Solar Energy and Clean Energy: Trends and Developmentsdownloads.hindawi.com/journals/specialissues/181840.pdf · evaluating the performance of solar drying in the Malaysian red chili

International Journal of Photoenergy

Solar Energy and Clean Energy: Trends and DevelopmentsGuest Editors: Ching-Song Jwo, Sih-Li Chen, Ho Chang, Yu-Shan Su, and Jen-Shiun Chen

Solar Energy and Clean Energy: Trends andDevelopments

International Journal of Photoenergy

Solar Energy and Clean Energy: Trends andDevelopments

Guest Editors: Ching-Song Jwo, Sih-Li Chen, Ho Chang,Yu-Shan Su, and Jen-Shiun Chen

Copyright © 2013 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “International Journal of Photoenergy.” All articles are open access articles distributed under theCreative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided theoriginal work is properly cited.

Editorial Board

M. S. Abdel-Mottaleb, EgyptNihal Ahmad, USAN. Alonso-Vante, FranceWayne A. Anderson, USAVincenzo Augugliaro, ItalyDetlef W. Bahnemann, GermanyM. A. Behnajady, IranI. R. Bellobono, ItalyRaghu N. Bhattacharya, USAPramod H. Borse, IndiaGion Calzaferri, SwitzerlandAdriana G. Casas, ArgentinaWonyong Choi, KoreaV. Cimrova, Czech RepublicVikram Dalal, USAD. Demetriou Dionysiou, USAM. M. El-Nahass, EgyptAhmed Ennaoui, GermanyChris Ferekides, USADavid Ginley, USAB. Glass, AustraliaShinya Higashimoto, JapanChun-Sheng Jiang, USA

Yadong Jiang, ChinaShahed Khan, USAC. H. Langford, CanadaY. Li, ChinaStefan Lis, PolandNiyaz M. Mahmoodi, IranD. Mantzavinos, GreeceUgo Mazzucato, ItalyJacek Miller, PolandJarugu N. Moorthy, IndiaFranca Morazzoni, ItalyF. Morlet-Savary, FranceE. B. Namdas, AustraliaM. da Graça P. Neves, PortugalLeonidas Palilis, GreeceLeonardo Palmisano, ItalyRavindra K. Pandey, USADavid Lee Phillips, Hong KongPierre Pichat, FranceGianluca Li Puma, UKXie Quan, ChinaTijana Rajh, USAPeter Robertson, UK

Avigdor Scherz, IsraelLukas Schmidt-Mende, GermanyP. Smirniotis, USAZofia Stasicka, PolandJ. Sworakowski, PolandN. Tamaoki, JapanGopal N. Tiwari, IndiaNikolai V. Tkachenko, FinlandVeronica Vaida, USARoel van De Krol, GermanyMark van Der Auweraer, BelgiumEzequiel Wolcan, ArgentinaMan S. Wong, Hong KongDavid Worrall, UKF. Yakuphanoglu, TurkeyMinjoong Yoon, KoreaHongtao Yu, USAJimmy C. Yu, Hong KongJun-Ho Yum, SwitzerlandKlaas Zachariasse, GermanyLizhi Zhang, ChinaJincai Zhao, China

Contents

Solar Energy and Clean Energy: Trends and Developments, Ching-Song Jwo, Sih-Li Chen, Ho Chang,Yu-Shan Su, and Jen-Shiun ChenVolume 2013, Article ID 749167, 2 pages

Fabrication of Large-Grain Thick Polycrystalline Silicon Thin Films via Aluminum-InducedCrystallization for Application in Solar Cells, Hsiao-Yeh Chu, Min-Hang Weng, and Chen LinVolume 2013, Article ID 245195, 4 pages

An Analysis and Research on the Transmission Ratio of Dye Sensitized Solar Cell Photoelectrodes byUsing Different Etching Process, Chin-Guo Kuo, Cheng-Fu Yang, Mu-Jung Kao, Wen-Pin Weng,Chi-Cheng Chang, Lih-Ren Hwang, and Jian-Lan NilVolume 2013, Article ID 151973, 8 pages

DSPACE Real-Time Implementation of MPPT-Based FLC Method, Abdullah M. Noman,Khaled E. Addoweesh, and Hussein M. MashalyVolume 2013, Article ID 549273, 11 pages

The Effect of Sputtering Parameters on the Film Properties of Molybdenum Back Contact for CIGS SolarCells, Peng-cheng Huang, Chia-ho Huang, Mao-yong Lin, Chia-ying Chou, Chun-yao Hsu,and Chin-guo KuoVolume 2013, Article ID 390824, 8 pages

Optical and Electrical Properties of the Different Magnetron Sputter Power 300◦C DepositedGa2O3-ZnO Thin Films and Applications in p-i-n α-Si : H Thin-Film Solar Cells, Fang-Hsing Wang,Chia-Cheng Huang, Cheng-Fu Yang, and Hua-Tz TzengVolume 2013, Article ID 270389, 7 pages

Effects of Titanium Oxide Nanotube Arrays with Different Lengths on the Characteristics ofDye-Sensitized Solar Cells, Chin-Guo Kuo, Cheng-Fu Yang, Lih-Ren Hwang, and Jia-Sheng HuangVolume 2013, Article ID 650973, 6 pages

A New Fuzzy-Based Maximum Power Point Tracker for a Solar Panel Based on Datasheet Values,Ali Kargarnejad, Mohsen Taherbaneh, and Amir Hosein KashefiVolume 2013, Article ID 960510, 9 pages

Fabrication of High Transparency Diamond-Like Carbon Film Coating on D263T Glass at RoomTemperature as an Antireflection Layer, Chii-Ruey Lin, Hong-Ming Chang, and Chien-Kuo ChangVolume 2013, Article ID 612163, 8 pages

Competing in the Global Solar Photovoltaic Industry: The Case of Taiwan, Yu-Shan SuVolume 2013, Article ID 794367, 11 pages

Fabrication of a Cu(InGa)Se2 Thin Film Photovoltaic Absorber by Rapid Thermal Annealing ofCuGa/In Precursors Coated with a Se Layer, Chun-Yao Hsu, Peng-Cheng Huang, Yu-Yao Chen,and Dong-Cherng WenVolume 2013, Article ID 132105, 7 pages

Drying of Malaysian Capsicum annuum L. (Red Chili) Dried by Open and Solar Drying, Ahmad Fudholi,Mohd Yusof Othman, Mohd Hafidz Ruslan, and Kamaruzzaman SopianVolume 2013, Article ID 167895, 9 pages

Experimental Investigation on an Absorption Refrigerator Driven by Solar Cells, Zi-Jie Chien,Hung-Pin Cho, Ching-Song Jwo, Chao-Chun Chien, Sih-Li Chen, and Yen-Lin ChenVolume 2013, Article ID 490124, 6 pages

Low-Voltage Ride-Through Capability of a Single-Stage Single-Phase Photovoltaic System Connected tothe Low-Voltage Grid, Yongheng Yang and Frede BlaabjergVolume 2013, Article ID 257487, 9 pages

Development of a Wind Directly Forced Heat Pump and Its Efficiency Analysis, Ching-Song Jwo,Zi-Jie Chien, Yen-Lin Chen, and Chao-Chun ChienVolume 2013, Article ID 862547, 7 pages

RF Magnetron Sputtering Aluminum Oxide Film for Surface Passivation on Crystalline Silicon Wafers,Siming Chen, Luping Tao, Libin Zeng, and Ruijiang HongVolume 2013, Article ID 792357, 5 pages

Solar Divergence Collimators for Optical Characterisation of Solar Components, D. Fontani, P. Sansoni,E. Sani, S. Coraggia, D. Jafrancesco, and L. MercatelliVolume 2013, Article ID 610173, 10 pages

Synthesis and Characterization of a Gel-Type Electrolyte with Ionic Liquid Added for Dye-SensitizedSolar Cells, Le-Yan Shi, Tien-Li Chen, Chih-Hao Chen, and Kun-Ching ChoVolume 2013, Article ID 834184, 7 pages

Performance of Bulk Heterojunction Solar Cells Fabricated Using Spray-DepositedPoly[[9-(1-octylnonyl)-9H-carbazole-2,7-diyl]-2,5-thiophenediyl-2,1,3-benzothiadiazole-4,7-diyl-2,5-thiophenediyl]/[6,6]-Phenyl C71 Butyric Acid Methyl Ester Blend ActiveLayers,Im-Jun No, Paik-Kyun Shin, Santhakumar Kannappan, Kumar Palanisamy, and Shizuyasu OchiaiVolume 2013, Article ID 202467, 5 pages

Theoretical Evidence for the Distance-Dependent Photoinduced Electron Transfer ofPorphyrin-Oligothiophene-Fullerene Triads, Shan Zhang, Yuanzuo Li, Jing Liu, Meiyu Zhao, Yueyi Han,Yong Ding, Peng Song, and Fengcai MaVolume 2012, Article ID 314896, 10 pages

Hindawi Publishing CorporationInternational Journal of PhotoenergyVolume 2013, Article ID 749167, 2 pageshttp://dx.doi.org/10.1155/2013/749167

EditorialSolar Energy and Clean Energy: Trends and Developments

Ching-Song Jwo,1 Sih-Li Chen,2 Ho Chang,3 Yu-Shan Su,4 and Jen-Shiun Chen5

1 Emission Reduction & Energy Conservation Center, National Taipei University of Technology, 1, Section 3, Chung-Hsiao East Road,Taipei 10608, Taiwan

2Department of Mechanical Engineering, National Taiwan University, No. 1, Section 4, Roosevelt Road, Taipei 10617, Taiwan3 Graduate Institute of Manufacturing Technology, National Taipei University of Technology, 1, Section 3, Chung-Hsiao East Road,Taipei 10608, Taiwan

4Department of Industrial Education, National Taiwan Normal University, No. 162, Section 1, Heping East Road, Taipei 106, Taiwan5Department of Electrical and Computer Engineering, Southern Illinois University Edwardsville, Campus Box 1801,Edwardsville, IL 62026-180, USA

Correspondence should be addressed to Ching-Song Jwo; [email protected]

Received 18 April 2013; Accepted 18 April 2013

Copyright © 2013 Ching-Song Jwo et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Solar energy and other clean energies are emerging andgrowing rapidly in the globe nowadays. Solar energy withless carbon emission is renewable and clean energy forour living environment. Solar energy can be converted toelectricity in photovoltaic (PV) devices, solar cells, or solarthermal/electric power plants.

It is a current trend that solar energy becomes theimportant renewable energy. This special issue addresses therole of the development of solar energy. The themes includefabrication methods, solar cell applications, theoretical anal-ysis, and solar cell development trend. From 28 submissions,19 papers are published in this special issue. Each paper wasreviewed by at least two reviewers and revised according toreview comments.

