• 한국태양에너지학회:학술대회논문집
• /
• 2012.03a
• /
• pp.243-247
• /
• 2012

The paper focuses on an anti-reflection (AR) coating deposited by PECVD in silicon solar cell fabrication. AR coating is effective to reduce the reflection of the light on the silicon wafer surface and then increase substantially the solar cell conversion efficiency. In this work, we carried out experiments to optimize double AR coating layer with silicon nitride and silicon oxide for the silicon solar cells. The p-type mono crystalline silicon wafers with $156{\times}156mm^2$ area, 0.5-3 ${\Omega}{\cdot}cm$ resistivity, and $200{\mu}m$ thickness were used. All wafers were textured in KOH solution, doped with $POCl_3$ and removed PSG before ARC process. The optimized thickness of each ARC layer was calculated by theoretical equation. For the double layer of AR coating, silicon nitride layer was deposited first using $SiH_4$ and $NH_3$, and then silicon oxide using $SiH_4$ and $N_2O$. As a result, reflectance of $SiO_2/SiN_x$ layer was lower than single $SiN_x$ and then it resulted in increase of short-circuit current and conversion efficiency. It indicates that the double AR coating layer is necessary to obtain the high efficiency solar cell with PECVD already used in commercial line.

• 한국태양에너지학회:학술대회논문집
• /
• 2011.11a
• /
• pp.245-248
• /
• 2011

Many researches have been carried out to improve light absorption in the crystalline silicon solar cell fabrication. The rear reflection is applied to increase the path length of light, resulting in the light absorption enhancement and thus the efficiency improvement mainly due to increase in short circuit current. In this paper, we manufactured the silicon solar cell using the mono crystalline silicon wafers with $156{\times}156mm^2$, 0.5~3.0 ${\Omega}{\cdot}cm$ of resistivity and p-type. After saw damage removal, the dielectric film ($SiN_x$)on the back surface was deposited, followed by surface texturing in the KOH solution. It resulted in single-side texturing wafer. Then the dielectric film was removed in the HF solution. The silicon wafers were doped with phosphorus by $POCl_3$ with the sheet resistance 50 ${\Omega}/{\Box}$ and then the silicon nitride was deposited on the front surface by the PECVD with 80nm thickness. The electrodes were formed by screen-printing with Ag and Al paste for front and back surface, respectively. The reflectance and transmittance for the single-sided and double-sided textured wafers were compared. The double-sided textured wafer showed higher reflectance and lower transmittance at the long wavelength region, compared to single-sided. The completed crystalline silicon solar cells with different back surface texture showed the conversion efficiency of 17.4% for the single sided and 17.3% for the double sided. The efficiency improvement with single-sided textured solar cell resulted from reflectance increase on back surface and light absorption enhancement.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2010.02a
• /
• pp.282-282
• /
• 2010

결정질 태양전지는 상대적으로 고효율이 보장되며, 낮은 공정 비용 등의 이유로 널리 사용되고 있는 기술이다. 결정질 태양전지의 효율을 증가시키는 공정 방법에는 표면 구조화, 도핑 농도, 반사방지막, 금속전극 형성 등이 있다. 특히, 도핑 공정에서 도핑 농도를 변화시킬 수 있으며, 이에 의하여 면 저항값을 변화 시킬 수 있다. 본 연구에서는 결정질 태양 전지에서 도핑 농도의 조절에 의한 이상적인 sheet resistance를 얻기 위한 실험을 진행하였다. 3개의 실험 set을 두고 각각의 경우를 실험하였다. 본 연구에서는 Pre-deposition과 drive-in 방법을 사용한 doping의 2가지 step으로 실험을 진행하였다. pre-deposition의 시간 condition은 21분으로 하였다. $N_2$ 분위기에서 $O_2$와 $POCl_3$ 의 비율을 각각 100sccm, 200sccm으로 하여 실험을 진행하였다. 변수인 온도의 경우는 각각의 set에 대하여, $830^{\circ}C$, $840^{\circ}C$, $850^{\circ}C$ 로 가변하였다. pre-deposition을 끝낸 뒤, sheet resistance의 값은 각각 $75{\sim}90\;\Omega/square$, $68{\sim}75\;\Omega/square$, $56{\sim}63\;\Omega/square$의 값을 나타내었다. 도핑의 경우에는 drive-in 방법을 사용하였으며, 모든 경우에서 20분에서 $890^{\circ}C$에서 진행하였다. 최종 sheet resistance의 값은 각각의 경우 최대 $33\;\Omega/square$, $34\;\Omega/square$, $30\;\Omega/square$의 값을 나타내었다. $40{\sim}45\;\Omega/square$ 정도의 sheet resistance가 많은 연구에서 이상적인 sheet resistance로 연구되고 있다. 본 연구에서 두 번째 조건이 이상적인 sheet resistance에 가장 접근 하였음을 확인 할 수 있다.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2013.02a
• /
• pp.568-568
• /
• 2013

