You are on page 1of 5

LOCAL INVESTIGATION OF MONOCRYSTALLINE SILICON SOLAR CELLS DEFECTS

1 1 1 1 2
Tománek P , Škarvada P , Grmela L Mack R and Smith, S.J.
1
Brno University of Technology, Brno, Czech Republic
2
South Dakota School of Mines and Technology, Rapid City, South Dakota, USA

broadening, masking many of the signatures of the


ABSTRACT individual defects.

Unfortunately for a quality and efficiency of solar cells,


We present results of microscale localization and
there is wide variety of defects of distinct nature there.
characterization of defects in monocrystalline silicon solar
Some of them are in particular cases troublesome for the
cells using LBIC and Scanning near-field optical
good function of solar cell, other not. For the detection and
microscopy (SNOM). Although etched silicon is still most
localization of defects different mapping and non-mapping
effective material for solar cells, some problems with their
electrical and optical techniques are used [3-6].
use in solar plant installation persist due to the defects.
The monocrystalline silicon is still most effective material
The basic characterization methods of silicon solar cells
for solar cells, some problems with their use in solar plant
are electrical measurements. They represent integral
installation persist due to the defects [7].
measurements on the whole cell and only inform us about
the presence of defects in the structure. Local defects in
With the development of charge coupled device detectors,
the p-n junction may be associated with structural
a lock-in thermography [8], a considerable number of fast
imperfections (such as grain boundaries, dislocations, and
imaging methods and applications in solar cell
scratches), chemical impurities, higher concentrations of
characterization has been found [9, 10, 11]. These
donors and acceptors. Therefore, to develop investigation
methods are mainly used in in-situ characterization and
methods, a number of existing apparent and hidden
process control for mass wafer production.
imperfections in the bulk, surface or edges of the cell have
to be examined, detected, localized with higher spatial
In addition to the experimental preparation of samples,
resolution, characterized and classified. Moreover, it is
which contain no or drastically reduced distributions,
important not only to find most harmful defects, but also to
experimental information on the optical properties of
understand their nature and identify the factors which
individual imperfections can be obtained only by using
adversely affect recombination properties and
optical techniques with subwavelength spatial resolution
consequently an overall efficiency of the cells.
[4, 12].
The suggested method combines an electric noise
Nevertheless, there are still many opened questions in the
measurement with local topography and near-field optical
field of local light emission, material defects and their
beam induced current. To prove the feasibility of this
relation. This paper presents experimental results showing
method, we have chosen one bulk and one edge defect
sequential method combining electrical noise
within the sample, which emitted light under low reverse-
measurements with scanning probe localization of
biased voltage.
luminous micro-spots, as well as a characterization of their
nature.
Light emission from edge spot exhibits strong positive I-U
METHOD
relation with temperature, while for a bulk spot this
dependence is slightly negative. Moreover, it was found
Electrical noise measurement on silicon solar wafer allows
that one emitting macroscopic spot can consist of several
detect a presence of low-voltage breakdowns in defect
small light emission spots which diameters are bellow
sites of p-n junctions.
3 μm.
Laser beam induced current (LBIC) is a method used for
INTRODUCTION defect localization in structure of exposed p-n junction
(Fig. 1). The method is based on interaction of laser beam
Solar cells are large p-n junction semiconductor devices, with matter [5]. Focused incident photons generate
and experimental studies of these systems using carriers in optically excited p-n junction area. This local
conventional optical techniques, yield in most cases current response is then measured. By scanning of laser
information only of the spatially averaged system beam the LBIC image can be acquired.
properties. Since real samples usually containdistributions
of imperfection sizes, shapes, or composition, [1,2] the The current can be measured as a voltage drops at
signals are often dominated by inhomogeneous sufficiently low resistance, but preferable is measurement
using trans-impedance amplifier (TIA) due to zero input

978-1-4244-9965-6/11/$26.00 ©2011 IEEE 001686


impedance. TIA has a high gain and together with input Figure 3 represents the defect sites distributed mainly in
capacity is decreasing the phase margin. So TIA usually sample edge area for a low reverse bias voltage.
inclines to oscillation. Especially for solar cell, which has
relatively high capacity compared to small photodiode, Spot 1
proper feedback capacitor has to been used to restore y-axis / mm log counts per sec.
phase margin. 20

EXPERIMENTAL 3.5

15
Noise current vs. reverse-bias voltage measurement
3
Spot 2
Figure 1 shows an I-U electrical signal if reverse-bias
2 10
voltage is applied on the 10 x 10 cm wafer. It is
interesting that noise current always appears at same bias 2.5
voltages [13].
As we shown elsewhere, this noise current is a first Spot 35
2
symptom which refers to the presence of defects in the
wafer [14].
0
0 5 10 15 20
x-axis / mm

Figure 3 Localization of defects in solar cell wafer


using reverse-bias light emission in combination with
LBIC (Ur = - 26.0V).