In Chu et al.’s paper, the authors presented the fabricationof large-grain 1.25 𝜇m thick polycrystalline silicon (poly-Si)films via two-stage aluminum-induced crystallization (AIC)for application in thin-film solar cells. In S. Chen et al.’spaper, the authors presented that the reactive sputtering isan effective technique of fabricating aluminum oxide surfacepassivation film for low-cost high efficiency crystallinesilicon solar cells. In I.-J. No et al.’s paper, the authorspresented the poly [[9-(1-Octylnonyl)-9H-carbazole-2,7-Diyl]-2,5-thiophenediyl-2,1,3-benzothiadiazole-4,7-diyl-2,5-thiophenediyl]/[6, 6]-phenyl C71 butyric acid methyl esterblend active layers prepared by spray deposition method.

The photovoltaic cells were fabricated using the spray coatedactive layers with and without TiO𝑥 layer. In L.-Y. Shi et al.’spaper, the authors presented that the dye-sensitized solarcells composed of the gel-type electrolyte with no ionicliquid added can acquire 4.13% photoelectric conversionefficiency. In C.-Y. Hsu et al.’s paper, the authors presentedthat the precursor with a Cu/(In + Ga) ratio of 0.95 exhibitslarger grains and lower resistance, which is suitable forits application to solar cells. In C.-G. Kuo et al.’s paper,the authors presented that the measured photovoltaicperformance of the dye-sensitized solar cells was dependenton the TNT-array length. In F.-H. Wang et al.’s paper, theauthors presented that the efficiencies of the 𝛼-Si thin-filmsolar cells increased from 2.83% to 3.38% as the depositionpower decreased from 150 W to 50 W when the GZO thinfilms were used as the front transparent conductive thinfilms. In C.-G. Kuo et al.’s paper, the authors presentedthat the electrochemical polishing-chemical etching andchemical etching processes were better for formation of TiO2nanotube arrays than the polishing-chemical etching processbecause they could fabricate TiO2 nanotube arrays with thebetter characteristics to be used as the photoelectrodes of dyesensitized solar cells. In C.-R. Lin et al.’s paper, the authorspresented depositing high transmittance diamond-likecarbon (DLC) thin films on D263T glass substrate at roomtemperature via a diamond powder target using the radio

2 International Journal of Photoenergy

frequency (RF) magnetron sputtering technique. DLC filmunder an RF power of 150 W possesses high transmissiveability (>81%) and low average reflectance ability (


Research ArticleFabrication of Large-Grain Thick Polycrystalline Silicon ThinFilms via Aluminum-Induced Crystallization for Application inSolar Cells

Hsiao-Yeh Chu,1 Min-Hang Weng,2 and Chen Lin3

1 Department of Mechanical Engineering, Kun Shan University, Tainan 71003, Taiwan2Metal Industries Research & Development Center, Kaohsiung 81160, Taiwan3 Taiwan Cement Corporation, Taipei 10448, Taiwan

Correspondence should be addressed to Hsiao-Yeh Chu; [email protected]

Received 24 December 2012; Revised 27 March 2013; Accepted 29 March 2013

Academic Editor: Ching-Song Jwo

Copyright © 2013 Hsiao-Yeh Chu et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The fabrication of large-grain 1.25 𝜇m thick polycrystalline silicon (poly-Si) films via two-stage aluminum-induced crystallization(AIC) for application in thin-film solar cells is reported. The induced 250 nm thick poly-Si film in the first stage is used as the seedlayer for the crystallization of a 1𝜇m thick amorphous silicon (a-Si) film in the second stage. The annealing temperatures in the twostages are both 500∘C. The effect of annealing time (15, 30, 60, and 120 minutes) in the second stage on the crystallization of a-Si filmis investigated using X-ray diffraction (XRD), scanning electron microscopy, and Raman spectroscopy. XRD and Raman resultsconfirm that the induced poly-Si films are induced by the proposed process.

1. Introduction

Silicon thin films can reduce the cost of solar cells and asso-ciated consumption of pure silicon. Polycrystalline silicon(poly-Si) film is a promising material for solar cell applicationbecause its carrier mobility is 10 to 100 times larger thanthat of a hydrogenated amorphous silicon (a-Si:H) film [1].Three methods are currently used for manufacturing poly-Si film on glass: catalytic chemical vapor deposition [2],excimer laser annealing (ELA) [3], and metal-induced crys-tallization (MIC) [4–10]. Aluminum-induced crystallization(AIC) induces the crystallization of a-Si below the eutectictemperature (577∘C) of Al and Si. Therefore, it can beeffectively applied to large-area glass substrates at processtemperatures below the glass transformation temperature [4,5]. In addition, AIC can create poly-Si film with a lateralgrain size that is larger than the film thickness [5]. However,commonly studied poly-Si films often have a thickness ofbelow 500 nm [6–10], which is insufficient for the activelayer of solar cells. For example, Hossain et al. [9] fabricatedpoly-Si films with a lateral grain size of up to 20𝜇m, butthe film thickness was as low as 300 nm. Subramanian

et al. [10] fabricated a poly-Si film via two-step solid-phasecrystallization. The process temperature was as high as 700∘C.The grain size and film thickness were 350 nm and 100 nm,respectively. Two-step AIC was used in the study of Tüzünet al. [11]. However, their process temperature was as high as1160∘C.

In this work, 1.25 𝜇m thick large-grain poly-Si films arefabricated via a two-stage AIC method. The process tem-perature is 500∘C, which is below the glass transformationtemperature.

2. Experiments

Figure 1 shows the experimental procedure of the two-stageannealing for fabricating thick poly-Si films. The substratematerial is a wafer. The standard RCA cleaning process wasapplied before film deposition. In order to simulate a glasssubstrate and prevent the crystal orientation of the wafersubstrate from affecting the crystallization of the a-Si:Hlayer [9], a 200 nm thick SiO2 film was deposited over thewafer surface in a wet oxide tube using atmospheric-pressure


Wet oxidation

Silicon wafer RCA cleaning

Al sputtering (250 nm)

PECVD a-Si deposition(250 nm)

PECVD a-Si deposition (1 𝜇m)

Al etching

Crystallinity Element analysisMorphology

XRDRaman

EDSSEM

Electricmeasurement

IV

1st annealing

2nd annealing

Figure 1: Flow chart of experimental procedure.

Intensity

(a.u.)

30 min

15 min

60 min

120 min

20 30 40 50 60

Si (1

11)

Si (2

20)

Si (3

11)

2 𝜃 (deg)

Figure 2: XRD patterns of specimens annealed for various dura-tions.

chemical vapor deposition (APCVD). Then, an aluminumfilm was deposited on the top surface of the SiO2 filmby sputtering. A layer of a-Si film was then deposited ontop of the Al film using plasma-enhanced chemical vapordeposition (PECVD). The specimen was then annealed at500∘C for 1 hour at a nitrogen flow rate of 2 slm in the firststage of annealing.

After the first annealing process, a 1𝜇m thick a-Si film wasdeposited by PECVD. The specimens were then annealed at500∘C for 15, 30, 60, and 120 minutes, respectively. The spec-imens were then wet-etched in order to remove the residualaluminum content.

Intensity

(a.u.)

400 450 500 550 600

Raman shift (cm−1)

30 min

15 min

60 min

120 min

Figure 3: Raman spectra of specimens annealed for various dura-tions.

X-ray diffraction (XRD, Rigaku RINT 2000) and Ramanspectroscopy (TRIAX 550) were used to evaluate the crys-tallinity of the induced poly-Si film. Scanning electron micro-scopy (SEM, Philips XL-40FEG) was used to observe themorphology of the film surface and cross section. The leakagecurrent density of the induced poly-Si thin film was measuredto evaluate the film quality.

3. Results and Discussion

Figure 2 shows the XRD patterns of the induced poly-Sifilms after the second annealing process. Three silicon peaks,corresponding to Si (111), (220), and (311), respectively, appear

International Journal of Photoenergy 3

(a) (b)

(c) (d)

(e) (f)

Figure 4: SEM images and EDS results of induced poly-Si thin films annealed for various durations: (a) 15 min, top view; (b) 15 min, cross-sectional view; (c) 15 min, EDS results; (d) 120 min, top view; (e) 120 min, cross-sectional view; and (f) 120 min, EDS results.

for all samples, confirming that poly-Si film was successfullyinduced in each case. No Si (100)-related peaks were detectedin any specimen even though a Si (100) wafer was used as thesubstrate. This is due to the glancing angle being very small(1∘) during the XRD measurement and the 200 nm thick SiO2film being sufficiently thick to prevent X-rays from reachingthe Si (100) wafer substrate [10]. Note that there is a fairlysmall peak at around 2𝜃 = 39∘ that corresponds to Al (111) forall specimens in Figure 2. This indicates that Al had not beencompletely etched off even though selective Al etching wasconducted on the surface of the crystallized Si film. The peakintensity increases with annealing time, but eventually levelsoff, indicating that a second-stage annealing time of over120 minutes will not further increase the crystalline siliconintensity.

Raman spectra measurements were performed on theinduced poly-Si thin films after the second stage to confirmcrystallinity. The results are shown in Figure 3. The maxi-mum-intensity peaks of the Raman shift are located between492 and 497 cm−1. This result verifies that poly-Si films wereinduced.

Figure 4 shows SEM images of the top surface and crosssection of the induced poly-Si thin films after the two-stageannealing AIC process. Figures 4(a) and 4(d) show thatthe circular grain size of the poly-Si film is about 1∼2 𝜇m.The grain size increases with increasing annealing time.The energy-dispersive spectrometer (EDS) results in Figures4(c) and 4(f) show that the Al residue intensity decreaseswith increasing the annealing time. It indicates that the Alprecipitation from the seed layer increases with increasing


0 2 4 6 8 10

15 min30 min

60 min120 min

1.00𝐸 − 09

3.00𝐸 − 09

5.00𝐸 − 09

Bias voltage (V)

Current d

ensity (A

/cm2)

Figure 5: Leakage current densities of the induced poly-Si thin filmsannealed under four different annealing time periods.

the annealing time. The cross-sectional SEM images inFigure 4 clearly show the first and second layers of theinduced poly-Si. The lateral grain size is larger than the filmthickness.

Figure 5 shows the leakage current densities of theinduced poly-Si thin films. The leakage current densityslightly increases with increasing annealing time. The leakagecurrent densities for all induced poly-Si thin films are below5 × 10−9 A/cm2, indicating that the induced films have goodquality without many defects or serious grain boundary effect[10].

