dc.description.abstract |
For the first time in 2015, monocrystalline silicon solar cell has been fabricated in one and only
monocrystalline solar cell fabrication laboratory established in Atomic Energy Research
Establishment (AERE), Savar, Bangladesh. The fabricated silicon solar cells in this laboratory
have the efficiency of only 6.89% so far. The aim of this PhD research is to investigate the
problems that result in the low efficiency and to find out the ways to improve the efficiency of
solar cells with the available facilities in Bangladesh. It is very important to identify the
problems of fabrication to enhance the efficiency. Through various experiment and minute
observation of the cell fabrication process of AERE and the world standard process, the reasons
of low efficiency have been determined. The reasons for achieving low efficiency: (i) are not
having clean room class 100-1000, (ii) Czochralski (CZ) silicon wafer has been used instead
of Float Zone (FZ) silicon wafer, usage of unknown edge isolation barrier paste, improper
temperature at surface passivation technique, (iii) low shunt resistance, (iv) low minority
carrier diffusion length (88 µm), (v) high series resistance (6.197 Ω.cm2), (vi) wafer bowing,
(vii) unsmooth busbars and grid fingers, (viii) micro cracks, (ix) less aspect ratio, (x) not
applying anti-reflection coating and (xi) lastly not utilizing oxygen and nitrogen gas in the
metallization process.
Any change in the solar cell fabrication steps is too costly, wastage of resource material and
the experimental procedure proves to be difficult and time consuming. For these reasons
simulation has gained importance in the last few years. PC1D is a commercially available
software most commonly used for solar cell modelling. Here simulation of monocrystalline
silicon solar cell has been done using PC1D software. From the simulation it is seen that, the
optimum value of P-type doping concentration is 1×1017 cm-3 and N-type doping concentration
is 1×1018 cm-3. Diffusion length of 200 µm gives optimum result. Both side textured wafer with
pyramid height of 2-3 micrometer and equal angles of 54.74 degrees produces the best result
in simulation.
Six different types of anti-reflection coating (ARC) layers have also been simulated using
PC1D software. Result shows that the range of 500 nm – 700 nm would be suitable for
designing an ARC. Single layer silicon nitride (Si3N4) ARC designed for 600 nm wavelength
and with 74.257 nm thickness shows 20.35% efficiency. Significant increase in efficiency has2 of 2 | P a g e
been observed by applying ARC in respect to not applying any kind of ARC. After efficient
solar cell modelling optimum efficiency of 20.67% is achieved by using SiO2 surface
passivation and Si3N4 ARC layer. The effects on voltage, current, photovoltaic efficiency,
reflectivity and external quantum efficiency for various ARCs are also represented in this work.
Texturization is a very important aspect of solar cell fabrication process. Optimum texturization
can increase the efficiency of solar cell. In this work practical experiments with different
texturization process have been done. Different concentrations of Na2CO3 - NaHCO3 solution,
KOH-IPA solution and TMAH solution with different time intervals have been investigated
for texturization process. Furthermore, saw damage removal process has been conducted with
10% NaOH solution, 20 wt% KOH - 13.33 wt% IPA solution and HF/Nitric/Acetic Acid
(HNA) solution.
The surface morphology of saw damage, saw damage removed surface and textured wafer have
been observed using optical microscope and Field Emission Scanning Electron Microscopy
(FE-SEM). Texturization causes pyramidal micro structures on the surface of (100) oriented
monocrystalline silicon wafer. The height of the pyramid on the silicon surface varies from 1.5
µm to 3.2 µm and the inclined planes of the pyramids form acute angles. Contact angle value
indicates that the textured wafers surface fall in between near-hydrophobic to hydrophobic
range. Less than 0.1% reflectance has been achieved when textured with 0.76 wt% KOH - 4
wt% IPA solution for 20 minutes. Furthermore, an alternative route of using 1 wt% Na2CO3 -
0.2 wt% NaHCO3 solution for 50 min has been exploited in the texturization process.
N-Type layer (Emitter) has been formed by diffusion process over an as-cut monocrystalline
P-type silicon wafer (Base). The diffusion process has been carried out at 875 °C in an
Atmospheric Pressure Chemical Vapor Deposition (APCVD) chamber. In the APCVD
chamber POCl3 (Phosphorus Oxychloride), N2 and O2 gas has been used. For the diffusion
process, the deposition time and drive time variations are 5, 10, 15, 20, 30 min and 10, 15, 20,
25 and 35 min, respectively. After diffusion process hot point probe experiment has been
carried out and it ensures that N-Type layer has been formed. Experimentally obtained values
of the emitter sheet resistance are 65.25 Ω/□, 44.62 Ω/□, 13.53 Ω/□, 32.75 Ω/□ and 55.38 Ω/□.
From the sheet resistance values, electron concentration has been theoretically calculated. From
sheet resistance values and considering low manufacturing cost, 5 min diffusion and 10 min
drive time provides the suitable recipe to create a N-type layer upon P-type silicon substrate.
Apart from fabrication of solar cell, characterizations of different silicon wafer and solar cell
have been done. The characterizations are: (i) surface morphology determination using optical
microscope and SEM, (ii) surface roughness and height determination using Surface
Profilometer, (iii) band gap measurement using Spectral Response Measurement Machine, (iv)
fill factor and efficiency measurement using LIV tester, (v) series and shunt resistance
measurement, (vi) sheet resistance determination using Four Point Probe System, (vii) P-type
and N-type determination using Hot Probe Method, (viii) thickness measurement of different
wafers using Dial Indicator, (ix) reflectance measurement different wafers using UV-VIS-NIR
Spectroscopy, (x) busbars and grid fingers height, width and aspect ratio measurement and (xi)
EDS measurement of different doped wafer. All the characterization processes and results are
discussed in details in this dissertation. Lastly, limitation and remedies of fabrication of solar
cell in Bangladesh has been discussed in this thesis. |
en_US |