Fuyuki 2005
Fuyuki 2005
Downloaded 04 May 2013 to 193.1.100.108. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions
APPLIED PHYSICS LETTERS 86, 262108 共2005兲
In recent years, the production scale of solar cells has carrier diffusion length and the detection of deteriorated
increased remarkably to meet with the pressing requirements areas.
of practical photovoltaic systems. Among installed systems, The schematic measurement setup is shown in Fig. 1. A
more than 90% are crystalline silicon cells, and especially sample cell biased at an appropriate forward voltage emitted
polycrystalline silicon 共poly-Si兲 shows the advantages of low infrared light, which was collected by the cooled CCD cam-
cost and large area with relatively high efficiency. In order to era using a selected objective lens. The cooled 共at around
get reliable high efficiency under a mass production process, −50 ° C兲 CCD could capture 1 ⫻ 1 cm2 cell area by 680
quick and precise evaluation of cell performance and feed- ⫻ 680 pixels in the sensitive wavelength region of
back to production lines are indispensable. Usually the fab- 300– 1100 nm. Through the data acquisition and smoothing
ricated cells and/or modules are inspected simply by current/ process, the spatial resolution was reduced, and the resolu-
voltage output performance under solar simulated light. For a tion limit became about 50 m in length.
detailed examination of cell performance, the most important The minority carrier diffusion length was calibrated by
material parameter to be monitored is the minority carrier the light-beam-induced current 共or voltage兲 analysis using
diffusion length 共or lifetime兲, which governs the collection four wavelengths of 660, 890, 950, and 980 nm. The spatial
efficiency. The photoconductivity decay method, using mi- resolution of the LBIC was 250 m. One measuring point of
crowave reflection,1 is widely used to check the minority the LBIC method included 5 ⫻ 5 points of the emission in-
carrier lifetime of substrates, but it requires good surface tensity measurement, so that an average of 25 values of
passivation in order to derive an accurate bulk lifetime. After emission intensity was used when the relationship between
the formation of p / n junctions, the spectroscopic light-beam- the diffusion length and the emission intensity was dis-
induced current2,3 共LBIC兲 or the electron-beam-induced cur- cussed.
rent methods4 are used, at the laboratory level, to elucidate Polycrystalline Si solar cells 共1 cm⫻ 1 cm兲 were fabri-
the minority carrier diffusion length and the effects of defects cated through conventional device processes using the cast
and/or grain boundaries. In all these methods, the probe tools silicon substrate. The average efficiency of measured
共light, electron beam, etc.兲 are required in order to acquire samples was 13%–15%. Detailed materials properties and
the spatial distribution of the diffusion length. cell configuration are not shown here since these data are not
In this letter, we propose a technique to analyze the mi- essential for the discussions in this work. The measurement
nority carrier diffusion length distribution under an as- was carried out at room temperature.
fabricated cell 共or even a module兲 structure by a simple and
quick photographic surveying method. Light emission from
solar cells under the forward bias was captured by a CCD
camera, and we found that the intensity distribution of emis-
sion clearly agreed with the mapping of the minority carrier
diffusion length in poly-Si active layers. Takamoto et al. re-
ported the application of electroluminescence in single-
crystalline InGaP / GaAs tandem solar cells, but they revealed
only the nonuniformity of the saturation current density and
the current leakage paths.5 Quantitative analysis of the rela-
tion between the emission intensity and the minority carrier
diffusion length was discussed. The feasibility of this tech-
nique was addressed showing the mapping of the minority
a兲
Electronic mail: [email protected] FIG. 1. Schematic diagram of an experimental setup.
N = no 冕 0
W
exp共− x/Le兲dx = noLe关1 − exp共− W/Le兲兴, 共2兲
fabricated cells, but also molded modules, and further devel-
opment correlating the analysis of emission with cell perfor-
mance will be needed.
where W is the thickness of the cell. When the term 1
exp共−W / Le兲 is considered to be much less than 1, N becomes J. A. Eikelboom, C. Leguijt, C. F. A. Frumau, and A. R. Burgers, Sol.
Energy Mater. Sol. Cells 36, 169 共1995兲.
nearly proportional to Le. The emission intensity will then be 2
O. Porre, M. Stemmer, and M. Pasquinelli, Mater. Sci. Eng., B 24, 188
proportional to the effective minority carrier diffusion length 共1994兲.
3
Le. The experimental results shown in Fig. 5 roughly fulfill N. Sakitani, K. Nishioka, T. Yagi, Y. Yamamoto, Y. Ishikawa, Y. Uraoka,
this relationship, but more accurate analysis is required. and T. Fuyuki, Solid State Phenom. 93, 351 共2003兲.
4
Figure 6 shows a typical emission spectrum measured by W. Seifert, M. Kittler, and J. Vanhellemont, Mater. Sci. Eng., B 42, 260
共1996兲.
an infrared-sensitive Ge photodetector at room temperature. 5
T. Takamoto, E. Ikeda, H. Kurita, and M. Yamaguchi, Proceedings of the
The dominant emission mechanism will be the band-to-band 14th European Photovoltaic Solar Energy Conference, Barcelona, Spain,
radiative recombination with the phonon assist, as the main 1997.
Downloaded 04 May 2013 to 193.1.100.108. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://apl.aip.org/about/rights_and_permissions