4. Conclusion

A thick large-grain poly-Si film for application in solar cellswas fabricated via AIC. The overall thickness of the poly-Sifilm was about 1.25 𝜇m. XRD patterns and Raman spectraconfirmed that the poly-Si film was induced using the pro-posed process at 500∘C.

Acknowledgments

This work was supported by National Nano Device Laborato-ries, Taiwan, under Project NDL 94S-C046. The authors alsowould like to thank Professor Ru-Yuan Yang in the GraduateInstitute of Materials Engineering, National Ping-Tung Uni-versity of Science and Technology. A lot of materials analyseswere finished in Professor Yang’s labs.

References

[1] S. W. Lee, Y. C. Jeon, and S. K. Joo, “Pd induced lateral crystal-lization of amorphous Si thin films,” Applied Physics Letters, vol.66, no. 13, pp. 1671–1673, 1995.

[2] O. Ebil, R. Aparicio, S. Hazra, R. W. Birkmire, and E. Sutter,“Deposition and structural characterization of poly-Si thinfilms on Al-coated glass substrates using hot-wire chemicalvapor deposition,” Thin Solid Films, vol. 430, no. 1-2, pp. 120–124, 2003.

[3] F. M. Zhang, X. C. Liu, G. Ni, and Y. W. Du, “Controlled growthof high-quality poly-silicon thin films with huge grains on glasssubstrates using an excimer laser,” Journal of Crystal Growth,vol. 260, no. 1-2, pp. 102–108, 2004.

[4] J. Schneider, R. Heimburger, J. Klein, M. Muske, S. Gall, andW. Fuhs, “Aluminum-induced crystallization of amorphous sil-icon: influence of temperature profiles,” Thin Solid Films, vol.487, no. 1-2, pp. 107–112, 2005.

[5] M. Shahidul Haque, H. A. Naseem, and W. D. Brown, “Inter-action of aluminum with hydrogenated amorphous silicon atlow temperatures,” Journal of Applied Physics, vol. 75, no. 8, pp.3928–3935, 1994.

[6] M. S. Ashtikar and G. L. Sharma, “Structural investigation ofgold induced crystallization in hydrogenated amorphous sili-con thin films,” Japanese Journal of Applied Physics, vol. 34, no.10, pp. 5520–5526, 1995.

[7] H. Kim, G. Lee, D. Kim, and S. H. Lee, “A study of polycrys-talline silicon thin films as a seed layer in liquid phase epi-taxy using aluminum-induced crystallization,” Current AppliedPhysics, vol. 2, no. 1, pp. 129–133, 2002.

[8] P. I. Widenborg and A. G. Aberle, “Surface morphology of poly-Si films made by aluminium-induced crystallisation on glasssubstrates,” Journal of Crystal Growth, vol. 242, no. 3-4, pp. 270–282, 2002.

[9] M. Hossain, H. M. Meyer, H. H. Abu-Safe, H. Naseem, andW. D. Brown, “Large-grain poly-crystalline silicon thin filmsprepared by aluminum-induced crystallization of sputter-depo-sited hydrogenated amorphous silicon,” Journal of MaterialsResearch, vol. 21, no. 3, pp. 761–766, 2006.

[10] V. Subramanian, P. Dankoski, L. Degertekin, B. T. Khuri-Yakub,and K. C. Saraswat, “Controlled two-step solid-phase crystal-lization for high-performance polysilicon TFT’s,” IEEE ElectronDevice Letters, vol. 18, no. 8, pp. 378–381, 1997.

[11] Ö. Tüzün, A. Slaoui, I. Gordon et al., “N-type polycrystallinesilicon films formed on alumina by aluminium induced crys-tallization and overdoping,” Thin Solid Films, vol. 516, no. 20,pp. 6892–6895, 2008.


Research ArticleAn Analysis and Research on the Transmission Ratio ofDye Sensitized Solar Cell Photoelectrodes by Using DifferentEtching Process

Chin-Guo Kuo,1 Cheng-Fu Yang,2 Mu-Jung Kao,3 Wen-Pin Weng,4 Chi-Cheng Chang,5

Lih-Ren Hwang,6 and Jian-Lan Nil7

1 Department of Industrial Education, National Taiwan Normal University, Taipei 10610, Taiwan2Department of Chemical and Materials Engineering, National University of Kaohsiung, Kaohsiung 81148, Taiwan3Department of Vehicle Engineering, National Taipei University of Technology, Taipei 10608, Taiwan4Department of Chemical and Materials Engineering, Lunghwa University of Science and Technology, Taoyuan 33306, Taiwan5Department of Information Management, Lunghwa University of Science and Technology, Taoyuan 33306, Taiwan6Department of Automation and Control Engineering, Chung Chou University of Science and Technology, Yuanlin 51003, Taiwan7Department of Mechatronic Technology, National Taiwan Normal University, Taipei 10610, Taiwan

Correspondence should be addressed to Mu-Jung Kao; [email protected]

Received 13 December 2012; Revised 5 February 2013; Accepted 22 February 2013

Academic Editor: Ho Chang

Copyright © 2013 Chin-Guo Kuo et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Classical photoelectrodes for Dye Sensitized Solar Cells (DSSCs) were fabricated by using the electrochemical method on thetitanium (Ti) template, for that the fabrication process would influence the characteristics of the DSSCs. In this study, at first threedifferent methods were used to etch Ti templates from 10 to 17 min, (1) polishing-chemical etching: Ti template was annealed at450∘C for 1 h, abraded using number 80 to 1500 SiC sheet, and then etched in a solution of 5% HF + 95% H2O; (2) electrochemicalpolishing-chemical etching: Ti template was annealed at 450∘C for 1 h, electrolytic polishing with 42% CH3OH + 5% HClO4 + 53%HOCH2CH2OC4H9 solution, and the chemical-etching in a solution of 5% HF + 95% H2O; (3) chemical etching: Ti template wasetched in a solution of 5% HF + 95% H2O and annealed at 450

∘C for 1 h. When the etching time was changed from 10 to 17 min,the thicknesses of Ti templates decreased from 75.3𝜇m to 14.8𝜇m, depending on the etching method. After etching process, theTiO2 nanotube arrays were fabricated as the photoelectrode of DSSCs by electrochemical process, in which the Ti as anode andplatinum (Pt) as cathode. The electrolyte solution included C2H4(OH)2, NH4F, and deionized water. After annealing the grownTiO2 nanotube arrays at 450

∘C for 3 h, we would show that the etching process had large effect on the structure and transmittanceratio of the TiO2 nanotube arrays.

1. Introduction

Tubular inorganic nanostructures offer great potential for usein heterojunction solar cells, fuel cells, molecular filtration,tissue engineering, and Dye Sensitized Solar Cells (DSSCs).In contrast to random nanoparticle systems where slowelectron diffusion typically limits device performance [1, 2],the precisely oriented nature of the crystalline (after anneal-ing) nanotube arrays makes them excellent electron percola-tion pathways for vectorial charge transfer between interfaces[2, 3]. Highly ordered vertically oriented nanotube arrays

fabricated by different metal-oxide can be used as photoelec-trode materials of the DSSCs, which include titanium dioxide(TiO2) [4], ferric oxide (Fe2O3) [5], zinc oxide (ZnO) [6], andso forth. Among those proposed photoelectrode materials,TiO2 is best because it has the merits of low cost, chemicalstability, and good charge transport [7].

As numerous major advances in research and technologyover the past decade have been made possible by the suc-cessful development of nanostructures, various avenues havebeen used to fabricate a diversity of TiO2 nanostructure pho-toelectrodes, including the sol-gel method [8], metalorganic


99.7% titanium plates 99.995% titanium plates

Annealing

Chemical etching

Abrading

Annealing

Electrolytic polishing

Chemical etching

99.7% titanium plates

Chemical etching

Annealing

Photoelectrodes growth SEM measurement

UV-Vis measurementSEM measurement

Figure 1: Experimental flow chart.

chemical vapor deposition (MOCVD) [9], templating [10],and electrochemical method [11]. Although many of thosefabrication routes are complicated due to the use of templatesor the nature of the involved chemical processes, it has beendemonstrated that self-organized vertically oriented titaniumdioxide (TiO2) nanotube arrays can be fabricated using asimple anodization technique. Using electrochemical methodto fabricate the TiO2 nanotube arrays as photoelectrodes isfirst one using hydrofluoric acid (HF) electrolyte by Zwillinget al. [12]. After that, many electrolytes are also developedto fabricate the TiO2 nanotube arrays as photoelectrodes,such as HF/H2O, HF/H2SO4/H2O [13], and EG (ethyleneglycol)/NH4F/H2O [14]. When the EG/NH4F/H2O elec-trolyte is used to fabricate TiO2 nanotube arrays, the lengthof TiO2 photoelectrodes has the value between 10 𝜇m and100 𝜇m [14].

When using as the photoelectrodes, the property of theanatase phase TiO2 is believed to be superior than that ofthe rutile phase TiO2. Because the band gap of anatase phaseTiO2 is 3.2 eV and rutile phase TiO2 is 3.0 eV, the light absorp-tion edge of anatase phase TiO2 is located at lower wave-length. When the ultraviolet light is irradiated, the pho-toelectrodes of TiO2 nanotube arrays are proceeded pho-tocatalytic activities to make light through photoelectrodesto dye layer [15]. Except the crystalline phase, when theTiO2 nanotube arrays are used as the photoelectrodes, thetransmission ratio of TiO2 nanotube arrays has large effecton the efficiency of the fabricated DSSCs. In this study,Dye Sensitized Solar Cells (DSSCs) photoelectrodes, TiO2nanotube arrays, were fabricated by using electrochemical

Figure 2: Cross-section observation of annealed titanium plate.

method and using UV-Vis examination transmission ratio inthe DSSCs structure. Therefore, the purpose of this study wasto investigate the effect of etching process on transmissionratio of the TiO2 nanotube arrays after treating by usingdifferent etching methods.

2. Experimental

The Ti metals were cut into Ti-plates with a size of 2.5 cm× 2.5 cm and thickness of ∼142𝜇m. The flow chart for thedifferent etching process, the process for formation of TiO2nanotube arrays, and the characteristic measurements of theTiO2 nanotube arrays was shown in Figure 1 and describedbelow.


(a) (b) (c)

(d) (e) (f)

(g) (h)

Figure 3: Cross-section morphologies of polishing-chemical etching under different times (a) 10 min, (b) 11 min, (c) 12 min, (d) 13 min, (e)14 min, (f) 15 min, (g) 16 min, and (h) 17 min.

