Simulation of multi-junction solar cells based on InGaN using AMPS-1D

Simulation of multi-junction solar cells based on InGaN using AMPS-1D

Benmoussa Dennai 1, A Hemmani 1,H-Ben Slimane 1 and A.Helmaoui 1

1Laboratory in semiconductor device University of Bechar, Algeria

Abstract

During the past few years a great variety of multi- junction solar cells has been developed with the aim of a further increase in efficiency beyond the limits of single junction devices. InxGa1-xN is one of a few alloys that can meet this key requirement. In this paper, we designed series of InxGa1-xN multi-junction solar cells. Key properties of InxGa1-XN tandem solar cells (for two junctions up to six junctions) are simulated by using AMPS-1D software, including I-V characteristic, conversion efficiency, band structure etc. Our calculation shows that the efficiency can be improved from 10.09% for a single junction up to 40.05% for six junctions obtained in 1-sun AM1.5 illumination and at room temperature, using realistic material parameters..

1. Introduction

   Semiconductors of the type III-N are of growing interest in the scientific world. This is justified by the fact that III-N semiconductors are robust, having a high thermal conductivity and a high melting point, and, moreover, a direct forbidden band gap. They currently represent ideal materials for the development of light emitting diodes (LEDs) operating in the green-blue and UV ranges of the electromagnetic spectrum. Among these semiconductors, we find mainly aluminum nitride (AlN), gallium nitride (GaN) and indium nitride (InN), respectively with a gap of 6.2eV, 3.4eV and 0.7eV [1].

   Tandem solar cells and multi-junction solar cells, which consist of a stack of p/n solar cells, are sometimes classified as third generation solar cells. They achieve the highest conversion efficiencies, even exceeding 40 % [6].

   Recently, InxGa1-XN alloys have become very potential for high performance MJ solar cells. Because the band gap of InxGa1-XN alloys can be varied continuously from 0.7 to 3.4 eV. This provides an almost perfect fit to the full solar spectrum offering a unique opportunity to design MJ solar cells using a single ternary alloy system. This will be technologically very significant because of easy

   fabrication, similarity in thermal expansion coefficient, electron affinity and lattice constant. In addition, InN- based alloys are predicted to show high nobilities and lifetime of charge carriers and superior resistance against irradiation damage. These all make InxGa1-xN alloys very promising for high performance solar cells.

   In order to evaluate the possibilities of these alloys, we tried, in this work, to model and simulate tandem cells made of two, three, four, five and six InxGa1-XN junctions with a one dimensional simulation program called a analysis of microelectronic and photonic structures (AMPS-1D).

   In this work, calculations were all performed under 1-sun AM1.5 illumination and a temperature of 300 K using the one diode ideal model, and for convenience, several simplifying assumptions were made, including no reflection losses and no surface recombination velocity.

2. Modelling and simulations

   2.1About AMPS-1D

   AMPS-1D is the first-principles simulation tool developed by the Penn State/Electric Power Research Institute (EPRI)[15]. AMPS software used in this study is based on the first-principles, basic equations of semiconductors and solar cells: Poissons equation, the continuity equation for free holes, and the continuity equation for free electrons.

   . Determining transport characteristics then becomes a task of solving the three coupled non-linear differential equations, each of which has two associated boundary conditions.

   In AMPS, these three coupled equations are solved simultaneously to obtain a set of three unknown state variables at each point in the device: the local vacuum level, the electron, and hole quasi-Fermi levels. From these three state variables, the free carrier concentrations, fields, currents, etc. can then be computed.

   Besides the classical continuity equations, other semi-classical or quantum transport equations such as the Boltzmann equation, Quantum Hydrodynamic

   (QHD) model, Wigner function method, and non- equilibrium Greens function method have also been applied to study the trans-port processes in photovoltaic devices so far [1618].

   four, five, six and seven junctions were simulated. The energy gap and indium fraction for InX Ga1X N alloys computed for a six junction are given in table II.

   AMPS-1D supplies two different approaches to the process of recombination/generation. One is the density	TABLE I I ENERGY GAPS FOR A SIX-JUNCTION TANDEM CELL
   of states (DOS)/capture cross section model and the Junction	indium fraction for InX Ga1X N	Band gap
   other is the carrier lifetime model. Zhang et al. [7] N°	alloys	(eV)
   In this study, we use the DOS model in the simulation of InGaN solar cells because the DOS 1	0.11718	2.25
   model could provide much more information about recombination/generation in semiconductors than the 2	0.613	1.79
   lifetime. 3	0.766659	1.475
   4	0,8535	1.19
   the InX Ga1X N SCs 5	0,921	0.927
   Band gap[2] 6	1	0.7

   AMPS-1D supplies two different approaches to the process of recombination/generation. One is the density	TABLE I I ENERGY GAPS FOR A SIX-JUNCTION TANDEM CELL
   of states (DOS)/capture cross section model and the Junction	indium fraction for InX Ga1X N	Band gap
   other is the carrier lifetime model. Zhang et al. [7] N°	alloys	(eV)
   In this study, we use the DOS model in the simulation of InGaN solar cells because the DOS 1	0.11718	2.25
   model could provide much more information about recombination/generation in semiconductors than the 2	0.613	1.79
   lifetime. 3	0.766659	1.475
   4	0,8535	1.19
   the InX Ga1X N SCs 5	0,921	0.927
   Band gap[2] 6	1	0.7

