EffectofhydrogenplasmapassivationonperformanceofHITsolarcells

Effect of hydrogen plasma passivation on performance of HIT solar cellsSeung Jik Lee a, Se Hwan Kim a, Dae Won Kim b , Ki Hyung Kim b , Beom Kyu Kim b, Jin Jang a, na Deptartment of Information Display, Kyung Hee University, Seoul, Republic of Koreab PV Technology Team, Hyosung Co. Ltd., Seoul, Republic of Koreaa r t i c l e i n f oArticle history:Received 5 November 2009Received in revised form16 February 2010Accepted 4 May 2010Available online 26 May 2010Keywords:Silicon heterojunction (SHJ)Surface passivationHydrogen pretreatmenta b s t r a c tWe studied the performance improvement of HIT solar cells by optimizing H 2 plasma exposure anddeposition of thin a-S:H layer on c-Si. With increasing H 2 treatment time, the V OC increases until 80 sand then decreases, indicating the optimum time is 80 s. It is found that the cell performance is almostthe same with and without a thin a-Si:H layer when 80 s plasma is treated on the c-Si before i-layerdeposition. The conversion ef?ciency of 14.04% was achieved at the substrate temperature of 160 1C.H layer thickness [10] . In thispaper, we report on the silicon heterojunction solar cells based onn-type Si, using the PECVD technique for the entire emitterdeposition process. We have studied the effect of H 2 plasmatreatment before the deposition of a-Si:H layer on the c-Si andcompared the performances with and without H 2 plasmatreatment.2. ExperimentalThe heterojunction solar cells were prepared by depositingp-type a-Si:H onto n-type c-Si (CZ, 0.5 –6 O cm, (100), 200 mm).The cleaning procedure of the c-Si prior to deposition was theimmersion of the samples in a (H 2 SO4:H 2 O2 ) (4:1) solution for10 min, followed by rinsing in a de-ionized water over?ow rinsebath (r ? 18 M O cm). Subsequently, the samples were immersedin either diluted HF (2%) or pure NH 4 F, yielding hydrophobicsurface.After this treatment the Si was immediately placed in thechamber and was pumped down. Passivation of c-Si surface wasprocessed in a PECVD chamber using a pretreatment of atomichydrogen. The exposure to atomic hydrogen before i-layerContents lists available at ScienceDirectjournal homepage: www.elsevier.com/locate/solmatSolar Energy Materials fax: +82 2 961 9154.E-mail address: jjang@khu.ac.kr (J. Jang).Solar Energy Materials & Solar Cells 95 (2011) 81– 83deposition helps to clean the wafer surface, remove residualoxide, and passivate surface defects.Intrinsic a-Si:H layers with the thickness of 3–5 nm weredeposited using a SiH 4 and H 2 mixture at the substratetemperature of 160 1C in a conventional PECVD chamber. Thep+a-Si:H with the thickness of 8 nm was deposited by PECVD at asubstrate temperature of 160 1C using a mixture of 1.5% B2 H6 and98.5% SiH 4 . On the other side of the c-Si, intrinsic and n+a-Si:Hlayers were deposited to obtain a back surface ?eld (BSF)structure. Transparent conducting oxide (TCO) layers are formedand ?nally, front and back contact metal electrodes were formedusing a dispenser or dc sputtering method.Fig. 1 shows the silicon heterojunction solar cell structure of afront electrode/TCO/p-a-Si:H/(buffer layer)/n c-Si/buffer layer/n-a-Si:H/TCO/back electrode, where a very thin buffer layer wasintrinsic a-Si:H, which was deposited after the hydrogen plasmatreatment on the c-Si.3. Results and discussionsFig. 2 shows the VOC of the HIT cell with and without 4 nma-Si:H layer plotted as a function of hydrogen plasma treatmenttime. It increases with H 2 treatment time until 80 s and thendecreases, indicating an optimum time of 80 s. This trend is thesame for the cells with and without intrinsic a-Si:H layer.Passivation of c-Si can be improved with the deposition of theintrinsic a-Si:H when the H 2 exposure time is less than 80 s, butthere is almost no effect of intrinsic layer when the exposure timeis 80 s. The excess treatment appears to generate the defects onthe Si wafer and increase surface roughness. It is due to theetching of the Si by H radicals, which is sensitive to the surfaceconditions. Exposure of H atoms can passivate the surface andalso etch the Si surface. It is found that the VOC of the HIT cellwithout a-Si:H layer can be the same as the one with intrinsic i-layer. The Si surface can be passivated with H atoms by using aproper exposure time, which is due to the reduction ofrecombination-generation centers at the interface.Fig. 3 shows the plot of JSC as a function of Voc for the HIT cellsmade with and without H passivation. Without hydrogen plasmatreatment, a thin intrinsic a-Si:H(i-layer) is critical to obtain goodcell performance. The intrinsic a-Si:H buffer layer