• Proceedings of the KIEE Conference
• /
• 1996.11a
• /
• pp.457-461
• /
• 1996

An optimum design methodology for the power MOSFET's with a predetermined mask is proposed and verified by comparing with the results of MEDICI simulation and the data of commercially available devices. Optimization is completed by determining a doping concentration and a thickness of the epitaxial layer which satisfy a specific voltage and current rating requirements as well as a minimum on-resistance for the mask set. The commercial HEX-1 mask set with a die area of $40.4{\times}10^{-3}\;cm^2$ and a T0-220 package has the on-resistance of $1.5{\Omega}$ at 200 V/2.5 A rating while the M-1 mask from this study exhibits $0.6{\Omega}$ on-resistance at 200 V/6 A. The 60 % reduction in the on-resistance and 58 % enhancement in the current rating have been obtained by the proposed method.

• Bulletin of the Korean Chemical Society
• /
• v.35 no.3
• /
• pp.705-708
• /
• 2014

We demonstrate the blue InGaN/GaN multiple quantum wells light-emitting diodes (LEDs) with an embedded air-gap photonic crystal (PC) which was fabricated by the lateral epitaxial overgrowth of GaN layer on the tungsten (W) nano-masks. The periodic air-gap PC was formed by the chemical reaction of hydrogen with GaN on the W nano-mask. The optical output power of LEDs with an air-gap PC was increased by 26% compared to LEDs without an air-gap PC. The enhanced optical output power was attributed to the improvement in internal quantum efficiency and light extraction efficiency by the air-gap PC embedded in GaN layer.

• 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
• /
• 2012.08a
• /
• pp.288-288
• /
• 2012

The oxidation characteristics of tungsten line pattern during the carbon-based mask layer removal process using oxygen plasmas and the reduction characteristics of the WOx layer formed on the tungsten line surface using hydrogen plasmas have been investigated for sub-50 nm patterning processes. The surface oxidation of tungsten line during the mask layer removal process could be minimized by using a low temperature ($300^{\circ}K$) plasma processing instead of a high temperature plasma processing for the removal of the carbon-based material. Using this technique, the thickness of WOx on the tungsten line could be decreased to 25% of WOx formed by the high temperature processing. The WOx layer could be also completely removed at the low temperature of $300^{\circ}K$ using a hydrogen plasma by supplying bias power to the tungsten substrate to provide an activation energy for the reduction. When this oxidation and reduction technique was applied to actual 40 nm-CD device processing, the complete removal of WOx formed on the sidewall of tungsten line could be observed.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.19 no.6
• /
• pp.538-546
• /
• 2006

We report on characteristics of specially designed inductively-coupled-plasma chemical vapor deposition (ICP-CVD) system for top-emitting organic light emitting diodes (TOLEDs). Using high-density plasma on the order of $10^{11}$ electrons/$cm^3$ generated by linear-type antennas connected in parallel and specially designed substrate cooling system, a 100 nm-thick transparent $SiN_{x}$ passivation layer was deposited on thin Mg-Ag cathode layer at substrate temperature below $50\;^{\circ}C$ without a noticeable plasma damage. In addition, substrate-mask chucking system equipped with a mechanical mask aligner enabled us to pattern the $SiN_x$ passivation layer without conventional lithography processes. Even at low substrate temperature, a $SiN_x$ passivation layer prepared by ICP-CVD shows a good moisture resistance and transparency of $5{\times}10^{-3}g/m^2/day$ and 92 %, respectively. This indicates that the ICP-CVD system is a promising methode to substitute conventional plasma enhanced CVD (PECVD) in thin film passivation process.

