• Journal of the Semiconductor & Display Technology
• /
• v.18 no.2
• /
• pp.48-52
• /
• 2019

Virtual metrology, which is one of APC techniques, is a method to predict characteristics of manufactured films using machine learning with saving time and resources. As the photoresist is no longer a mask material for use in high aspect ratios as the CD is reduced, hard mask is introduced to solve such problems. Among many types of hard mask materials, amorphous carbon layer(ACL) is widely investigated due to its advantages of high etch selectivity than conventional photoresist, high optical transmittance, easy deposition process, and removability by oxygen plasma. In this study, VM using different machine learning algorithms is applied to predict the thickness of ACL and trained models are evaluated which model shows best prediction performance. ACL specimens are deposited by plasma enhanced chemical vapor deposition(PECVD) with four different process parameters(Pressure, RF power, $C_3H_6$ gas flow, $N_2$ gas flow). Gradient boosting regression(GBR) algorithm, random forest regression(RFR) algorithm, and neural network(NN) are selected for modeling. The model using gradient boosting algorithm shows most proper performance with higher R-squared value. A model for predicting the thickness of the ACL film within the abovementioned conditions has been successfully constructed.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2015.08a
• /
• pp.132-132
• /
• 2015

Virtual metrology (VM) model based on plasma information (PI) parameter for C4F8 plasma-assisted oxide etching processes is developed to predict and monitor the process results such as an etching rate with improved performance. To apply fault detection and classification (FDC) or advanced process control (APC) models on to the real mass production lines efficiently, high performance VM model is certainly required and principal component regression (PCR) is preferred technique for VM modeling despite this method requires many number of data set to obtain statistically guaranteed accuracy. In this study, as an effective method to include the 'good information' representing parameter into the VM model, PI parameters are introduced and applied for the etch rate prediction. By the adoption of PI parameters of b-, q-factors and surface passivation parameters as PCs into the PCR based VM model, information about the reactions in the plasma volume, surface, and sheath regions can be efficiently included into the VM model; thus, the performance of VM is secured even for insufficient data set provided cases. For mass production data of 350 wafers, developed PI based VM (PI-VM) model was satisfied required prediction accuracy of industry in C4F8 plasma-assisted oxide etching process.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.29A no.4
• /
• pp.38-48
• /
• 1992

OEIC(Optoelectronic Integrated Circuit)'s can be integrated horizontally or vertically. Horizontal integration approach is, however, more immune to parasitic and more universally applicable. In this paper, a structural modeling, fabrication and characterization of PIN photodiodes which can be used in the horizontal integration are performed. For device modeling, we build a transmission line model from 2-D device simulation, from which lumped model parameters are extracted. The speed limits of the PIN photodiodes can also be calculated under various structural conditions from the model. Thus optimum design of horizontally integrated PIN photodiodes for high speed operation are possible. Such InGaAs/InP PIN photodiodes for long-wavelength communications are fabricated using pit etch, epi growth, planarization, diffusion and metallization processes. Planarization process using both RIE and wet etching and diffusion process using evaporated Zn$_{3}P_{2}$ film are developed. Characterization of the fabricated devices is performed through C-V and I-V measurements. At a reserve bias of 10V, the dark current is less than 5nA and capacitance is about 0.4pF. The calculated bandwidth using the measured series resistance and capacitance is about 4.23GHz.

• Journal of the Korean Society for Precision Engineering
• /
• v.22 no.9 s.174
• /
• pp.60-68
• /
• 2005

To investigate the effects of beam focusing in the etching of polymers with short pulse Excimer lasers, a polymer etching model of SSB's is combined with a beam focusing model. Through the numerical simulation, it was found that in the high laser fluence region, SSB model considering both photochemical and thermal contribution is considered to be suitable to predict the etched hole shape than a simple photochemical etching model. The average temperature distribution into the substance obtained by assuming 1-D heat transfer is found to be fairly similar to the fluence distribution on the ablated surface. The experimental etching data fur polymers are used to give material properties for ablation model. The fitted etch depth curve gives a nice agreement with the experimental data.

• ETRI Journal
• /
• v.26 no.4
• /
• pp.297-306
• /
• 2004

A plasma is a collection of charged particles and on average is electrically neutral. In fabricating integrated circuits, plasma etching is a key means to transfer a photoresist pattern into an underlayer material. To construct a predictive model of plasma-etching processes, a polynomial neural network (PNN) is applied. This process was characterized by a full factorial experiment, and two attributes modeled are its etch rate and DC bias. According to the number of input variables and type of polynomials to each node, the prediction performance of the PNN was optimized. The various performances of the PNN in diverse environments were compared to three types of statistical regression models and the adaptive network fuzzy inference system (ANFIS). As the demonstrated high-prediction ability in the simulation results shows, the PNN is efficient and much more accurate from the point of view of approximation and prediction abilities.

• Journal of the Semiconductor & Display Technology
• /
• v.10 no.4
• /
• pp.21-23
• /
• 2011

Uniformity of plasma etching processes critically depends on the wafer temperature and its distribution. The wafer temperature is affected by plasma, chucking force, He back side pressure and the surface temperature of ESC(electrostatic chuck). In this work, 3D mathematical modeling is used to investigate the influence of the geometry of coolant path and the temperature distribution of the ESC surface. The model that has the coolant path with less change of the cross-sectional area and the curvature shows low standard deviation of the ESC surface temperature distribution than the model with the coolant path of the larger surface area and more geometric change.

• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.403-404
• /
• 2007

소자 수율을 향상시키기 위해서는 웨이퍼 전체에 걸쳐 플라즈마 공정특성이 균일하게 분포되어야 한다. 본 연구에서는 Actinomeric 광 반사분광기 (Otical Emission Spectroscopy) 정보를 이용하여 식각률 비균일도에 대한 모델을 개발하였다. 제안된 기법은 Oxide 식각공정에서 수집한 데이터에 적용하였으며, 체계적인 모델링을 위해 공정데이터는 통계적 실험계획 법을 적용하여 수집되었다. 신경망의 예측성능은 유전자 알고리즘을 이용해서 증진시켰다. OES의 차수를 줄이기 위해 주인자 분석을 세 종류의 분산(100, 99, 98%)에 대해서 적용하였다. 개발된 모델은 발표된 이전의 모델에 비해 17% 증진된 예측성능을 보였다.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2582-2584
• /
• 2003

역전파 신경망 (BPNN)은 반도체 공정 모델링에 효과적으로 응용되고 있다. 뉴런의 활성화 함수는 동일한 값을 가지며, 이로 인해 예측정확도를 증진하는 데에는 한계가 있었다. 본 연구에서는 난수발생기(Random generator-RG)를 이용하여 뉴런 경사들이 다중값을 가지도록 최적화하였다. 본 기법은 은닉충의 뉴런수의 함수로 고찰하였으며, 종래의 고정된 경사를 갖는 모델과 그 성능을 비교 평가하였다. 평가에 이용된 데이터는 플라즈마 식각 공정데이터이며, 모델에 이용된 응답은 식각률과 프로파일 각이다. 비교결과 종래의 모델에 비해 예측정확도가, 식각률의 경우 19%-43%, 프로파일의 경우 10%-56% 정도 향상하였으며, 이는 제안된 기법이 모델개발에 매우 효과적으로 적용될 수 있음을 보여준다.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.239-240
• /
• 2012

With advancements in semiconductor device technologies, manufacturing processes are getting more complex and it became more difficult to maintain tighter process control. As the number of processing step increased for fabricating complex chip structure, potential fault inducing factors are prevail and their allowable margins are continuously reduced. Therefore, one of the key to success in semiconductor manufacturing is highly accurate and fast fault detection and classification at each stage to reduce any undesired variation and identify the cause of the fault. Sensors in the equipment are used to monitor the state of the process. The idea is that whenever there is a fault in the process, it appears as some variation in the output from any of the sensors monitoring the process. These sensors may refer to information about pressure, RF power or gas flow and etc. in the equipment. By relating the data from these sensors to the process condition, any abnormality in the process can be identified, but it still holds some degree of certainty. Our hypothesis in this research is to capture the features of equipment condition data from healthy process library. We can use the health data as a reference for upcoming processes and this is made possible by mathematically modeling of the acquired data. In this work we demonstrate the use of recurrent neural network (RNN) has been used. RNN is a dynamic neural network that makes the output as a function of previous inputs. In our case we have etch equipment tool set data, consisting of 22 parameters and 9 runs. This data was first synchronized using the Dynamic Time Warping (DTW) algorithm. The synchronized data from the sensors in the form of time series is then provided to RNN which trains and restructures itself according to the input and then predicts a value, one step ahead in time, which depends on the past values of data. Eight runs of process data were used to train the network, while in order to check the performance of the network, one run was used as a test input. Next, a mean squared error based probability generating function was used to assign probability of fault in each parameter by comparing the predicted and actual values of the data. In the future we will make use of the Bayesian Networks to classify the detected faults. Bayesian Networks use directed acyclic graphs that relate different parameters through their conditional dependencies in order to find inference among them. The relationships between parameters from the data will be used to generate the structure of Bayesian Network and then posterior probability of different faults will be calculated using inference algorithms.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.241-241
• /
• 2012

Semiconductor industry has been taking the advantage of improvements in process technology in order to maintain reduced device geometries and stringent performance specifications. This results in semiconductor manufacturing processes became hundreds in sequence, it is continuously expected to be increased. This may in turn reduce the yield. With a large amount of investment at stake, this motivates tighter process control and fault diagnosis. The continuous improvement in semiconductor industry demands advancements in process control and monitoring to the same degree. Any fault in the process must be detected and classified with a high degree of precision, and it is desired to be diagnosed if possible. The detected abnormality in the system is then classified to locate the source of the variation. The performance of a fault detection system is directly reflected in the yield. Therefore a highly capable fault detection system is always desirable. In this research, time series modeling of the data from an etch equipment has been investigated for the ultimate purpose of fault diagnosis. The tool data consisted of number of different parameters each being recorded at fixed time points. As the data had been collected for a number of runs, it was not synchronized due to variable delays and offsets in data acquisition system and networks. The data was then synchronized using a variant of Dynamic Time Warping (DTW) algorithm. The AutoRegressive Integrated Moving Average (ARIMA) model was then applied on the synchronized data. The ARIMA model combines both the Autoregressive model and the Moving Average model to relate the present value of the time series to its past values. As the new values of parameters are received from the equipment, the model uses them and the previous ones to provide predictions of one step ahead for each parameter. The statistical comparison of these predictions with the actual values, gives us the each parameter's probability of fault, at each time point and (once a run gets finished) for each run. This work will be extended by applying a suitable probability generating function and combining the probabilities of different parameters using Dempster-Shafer Theory (DST). DST provides a way to combine evidence that is available from different sources and gives a joint degree of belief in a hypothesis. This will give us a combined belief of fault in the process with a high precision.