HPLC-UV, ELSD, RI를 이용한 자일리톨 분석법 비교

서은빈 서울과학기술대학교 2022 국내석사

고려한 전처리법과 분석 조건을 조사하고 국내 유통중인 제품의 xylitol의 함량을 구하였다. 사용된 분석법은 우선적으로 유효성 검증을 진행하여 확립하고 측정불확도를 산출하였다. 또한, 분석법의 적용성을 확인하기 위해 국내에 유통되는 츄잉껌, 캔디류, 혼합음료, 고형차 중 xylitol을 정량하여 그 결과를 확인하였으며, 결과를 비교함으로써 보다 합리적인 분석법을 선정하여 xylitol이 사용되는 제품을 추가적으로 구매하여 모니터링을 진행하였다. HPLC-UV를 이용한 xylitol 분석법의 전처리는 p-nitrobenzoyl chloride 시약을 이용해 유도체화를 진행하였다. 유효성 검증을 통해 특이성을 확보하였으며, 직선성은 r2=1.0000 이었다. LOD와 LOQ는 각각 0.01, 0.04 mg/kg이었으며, MDL과 MQL은 0.36, 1.10 mg/kg이었다. 정확성과 정밀성은 94.46~99.52%, 0.17~1.81%으로 양호한 결과를 나타냈으며, AOAC guideline에서 제시된 범위에 적합하였다. 측정불확도를 산출한 결과, 상대불확도는 1.12~3.98%로 확인하였으며, 이는 Codex의 농도별 상대불확도 범위에 적합하였다. 또한, 적용성 검토 결과 모든 제품에서 xylitol을 정량할 수 있었다. HPLC-ELSD를 이용한 xylitol 분석법은 물 추출을 통한 전처리를 진행하였으며, 유효성 검증을 통해 특이성을 확보하였다. 검량선은 이차함수로 나타났으며, r2=0.9994이었다. LOD와 LOQ는 2.85, 8.63 mg/kg이었으며, MDL과 MQL은 56.93, 172.52 mg/kg이었다. 정확성과 정밀성은 94.25~105.73%, 0.52~5.97%으로 AOAC guideline에서 제시된 범위에 적합하여 양호한 결과를 나타냈다. 산출된 상대불확도는 0.87~11.98%였으며, Codex에 제시된 농도별 상대불확도 범위에 적합하였다. 또한, 적용성 검토 결과 모든 제품에서 xylitol을 정량할 수 있었다. HPLC-RI를 이용한 xylitol 분석법은 ELSD와 동일하게 물 추출을 통한 전처리를 진행하였으며, 유효성 검증을 통해 특이성을 확보하였다. 직선성은 r2=1.0000 이었고, LOD와 LOQ는 각각 40, 110 mg/kg이었으며, MDL과 MQL은 377.53, 1,144.02 mg/kg이었다. 정확성과 정밀성 결과는 96.87~101.52%, 0.65~2.59%으로 AOAC guideline에서 제시된 범위에 적합하였다. 측정불확도를 산출한 결과, 상대불확도는 0.78~4.78%로 나타났다. 이는 Codex의 농도별 상대불확도 범위에 적합하였다. 적용성 검토 결과, 2건의 제품에서 불검출로 나타났다. HPLC-UV, ELSD, RI을 이용한 xylitol 분석법의 유효성 검증 및 측정불확도의 결과를 확인하였다. HPLC-UV의 경우, LOD, LOQ의 값이 다른 두가지 분석법과 비교하였을 때 낮은 결과를 보였고, 낮은 상대불확도의 범위를 나타냈기 때문에 식품 중 미량의 xylitol까지 정량 가능할 것으로 판단하였다. 따라서, HPLC-UV를 이용하여 모니터링을 실시하였고 3반복의 전처리를 실시하고 각각의 분석을 진행하여 데이터 결과의 평균값 (g/kg)을 구하였다. HPLC-UV를 이용한 모니터링 결과(n=160), 츄잉껌 (n=50)의 자일리톨 함량은 13.823~749.510 g/kg의 범위에서 평균함량은 409.399 g/kg이었으며, 캔디류 (n=45)는 1.474~979.472 g/kg의 범위에서 104.229 g/kg, 혼합음료 (n=21)은 0.469~379.238 g/kg의 범위에서 37.079 g/kg, 고형차 (n=20)는 0.053~144.033 g/kg의 범위에서 24.373 g/kg, 기타가공품 (n=14)은 2.455~398.956 g/kg의 범위에서 75.259 g/kg, 음료 베이스 (n=10)에서 0.315~96.017 g/kg의 범위에서 16.462 g/kg이었다. In this study, xylitol, which can replace sucrose in low-calorie foods as an additive among various foods as a sweetener, was quantified using HPLC that can be universally used. Using UV, ELSD and RI, which are representative detectors of HPLC, analysis methods of xylitol were set up and applied to obtain the content of xylitol in foods distributed in Korea. The proposed methods were established by conducting validation and evaluated measurement uncertainty. The application of the methods was confirmed by quantifying xylitol in chewing gum, candy, beverage and tea distributed in Korea. The optimal analysis method was selected by comparing the results of application, and food items containing xylitol were additionally purchased for monitoring. For the pretreatment of xylitol analysis method using HPLC-UV, derivation was performed using PNBC. Specificity was confirmed, and