현재 태양전지 도핑 공정은 퍼니스와 레이저 도핑공정이 주요공정으로 사용되고 있다. 퍼니스 도핑 공정은 POCl3 가스를 도펀트로 사용하여 확산 공정으로 진행한다. 퍼니스 도핑공정은 고가의 장비와 유독 가스사용으로 인한 처리 문제, 웨이퍼의 국부적인 부분에 고농도 도핑을 하는데는 제한적이다. 레이저를 사용한 선택적 도핑의 경우 고가의 레이저장비가 요구되어진다. 본 연구는 기존 도핑공정 문제점을 보완한 저가이면서 새로운 구조의 대기압 플라즈마 제트를 개발하였고, 이를 통한 인산을 사용하여 선택적 도핑에 관한 연구를 하였다. 대기압 플라즈마 제트는 Ar 가스를 주입하여 저주파(1 kHz~100 kHz) 전원을 인가하여 플라즈마를 발생시키는 구조로 제작하였다. 웨이퍼는 태양전지용 P-type shallow 도핑된(120 Ohm/square) 웨이퍼를 사용하였고, 도펀트는 스핀코터를 사용하여 도포를 하였다. 인산의 농도는 10%, 50%, 85%를 사용하였다. 플라즈마 발생 전류는 70 mA, 120 mA에서 실험을 하였다. 대기압 플라즈마 처리시간은 30 s, 90 s, 150 s 처리하여 도핑공정을 진행하였고, 도핑 프로파일은 SIMS (Secondary Ion Mass Spectroscopy)측정을 통하여 분석을 진행하였다. 도펀트의 농도와 전류가 높아짐에 따라서, 도핑 처리시간이 길어짐에 따라서 도핑 깊이가 깊어짐을 확인하였다. 도핑 프로파일을 분석하여 Effective carrier lifetime을 얻었으며, 도펀트 농도가 증가하거나 도핑 처리시간이 길어짐에 따라서 Effective carrier lifetime 낮아짐을 확인하였다.

• Transactions on Electrical and Electronic Materials
• /
• v.4 no.6
• /
• pp.17-20
• /
• 2003

This paper describes the effect of halogenic gettering during oxide passivation of commercial solar cell with the $N^{+}$-P-$N^{+}$ structure. In order to study the effect of halogenic gettering on $N^{+}$-P-$N^{+}$ structure mono-crystalline silicon solar cell, we performed conventional POCl$_3$ diffusion for emitter formation and oxide passivation in the presence of HCl vapors. The $N^{+}$-P-$N^{+}$ structure based silicon solar cells were found to have higher short circuit current and minority carrier lifetime. Their performance was also found to be superior than the conventional $N^{+}$-P-$N^{+}$ structure based mono-crystalline silicon solar cell. The cell parameters of the $n^{+}$-p-$p^{+}$ and $n^{+}$-p-$n^{+}$ structure based cells, passivated by HCl assisted oxidation were measured. The improvement in $I_{sc}$ was attributed to the effect of the increased diffusion length of minority carriers, which came from the halogenic gettering effect during the growth of passivating oxide. The presence of chlorine caused gettering of the cells by removing the heavy metals, if any. The other advantage of the presence of chlorine was the removal of the diffusion induced (in oxygen environment) stacking faults and line defects from the surfaces of the silicon wafers. All these effects caused the improvement of the minority carrier lifetime, which in-turn helped to improve the quality of the solar cells.