2.1. Preprocess

2.1.1. Polishing-Chemical Etching. A pure Ti plate with 99.7%purity was used as the template for the polishing-chemicaletching process. The template was first annealed 1 h at 450∘Cand cooled in air, then the template was abraded using SiCsheets (Nos 80 to 1500), and then it was ultrasonicated for30 min in distilled water. After that, the mixture of 5% HF +95% H2O was used as the etching solution and the etchingtime was changed from 10 to 17 min.

2.1.2. Electrochemical Polishing-Chemical Etching. Anothertitanium plate with higher purity 99.995% was also usedas the template, too. The sample was first annealed 1 h at450∘C and then it was ultrasonicated for 30 min in distilledwater. The polishing electrolytic solution was mixed with 42%methanol (CH3OH) + 5% perchloric acid (HClO4) + 53%ethylene glycol monobutylether (HOCH2CH2OC4H9), andthe etching solution was mixed by 5% HF + 95% H2O; theetching time was changed from 10 to 17 min.

2.1.3. Chemical Etching. A pure titanium plate with 99.7%purity was also used as the template for the chemical etchingprocess. The mixture of 5% HF + 95% H2O was used as theetching solution and the etching time was changed from 10 to16 minutes. After the etching process the titanium plate wasannealed at 450∘C for 1 h.

2.2. Preparation of Photoelectrodes. We fabricated orderednanochannel TNT arrays at 25∘C on the prepared titanium(Ti) square foils (2 cm × 2 cm) at a constant voltage of50 V and Ti foils as an anode. The electrolyte solutionscontained 0.2 M ammonium fluoride (NH4F) + 2% H2O+ ethylene glycol (C2H4(OH)2) with anodization. Ti foilswere degreased by ultrasonication in acetone and then iso-propanol, respectively, for about 30 min, followed by rinsingwith deionized (DI) water, and finally dried in the air beforeused. Highly ordered TiO2 nanotube arrays over large areawere prepared by a potentiostatic anodization in a two-electrode electrochemical cell with a platinum (Pt) sheet ascounter electrode. All anodization experiments were carried


(a) (b) (c)

(d) (e) (f)

(g) (h)

Figure 4: Cross-section morphologies of electrochemical polishing-chemical etching under different times (a) 10 min, (b) 11 min, (c) 12 min,(d) 13 min, (e) 14 min, (f) 15 min, (g) 16 min, and (h) 17 min.

out at room temperature. After the electrochemical process,the foils were annealed at 450∘C for 3 h.

3. Results and Discussion

3.1. Morphologies for the Cross-Section of Nonetched andEtched Ti Specimens. Figure 2 shows the cross-section of thenonetched and annealed Ti specimen, the nonetched Ti spec-imen showed a densified structure and no defect and porouswere observed. After different etching processes were used,the etch-treated Ti plates were thoroughly rinsed with dis-tilled water for clean at room temperature and morphologiesfor the cross-section were observed. In Figures 3–5, a series ofmicrographs of titanium plate after various etching processesand at different etching time, from 10 to 17 min, are observed.From Figures 2 and 5, the thickness of titanium plates linearlydecreased with etching time and, however, the nonuniformityin thickness became apparently at a longer etching time.For the polishing-chemical etching specimens, as the etchingtime was equal to and longer 13 min, the thickness becomeobviously asymmetrical. As shown in Figure 3, the polished

surface of Ti plate was changed to be a porous structure afteralkali treatment. Apparently, the porosity of the structure wasobserved to be increased after chemical etching treatment.

For the electrochemical polishing-chemical etching spec-imens, as Figure 4 shows, asymmetrical in the thickness wasnot observed even the etching time was 17 min. Figure 4also shows that even with the electrolytic publishing andchemical etching the Ti plate had a smooth surface. Forthe results shown in Figure 5, after annealing treatment, theporous structure seemed to become more compact and rigidthan before the annealing treatment, and asymmetrical in thethickness was not observed even the etching time was 17 min.A cross-section of the etched Ti plates was observed by usingscanning electron microscope (SEM) and the thickness of theTi plates was measured by SEM. The variation in the thicknessof Ti plates under different etching processes as a function ofetching time is displayed in Figure 6.

When the etching time was increased from 10 to 17 min,the thicknesses of the polishing-chemical etching, electro-chemical polishing-chemical etching, and chemical etchingTi plates decreased from 75.3𝜇m to 33 𝜇m, from 60.4 𝜇m


(a) (b) (c)

(d) (e) (f)

(g)

Figure 5: Cross-section morphologies of chemical etching under different times (a) 10 min, (b) 11 min, (c) 12 min, (d) 13 min, (e) 14 min, (f)15 min, and (g) 16 min.

020406080100120140

0 10 11 12 13 14 15 16 17

Polishing-chemical etchingElectrochemical polishing-chemical etchingChemical etching

Time (min)

Thic

knes

s (𝜇

m)

Figure 6: Titanium plates thickness curve chart of different meth-ods and etching times.

to 23.6 𝜇m, and from 50 𝜇m to 14.8 𝜇m, respectively. AsFigure 6 shows, as the same etching time is used, the thick-nesses of the chemical-etching Ti plates are thinner than

those of the other two processes-treated Ti plates, and theabrading process is the reason to cause this result. From theresults shown in Figures 3–5, the porous structure on the sur-faces of Ti plates only existed in polishing-chemical etching Tiplates but not existing in electrochemical polishing-chemicaletching and chemical etching Ti plates. A similar porousstructure was reported to be produced by the alkali (KOH)treatment of a commercial Ti plate; the porous structurewas composed of nanowires with a diameter of less than30 nm, and the thickness of nanowire layer was estimated tobe approximately 500 nm on the Ti substrate [16]. For that,the different etching processes would have large effect on theformation of the TiO2 nanotube arrays. If the thickness of Tiplates is too thin, the photoelectrodes of the TiO2 nanotubearrays are not easy to be grown; therefore, the 11 min treatedTi plates are chosen for producing the TiO2 nanotube arrays.

3.2. Morphology of TiO2 Nanotube Arrays. From the X-raydiffraction patterns (not shown here), the main crystallinephase of the as-prepared TiO2 nanotube arrays was amor-phous, the diffraction peaks of Ti and rutile TiO2 phases


(a) (b)

Figure 7: Morphologies of the TiO2 nanotube arrays grown on polishing-chemical etching Ti plate (a) top view and (b) side view.

(a) (b)

Figure 8: Morphologies of the TiO2 nanotube arrays grown on electrochemical polishing-chemical etching Ti plates (a) top view and (b)side view.

could be found, and these diffraction peaks were notenhanced as the TiO2 nanotube array length was changed.It is noted that the anatase TiO2 peaks are much worse thanthe Ti and rutile TiO2 peaks, so it is reasonable to neglectthe influences of such trace anatase TiO2 content in the as-prepared TiO2 nanotube arrays. As TiO2 nanotube arrayswere annealed at 450∘C for 1 h, the diffraction peaks of anataseTiO2 were clearly observed. Figures 7–9 show the top viewand side view of TiO2 nanotube arrays. The top views inFigures 7(a), 8(a), and 9(a) show that all of the TiO2 nan-otube arrays had the diameters between 100 nm and 140 nm.The surfaces of TiO2 nanotube arrays with electrochemicalpolishing-chemical etching and chemical etching processeswere smoother than that of TiO2 nanotube arrays withpolishing-chemical etching process. The side views in Figures7(b), 8(b), and 9(b) show that the lengths of TiO2 nanotubearrays with the polishing-chemical etching, electrochemicalpolishing-chemical etching, and chemical etching processesare 19.6 𝜇m, 18.2 𝜇m, and 22.6𝜇m, respectively. From the sideviews in Figures 7–9 TiO2 nanotube arrays with the polish-ing-chemical etching process are not good as the photoelec-trodes of DSSCs because TiO2 nanotube arrays are easilypeeled off. Figures 8(b) and 9(b) show that the TiO2 nanotubearrays with polishing-chemical etching and chemical etching

processes are not easily to be peeled off. Figure 7(b) alsoshows that the TiO2 nanotube arrays reveal an unsmoothsurface, which will influence the transmission ratio and willbe proven in Figure 10.

3.3. Transmission of Photoelectrodes Nanotubes. The trans-mission ratios of TiO2 nanotube arrays are shown in Figure 10as a function of different polishing process and optical wave-length. The measured structure includes the layers ofTiO2 nanotube arrays (photoelectrode), dye, electrolyte, andcounter electrode. When the sunlight passes through the pho-toelectrode layer to dye layer, the transmission ratio has noapparent change. The results in Figure 10 show that the struc-tures with the polishing-chemical etching (70–75%) and ele-ctrochemical polishing-chemical etching processes (80–85%)have high transmission ratio in the range of visible light.The structure with the polishing-chemical etching processhas the lower transmission ratio which is caused by thebeing peeled off and unsmooth surfaces. Because TiO2 nano-tube arrays with the chemical etching process have the thickerthickness, the transmission ratio (20–25%) is very low. Fromthose results, the electrochemical polishing-chemical etchingand chemical etching processes are the better for formationof TiO2 nanotube arrays than polishing-chemical etching


(a) (b)

Figure 9: Morphologies of the TiO2 nanotube arrays grown on chemical etching Ti plates (a) top view and (b) side view.

0102030405060708090100

300 400 500 600Wavelength (min)

Tran

smiss

ion

ratio

(%)

Polishing-chemical etchingElectrochemical polishing-chemical etchingChemical etching

Figure 10: Transmission ratio curves of the TiO2 nanotube arraysgrown on etching Ti plates.

process because they have the better structure and mor-phology factors to get the better characteristics of TiO2 nano-tube arrays as the photoelectrodes of DSSCs.

4. Conclusions

In this study, three different methods were used to etch Titemplate from 10 to 17 min, the experimental results weresummarized as follows.

(i) The thicknesses of the polishing-chemical etch-ing, electrochemical polishing-chemical etching, andchemical etching Ti plates decreased from 75.3 𝜇m to33 𝜇m, from 60.4𝜇m to 23.6 𝜇m, and from 50𝜇m to14.8 𝜇m, respectively.

(ii) The TiO2 nanotube arrays with polishing-chemicaletching, electrochemical polishing-chemical etching,and chemical etching processes have the lengths of19.6 𝜇m, 18.2 𝜇m, and 22.6𝜇m, respectively, and theirdiameters were between 100 nm and 140 nm.

(iii) In the range of visible light, the TiO2 nanotubearrays with polishing-chemical etching (70–75%)

and electrochemical polishing-chemical etching (80–85%) processes had the higher transmission ratiothan that of the TiO2 nanotube arrays with chemicaletching process (20–25%).

(iv) The electrochemical polishing-chemical etching andchemical etching processes were better for formationof TiO2 nanotube arrays than the polishing-chemicaletching process because they could fabricate TiO2nanotube arrays with the better characteristics to beused as the photoelectrodes of DSSCs.