• Journal of the Korean Crystal Growth and Crystal Technology
• /
• v.9 no.2
• /
• pp.168-172
• /
• 1999

The 9 $\mu\textrm{m}$ GaN films on sapphire substrate were grown by Hydride vapor phase epitaxy. Dislocation density of these GaN films was measured by TEM. GaN film with crack free and mirror surface was directly grown on sapphire substrate. The dislocation density of this GaN film was $2{\times}10^9/cm^2$. The surface of GaN film on patterned GaN layer also presented a smooth mirror. But a part of GaN surface included holes because of incomplete coalescence. The dislocation density of GaN film above the mask region was lower than that in the window region. Especially, the dislocation density in the region between mask center and window region was close to dislocation free. The average dislocation density of ELO GaN was $8{\times}10^7/cm^2$.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.32A no.8
• /
• pp.94-106
• /
• 1995

A three-dimensional(3D) Monte Carlo simulator for boron ion implantation into <100>single-crystal silicon considering the mask structure has been developed to predict the mask-dependent impurity doping profiles of the implanted boron at low energies into the reduced area according to the trend of a reduction in the size of semiconductor devices. All relevant important parameters during ion implantation have been taken into account in this simulator. These are incident energy, tilt and rotation of wafer, orientation of silicon wafer, presence of native silicon dioxide layer, dose, wafer temperature, ion beam divergence, masking thickness, and size and structure of open window in the mask. The one-dimensional(1D) results obtained by using the 3D simulator have been compared with the SIMS experiments to demonstrate its capabilities and confirem its reliability, and we obtained relatively accurate 1D doping profiles. Through these 3D simulations considering the hole structure and its size, we found the mask effects during boron ion implantation process.

• Journal of Information Processing Systems
• /
• v.18 no.3
• /
• pp.411-427
• /
• 2022

Digging historical and cultural information from seals in ancient books is of great significance. However, ancient Chinese seal samples are scarce and carving methods are diverse, and traditional digital image processing methods based on greyscale have difficulty achieving superior segmentation and recognition performance. Recently, some deep learning algorithms have been proposed to address this problem; however, current neural networks are difficult to train owing to the lack of datasets. To solve the afore-mentioned problems, we proposed an SEL-RefineMask which combines selector of feature pyramid network (SEL-FPN) with RefineMask to segment and recognize seals. We designed an SEL-FPN to intelligently select a specific layer which represents different scales in the FPN and reduces the number of anchor frames. We performed experiments on some instance segmentation networks as the baseline method, and the top-1 segmentation result of 64.93% is 5.73% higher than that of humans. The top-1 result of the SEL-RefineMask network reached 67.96% which surpassed the baseline results. After segmentation, a vision transformer was used to recognize the segmentation output, and the accuracy reached 91%. Furthermore, a dataset of seals in ancient Chinese books (SACB) for segmentation and small seal font (SSF) for recognition were established which are publicly available on the website.

• Proceedings of the KAIS Fall Conference
• /
• 2002.11a
• /
• pp.201-204
• /
• 2002

Overlay margin 확보를 위한 oversizing과, design rule checking, jog filtering를 통하여 side-lobe를 추출하였다. 이렇게 추출한 side-lobe를 extent하고, Cr pattern을 정의하여 side-lobe 현상을 해결할 수 있었다. 하지만 이 방법은 mask제조 공정이 복잡하므로 Cr shield방식의 단점인 복잡한 mask제작공정과 구조를 단순화하기 위하여 scattering bar를 이용하였다. 따라서, scattering bar를 삽입하기 위한 rule을 생성하여 metal layer에 적용하고 aerial image simulation을 통해 side-lobe 현상이 억제되었음을 확인하였다. 그리고 앞에서와는 반대로 background clear의 경우에 발생하는 side-lobe에 scattering bar를 적용하여 억제됨을 확인하였다.

• Proceedings of the IEEK Conference
• /
• 2002.06c
• /
• pp.87-90
• /
• 2002

This paper has been studied a new approach using neural network to detect false laser marks. In the proposed approach, input images are segmented into R, G and B colors and implements mask areas respectively. And then average and variation values of the each mask area are extracted for the learning process to minimize input nodes. Using this technique, the new input data is obtained and implemented to the back-propagation algorithm using multi layer perception. This paper reduces the computational complexity necessary and shows better effectiveness to inspect false laser marks.