• Korean Journal of Materials Research
• /
• v.28 no.11
• /
• pp.680-684
• /
• 2018

We investigate the effect of light intensity and wavelength of a solar cell device using photoconductive atomic force microscopy(PC-AFM). A $POCl_3$ diffusion doping process is used to produce a p-n junction solar cell device based on a polySi wafer, and the electrical properties of prepared solar cells are measured using a solar cell simulator system. The measured open circuit voltage($V_{oc}$) is 0.59 V and the short circuit current($I_{sc}$) is 48.5 mA. Moreover, the values of the fill factors and efficiencies of the devices are 0.7 and approximately 13.6 %, respectively. In addition, PC-AFM, a recent notable method for nano-scale characterization of photovoltaic elements, is used for direct measurements of photoelectric characteristics in limited areas instead of large areas. The effects of changes in the intensity and wavelength of light shining on the element on the photoelectric characteristics are observed. Results obtained through PC-AFM are compared with the electric/optical characteristics data obtained through a solar simulator. The voltage($V_{PC-AFM}$) at which the current is 0 A in the I-V characteristic curves increases sharply up to $18W/m^2$, peaking and slowly falling as light intensity increases. Here, $V_{PC-AFM}$ at $18W/m^2$ is 0.29 V, which corresponds to 59 % of the average $V_{oc}$ value, as measured with the solar simulator. Furthermore, while the light wavelength increases from 300 nm to 1,100 nm, the external quantum efficiency(EQE) and results from PC-AFM show similar trends at the macro scale but reveal different results in several sections, indicating the need for detailed analysis and improvement in the future.

• YAKHAK HOEJI
• /
• v.36 no.1
• /
• pp.1-6
• /
• 1992

Irradiation of phenolbetaine in a stream of nitrogen produced 8,14-cycioberbine[1]. Compound[1] was treated with 10% HCl solution to give the 8-hydroxycycloberbine[2] in 67.7% yield. Subsequently addition of ethylchloroformate to the compound[2] gave rise to the 8-hydroxy-7-ethylcarboxy-9, 10-dimethoxy-2, 3-methylenelioxy-13-oxo-norochotensane[3] in 78% yield. Treatment of the compound[3] with bis-(2-chloroethyl)amine then lead to the 7-bis(2-chloroethyl)carbamyl-norochoteneare[4]. On the other hand the compound[5], which is the 8-methoxynorochotensane, was derived when compound[1] was treated with methanol in a few drops of BF. Treatment of the compound[6], and the compound[7], 7-bis(2-chloroethyl)-carbanyl-8-methoxy-norocheyensane, was then synthesized by reaction of the compound[6] with bis(2-chloroethyl) amine. In the other synthetic pathway when compound[5] was treated with $POCl_3$ in dried benzene, 13-chloro-6-ene-norochetensane[8] with 42% yield was formed. Finally the 13-bis-(2-chloroethyl) amino-8-methoxy-norochotensane[9] was produced when we treated the compound[8] with bis-(2-chloroethyl) amine. In another pathway, reaction between phenolbetaine which is the precursor of the compound[1] and benzoylchloride in dried chloroform gave us the 5,6,7 trihydro-2, 3-methylene-dioxy-9-chloromethyl-10, 11-dimethoxyphenylisoquinoline-8-benzoate[10] in 73% yield. The results of biological activities for these compounds are also presented in Table I and II.

• Journal of the Korean Chemical Society
• /
• v.59 no.5
• /
• pp.373-378
• /
• 2015

The solvolysis rate constants of di-n-butyl phosphorochloridate ((CH₃CH₂CH₂CH₂O)₂POCI, 1) in 28 different solvents are well correlated with the extended Grunwald-Winstein equation, using the N_T solvent nucleophilicity scale and Y_Cl solvent ionizing scale, with the sensitivities values of 1.40 and 0.42 for l and m, respectively. These l and m values can be considered to support an S_N2 reaction pathway. This interpretation is further supported by the activation parameters, i.e., relatively small positive ΔH^≠ (8.0 to 15.9 kcal·mol⁻¹ ) values and large negative ΔS^≠ (−25.8 to −53.1 cal·mol⁻¹ ·K⁻¹ ) values, the Kivinen’s n values (0.9~1.7), and the solvent kinetic isotope effect (1.62).