shows excellentpassivating property, which is due to the reduction of carriersrecombination at the interface [11] .On the other hand, we found that there is no obviousadvantage of using the intrinsic a-Si:H layer in cells that aresubjected to hydrogen plasma treatment. Since hydrogen atomcan passivate Si surface defects, there is little effect of adding athin intrinsic a-Si:H layer.Fig. 4 shows the current density versus voltage characteristics forthe a-Si/n c-Si SHJ solar cells with an 8-nm-thick p layer and 80 shydrogen plasma treatment, exhibiting 0.71 of FF, 34.4 mA/cm 2 ofshort-circuit current density and 0.575 V of open-circuit voltage andthus 14.04% of conversion ef?ciency.4. ConclusionWe have studied the effects of H 2 plasma on c-Si anddeposition of a thin intrinsic a-Si:H layer on c-Si for the HIT solarcells. The conversion ef?ciency can be enhanced by the depositionof a very thin a-Si:H on c-Si and also by the exposure of H2 plasmaFig. 1. Structure of a-Si:H/c-Si heterojunction solar cells: (a) with front intrinsicthin ?lm layer and (b) without front intrinsic thin ?lm layer.Fig. 2. Open-circuit voltage plotted as a function of hydrogen plasma treatment time.Fig. 3. The short-circuit current density as a function of VOC for the HIT solar cellswith/without hydrogen plasma treatment and intrinsic a-Si:H layer.S. Jik Lee et al. / Solar Energy Materials & Solar Cells 95 (2011) 81– 8382on the c-Si. With an 80 s atomic hydrogen plasma treatment theconversion ef?ciency is almost the same with and without anintrinsic a-Si:H layer on the n-type c-Si. However, the prolongedhydrogen plasma treatment ( 4 100 s) reduced VOC . These resultsshow that good junction properties are obtained with a devicequality a-Si:H buffer layer or by proper hydrogen passsivation onc-Si surface. We achieved the conversion ef?ciency of 14.04% for aHIT cell with an 8 nm p-type a-Si:H emitter or by 80 s hydrogenplasma treatment on the c-Si.AcknowledgementThis work was supported in part by Korea Science andEngineering Foundation through the Nano Technology Develop-ment Program (Grant no. 2007-02775).References[1] A. Froitzheim, K. Brendel, L. Elstner, W. Fuhs, K. Kliefoth, M. Schmidt, Interfacerecombination in heterojunctions of amorphous and crystalline silicon,J. Non-Cryst. Solids 299– 302 (2002) 663– 667.[2] Q. Zhang, M. Zhu, F. Liu, Y. Zhou, The optimization of interfacial properties ofnc-Si:H/c-Si solar cells in hot-wire chemical vapor deposition process,J. Mater. Sci. Mater. Electron. 18 (2007) S33– S36.[3] Y. Tsunomura, Y. Yoshimine, M. Taguchi, T. Baba, T. Kinoshita, H. KannoH. Sakata, E. Maruyama, M. Tanaka, Twenty-two percent ef?ciency HIT solarcell, Sol. Energy Mater. Sol. Cells 93 (2009) 670– 673.[4] H. Angermann, W. Henrion, A. Roseler, M. Rebien, Wet-chemical passivationof Si(1 1 1)- and Si(1 0 0)-substrates, Mater. Sci. Eng. B73 (2000) 178– 183.[5] A.G. Aberle, Surface passivation of crystalline silicon solar cells A Review,Prog. Photovolt.: Res. Appl. 8 (2000) 473– 487.[6] Z. Chen, S.K. Pang, K. Yasutake, A. Rohatgi, Plasma-enhanced chemical-vapor-deposited oxide for low surface recombination velocity and high effectivelifetime in silicon, J. Appl. Phys. 74 (1993) 2856– 2860.[7] M. Taguchi, H. Sakata, Y. Yoshimine, E. Maruyama, A. Terakawa, M. Tanaka,An approach for the higher ef?ciency in the HIT cells, in: Proceedings of the31st IEEE Photovoltaic Specialists Conference, FL, 2005, pp. 866– 871.[8] I. Martin, M. Vetter, A. Orpella, J. Puigdollers, A. Cuevas, R. Alcubilla, Surfacepassivation of p-type crystalline Si by plasma enhanced chemical vapordeposited amorphous SiCx:H ?lms, Appl. Phys. Lett. 79 (2001) 2199– 2201.[9] S. Dauwe, J. Schmidt, R. Hezel, Very low surface recombination velocities onp- and n-type silicon wafers passivated with hydrogenated amorphoussilicon ?lms, in: Proceedings of the 29th IEEE Photovoltaic SpecialistsConference, New Orleans, 2002, pp. 1246– 1249.[10] L. Korte, E. Conrad, H. Angermann, R. Stangl, M. Schmidt, Advances in a-Si:H/c-Si heterojunction solar cell fabrication and characterization, Sol. EnergyMater. Sol. Cells 93 (2009) 905– 910.[11] M. Tanaka, M. Taguchi, T. Matsuyama, T. Sawada, S. Tsuda, S. NakanoH. Hanafusa, Y. Kuwano, Development of new a-Si/c-Si heterojunction solarcells: ACJ-HIT (arti?cially constructed junction – heterojunction with intrinsicthin-layer), Jpn. J. Appl. Phys. 31 (1992) 3518– 3522.Fig. 4. Current density vs voltage characteristics for the HIT solar cells underillumination for one of the cells treated with hydrogen plasma. The p-type emitterthickness was 8 nm.S. Jik Lee et al. / Solar Energy Materials & Solar Cells 95 (2011) 81– 83 83