimentally confirm that the grown crystal is a Yb:LiNbO₃ single crystal. We also measured the absorption and emission spectra of the grown crystal. It was found out that there is a near-flat broad emission within a spectral range of 1004–1030 nm when excited at 980 nm for this high doping concentration Yb:LiNbO₃ crystal. Such a near-flat broad emission can be very useful for realizing high slope efficiency ultrafast (femtosecond) lasing in the Yb:LiNbO₃ crystal due to the low quantum defect of the Yb:LiNbO₃ crystal. We also investigated the electro-optic effect of the Yb:LiNbO₃. The experimental result confirms that the electro-optic (EO) effect of a highly doped (2.1 at%) lithium niobate crystal is close to the EO value of the pure lithium niobate. Thus, the highly doped Yb:LiNbO₃ crystal can still be an effective electrically tunable lasing medium. It can enable electrically tunable, high slope efficiency femtosecond lasing due to the combined features, including (1) a near flat broad emission spectrum at the spectral range of 1004–1030 nm, (2) a non-compromised electro-optic effect at high doping concentration Yb:LiNbO₃ crystal, and (3) a low quantum defect. Full article

Piezoelectric materials have garnered significant attention due to their diverse applications in technologies such as sensors, actuators, and energy-harvesting systems. This study focuses on the growth and characterization of Tb³⁺-doped La₃Ga₅SiO₁₄ (LGS) crystals. A novel 10% Tb³⁺-doped single LGS crystal was successfully grown using the Czochralski method. The crystal structure and fluorescence properties were determined, and the electro-elastic properties were evaluated by the impedance method, which assessed dielectric, piezoelectric, and elastic constants. The Tb³⁺-doped crystal was observed to crystallize in the trigonal system, with the concentration of the Tb³⁺ ion in the crystal determined to be 2.50 wt%. The piezoelectric coefficients were measured as d₁₁ = 5.41 pC/N and d₁₄ = −5.52 pC/N, and the dielectric constants were found to be 19.60 and 52.75, respectively. The temperature-dependent behavior of Tb:LGS crystals was investigated, particularly concerning their elastic constants, demonstrating favorable thermal stability. This study provides valuable insights into the relationship between the crystals’ structural characteristics and performance. Additionally, the fluorescence properties were measured; a long lifetime (τ = 1.655 ms) indicated the potential applications of Tb:LGS crystals in laser technology. Full article

Herein we present a roadmap for tailoring the crystal growth conditions, near-infrared (NIR) laser emission, and self-frequency doubling (SFD) performances of newly developed Nd-doped La_xGd_ySc_4−x−y(BO₃)₄ (Nd:LGSB) crystals. Three different Nd³⁺ doping concentrations of 2.3 at.%, 3.5 at.%, and 4.6 at.% were investigated. Considering their incongruent melting, special conditions were employed for the growth using the Czochralski technique. Laser emission performances at 1062 nm in the CW regime were evaluated for uncoated crystal samples with different orientations (a-cut, c-cut, and SFD-cut). The highest slope efficiency η_sa = 0.68 was obtained for the 4.6 at.% c-cut Nd:LGSB crystal, with a randomly polarized emission. The a-cut 4.6 at.% Nd:LGSB crystal delivered a linearly polarized beam with a slope efficiency η_sa = 0.63. The SFD-cut 2.3 at.% and 3.5 at.% Nd:LGSB crystals achieved slightly lower efficiencies of ~ 0.56. The SFD capabilities of 2.3 at.% and 3.5 at.% Nd:LGSB crystals were also explored. Green laser emission at ~531 nm was achieved with a diode-to-green conversion efficiency increasing significantly from 0.17% to 1.44%, respectively. These results demonstrate that the Nd-doping concentration, crystal orientation, and sample length of Nd:LGSB crystals, must be carefully selected depending on the specific requirements of the intended application. Full article

The accurate real-time prediction of the crystal quality index v/G is an important reference for the real-time monitoring of the growth quality status and the process optimization adjustment of semiconductor silicon single crystals. This paper proposes a data-driven crystal quality indicator v/G soft sensor prediction model based on multi-timescale feature fusion to achieve the effective prediction of the crystal quality indicator v/G. Firstly, the characteristics of the crystal quality index v/G in the growth process of Czochralski silicon single crystal are analyzed. Secondly, the crystal quality index v/G is broken down into several natural components using something called complete ensemble empirical mode decomposition with adaptive noise (CEEMDAD), which provides more stable data. On this basis, each intrinsic mode component is reconstructed according to the sample entropy. Then, the maximum mutual information coefficient (MIC) method is applied to identify the characteristic variables most closely associated with each reconstructed component of the crystal quality index v/G from the process-influencing factors. Then, a long short-term memory network with a self-attention mechanism is used to establish a prediction model of the reconstructed components to extract the multi-timescale feature information from the different components of the crystal quality index v/G. Finally, the prediction results of the crystal quality index v/G are obtained by fusing each subsequent prediction model. According to the actual field data, the comprehensive experimental results validate the efficacy of the proposed soft sensor modeling method for the crystal quality index v/G. Compared with the single model, the proposed prediction model has a smaller MAE, RMSE, and prediction performance index and a higher HR prediction hit rate. Full article

Precise temperature and heating power control are crucial for crystal quality and production efficiency in the Czochralski single-crystal growth process. Existing sensor technologies can only monitor these parameters in real time, lacking the ability to predict future trends, which limits the ability to implement preventive control before issues arise. To address this, a temperature and heating power prediction model based on Long Short-Term Memory (LSTM) is proposed and developed using extensive production data. Spearman’s rank correlation coefficient is applied to identify the key parameters related to temperature and heating power. Hyperparameter optimization uses Particle Swarm Optimization (PSO) to improve prediction accuracy. The performance of the PSO-LSTM model is compared with two other widely used prediction models, demonstrating its superior predictive capability. The results show that the PSO-LSTM model achieves highly accurate temperature and heating power predictions in the crystal growth process, with a Mean Absolute Error (MAE) of 0.0295 for temperature and 0.0392 for heating power, further validating its effectiveness for real-time predictive control. Full article

The Czochralski method is the primary technique for single-crystal silicon production. However, anomalous states such as crystal loss, twisting, swinging, and squareness frequently occur during crystal growth, adversely affecting product quality and production efficiency. To address this challenge, we propose an enhanced multimodal fusion classification model for detecting and categorizing these four anomalous states. Our model initially transforms one-dimensional signals (diameter, temperature, and pulling speed) into time–frequency domain images via continuous wavelet transform. These images are then processed using a Dense-ECA-SwinTransformer network for feature extraction. Concurrently, meniscus images and inter-frame difference images are obtained from the growth system’s meniscus video feed. These visual inputs are fused at the channel level and subsequently processed through a ConvNeXt network for feature extraction. Finally, the time–frequency domain features are combined with the meniscus image features and fed into fully connected layers for multi-class classification. The experimental results show that the method can effectively detect various abnormal states, help the staff to make a more accurate judgment, and formulate a personalized treatment plan for the abnormal state, which can improve the production efficiency, save production resources, and protect the extraction equipment. Full article