Acknowledgments

The authors acknowledge financial support from NSC 101-2514-S-003-004-, NSC 101-3113-S-262-001-, and NSC 101-2622-E-003-002-CC3.

References

[1] M. Paulose, K. Shankar, S. Yoriya et al., “Anodic growth of highlyordered TiO2 nanotube arrays to 134 𝜇m in length,” The Journalof Physical Chemistry B, vol. 110, no. 33, pp. 16179–16184, 2006.

[2] M. Law, L. E. Greene, J. C. Johnson, R. Saykally, and P. D. Yang,“Nanowire dye-sensitized solar cells,” Nature Materials, vol. 4,no. 6, pp. 455–459, 2005.

[3] G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese, and C. A.Grimes, “Use of highly-ordered TiO2 nanotube arrays in dye-sensitized solar cells,” Nano Letters, vol. 6, no. 2, pp. 215–218,2006.

[4] A. Fujishima and K. Honda, “Electrochemical photolysis ofwater at a semiconductor electrode,” Nature, vol. 238, no. 5358,pp. 37–38, 1972.

[5] M. A. Gondal, M. N. Sayeed, and A. Alarfaj, “Activity compari-son of Fe2O3, NiO, WO3, TiO2 semiconductor catalysts inphenol degradation by laser enhanced photo-catalytic process,”Chemical Physics Letters, vol. 445, no. 4–6, pp. 325–330, 2007.

[6] Y. Li, W. Xie, X. Hu et al., “Comparison of dye photodegradationand its coupling with light-to-electricity conversion over TiO2and ZnO,” Langmuir, vol. 26, no. 1, pp. 591–597, 2010.

[7] N. K. Allam and M. A. El-Sayed, “Photoelectrochemical wateroxidation characteristics of anodically fabricated TiO2 nan-otube arrays: structural and optical properties,” Journal of Phy-sical Chemistry C, vol. 114, pp. 12024–12029, 2010.


[8] M. Gotic, M. Ivanda, A. Sekulic, S. Popovic, A. Turkovic, and K.Furic, “Microstructure of nanosized TiO2 obtained by sol-gelsynthesis,” Materials Letters, vol. 28, pp. 225–229, 1996.

[9] W. Li, S. I. Shah, C. P. Huang, O. Jung, and C. Ni, “Metallor-ganic chemical vapor deposition and characterization of TiO2nanoparticles,” Materials Science and Engineering B, vol. 96, no.3, pp. 247–253, 2002.

[10] J. Choi, R. B. Wehrspohn, J. Lee, and U. Gösele, “Anodizationof nanoimprinted titanium: a comparison with formation ofporous alumina,” Electrochimica Acta, vol. 49, no. 16, pp. 2645–2652, 2004.

[11] D. Gong, C. A. Grimes, O. K. Varghese et al., “Titanium oxidenanotube arrays prepared by anodic oxidation,” Journal of Mat-erials Research, vol. 16, no. 12, pp. 3331–3334, 2001.

[12] V. Zwilling, M. Aucouturier, and E. Darque-Ceretti, “Anodicoxidation of titanium and TA6V alloy in chromic media. Anelectrochemical approach,” Electrochimica Acta, vol. 45, no. 6,pp. 921–929, 1999.

[13] R. Beranek, H. Hildebrand, and P. Schmuki, “Self-organizedporous titanium oxide prepared in H2SO4/HF electrolytes,”Electrochemical and Solid-State Letters, vol. 6, no. 3, pp.B12–B14,2003.

[14] S. Sobieszczyk, “Self-organized nanotubular oxide layers on Tiand Ti alloys,” Journal of Materials Science, vol. 9, no. 20, pp. 25–41, 2009.

[15] K. M. Reddy, S. V. Manorama, and A. R. Reddy, “Bandgapstudies on anatase titanium dioxide nanoparticles,” MaterialsChemistry and Physics, vol. 78, no. 1, pp. 239–245, 2003.

[16] J. I. Kim, S. Y. Lee, and J. C. Pyun, “Characterization of photo-catalytic activity of TiO2 nanowire synthesized from Ti-plate bywet corrosion process,” Current Applied Physics, vol. 9, no. 4, pp.e252–e255, 2009.


Research ArticleDSPACE Real-Time Implementation ofMPPT-Based FLC Method

Abdullah M. Noman,1 Khaled E. Addoweesh,1 and Hussein M. Mashaly2

1 Electrical Engineering Department, King Saud University, Saudi Arabia2 Sustainable Energy Technologies Center, King Saud University, Saudi Arabia

Correspondence should be addressed to Abdullah M. Noman; [email protected]

Received 22 December 2012; Revised 23 February 2013; Accepted 27 February 2013

Academic Editor: Sih-Li Chen

Copyright © 2013 Abdullah M. Noman et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Maximum power point trackers are so important in photovoltaic systems to improve their overall efficiency. This paper presents aphotovoltaic system with maximum power point tracking facility. An intelligent fuzzy logic controller method is proposed in thispaper to achieve the maximum power point tracking of PV modules. The system consists of a photovoltaic solar module connectedto a DC-DC buck-boost converter. The system is modeled using MATLAB/SIMULINK. The system has been experienced underdisturbance in the photovoltaic temperature and irradiation levels. The simulation results show that the proposed maximum powertracker tracks the maximum power accurately and successfully in all conditions tested. The MPPT system is then experimentallyimplemented. DSPACE is used in the implementation of the MPPT hardware setup for real-time control. Data acquisition andcontrol system is implemented using dSPACE 1104 software and digital signal processor card. The simulation and practical resultsshow that the proposed system tracked the maximum power accurately and successfully under all atmospheric conditions.

1. Introduction

Energy is important for the human life and economy. Con-sequently, due to the increase in the industrial revolution,the world energy demand has also increased. In the lateryears, irritation about the energy crisis has been increased.Fossil fuels have started to be gradually depleted. On theother hand, people are more concerned about the fossilfuel exhaustion and other environment problems which area result of conventional power generation. It is a globalchallenge to generate a secure, available, and reliable energyand at the same time reduce the greenhouse gas emission [1].Energy saving was suggested by the researchers to meet theworldwide energy demand.But this method is a cost-effectivesolution. One of the most effective and suitable solutions isthe renewable energy supplies. Renewable energy sources areconsidered as a technological option for generating clean,green, environment-friendly, and sustainable energy [1, 2].

Photovoltaic (PV) system has taken a great attentionsince it appears to be one of the most promising renewableenergy sources. The photovoltaic (PV) solar generation is

preferred over the other renewable energy sources due toadvantages such as the absence of fuel cost, cleanness, beingpollution-free, little maintenance, and causing no noise dueto absence of moving parts. However, two important factorslimit the implementation of photovoltaic systems. These arehigh installation cost and low efficiency of energy conversion[1]. In order to reduce photovoltaic power system costsand to increase the utilization efficiency of solar energy,the maximum power point tracking system of photovoltaicmodules is one of the effective methods [3]. Maximum powerpoint tracking, frequently referred to as MPPT, is a systemused to extract the maximum power of the PV moduleto deliver it to the load [4]. Thus, the overall efficiency isincreased [4].

Since the power generated from the photovoltaic moduledepends on the temperature and the solar radiation, thesefactors must be taken into account while designing themaximum power point tracker. The main goal of the MPPTis to move the module operating voltage close to the voltageat which the PV produces the maximum power under all


atmospheric conditions. MPPT is very important in PV sys-tems. Different techniques have been developed to maximizethe output power of the photovoltaic module. They haveadvantages and limitations over the others. These techniquesvary in complexity, in the number of sensors required, in theirconvergence speed, and in their cost. In the literature someof MPPT methods are introduced such as feedback voltagemethod, and incremental conductance method, perturbationand observation method [1, 4–8]. The open-circuit voltagemethod is based on (1) which states that the voltage of thePV module at maximum power point is linearly proportionalto the open circuit voltage [9–12]. The proportional constant𝐾 depends on the meteorological conditions, fabrication ofthe PV cell and on the fill factor of the PV cell [8]:

𝐾 = 𝑉MPP𝑉oc ≅ constant < 1. (1)The proportional constant𝐾 has been reported to be between0.71 and 0.78 [13]. The common value of 𝐾 is about 0.76(within ±2%) [12].

In order to implement the constant voltage algorithm,PV modules must be interrupted with a certain frequencyto measure the open-circuit voltage of the PV module. Themeasured voltage is then multiplied by the factor𝐾 to obtainthe voltage at maximum power point. Then the operatingvoltage of the PV module is adjusted to the calculated voltagein order to obtain the maximum power. This process must berepeated periodically [8]. Although this method is simple toimplement, it has a drawback which is high power losses dueto periodically interrupting the system operation. Anotherdrawback is that it is difficult to choose an optimal value ofthe constant K [8, 10].

The other method is the constant voltage (current)method. The constant voltage (current) method compares themeasured voltage (current) of the PV module with a referencevoltage (current) to continuously adjust the duty cycle ofthe DC-DC converter and hence operate the PV module atthe predetermined point close to the MPP [8]. Although theconstant voltage (current) tracking method is very simple,this method is not able to track the maximum power pointwith changing environment conditions specially when thetemperature changes. That means it cannot be applied in ageneralized fashion in systems which do not consider theeffect of variations of the irradiation and temperature of thePV panels [8, 14].

Perturbation and observation (P&O) method is an alter-native method to obtain the maximum power point of thePV module. It measures the voltage, current, and power ofthe PV module and then perturbs the voltage to encounterthe change direction [9]. Figure 1 shows the 𝑃-𝑉 curve of thePV module. As shown in the left-hand-side MPP the powerof the PV is increased with increasing the voltage of the PVmodule until the MPP is reached. In the right-hand-side ofthe MPP with increasing the voltage the power is decreased.That means if there is an increase in the power, the subsequentperturbation should be kept in the same direction until MPPis reached. If there is a decrease in the power, the perturbationshould be reversed [4, 5, 8, 15, 16].

Table 1: The operation of P&O algorithm.

Δ𝑃PV Δ𝑉PV Perturbation>0 >0 Increase 𝑉>0


Start

Return

No

No

NoNo

Yes

Yes

YesYes

Measure ( ), ( )

Calculate power ( )

( )− ( )= 0

( )− ( − 1)> 0

( )− ( − 1)> 0( )− ( − 1)< 0

= + = − = + = −

Figure 2: The flow chart of the P&O algorithm.

+

−

PV

Figure 3: Equivalent circuit of PV cell simulation.

with a small leakage of current through a resistive path inparallel with the intrinsic device. Parallel resistance is largeand its effect is negligible. The equivalent circuit of the PVcell is shown in Figure 3.

