HIT电池_HITcellsbyQiWang

A national laboratory of the U.S. Department of EnergyOffice of Energy Efficiency or grid/ITO/a-Si:H/c-Si(n or p)/ a-Si:H/metal in Fig 2 b). The finished cell is about 1 cm 2 in area with a 5%-coverage metal grid on top of the ITO. Confirming JV measurements were done at NREL by the accredited PV Performance Characterization Team. Standard JV measurement was done by NREL ’ s XT-10 solar simulator. High-resolution transmission electron microscopy (HRTEM) was used to study the interfaces, the thickness, and conformal coverage of the a-Si:H and ITO layer. RESULTS Figure 3 shows our record 19.1% SHJ cells on p-type FZ wafer. This cell has a V oc of 0.678 V, FF of 0.779, J scof 35.9 mA/cm 2. The cell structure is grid/ITO/n/i/c-Si(p)/i/p/Al. The thickness of i/n layer is about 10 nm (confirmed by TEM measurement). The c-Si wafer size is about 2.5 x 4.5 cm^2. Front ITO thickness is about 900 nm and was photolithographically defined and etched to 1 cm 2 for two cells per wafer. A 0.9 cm 2 mask covers the cell during the JV measurement. We used a standard RCA 1 and 2 clean procedure after in-house texturing the wafer before deposition of the a-Si:H emitter and back contact. We found this clean procedure was enough to remove the residual impurities and have a good V oc on textured wafers. Earlier, we have used our extensive clean procedure for flat wafers on textured ones with less success. We found that some of the textured features were removed after the chemical wet clean. Therefore, we may keep the surface clean with high V oc but lost the optical enhancement. In searching for a new cleaning procedure for textured wafers, it is helpful that the standard RCA process gives good results. We will start from here and eventually improve the cleaning process. We found that the double-heterojunction was essential for high V oc [3]. With a-Si:H double heterojunctions, the lifetime of the minority carriers is in the range of 1 ms and the back surface recombination velocity can be reduced to ~15 cm/s. high lifetime and low surface recombination velocity are the factors to improve cell performance. Many other groups [10-13] also reported the effective surface passivation of a-Si:H to c-Si. In addition to high lifetime, we found that contact formation to the thin doped a-Si:H layer was also critical [7]. We reported before that Ti directly deposited on thin doped a-Si:H layers caused a decrease in V oc. Our early SHJ cells suffered from a low FF. The ITO on the back contact of the heterojunction solar cells was considered the cause. We tried to use a metal with much higher conductivity than the TCO to directly contact the a-Si:H doped layer to reduce the FF loss. We succeeded in improving the FF but discovered complications. Figure 2 describes two structures with techniques for forming back 2Figure 3: Performance of our best SHJ p-type c-Si solar cell contacts: a) with an ITO intermediate layer to enhance optical back reflection, and b) with simpler direct metal contacts such as Al or Ti. With an ITO contact to the back doped layer, we found that we can reach a high Voc of 0.680 V. When we use Ti direct contact to the same doped-layer the Voc decreases by 56 mV. However, when we deposit a much thicker doped a-Si:H layer (6 times increased deposition time) with the Ti contact, Voc is restored to about 0.680 V. This effect can be very important to avoid possible metal indiffusion at the back and make a high Voc cell. An Al contact to a p-type a-Si:H layer seems to form a good contact with a high Voc. Notice that our record cell used an Al direct contact to the p-layer in the back. This did not cause a rapid decrease of V oc as we mention before with the Ti contact. It seems that some metals are easier diffusers than others. Table I. Summary of the best double-sided SHJ 1 cm 2 cells by HWCVD, fabricated at NREL. Voc(V) Jsc(mA/cm 2) FF(%) η(%) p-type** (planar) 0.688 31.6 81.3 17.7 n-type (planar)** 0.691 33.6 72.1 16.7 p-type (textured)** 0.678 35.9 78.6 19.1 n-type (textured)* 0.664 35.3 74.5 17.2 ** independently confirmed by NREL PV Performance Characterization Team * measured by using calibrated NREL XT-10 solar simulator. In summary of our SHJ cells research, we use planar wafers as our baseline because the cleaning