Spot D Spot B

Figure 1 Noise current vs reverse-bias voltage of the


solar cell defects [14]. Spot A
Laser Beam Induced Current
Spot C
Second step is the use of laser beam induced current
(LBIC) measurement for rough localization of defects
(Fig.2) [14].

Figure 4 Light emission from reverse biased sample


(UR = - 22.0V, T = 300 K)

The light emission at reverse voltage UR = -22.0V can be


seen in Fig. 4. In the left side there is light emission before
locale damage by repetitive electrical connection on the
top side contacts. Sample was repetitively connected in
the green circles area (17 points). In the right part of
Fig. 4 there is the same sample after described connection
Figure 2 Computer controlled measurement system procedure. The new nine lighting spots arise in areas of
for low radiant flux detection using photon counting probe connection. Also in this case there is a proof that
mode. mechanical surface damage can lead to light emission

978-1-4244-9965-6/11/$26.00 ©2011 IEEE 001687


from sample under reverse bias condition. Four spots A, Once a rough position of the defect has been known, we
B, C, D (Fig.4) were selected for the further testing. Note can proceed to the local investigation of its electrical,
that spots A and D were presented on the sample before optical and thermal characteristics.
local mechanical damage whilst spots B and C are
artificial defects. Near-field Optical Beam Induced Current

Light emission thermal dependence The resolution of LBIC technique is limited similarly to this
of optical microscopy due to the diffraction. Therefore to
The temperature of solar cell sample was controlled by circumvent the Rayleigh resolution limit, very promising
computer in the range from 305K to 335K during tool for high spatial resolution optical measurements is
application of bias voltage on sample from Fig.3. Optical scanning near-field optical microscope (SNOM). [4, 12, 14]
probe scanned over the sample and locally collected the Here, modifications of electromagnetic radiation are
emitted light (Fig.5.) The light was guided to the cooled detected which are caused by the interaction of the
PMT in photon counting regime. Wavelength range of microscope tip with an object smaller than the wavelength
PMT was set from 350 to 800nm. of light. Hence SNOM seem to be ideally suited to
investigate the properties of individual shining spots within
a defect in realistic structure.

One of the optoelectronic techniques allowing overcome


the diffraction limitation is near-field optical beam-induced
current (NOBIC) [7]. Using this method it is possible to
study photoelectric properties of devices in sub-
wavelength resolution. This method is based on coaction
of the scanning near-field optical microscope (SNOM) and
optical (or laser) beam induced current (OBIC) methods.

Here measured current signal has been detected by a


lock-in nanovoltmeter while the solar cell was reverse-
biased or unbiased. The relative electric response on the
locally induced photocurrent has been mapped by color
Figure 5 Thermal characteristics of I-U curves and scale onto original topography of the sample is shown in
light emission from solar cell sample. Fig. 7.

Light emission versus sample temperature has been


measured also for 4 selected spots from Fig. 4. Solar cell
sample was biased from voltage source UR = -22.0V
during the measurement. Light emission from spots A and
D reflects a negative thermal dependence while spots B
and C have positive thermal dependence (Fig. 6). Since
the light emissions from selected spots have considerably
different values, measurement curves were normalized to
fit one plot.
1

0.9

0.8
Rel. light emission /arb.

0.7
Figure 7 NOBIC image mapped onto topography in the
vicinity of local spot (red).
0.6
The red spot corresponds to the defect near the cell edge.
0.5
Spot A The derived local optical signal from this defect is then
Spot B represented in Fig.8. The lateral resolution of this result is
0.4 Spot C
Spot D
better than 3Pm.
0.3
305 310 315 320 325 330 335
Temperature /K

Figure 6 Thermal properties of light emission of


selected spots (UR = -22.0V).

978-1-4244-9965-6/11/$26.00 ©2011 IEEE 001688


dependence of breakdown voltage the breakdown process
can be determined.

The similarity of partial breakdown values and light


emission threshold show that the defect has direct
connection with local breakdown. The question persists,
which defects are responsible for local breakdown in I-U
measurement. The voltage values of partial breakdown
and light emission are the same for many of samples.
Although there are samples without local breakdown, they
can harbor light emitting imperfections, and vice versa.
This type of imperfections was not inspected in terms of
this work.