The output current delivered to the load can be expressedas follows [4, 17, 18]:

𝐼 = 𝐼PV − 𝐼𝑜 (𝑒(𝑞(𝑉+𝐼𝑅𝑠)/(𝑛𝑘𝑇𝑎)) − 1) , (2)where 𝐼 is the output current of the solar module (A) and𝑉 isthe output voltage of the solar cell (V), which can be obtainedby dividing the output voltage of the PV module by thenumber of cells in series, 𝐼PV is the current source of the solarmodule by solar irradiance (A), 𝐼𝑜 is the reverse saturationcurrent of a diode (A), 𝑁𝑆 is the series connection numberof the solar module, n is the ideality factor of the diode (𝑛 =1∼2), q is the electric charge of an electron (1.6 × 𝑒−19c), k is

Boltzmann’s constant (1.38 × 10−23 j/K), and T is the absolutetemperature of the solar cell (∘K).

To model the PV module using MATLAB, the currentgenerated by the incident light which is also called short-circuit current (𝐼sc) at a given temperature (𝑇𝑎) must becalculated as follows [17–19]:

𝐼PV = 𝐼scn (1 + 𝑎 (𝑇𝑎 − 𝑇𝑛)) 𝐺𝐺𝑛 , (3)where 𝐼scn is the short-circuit current at normal conditions(25∘C, 1000 W/m2), 𝑇𝑎 is the given temperature (∘K), 𝐼PV isthe short-circuit current at a given cell temperature (𝑇𝑎), 𝑎 isthe temperature coefficient of 𝐼sc, and𝐺𝑛 is the nominal valueof irradiance, which is normally 1000 W/m2.

On the other hand, the reverse saturation current of diode(𝐼𝑜) at the reference temperature (𝑇𝑛) is given as follows [17,18]:

𝐼𝑜𝑛 = 𝐼scn𝑒(𝑞𝑉ocn/(𝑛𝑘𝑇𝑛)) − 1 , (4)where 𝑉ocn is the open-circuit voltage at normal conditions.The reverse saturation current at a given cell temperature (𝑇𝑎)can be expressed as follows [18]:

𝐼𝑜 = 𝐼𝑜𝑛(𝑇𝑎𝑇𝑛)(3/𝑛)𝑒((−𝑞𝐸𝑔/𝑛𝐾)(1/𝑇𝑎−1/𝑇𝑛)). (5)


Table 2: PV module parameters.

Maximum power (𝑃max) 115 WVoltage at 𝑃max (𝑉mp) 17.1 VCurrent at 𝑃max (𝐼mp) 6.7 AOpen-circuit voltage (𝑉oc) 21.8 VShort-circuit current (𝐼sc) 7.5 ATemperature coefficient of 𝐼sc 0.065 ± 0.015%/∘C

0 5 10 15 20 250

20

40

60

80

100

120

(W)

1000 W/m2

800 W/m2

600 W/m2

400 W/m2

200 W/m2

(V)

Figure 4: 𝑃-𝑉 curves under changing solar radiation.

TheBP3115 PV module is used in this paper. The PV mod-ule parameters under the reference conditions (1000 W/m2,25∘C) are listed in Table 2. The PV module is simulated usingMATLAB. Figure 4 shows the simulated P-V curves of thePV module under changing solar radiation from 200 W/m2to 1000 W/m2 while keeping the temperature constant at25∘C. On the other hand, Figure 5 shows the simulationresults of the 𝑃-𝑉 curves of the PV module under changingtemperature from 10∘C to 50∘C while keeping the solarradiation constant at 1000 W/m2.

3. DC-DC Buck-Boost Converter

DC conversion has gained the great importance in manyapplications, starting from low-power applications to high-power applications. In this paper, buck-boost converter ischosen to be used in the MPPT system. Buck-boost converteris used to step down and step up the DC voltage by changingthe duty ratio of the MOSFET. If the duty ratio is less than 0.5,the output voltage is less than the input voltage; while if theduty ratio is greater than 0.5, the output voltage is greater thanthe input voltage. Duty ratio is the time at which the MOSFETis on to the total switching time. The buck-boost converter isshown in Figure 6.

The relation between the input and the output voltages ofthe buck-boost converter is given as follows:

𝑉out = −𝐷1 − 𝐷𝑉in, (6)

Table 3

Buck-boost converter parameters𝐿 1 mH𝐶1 1000𝜇F𝐶2 330 𝜇F𝑓𝑠 40 KHZResistive load RL 5Ω

Controller type: dSPACE 1104 DSPMOSFET type: IRF3710Diode type: BYV32-200

Components used in the measurement circuitCurrent transducer LTS 25-NP

Voltage divider

Two 120 KΩ and 39 KΩresistors are connected inseries. The voltage is takenacross 39 KΩ resistor.

0 5 10 15 20 250

20

40

60

80

100

120

140

(W)

(V)

10 C25 C

40 C

55 C

Figure 5: 𝑃-𝑉 curves under changing temperature.where 𝐷 is the duty cycle of the converter which is given asfollows:

𝐷 = 𝑇on𝑇𝑆 , (7)where 𝑇on is the on-state time of the MOSFET while 𝑇𝑆 is theswitching time.

The buck-boost converter is designed and simulatedusing MATLAB/SIMULINK. The converter componentsused in the simulation and in the hardware setup are shownin Table 3.

4. MPPT-Based FLC Method

PV systems have relatively high initial cost. Approximately57% is spent on the PV modules, 30% on the batteries,7% on the MPPT controllers and inverters, and 6% onthe installation [20]. Therefore, introducing a high-efficientMPPT controller can help in decreasing the total cost of thePV systems. FLC can be used as a controller to obtain themaximum power that the PV modules capable of producing


MOSFET

Gate drive circuit

Diode

+

++

+

+

−

5 ohm outin

[ ]

2

Figure 6: The buck-boost converter circuit.

Fuzzification

Inference

Defuzzification

Rules

Crisp inputs

Crisp outputs

Figure 7: The stages of the FLC.

under changing weather conditions. The use of fuzzy logiccontrollers has been increased over the last decade becausethey are simplicity, deal with imprecise inputs, do not need anaccurate mathematical model, and can handle nonlinearity[21]. The nonlinear nature of the PV modules and the envi-ronment conditions make the tracking behavior so difficult.Thus, the FLC is an interesting tool to achieve the maximumpower and eliminate the complexity in the computation sinceit is simple, does not need the mathematical model, and doesnot need any reference MPP parameters [20].

The process of FLC can be classified into three stages,fuzzification, rule evaluation, and defuzzification. These com-ponents and the general architecture of an FLC are shown inFigure 7. The fuzzification step involves taking a crisp input,such as the change in the voltage reading, and combining itwith stored membership function to produce fuzzy inputs.To transform the crisp inputs into fuzzy inputs, membershipfunction must be first assigned for each input. The numberof membership functions used depends on the accuracy ofthe controller, but it usually varies between 5 and 7 [10]. Thesecond step of fuzzy logic processing is the rule evaluation inwhich the fuzzy processor uses linguistic rules to determinewhat control action should occur in response to a given setof input values. The result of rule evaluation is a fuzzy outputfor each type of consequent action.

The last step in fuzzy logic processing is defuzzificationin which the expected value of an output variable is derivedby isolating a crisp value in the universe of discourse of theoutput fuzzy sets. In this process, all of the fuzzy outputvalues effectively modify their respective output membershipfunction. One of the most commonly used defuzzificationtechniques is called center of gravity (COG) or centroidmethod.

Fuzzy logic was applied in designing different MPPTcontrollers [14, 20–27]. They apply a set of linguistic rules toobtain the required duty cycle. The input variables of the FLCdiffer from one paper to another. In [22–25] the inputs to theFLC are the error (E) and the change in error (Δ𝐸). The error(E) is calculated as the change in the PV power to the changein the PV voltage (Δ𝑃/Δ𝑉). The change in the duty cycle is theoutput from FLC. In other cases, the change in current insteadof the change in the PV voltage to calculate the error (E) isused as in [26]. Some other papers use other inputs to the FLCsuch as the change in the voltage (ΔV) and the change in thepower (Δ𝑃) while the output from FLC is either the changein the duty ratio of the power converter (Δ𝐷) or the changein the reference voltage (Δ𝑉). Li and Wang use Δ𝑉 and Δ𝑃 asthe input variables while the output variable is the change inthe reference voltage [14]. An adaptive fuzzy logic controllerfor MPPT was presented in [27] to adjust the duty cycle of thedefuzzification to enhance the controller performance underchanging atmospheric conditions.

In this paper, a new method-based FLC is proposed toachieve tracking the maximum power of the PV moduleunder changing weather conditions. The proposed inputs ofthe FLC are the change in the voltage of the PV module (Δ𝑉)and the change in the power of the PV module (Δ𝑃). Theproposed output from FLC is (Δ𝑈) which corresponds to themodulation signal which is applied to the PWM modulatorin order to produce the switching pulses.

5. The Proposed MPPT Fuzzy LogicBase Method

The input variables are defined as in (8). During fuzzification,the numerical input variables which are converted into


0.04

0.00

80 0.2

0

1

−0.2

−0.08

−0.04

−0.001

0.001

Figure 8: The membership function of the input variable (Δ𝑃).0

0

1

−0.04

−0.0135

−0.0067

0.001

−0.001

0.00

67

0.01

35

0.04

Figure 9: The membership function of the input variable (Δ𝑉).

linguistic variables are based on the membership func-tions. Figures 8, 9, and 10 show the membership of Δ𝑃,Δ𝑉, and Δ𝑈, respectively. Five fuzzy levels are used for all theinput and output variables: NB (negative big), NS (negativesmall), ZE (zero), PS (positive small), and PB (positive big):

Δ𝑉 = 𝑉 (𝐾) − 𝑉 (𝐾 − 1) ,Δ𝑃 = 𝑃 (𝐾) − 𝑃 (𝐾 − 1) . (8)

The theoretical design of the rules is based on the factthat if the change in the voltage causes the power to increase,the moving of the next change is kept in the same direction;otherwise the next change is reversed. After the theoreticaldesign, all the MFs and the rules were adjusted by the trialand error to obtain the desired performance.

The proposed rules are shown in Table 4. The fuzzyrules are designed to track the maximum power point ofthe photovoltaic system under changing weather conditions.Rapidly changing solar radiation is taken into account whiledesigning these rules.

6. Simulation Results

In order to verify that the proposed MPP tracker tracksthe maximum power point successfully, the controller istested under changing weather conditions. It is importantto test the proposed MPPT system under different ambientconditions in order to validate the designed system. Thedesign and the simulation performance are done using MAT-LAB/SIMULINK. The model used for simulation is shownin Figure 11. In this system, the PV module is connected

Table 4: Rule base used in the fuzzy logic controller.

