**6. Solar cell fabrication**

350 Solar Cells – New Aspects and Solutions

The annealed film shows a linear lnσ vs 1/*T* relationship. The activation energy of the dark conductivity was estimated to be 0.68 eV from the temperature dependence of the conductivity curve for AlSb film. This value is in good agreement with work done by Chen et al. (Chen et al., 2008). This curve also confirms the semiconducting property of the AlSb (3:7) film because the conductivity of the film was seen to be increasing with increasing the

AMPS 1D beta version (Penn State Univ.) was used to simulate the current voltage characteristics of p-*i*-n junction AlSb solar cells. The physics of solar cell is governed by three equations: Poisson's equation (links free carrier populations, trapped charge populations, and ionized dopant populations to the electrostatic field present in a material system), the continuity equations (keeps track of the conduction band electrons and valence band holes) for free holes and free electrons. AMPS has been used to solve these three coupled nonlinear differential equations subject to appropriate boundary conditions. Following

**Contact Interface** 

*\* S* surface recombination velocity of electrons or holes. **Semiconductor Layers** 

Density of electrons on conduction band, *N***<sup>C</sup>** (cm-3) 1.80 × 1018 7.80 × 1017 2.22× 1018 2.22× 1018 Density of holes on valence band, *N***<sup>V</sup>** (cm-3) 2.20× 1019 1.80× 1019 1.80× 1019 1.80× 1019 Electron mobility, *µ***<sup>e</sup>** (cm2/Vs) 100 80 100 100 Hole mobility, **µp** (cm2/Vs) 25 420 25 25

> \* *ε*0 = 8.85×10-12 *F*/m electric constant; TCO is In2O3: SnO2. **Gaussian Midgap defect states**

*\* N*DG/AG the donor-like or acceptor-like defect density, *W*G the energy width of the Gaussian distribution

for the defect states, *τ* carrier lifetime, and *σ* capture cross section of electrons (*σ***e**) or holes (*σ***p**). Table 2. Parameters of the simulating the IV behavior or p-*i*-n junction solar cells.

*<sup>N</sup>***DG,** *N***AG** (cm-3) A = 1×

Barrier Height (eV) 0.1 (EC-EF) 0.3 (EF-EV) *S***e** (cm/s) 1.00 × 108 1.00 × 108 *S***<sup>h</sup>** (cm/s) 1.00 × 108 1.00 × 108

Thicknesses, **d** (nm) 100 1000 45 200

Band gap, **Eg** (eV) 3.6 1.6 2.4 3.6

1× 1018

1019

*W***<sup>G</sup>** (eV) 0.1 0.1 0.1 0.1 *σ***<sup>e</sup>** (cm2) 1.00 × 10-13 1.00 × 10-8 1.00 × 10-16 1.00 × 10-11 *σ***<sup>p</sup>** (cm2) 1.00 × 10-131.00 × 10-111.00 × 10-13 1.00 × 10-14

Electron affinity, *Χ* (eV) 4.5 4.5

**<sup>0</sup>** 10 9.4 10 9

NA = 1× 1014

D = 9× 1010

ND = 1.1× 1018

> A = 1× 1019

ND = 1× 1018

D = 1× 1016

**CuSCN AlSb ZnO TCO** 

simulation parameters was used for the different layers of films.

Permittivity,

**/**

Acceptor or donor density, *<sup>N</sup>***<sup>A</sup>** or *N***<sup>D</sup>** (cm-3) NA =

excitation.

**5. Simulation of solar cell** 

Both p-n and p-*i*-n junction solar cells were designed and fabricated in 1cm x 2 cm substrate with AlSb as a p type and an absorber material respectively. Variety of n type materials including TiO2 and ZnO were used to check the photovoltaic response of AlSb thin film. Fig. 9 shows the p-n and p-*i*-n based solar cell design with ZnO and TiO2 are an n-type layer and CuSCN as a p-type layer.

Fig. 9. Solar Cell Design (a) p-n and (b) p-*i*-n structure.