good linearity was also represented (r2=1.0000). LOD and LOQ were 0.01 and 0.04 mg/kg, respectively. MDL and MQL were 0.36 and 1.10 mg/kg, respectively. Accuracy and precision showed satisfactory results with 94.46~99.52% and 0.17~1.81% within the range suggested by the AOAC guideline. As a result of the measurement uncertainty, the relative uncertainty was showed 1.12~3.98% satisfied with the range suggested by Codex guideline. The application was confirmed to quantify xylitol in all food items. Analysis method of xylitol quantification using HPLC-ELSD was pre-treated through water extraction, and specificity was satisfied. The calibration curve was shown as a quadratic function, and r2=0.9994. LOD and LOQ were 2.85, 8.63 mg/kg, and MDL and MQL were 56.93, 172.52 mg/kg, respectively. The accuracy and precision were 94.25~105.73% and 0.52~5.97% shown as satisfactory results with the range suggested by the AOAC guideline. The relative uncertainty was 0.87~11.98% suitable for the Codex guideline. As a result of application, it was also possible to quantify xylitol in all food items. The xylitol analysis method using HPLC-RI was pre-treated through water extraction in the same way as ELSD. Specificity was satisfied, and linearity also satisfied with r2=1.0000. LOD and LOQ were 40, 110 mg/kg, and MDL and MQL were 377.53, 1,144.02 mg/kg, respectively. The results of accuracy and precision were 96.87~101.52% and 0.65~2.59%, which were shown within the range suggested by AOAC guideline. As results of the measurement uncertainty, the relative uncertainty was shown to 0.78~4.78%. This was satisfied with the Codex guideline. The application results were shown that non-detected (N.D.) was confirmed in 2 food items. As the xylitol analysis methods of HPLC-UV, ELSD and RI, the results of validation, measurement uncertainty and application were compared. In the case of HPLC-UV, LOD, LOQ, MDL, and MQL were lower than the other methods, and also it showed a low range of relative uncertainty. It was determined that the analysis method using HPLC-UV could quantify even trace amounts of xylitol. Therefore, a total of 160 food items, including chewing gum (n=50), candy (n=45), beverage (n=21), tea (n=20), other processed products (n=14), and beverage base (n=10), were analyzed using method using HPLC-UV. The average amount (g/kg) of the results was obtained by carrying out in 3 replicates. As a result of monitoring using HPLC-UV, the average content of xylitol in chewing gum was 409.399 g/kg in the range of 13.823 to 749.510 g/kg, 104.229 g/kg in the range of 1.474 to 979.472 g/kg for candy, and 37.079 g/kg in the range of 0.469 to 379.238 g/kg for beverage, 24.373 g/kg in the range of 0.053 to 144.033 g/kg for tea, 75.259 g/kg in the range of 2.455 to 398.956 g/kg for other processed products, 16.462 g/kg in the range of 0.315 to 96.017 g/kg for beverage base.