Δ𝑉 Δ𝑃NB NS Z PS PB

NB PB PS NB NS NSNS PS PS NB NS NSZ NS NS NS PB PBPS NS PB PS NB PBPB NB NB PB PS PB

0.48

0.61

0.22

0.09

0.36

3

0.337 1

0

1

0.350

−0.3

Figure 10: The membership function of the output variable (Δ𝑈).

to the DC-DC buck-boost converter. The output of theconverter is the 5Ω load resistor. In order to start trackingthe maximum power, the output voltage and the outputcurrent of the PV module must be measured to be usedas an input to the MPPT control block. The output of theMPPT control block is the gating signal which is used to drivethe MOSFET. The proposed FLC MPPT method is testedunder changing weather conditions as shown in Figure 12.As shown in this figure the solar radiation is changed as aconstant value 300 W/m2 until 0.03 sec. The solar radiationis then assumed to be changed as a ramp function withpositive slope to account for changing the solar radiationin the sunrise periods. Then the irradiance is changed as aunit step function to account for changing the solar radiationrapidly. In practical point of view, the solar radiation isdecreased as a ramp function with a negative slope duringthe sunset periods. On the other hand, the temperature iskept constant at 25∘C and then raised up rapidly to 50∘Cat 0.06 sec. It is clear from Figure 12 that the proposedsystem will be tested under all expected ambient conditions.These are constant solar radiation, rapidly changing solarradiation, and changing solar radiation as a ramp function.The FLC-based MPPT method is tested under these ambientconditions. The proposed method tracked the maximumpower effectively and accurately as shown in Figure 13. Thecontroller tracked the maximum power under all ambientconditions listed above. The proposed system tracked themaximum power under changing solar radiation as a positiveand negative ramp function. On the other hand, it followsthe maximum power under rapidly changing solar radiationaccurately. The tracking efficiency using FLC is 98.13%. Thetracking efficiency can be calculated as the energy generatedfrom the PV module divided by the theoretical maximumenergy. Comparing the tracking behavior of the proposedFLC MPPT method shown in Figure 13 with the tracking


[ PV ]

[ PV ]

[ ]

[ ]

[ ]

[ ]

+−+

−

+

+

−

+−

In1

PV module

In2

[ max ]

Out1 [ ]

[ ]

Temperature

Irradiation

MeasurementsFL controller

Goto5

7

4

MOSFETPV

[ PV ]

PV

[ PV ]

PV1−

1

2

PV

PV

[ PV ]

[ ]

[ ]

Figure 11: MPPT system used for simulation.

0

500

1000

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)

(W/m

2)

(a)

0204060

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)

(C)

(b)

Figure 12: Changing ambient condition.

behavior of the P&O MPPT method shown in Figure 14,the tracking efficiency using MPPT fuzzy logic controller is98.13% which is higher than the tracking efficiency obtainedusing P&O MPPT method which is 97.15%. On the otherhand the oscillation around the MPP when the P&O is usedis much higher than that when the FLC is used for MPPtracking. The system performance shows that the proposedsystem is well functioning to obtain the maximum powerthat the PV module is capable of producing under differentambient conditions.

7. Experimental Setup

The implementation of the MPPT hardware setup is done byusing dSPACE real-time control. Figure 15 shows the blockdiagram of the hardware step while Figure 16 shows thehardware setup of the MPPT system. In the hardware setup,one BP 3115J PV module is connected to the DC-DC buck-boost converter. Data acquisition and the control system

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.100

50100150

(W)

Time (s)

(a)

05

101520

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10

(V)

Time (s)

(b)

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1026

10

(A)

Time (s)

(c)

Figure 13: MPP tracking with FLC.

are implemented by using dSPACE 1104 software and digitalsignal processor card on PC. The PV voltage and the PVcurrent must be initially measured. In this system, the voltageis measured by using the voltage divider while the PV currentis measured by using the LTS 25-NP current sensor. Theanalog measured quantities of the PV voltage and PV currentwhich are fed to the A/D converter of the dSPACE in order tobe used in the SIMULINK MPPT control block. The MPPTcontrol which is constructed on MATLAB/SIMULINK isshown in Figure 17.

The signal applied to a dSPACE A/D channel must be inthe range from −10 V to +10 V. A signal of +10 V gives aninternal value of 1.00 within SIMULINK.


0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)

050

100150

(W)

(a)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)

01020

(V)

(b)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Time (s)

26

10

(A)

(c)

Figure 14: MPP tracking with P&O method.

IRF3710

1000

BYV32-200

Current sensor

dSPACE 1104 system

Gate drivecircuit

25+ 273

PV moduleBP 3115 J

39 K

120 K 100 V, 57 A16 A

5 ohm120 V

+

+

+ ++

+

Solar radiation (W/m2) 1 2

Temperature ( K)

Figure 15: Block diagram of the hardware setup.

Figure 16: The hardware setup of the system.

Every signal came from A/D converter must be multipliedby 10. The filters are used for removing any high-frequencynoise or any switching noise that appears in the signals. Asshown in Figure 17 the instantaneous measured voltage andcurrent are then multiplied by each other to obtain the PVinstantaneous power. The PV voltage and the PV power arethen applied to the MPPT algorithm to generate the requiredduty cycle. The output signal of the MPPT algorithm is then

applied to the DS1104SL DSP PWM block which is used togenerate the required switching signal to drive the MOSFET.The generated PWM signals should not be connected directlyto the MOSFET since the maximum current drown fromdSPACE board must not exceed 13 mA. For this reason andfor the isolation purposes a 6N137 optocoupler is used. ThePWM generated signal from the dSPACE is connected to the6N137 optocoupler and the output of the optocoupler is thenconnected to the MOSFET gate on the buck-boost converterand manage the on-off time of the switch.

To verify the function and the performance of theproposed FLC MPPT method, the method is experimentallyimplemented by using dSPACE 1104 data acquisition system.In order to start real-time tracking of the MPP of thePV module, the SIMULINK MPPT control block, must bedownloaded to the dSPACE board to generate C code ofthe MPPT control block. To successfully track the MPP,some modifications were taken into consideration whenthe proposed method is experimentally implemented. Themembership function of the input variable Δ𝑃 is modified asshown in Figure 18. On the other hand, some of rules are also


Voltage sensorcalibration

3

1

MPPTalgorithm

Product

10Gain3

10Gain2

In1

Current sensorcalibration

0

1

0

Butter

Analogfilter design5

Butter

Analogfilter design1

9

3

In1ADCchannel

ADCchannel

Out1

Out2

2

[ ]

[ ]

+−

DS1104ADC C6

DS1104ADC C5

DS1104SL DSP PWM

PVa

PVa

[ PV ]

[ PV ]

PV

×

Figure 17: MPPT SIMULINK model implemented in dSPACE 1104.

0.2

0.40 0.1 1

0

1

−0.1

−0.2

−0.4

−1

Figure 18: Modified membership function of the input variable Δ𝑃.

0400800

(W/m

2)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (s)

08:38AM

02:38PM

12:00PM

2.2×104

(a)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20

204060

Time (s)

08:38AM

02:38PM

12:00PM

PV(C)

2.2×104

(b)

Figure 19: Changing the ambient conditions on October 2, 2012.

tuned in order to obtain a better performance. Table 5 showsthe modified FLC rules.

After doing the above-mentioned modifications, FLCMPPT method is tested under some different ambientconditions. Figure 19 shows the changing in the ambient

Table 5: Modified rule base used in the fuzzy logic controller.

Δ𝑉 Δ𝑃NB NS Z PS PB

NB PB NB NB NS NSNS PB NB NB NS NSZ NS NS NS PB PBPS NS PB PS NB PBPB NB NB PB PS PB

conditions on 02-10-2012 starting from 08:38 AM to 02:38PM. Figure 20(a) shows the changing in the solar radia-tion while the lower plot shows the changing in the PVtemperature. FLC MPPT method tracked the MPP of thePV module successfully as shown in Figure 20. The upperplot in this figure shows the maximum power tracked.Figures 20(b) and 20(c) show the PV voltage and the PVcurrent at the maximum power. Figure 20(d) the duty ratiowhich is generated by the proposed FLC method. Duty ratiois measured at the output of the MPPT block which is thendirected to the PWM in order to generate the switchingpulses of the MOSFET. It is noted that the proposed FLCMPPT method tracked the maximum power successfully andaccurately with fast response.

Having a deep investigation on the proposed MPPTsystem performance under rapidly changing solar radiation,the PV module is covered by an opaque cloth to prevent theincidence of the solar radiation on the PV module. Variationof the power, the voltage, and the current of the system isshown in Figure 21. As shown in this figure, the proposedMPPT method has tracked the maximum power effectivelyand accurately under rapidly changing solar radiation.


0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20

50100

Time (s)

02:38PMPM

12:00AM

08:38

PVmax

(W)

2.2×104

(a)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20

1018

Time (s)

02:38PMPM

12:00AM

08:38

PVmax

(V)

2.2×104

(b)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2058

Time (s)

02:38PMPM

12:00AM

08:38

PVmax

(A)

2.2×104

(c)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (s)

00.5

1

Duty ratio

02:38PMPM

12:00AM

08:38

2.2×104

(d)

Figure 20: Experimental tracking behavior of the FLC MPPT. 02-10-2012.

Time (s)0 50 100 150 200 250 300 350 400

050

100

PVmax

(W)

(a)

Time (s)0 50 100 150 200 250 300 350 400

01020

PVmax

(V)

(b)

Time (s)0 50 100 150 200 250 300 350 400

05

PVmax

(A)

(c)

Time (s)0 50 100 150 200 250 300 350 400

00.5

1

Duty ratio

(d)

Figure 21: Performance of the FLC MPPT method under rapidly changing solar radiation.

8. Conclusion

Photovoltaic model using MATLAB/SIMULINK and thedesign of appropriate DC-DC buck-boost converter with amaximum power point tracking facility are presented in thispaper. MPPT is achieved using fuzzy logic controller whichenhanced the performance of the MPPT and eliminated thecomplexity in the computation needed. The proposed systemis simulated using MATLAB/SIMULINK and tested underdifferent ambient conditions to show the tracking behavior.The tracking behavior shows that the proposed systemsuccessfully and accurately tracked the maximum powerpoint with better performance than that of conventionalmethod. Experimental implementation of the MPPT systemis presented in this paper where data acquisition and thecontrol of the proposed FLC MPPT method are achieved bydSPACE 1104. The practical results show that the proposedmethod tracked the MPP effectively and accurately with fastresponse. Furthermore, tests verified that the proposed FLCmethod is well functioning with a good performance onrapidly changing atmospheric conditions. The results indicatethat the designed MPP tracker is capable of tracking the PVmodule maximum power and hence improves the efficiencyof the PV system.