Figure 8 Derived optical signal from the defect (red As conclusion we have discovered, that observed and
spot in Fig.7). measured defects could be classify into several groups.
The first group consists of defect having a correlation with
Microscale study of bulk defect the local electric breakdown. Second group contains
thermally dependent imperfections – hot spots. Third
The area near spot 3 from Fig. 3 was measured using group of imperfections can be only localized using near-
Scanning Probe Microscope to visualize the imperfection field optical beam induced current techniques. Moreover,
surface in microscale (Fig.9). Unique structures that are there are a number of defects with very weak optical
probably inclusions causing local change of electric field response, which nature is not still evident.
distribution can be resolved in this figure. Hence the
This work has been supported by the Czech Ministry of
breakdown will occur in this sample area when the sample
Education in the frame of MSM 0021630503 Research
is biased enough. For the light detection scanning near-
Intention MIKROSYN “New Trends in Microelectronic
field optical microscope can be used [15]. Light emission
System and Nanotechnologies”, by grant LH11060 “Study
comes out from small area near the center of the image.
of local electric and optical characteristics of solar cells”
and GACR Grant P102/10/2013 “Fluctuation process in
PN junctions of solar cells”. Smith acknowledges funding
provided by NASA award #NNX09AP67A, NSF award #’s
0619890 (DMR), 0903804 (EPSCoR), and the State of
South Dakota.

REFERENCES
[1] L. Kronik and Y. Shapira, “Surface Photovoltage
Spectroscopy of Semiconductor Structures: at the
Crossroads of Physics, Chemistry and Electrical
Engineering”, Surf. Interface Anal., 31, 2001, pp. 954–965.
[2] D.C.Coffey,et al., “Mapping Local Photocurrents in
Polymer/Fullerene Solar Cells with Photo-conductive
Atomic Force Microscopy”, Nano Lett., 7, 2007, pp. 738–
Figure 9 SNOM image of the sample surface (spot 3). 744
[3] L. Zhang, et al., “Shunt Removal and Patching for
SUMMARY
Crystalline Silicon Solar Cell Using Infrared Imaging and
Laser Cutting”, Prog. Photovolt: Res. Appl. 18, 2010, pp.
This paper presents novel method of microscale
54-60.
localization of solar cell defects based on the combination
of local electrical and optical measurement. It was found [4] M.J. Romero, et al., “Nanoscale Measurements of
that one macro-spot can consist of several micro-spots Local Junction Breakdown in Epitaxial Film Silicon Solar
with diameters bellow d = 3μm. For two spots shown in Cells”, Appl. Phys. Lett. 97, 2010 092107.
Fig.6 the voltage values, at which the spots start to emit
the visible light, are equal. Similarly, the voltage values [5] O. Breitenstein, et al., “EBIC and Luminescence
necessary for the light emission and the partial breakdown Studies of Defects in Solar Cells”, Scanning, 30, 2008, pp.
are the same. No difference has been found when 331-338.
measuring at various temperatures (Fig.4) although the [6] S. Rein, “Lifetime Spectroscopy: A Method of Defect
breakdown voltage changes (for some samples) in Characterization”, Silicon for Photovoltaic Applications,
measured range of temperatures (light emission threshold 85, Spinger Berlin, 2005.
follows the breakdown voltage). From thermal

978-1-4244-9965-6/11/$26.00 ©2011 IEEE 001689


[7] D.K. Schroder, “Semiconductor Material and Device
Characterization”, Wiley-IEEE Press, 3rd edition, 2006.
[8]http://www.renewableenergyworld.com/rea/news/article/
2010/01/lock-in-thermography-enables-solar-cell-
development, last access (15/02/2011).
[9] B. Sopori, et al., “A Reflectance Spectroscopy-Based
Tool for High-Speed Characterization of Silicon Wafers
and Solar Cells in Commercial Production”, 35th IEEE
PVSC, 2010, pp. 2238-2241.
[10] J. Haunschild, et al., “Fast Series Resistance Imaging
for Silicon Solar Cells Using Electroluminescence”, Phys.
Status Solidi RRL 3, 2009, pp. 227– 229.
[11] M. Kasemann, et al. “Progress in Silicon Solar Cell
rd
Characterization with Infrared Imaging Methods”, 23 EU
PSEC, 2008, pp. 965-973.
[12] M.S. Unlü, et al., “Near-field Optical Beam-Induced
Current Measurements on Heterostructures”, Appl. Phys.
Letts., 67, 1995, pp. 1862-1864.
[13] P. Koktavy, et al., “Noise Diagnostics of Solar Cells”,
AIP Conf. Proc., 922, 2007, pp. 141-144.
[14] P. Tománek, et al., “Detection and Localization of
Defects in Monocrystalline Silicon Solar Cell”, Adv. Opt.
Technol., 2010, 805325, 1-5.
[15] Grmela, L., et al., “Local Investigation of Thermal
Dependence of Light Emission from Reverse-Biased
Monocrystalline Silicon Solar Cells”, Sol. Energ. Mat.
Sol.C., (submitted)
[16] E. Esposito, F.-J. Kao, and G. McConnell, “Confocal
optical beam induced current microscopy of light-emitting
diodes with a white-light supercontinuum source,” Appl.
Phys. B, 88, 2007, pp. 551–555.

978-1-4244-9965-6/11/$26.00 ©2011 IEEE 001690

You might also like