무기력함의 구조에 대하여 : 본인 작품을 중심으로

서은빈 중앙대학교 대학원 2025 국내석사

현대 사회에서 무기력감은 개인과 사회 모두에게 만연한 문제로, 이는 경제적 양극화와 자극적 콘텐츠 소비로 인해 더욱 심화되고 있다. 본 논문은 무기력과 자극 추구의 악순환을 탐구하고 그 원인을 알고리즘에 의한 노력 없는 성공 사례에 대한 부정과 경제적 불평등에 대한 반감에서 찾는다. 특히 무기력을 용납하지 않는 현대 사회에서 자극적인 콘텐츠 소비로 도피하려는 경향과 반복적 자극에 대한 내성이 어떻게 더 강한 자극 추구로 이어지는지를 사례를 통해 논의한다. 이를 바탕으로, 사회의 무기력함의 구조에 대하여 본인의 작품을 통해 분석하였다. <알고리즘>과 <‘이것’도 알고리즘>에서는 선택적인 채색을 통해 알고리즘의 메커니즘을 통한 쉬운 성공 때문에 생기는 근본적인 무기력의 원인에 대해서 파악한다. 노력이나 능력이 없이 외부 요인으로 이룬 성공은 과시적이며 반감을 일으키고 개인의 좌절로 이어진다. <SNS Series>에서는 무기력을 감추기 위해 만드는 거짓된 모습은 자기 비하로 이어지고 <Flunkeyism>1, 2를 통해 타인과의 비교는 자신의 가치를 상실하게 만들며 또 다른 무기력의 원인임을 드러낸다. <설탕물을 찾아서>와 <도파민>은 무력감에서 회피하고자 자극을 찾아 몰려드는 사람들을 개미에 비유하여 자극의 중독성과 일시적 만족 뒤에 찾아오는 공허함을 보여준다. <무기력함의 구조에 대하여>는 화려한 자극으로 구성된 현 사회를 표현하고 있으며 무기력이라는 문제를 인식하지 못하면 세상의 본질을 알 수 없으며 문제를 직면하여 해결하는 능력을 상실할 수 있음을 경고한다. <붕괴>에서는 무기력함의 구조가 지속 될 시 모든 것이 의미를 잃고 허상만이 남는 사회에서의 사람들은 <무기력 표본>의 풍뎅이 표본처럼 개인의 자아를 상실하고 수동적 인간으로 전락할 것을 말하고 있다. 본 논문은 사회의 무기력 구조 속에서도 자극으로의 회피를 지양하고 무기력한 자신을 인식하여, 무기력함을 극복하기 위해 개인의 내적 성취를 추구하고 스스로 가치판단을 할 수 있어야 한다는 것을 재차 말한다. In contemporary society, lethargy has become a pervasive issue for both individuals and the community at large, exacerbated by economic polarization and the consumption of provocative content. This paper explores the vicious cycle of lethargy and the pursuit of stimulation, identifying its roots in resentment toward algorithm-driven cases of effortless success and reactions against economic inequality. In particular, the paper discusses the tendency to escape into provocative content consumption in a society that does not tolerate lethargy, as well as how repeated exposure to such stimulation leads to desensitization and a demand for even stronger stimuli. These issues are analyzed through the lens of the author’s own artwork. In the works <Algorithm> and <This Too Is Algorithm>, selective coloring is used to examine the fundamental causes of lethargy arising from effortless success achieved through algorithmic mechanisms. Success achieved without effort or skill often fosters ostentation, engenders resentment, and leads to personal disillusionment. The <SNS Series> reveals how fabricated appearances, created to mask lethargy, can result in self-deprecation. Through <Flunkeyism> 1 and 2, the works depict how comparisons with others lead to a loss of self-worth, serving as another source of lethargy. <In Search of Sugary Water> and <Dopamine> liken those who seek stimulation as a means of escaping their sense of powerlessness to ants, illustrating the addictive nature of stimulation and the emptiness that follows temporary satisfaction. <About the Structure of Apathy> portrays a society saturated with dazzling stimuli, warning that failing to recognize the issue of lethargy can render individuals unable to grasp the essence of reality or develop the ability to address challenges. In <Collapse>, the continuation of the structure of lethargy suggests that everything loses its meaning, leaving only illusions behind, turning people into passive beings who, like beetle specimens in <The Specimen of Lethargy>, lose their sense of self and intrinsic value. This paper underscores the need to move away from escapism through stimulation within the structure of societal lethargy. It reiterates that individuals must recognize their own lethargy and strive for intrinsic achievements, fostering the ability to make their own value judgments as a means of overcoming it.