Conflict of Interests

The authors of this paper assure that they do not have anyinterest in dSPACE 1104 board and its software. They useddSPACE 1104 for research purposes only without any relationswith the manufacturer or dealer.

Acknowledgment

This work was financially supported by NPST program, KingSaud University, Project no. 09 ENE 741-02.

References

[1] J. L. Santos, F. Antunes, A. Chehab, and C. Cruz, “A maximumpower point tracker for PV systems using a high performanceboost converter,” Solar Energy, vol. 80, no. 7, pp. 772–778, 2006.

[2] B. Khan, Non-Conventional Energy Resources, Tata McGraw-Hill Education, Noida, India, 2006.

[3] V. A. Chaudhari, Automatic peak power tracker for solar pvmodules using dspacer software [thesis], Maulana Azad NationalInstitute of Technology (Deemed University), 2005.

[4] T. C. Yu and T. S. Chien, “Analysis and simulation of charac-teristics and maximum power point tracking for photovoltaic


systems,” in Proceedings of the International Conference onPower Electronics and Drive Systems (PEDS ’09), pp. 1339–1344,January 2009.

[5] A. M. Bazzi and S. H. Karaki, “Simulation of a new maxi-mum power point tracking technique for multiple photovoltaicarrays,” in Proceedings of the IEEE International Conferenceon Electro/Information Technology (EIT ’08), pp. 175–178, May2008.

[6] C. Liu, B. Wu, and R. Cheung, “Advanced algorithm for MPPTcontrol of photovoltaic systems,” in Canadian Solar BuildingsConference Montreal, 2004.

[7] A. Yafaoui, B. Wu, and R. Cheung, “Implementation of max-imum power point tracking algorithm for residential pho-tovoltaic systems,” in Proceedings of the 2nd Canadian SolarBuilding Conference, 2007.

[8] V. Salas, E. Oĺıas, A. Barrado, and A. Lázaro, “Review of themaximum power point tracking algorithms for stand-alonephotovoltaic systems,” Solar Energy Materials and Solar Cells,vol. 90, no. 11, pp. 1555–1578, 2006.

[9] D. Hohm and M. Ropp, “Comparative study of maximumpower point tracking algorithms using an experimental, pro-grammable, maximum power point tracking test bed,” inProceedings of the 28th IEEE Photovoltaic Specialists Conference,pp. 1699–1702, 2000.

[10] T. Esram and P. L. Chapman, “Comparison of photovoltaic arraymaximum power point tracking techniques,” IEEE Transactionson Energy Conversion, vol. 22, no. 2, pp. 439–449, 2007.

[11] J. H. R. Enslin, M. S. Wolf, D.B. Snyman, and W. Swiegers, “Inte-grated photovoltaic maximum power point tracking converter,”IEEE Transactions on Industrial Electronics, vol. 44, no. 6, pp.769–773, 1997.

[12] D. Y. Lee, H. J. Noh, D. S. Hyun, and I. Choy, “An improvedMPPT converter using current compensation method for smallscaled PV-applications,” in Proceedings of the 18th Annual IEEEApplied Power Electronics Conference and Exposition, vol. 1, pp.540–545, February 2003.

[13] Y. E. Wu, C. L. Shen, and C. Y. Wu, “Research and improvementof maximum power point tracking for photovoltaic systems,” inProceedings of the International Conference on Power Electronicsand Drive Systems (PEDS ’09), pp. 1308–1312, January 2009.

[14] J. Li and H. Wang, “Maximum power point tracking of pho-tovoltaic generation based on the fuzzy control method,” inProceedings of the 1st International Conference on SustainablePower Generation and Supply (SUPERGEN ’09), pp. 1–14, April2009.

[15] I. Houssamo, F. Locment, and M. Sechilariu, “Maximumpower tracking for photovoltaic power system: developmentand experimental comparison of two algorithms,” RenewableEnergy, vol. 35, no. 10, pp. 2381–2387, 2010.

[16] J. Zhang, T. Wang, and H. Ran, “A maximum power pointtracking algorithm based on gradient descent method,” inProceedings of the IEEE Power and Energy Society GeneralMeeting (PES ’09), pp. 1–5, July 2009.

[17] M. G. Villalva, J. R. Gazoli, and E. Ruppert Filho, “Modeling andcircuit-based simulation of photovoltaic arrays,” in Proceedingsof the Brazilian Power Electronics Conference (COBEP ’09), pp.1244–1254, October 2009.

[18] G. Walker, “Evaluating MPPT converter topologies using a mat-lab PV model,” Journal of Electrical and Electronics Engineering,Australia, vol. 21, no. 1, pp. 49–55, 2001.

[19] X. Liu, An improved perturbation and observation maximumpower point tracking algorithm for PV panels [M.S. thesis],Concordia University, 2004.

[20] A. M. Z. Alabedin, E. F. El-Saadany, and M. M. M. SalamaA, “Maximum power point tracking for Photovoltaic systemsusing fuzzy logic and artificial neural networks,” in Power andEnergy Society General Meeting, pp. 1–9, 2011.

[21] C. Larbes, S. M. Aı̈t Cheikh, T. Obeidi, and A. Zerguerras,“Genetic algorithms optimized fuzzy logic control for the max-imum power point tracking in photovoltaic system,” RenewableEnergy, vol. 34, no. 10, pp. 2093–2100, 2009.

[22] F. Chekired, C. Larbes, D. Rekioua, and F. Haddad, “Imple-mentation of a MPPT fuzzy controller for photovoltaic systemson FPGA circuit,” Proceedings of the Impact of IntegratedClean Energy on the Future of the Mediterranean Environment(MEDGREEN ’11), pp. 541–549, April 2011.

[23] M. Adly, H. El-Sherif, and M. Ibrahim, “Maximum powerpoint tracker for a PV cell using a fuzzy agent adapted by thefractional open circuit voltage technique,” in IEEE InternationalConference on Fuzzy Systems, pp. 1918–1922, 2011.

[24] M. A. Islam, A. B. Talukdar, N. Mohammad, and P. K. S.Khan, “Maximum power point tracking of photovoltaic arraysin matlab using fuzzy logic controller,” in Proceedings of theAnnual IEEE India Conference: Green Energy, Computing andCommunication (INDICON ’10), pp. 1–4, December 2010.

[25] M. S. Ngan and C. W. Tan, “A study of maximum power pointtracking algorithms for stand-alone photovoltaic systems,” inProceedings of the IEEE Applied Power Electronics Colloquium(IAPEC ’11), pp. 22–27, April 2011.

[26] I. Purnama, Y.-K. K. Lo, and H.-J. J. Chiu, “A fuzzy controlmaximum power point tracking photovoltaic system,” in IEEEInternational Conference on Fuzzy Systems, pp. 2432–2439, 2011.

[27] Y. H. Chang and W. F. Hsu, “A maximum power point trackingof PV system by adaptive fuzzy logic control,” in InternationalMultiConference of Engineers and Computer Scientists (IMECS’11), pp. 975–980, March 2011.


Research ArticleThe Effect of Sputtering Parameters on the Film Properties ofMolybdenum Back Contact for CIGS Solar Cells

Peng-cheng Huang,1 Chia-ho Huang,1 Mao-yong Lin,2 Chia-ying Chou,1

Chun-yao Hsu,1 and Chin-guo Kuo3

1 Department of Mechanical Engineering, Lunghwa University of Science and Technology, Taoyuan, Taiwan2Department of Mechanical and Electrical Engineering, Fujian Polytechnic of Information Technology, Fujian, Taiwan3Department of Industrial Education, National Taiwan Normal University, Taipei, Taiwan

Correspondence should be addressed to Chin-guo Kuo; [email protected]

Received 6 January 2013; Revised 9 February 2013; Accepted 9 February 2013

Academic Editor: Ho Chang

Copyright © 2013 Peng-cheng Huang et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Molybdenum (Mo) thin films are widely used as a back contact for CIGS-based solar cells. This paper determines the optimalsettings for the sputtering parameters for an Mo thin film prepared on soda lime glass substrates, using direct current (dc)magnetron sputtering, with a metal Mo target, in an argon gas environment. A Taguchi method with an L9 orthogonal array, thesignal-to-noise ratio, and an analysis of variances is used to determine the performance characteristics of the coating operation. Themain sputtering parameters, such as working pressure (mTorr), dc power (W), and substrate temperature (∘C), are optimized withrespect to the structural features, surface morphology, and electrical properties of the Mo films. An adhesive tape test is performedon each film to determine the adhesion strength of the films. The experimental results show that the working pressure has thedominant effect on electrical resistivity and reflectance. The intensity of the main peak (110) for the Mo film increases and the fullwidth at half maximum decreases gradually as the sputtering power is increased. Additionally, the application of an Mo bilayerdemonstrates good adherence and low resistivity.

1. Introduction

Copper indium gallium selenium (CIGS) solar cells are highlyefficient, low-cost thin film solar cells, and efficiencies of20.0% have been reported, so the commercial productionof CIGS films is growing rapidly [1]. The CIGS cell uses asubstrate, Mo (back contact), CIGS (absorber layer), CdS(buffer layer), ZnO, and Al2O3-doped ZnO (window layer),in which each layer has a different role in the workingcell [2]. The back contact layer functions as a barrier thathinders the diffusion of impurities from the substrate into theabsorber. Molybdenum (Mo) thin films are widely used as aback contact for CIGS-based solar cells. Substrates of Mo-coated soda lime glass can be purchased from commercialsources. An Mo back contact with good adherence and lowresistivity is essential, because the properties of the Mo thinfilms significantly affect the performance of CIGS solar cells[3].

Several materials have been with the subjects of exper-iment for use as a back contact for CIGS thin film solarcells. In order to ensure good electronic device properties,the formation of an ohmic contact for the majority carriers(holes) and a low recombination rate for the minority carriers(electrons) at the CIGS/back contact interface is essential [4].Compared with other materials, such as W, Ta, Nb, Cr, V, orTi, Mo metal is an ideal back contact material for CIGS solarcells, because of its inertness and high conductivity [5]. Theadvantages of this refractory metal are its low susceptibilityto corrosion in a selenium atmosphere and its ability to allowNa diffusion from a soda lime glass substrate to a CIGSlayer.

Traditional experimental methods are too complicatedand difficult. These methods require a large number of exper-iments, when the number of process parameters inc

Documents

Solar Energy and Clean Energy: Trends and Developmentsdownloads.hindawi.com/journals/specialissues/181840.pdf · evaluating the performance of solar drying in the Malaysian red chili