and coverage of the films is relatively easy. This gave us a good starting point. We found that the wafer with a carrier lifetime over 700 μ s after a-Si:H emitter and back contact, in general, will lead to a V oc over 0.680 V. Once the wafer can have a high V oc, we texture the wafer and make high efficiency cells. a) b) Figure 4 SEM (a) and TEM (b) pictures of 19% SHJ cell Table 1 summarizes our best double-side SHJ solar cells ’ performance. For cells with planar surfaces, it seems that n-type c-Si cells have a slightly higher Vocbut a worse Fill factor (FF) than the p-type c-Si cells. Both 0.691 and 0.688 V are the highest Voc of SHJ cells ever made by HWCVD. We attribute this high Voc to proper surface cleaning before the Si deposition and the a rapid transition from the c-Si surface to growth of a-Si:H by HWCVD. For the textured cell, we have better performance on our p-type cells. In n-type c-Si, the lower Voc in textured one may be related to the difficulty in cleaning the textured surface. We are still optimizing cleaning with a goal of a Voc close to the planar one of 0.69 V. We are also optimizing our texture to gain more 3in the short circuit current ( Jsc). Our 36 mA/cm 2 is much lower than Sanyo ’ s standard textured cells (37-39 mA/cm 2). We use SEM and TEM to further investigate our textured SHJ solar cell. Figure 4a shows SEM picture of our 19% SHJ cell ’ s textured surface. It appears that the surface was textured with less than 10 μ m pyramids. Our texturing still needs some improvement. The distribution of the pyramid ’ s size is still too large. Figure 4b of TEM picture shows that our a-Si:H thin layers and ITO are conformally covering the textured surface even at the tip and the valley of the pyramids. Experimentally, TEM confirms that HWCVD has provided high quality conformal layers even at a few nm in thickness and so does our ITO coating. We believe that the conformal coating of HWCVD is one of the keys to making high V oc textured SHJ cells. Figure 5. Efficiency of a p-type CZ wafer SHJ cell as a function of light exposure time. Figure 5 shows light soaking data on our CZ p-type SHJ solar cell. Our data show that there is little degradation on CZ p-type c-Si regard to a-Si:H related issues. We have not excluded B-O related degradation but the stable efficiency is high. We concluded that widely available p-type CZ c-Si can be used for SHJ solar cell. Currently, only n-type c-Si SHJ solar cells are in production. DISCUSSIONS We used Figure 6 to discuss the issues with SHJ solar cells and a pathway to even higher efficiency. Figure 6 shows an Internal quantum efficiency (IQE) data of PERL cell from University of New South Wales [14], HIT cell from Sanyo [15], and SHJ cell from NREL. It clearly shows that Sanyo and NREL cell have a similar IQE. However, both IQE have a less response in the infrared region, from 1000 to 1200 nm, compared to the PERL cell. This leads to a J sc deficit about 4 mA/cm 2. A record PERL cell has J sc about 43 mA/cm 2 and the best HIT cell has a J sc of 39 mA/cm 2. Increase red response in SHJ cell will be the key for further improvement of cell efficiency and ultimately a record efficiency over 25 %. Figure 6: Internal quantum efficiency comparison between SHJ cells and PERL cells 020406080100300 500 700 900 1100Wavelength (nm)UNSW-PERLSanyo-HITNREL-SHJ0481216200 400 800 1200Light exposure time (hr)SUMMARY The advantages of using HWCVD in comparison to plasma-enhanced CVD are the fast deposition rate and, more important, the wide deposition parameters for forming conformal covered heterojunctions of high V oc. Additionally, the low temperature (below 200 ° C for entire cell process) makes HWCVD one of only a few promising methods for the production of next generation ultra-thin Si wafer solar cells with low stress. ACKNOWLEDGMENT Authors like to thank Richard Crandall, Howard Branz, Dean Levi, and Scott Ward for helpful discussions. This work was supported by the U.S. Department of Energy under Contract DE-AC39-98-GO10337. REFERENCES [1] S. Taira, Y. Yoshimine, T. Baba, M. Taguchi, H. Kanno, T. Kinoshita, H. Sakata, E. Maruyama, and M. Tanaka, Proceedings of the 22nd EU PVSEC: p. 932. 