AlSb Compound Semiconductor as Absorber Layer in Thin Film Solar Cells 353

separation at the junction and decreases VOC of the device. A better material needs to be explored to dope AlSb n type to increase the built in field. The field could also be extended

AlSb/TiO2 80 12x10-3 0.23 0.001

AlSb/ZnO 120 76x10-3 0.24 0.009

Interesting results were obtained with a p-*i*-n junction, CuSCN/AlSb/ZnO. The used cell has an active cell area of this cell was 0.36 cm2 and fabricated on Mo coated glass surface. Charge was collected from the silver epoxy fingers casted on top of ITO surface and Mo back contact. The cell showed a *V*OC of ~ 500 mV and a *J*SC of 1.5 mA/cm2. With a FF value of 0.5, the efficiency of this cell was calculated to be 0.32%. This observation may be attributed to the more efficient charge separation than that in the p-n junction devices due to a strong build-in field. However, the efficiency of the p-*i*-n junction device is very low in comparison to other available thin film solar cells devices. There are still many unknown factors including the interfaces in the junction. Such a low efficiency could be attributed to the defects along the AlSb interface with both the p- and n-type of layers. Interfaces between AlSb and other layers needed to be optimized for a better

AlSb thin film has been prepared by co-sputtering aluminum and antimony. The deposition rate of Al:Sb was required to be 3:7 to produce the stoichiometric AlSb film with optical band gap of 1.44 eV. After annealing the film at 200 0C in vacuum for two hours, the film likely formed crystalline structures with a size of ~200 nm and has strong absorption coefficient in the range of 105 cm-1 in the visible light. p-n and p-*i*-n heterojunction solar cells were designed and fabricated with AlSb as a p-type material and an intrinsic absorber layer. The simulation of the p-i-n junction solar cell with CuSCN/AlSb/ZnO using AMPS at AM1.5 illumination shows efficiency of 14% when setting ~1 m-thick absorber layer. The p-n junction solar cells were fabricated with different types of n layers shows the photovoltaic responses. The p-*i*-n showed better photovoltaic performance than that of p-n junction cells. All the preliminary results have demonstrated that AlSb is promising photovoltaic material. This work is at the early stage. More experiment is needed for the understanding of the crystallization and

Support for this project was from NSF-EPSCoR Grant No. 0554609, NASA-EPSCoR Grant NNX09AU83A, and the State of South Dakota. Simulation was carried out using AMPS 1D beta version (Penn State University). Dr. Huh appreciates AMES Lab for providing

properties of the AlSb films and the interface behaviors in the junctions.

sputtering facility. We appreciate AMPS 1D beta version (Penn State Univ.)

**Cell VOC (mV) ISC (mA) FF Efficiency %** 

using the p-*i*-n structure to design the solar cells.

performance.

**8. Summary** 

**9. Acknowledgments** 

Table 3. Current-voltage characteristics of p-n junction solar cells

ZnO thin film was prepared by RF sputtering of 99.999% pure ZnO target (Kurt J. Lesker, PA, diameter 2 inches and thickness 0.25 inches). ZnO intrinsic film was deposited by RF power of 100 W at 0.7 Å/s and subsequently annealed in air at 150 0C. ITO film was also prepared from 99.99% pure ITO target (Kurt J. Lesker, PA, diameter 2 inches and thickness 0.25 inches) on the similar fashion using dc magnetron sputtering. Transparent ITO film was deposited at plasma pressure of 4.5 mTorr. The sputtering power of 20 W yields deposition rate 0.3 Å/s. The film was then annealed at 150 0C in air for 1 hour. The highly ordered mesoporous TiO2 was deposited by sol gel technique as described by Tian et al. (Tian et al. 2005). CuSCN thin film was prepared by spin coating the saturated solution of CuSCN in dipropyl sulphide and dried in vacuum oven at 80 0C (Li et al., 2011). The thickness of all three films ZnO, ITO and TiO2 film was about 100 nm and the thickness of AlSb layer is ~1 micron. The active layer was annealed.