2007 개정 『중국어 I』 교과서 문화란 내용 분석

서은빈 韓國外國語大學校 敎育大學院 2012 국내석사

중공형 STS304 스테인리스 강재 부품 제작을 위한 다단 플러그 인발 공정의 패스 스케줄 설계

서은빈 국립한국해양대학교 해사산업대학원 2025 국내석사

The plug drawing process, one of the cold tube drawing processes, uses a pair of drawing dies and a plug to simultaneously reduce the outside diameter and thickness of the initial tube while increasing its length. The use of plugs enables relatively high dimensional accuracy and excellent straightness of the drawn tube to be secured compared to the tube sinking process. In this study, materials and parts manufactured through the plug drawing process are mainly used in transportation machinery and robotics industrial equipment as housings for servo motors. Housing materials and parts for servo motors are facing problems such as noise and reduced product life due to dimensional errors in the cutting and assembly processes after drawing processing, and the importance of dimensional accuracy and roundness for the outer diameter and thickness of the drawn tube is increasing. Therefore, this study aims to develop a theoretical design method for stable multi-pass plug drawing processing of housing materials and parts for servo motors. A pass schedule design method for a multi-pass plug drawing process is proposed based on a strain control model that considers the dimensional change of the tube and a flow stress prediction model using the A. Geleji formula. To verify the proposed pass schedule, elastic-plastic finite element analysis and field experiments were performed for a 3 pass plug drawing process. The initial outer diameter and thickness dimensions of the hollow STS304 stainless steel used for verification were 44.5 mm and 1.1 mm, respectively. The drawing load for each pass, tube dimensions, and fracture during the drawing process were analyzed through elastic-plastic finite element analysis. Subsequently, through field experiments, we successfully manufactured a precision housing part with a final outer diameter of 39.9 mm and a thickness of 0.67 mm, which met the required specifications for housing materials and parts for servo motors. Through this, it was confirmed that the pass schedule design method of the multi-pass plug drawing process proposed in this study can be applied to the manufacturing of high-precision hollow steel materials and parts. Keywords : Multi-pass plug drawing process, STS304 stainless steel pipe, Pass schedule design, Advanced strain control model, Finite element analysis

(A) study of catalyst and scavenger for tunable fenton reaction in tungsten chemical mechanical planarization