2008 [2] Q. Wang, M. R. Page, E. Iwaniczko, E. Williams, Y. Yan, T. H. Wang, and T. F. Ciszek, Proceedings 45of 3rd World Conference on Photovoltaic Energy Conversion Vol. B, 2003, pp. 1427-1430 [3] T.H. Wang, E. Iwaniczko, M.R. Page, D.H. Levi, Y. Yan, V. Yelundur, H.M. Branz, A. Rohatgi, and Q. Wang, 31 st IEEE Photovoltaic Specialists Conference , Orlando, Florida, 2005, pp. 955-958 [4] D. H. Levi, C. W. Teplin, E. Iwaniczko, Y. Yan, T. H. Wang and H. M. Branz, Mat. Res. Soc. Proc ., Vol. 862. 2005, pp. 159-170 [5] M.R. Page, E. Iwaniczko, Y. Xu, Q. Wang, Y. Yan, L. Roybal, H.M. Branz, and T. H. Wang.the 2006 IEEE 4th World Conference on Photovoltaic Energy Conversion (WCPEC-4) p. 6. 2006 [6] Y.Yan, M. Page, T. H. Wang, M.M. Al-Jassim, H. M. Branz, and Q. Wang, Appl. Phys. Lett., 88: p. 3. 2006 [7] Qi Wang, M.R. Page, E. Iwaniczko, Yueqin. Xu, L. Roybal, R. Bauer, D. Levi, Y.F. Yan, T.H. Wang, H.M. Branz, Mat. Res. Soc. Proc ., Vol. 989. 2007, pp. 41 [8] E. Yablonovitch, T. Gmitter, R. M. Swanson, and H. T, Kwark, Appl. Phys. Lett., 47(11): p. 1211. 1985[9] Y. Xu, B. Yan, B. P. Nelson, E. Iwaniczko, R. C. Reedy, A. H. Mahan, and H. M. Branz Mat. Res. Soc. Proc .,Vol. 808. 2004, pp. 617-621 [10] E. Conrad, K. v. Maydell, H. Angermann, C. Schubert and M. Schmidt, in press WCPEC-4 , Hawaii, US, May, 2006 [11] H. Fujiwara, M. Kondo, in press, WCPEC-4 , Hawaii, US, May, 2006 [12] Stefaan De Wolf and Guy Beaucarne, APL , 88 , 022104, 2006 [13] U. K. Das, M. Z. Burrows, M. Lu, S. Bowden, and R.W. Birkmire, Appl. Phys. Lett., 92: p. 063504. 2008 [14] A. Wang, J. Zhao, and M.A. Green, Appl. Phys. Lett., 57(6): p. 602. 1990 [15] Eiji Maruyama, Akira Terakawa, Mikio Taguchi, Yukihiro Yoshimine, Daisuke Ide, Toshiaki Baba, Masaki Shima, Hitoshi Sakata and Makoto Tanaka in press, WCPEC-4 , Hawaii, US, May, 2006 F1147-E(09/2007) REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing the burden, to Department of Defense, Executive Services and Communications Directorate (0704-0188). Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ORGANIZATION.1. REPORT DATE (DD-MM-YYYY)May 2008 2. REPORT TYPEConference Paper 3. DATES COVERED (From - To)11-16 May 2008 5a. CONTRACT NUMBERDE-AC36-99-GO10337 5b. GRANT NUMBER4. TITLE AND SUBTITLECrystal Silicon Heterojunction Solar Cells by Hot-Wire CVD: Preprint5c. PROGRAM ELEMENT NUMBER5d. PROJECT NUMBERNREL/CP-520-42554 5e. TASK NUMBERPVA74101 6. AUTHOR(S)Q. Wang, M.R. Page, E. Iwaniczko, Y.Q. Xu, L. Roybal, R. Bauer, B. To, H.C. Yuan, A. Duda, and Y.F. Yan 5f. WORK UNIT NUMBER7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)National Renewable Energy Laboratory 1617 Cole Blvd. Golden, CO 80401-3393 8. PERFORMING ORGANIZATION REPORT NUMBERNREL/CP-520-42554 10. SPONSOR/MONITOR S ACRONYM(S)NREL 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)11. SPONSORING/MONITORING AGENCY REPORT NUMBER12. DISTRIBUTION AVAILABILITY STATEMENTNational Technical Information Service U.S. Department of Commerce 5285 Port Royal Road Springfield, VA 22161 13. SUPPLEMENTARY NOTES14. ABSTRACT (Maximum 200 Words)Hot-wire chemical vapor deposition (HWCVD) is a promising technique for fabricating Silicon heterojunction (SHJ) solar cells. In this paper we describe our efforts to increase the open circuit voltage (Voc) while improving the efficiency of these devices. On p-type c-Si float-zone wafers, we used a double heterojunction structure with an amorphous n/i contact to the top surface and an i/p contact to the back surface to obtain an open circuit voltage (Voc) of 679 mV in a 0.9 cm2 cell with an independently confirmed efficiency of 19.1%. This is the best reported performance for a cell of this configuration. We also made progress on p-type CZ wafers and achieved 18.7% independently confirmed efficiency with little degradation under prolong illumination. Our best Voc for a p-type SHJ cell is 0.688 V, which is close to the 691 mV we achieved for SHJ cells on n type c-Si wafers. 15. SUBJECT TERMSPV; silicon heterojunction; SHJ; hot-wire chemical vapor deposition; silicon heterojunction; solar cells; float-zone wafer; 16. SECURITY CLASSIFICATION OF: 19a. NAME OF RESPONSIBLE PERSON a. REPORT Unclassified b. ABSTRACT Unclassifiedc. THIS PAGE Unclassified17. LIMITATION OF ABSTRACTUL 18. NUMBER OF PAGES 19b. TELEPHONE NUMBER (Include area code)Standard Form 298 (Rev. 8/98) Prescribed by ANSI Std. Z39.18