서은빈 한양대학교 대학원 2020 국내박사

In tungsten (W) film chemical-mechanical-planarization (CMP), the chemical and mechanical reaction behaviors of the W film surface play a critical role in the CMP performance, as determined by oxidation (i.e.,WO3), corrosion (i.e., WO42-), and the electrostatic force at the interface between abrasives and the surface. The Fenton reaction is used to oxidize the W film surface to WO3. The Fenton reaction is a reaction in which radicals and oxygen are generated by the chemical reaction of iron ions and hydrogen peroxide. The radicals and oxygen formed by the Fenton reaction oxidize the surface of W to enhance the polishing rate of W film, but deteriorate the abrasive stability of the slurry. Therefore, in this thesis, the chemical and mechanical properties of the zirconia (ZrO2)-abrasive based slurry using potassium ferric oxalate [K3Fe(C2O4)3] as catalyst and using trilithium tetrahydrate (TCT-Li) as scavenger were studied, respectively. The concentration of K3Fe(C2O4)3 determines the mechanism by which oxidation- or corrosion-dominant of W surface. It was confirmed that the oxidation reaction dominates in the low concentration region (i.e., K3Fe(C2O4)3 concentration 0.001 ~ 0.03 wt%) and the corrosion reaction dominates in the high concentration region (i.e., K3Fe(C2O4)3 concentration 0.03 ~ 1 wt%). To analyze the polishing mechanism, chemical properties such as static etch rate and potentiodynamic polarization curves were investigated. As a result, the ZrO2-based slurry using K3Fe(C2O4)3 as a catalyst showed a maximum W-film polishing rate at a specific concentration (i.e., 0.03 wt%), and corrosion did not occur on the W-film surface after CMP. In addition, the electrostatic force between the abrasives and films affects the polishing rate and the number of remaining particle on the surface after CMP. By measuring the zeta potential of the particles according to the concentration, the electrostatic force between the abrasives and the films is calculated using Coulomb’s law and the number of remaining particle on the surface after CMP was confirmed through surface analysis. In particular, compared to the colloidal-silica based slurry, it showed excellent dishing characteristics of less than ± 4 Å at all concentrations. By selecting the appropriate catalyst concentration, it is possible to achieve sufficient polishing rate, improve abrasive stability, and achieve dishing-less characteristics. In addition, a ZrO2-based W slurry with improved abrasive stability was designed by adding TCT-Li, a scavenger. The formation of Fe(III)-citrate complex improved the abrasive stability by inhibiting the generation of the radical ions that promote the degradation of the dispersant. The characteristics of the W polishing mechanism was studied according to the concentration. Resultantly, the W-film polishing rate and abrasive stability are a trade-off relation. It was found that the tendency for the SiO2 polishing rate to be divided into a region in which the polishing rate increases and a region in which the polishing rate decreases according to the concentration of TCT-Li. The polishing mechanism was understood by analyzing the binding energy of the SiO2 surface and the electrostatic force between the abrasives and the film. I conducted a basic mechanism analysis of a new catalyst and scavenger for controlloing a Fenton reaction based on crystalline-ZrO2 abrasives in W CMP slurry.

Omnidirectional Depth Estimation for Multi-Camera Systems: Addressing Distortion and Leveraging Multi-Scale Features

서은빈 서울대학교 대학원 2025 국내석사

This paper propose the new framework for omnidirectional depth estimation in multi-camera systems, specifically designed to address the challenges posed by fisheye lens distortion and inconsistency of depth map. The proposed method introduces distortion-aware patchfication, leveraging spherical coordinates to mitigate distortion and ensure geometrically consistent feature extraction. Multi-scale tokens are generated to integrate both global and local information effectively, while deformable attention is employed to fuse multi-view features seamlessly. A feature aggregation decoder hierarchically integrates multi-scale features, enabling robust and accurate dense depth predictions. The framework is evaluated on synthetic indoor datasets, OmniThings and OmniHouse, as well as outdoor datasets, including Sunny, Cloudy, and Sunset. Experimental results demonstrate superiority of the proposed method over existing baselines in terms of both accuracy and generalization, especially after fine-tuning. This paper highlights the significance of distortion mitigation, multi-scale feature extraction and integration for advancing omnidirectional depth estimation in multi-camera systems. 본 논문은 멀티 카메라 시스템에서 어안 렌즈의 왜곡과 비일관적인 깊이 추정 문제를 해결하기 위한 새로운 전방위 깊이 추정 프레임워크를 제안한다. 제안된 방법은 왜곡 완화를 위해 구면 좌표를 활용한 왜곡 인지 패치화를 도입하여 기하학적으로 일관된 특징 추출을 가능하게 한다. 또한, 다중 스케일 토큰을 생성하여 전역 및 지역 정보를 효과적으로 통합하며, deformable attention을 활용해 멀티뷰 특징을 원활히 융합한다. 특징 통합 디코더는 다양한 해상도의 특징을 계층적으로 통합하여 견고하고 정확한 깊이 지도를 생성한다. 제안된 프레임워크는 대규모 실내 합성 데이터셋인 OmniThings와 OmniHouse 및 실외 합성 데이터셋인 Sunny, Cloudy, Sunset에서 평가되었으며, 실험 결과 기존 방식 대비 정확도 측면에서 우수한 성능을 보여주었다. 특히, fine-tuning 후 대부분의 평가 항목에서 기존 방법을 능가하는 결과를 보여주고 있다. 본 연구는 왜곡 완화와 다중 스케일 특징 통합이 멀티 카메라 시스템에서의 전방위 깊이 추정 성능 향상에 중요한 요소임을 강조하고 있다.