   2.2. Parameters for the simulation

   Material parameter equations used for the simulation of

   Eg x = 0.7x + 3.4 1 x 1.43(1 x) (1)

   Electron affinity [8,9]:

   = 4.1 + 0.7(3.4 Eg ) (2)

   Absorption coefficient

   = 2.2 × 105 1.24/ (3) Effective density of states in the conduction band [8]

   = [0.9 + 1 2.3] × 1018 (4)

   Effective density of states in the valence band [4]

   Nv = [5.3x + 1 x 1.8] × 1019 (5)

   Relative permittivity [8]:

   r = 14.6x + 1 x 10.4 (6)

   Carrier mobility [10] :

3. Results and discussions

   1. Simulations for a six-junction tandem cell

      InxGa1-xN tandem cells comprising two, three, four, five and six junctions were simulated.

      Hereaftr, the results computed for a InxGa1-xN tandem structure comprising six junctions are given.

      The following table (Table III) shows the energy gaps of the identified materials, the thicknesses of the junctions and that of the n-side.

      TABLE III

      =

      + ,+ ,

      (7)

      ENERGY GAPS AND THICKNESSES FOR A SIX-JUNCTION TANDEM

      ,

      1+ ,

      CELL

      Junction	Band gap	n-side thickness	Junction thickness
      N°	(eV)	(m)	(m)
      1	2.25	0.1	0.3
      2	1.79	0.1	0.3
      3	1.475	0.1	0.3
      4	1.19	0.1	0.5
      5	0.927	0.1	0.5
      6	0.7	0.1	0.5

      Junction	Band gap	n-side thickness	Junction thickness
      N°	(eV)	(m)	(m)
      1	2.25	0.1	0.3
      2	1.79	0.1	0.3
      3	1.475	0.1	0.3
      4	1.19	0.1	0.5
      5	0.927	0.1	0.5
      6	0.7	0.1	0.5

      The above formulae with asterisk are obtained from the linear fitting of the corresponding parameters of InN and GaN. The carrier mobility of InGaN is assumed to be similar to GaN, where i= n, p denotes electrons and holes, respectively, and N the doping concentration, while the model parametersmin ,i , max ,i

      , Ng,i and i depend on the type of semiconductor [10].

      TABLE I

      MODEL PARAMETERS USED IN THE CALCULATIONS OF THE CARRIER MOTILITIES

      Type of

      ,

      max ,i

      2 1 1

      Ng,i

      3 i

      carriers

      ()

      (cm V S )

      cm

      We can see from Table III that the energy gaps of the junctions decrease from the top to the bottom of the tandem cell.

      Electrons 100	55		1
      Holes 170	3		2

      Electrons 100	55		1
      Holes 170	3		2

      . InxGa1-xN tandem cells comprising two, three,

      Table IV gives the computations of the photocurrent

      60

      60

      densities, the open-circuit voltages and the output peak power for a six-junction InxGa1xN tandem cell.

      50

      50

      Simulations show that the six-junctions InxGa1-xN tandem cell could reach an efficiency of more than 40% with a short-circuit current density of 52.27mA/cm2 and

      current density(mA/cm2)

      current density(mA/cm2)

      an open-circuit voltage of 0.51V. 40

      TABLE IV

      SIMULATION RESULTS FOR A SIX-JUNCTIONS TANDEM CELL

      30

      Junction N Isc mA/cm2 Voc

      Fill

      Efficiency (%)

      		(V)	factor(%)
      1	5.79	1.87	0.92	10.09
      2	19.66	1.31	0.90	23.35
      3	27.29	1.00	0.88	24.30
      4	40.74	0.86	0.86	30.64
      5	46.32	0.63	0.83	36.57
      6	52.27	0.51	0.80	40.05

      		(V)	factor(%)
      1	5.79	1.87	0.92	10.09
      2	19.66	1.31	0.90	23.35
      3	27.29	1.00	0.88	24.30
      4	40.74	0.86	0.86	30.64
      5	46.32	0.63	0.83	36.57
      6	52.27	0.51	0.80	40.05

      20

      10

      0

      1 2 3 4 5 6

      The junction number

      Fig.2: Variation of the short-circuit current density versus the number of junctions in the cell

      We can notice from Table VI and Fig.1 that the open-circuit voltages produced by the junctions of the tandem cell decrease almost linearly from the top to the bottom.

      2,0

      1,8

      0,92

      0,90

      Fill factor(%)

      Fill factor(%)

      0,88

      0,86

      0,84

      0,82

      0,80

      Open circuit voltage (V)

      Open circuit voltage (V)

      1,6

      1,4

      1,2

      1,0

      0,8

      0,6

      0,4

      1 2 3 4 5 6

      Number of the junction

      1 2 3 4 5 6

      Number of junctions

      Fig.3. Variation of Fill factor versus the number of junctions in the cell

      Fig. 1: Variation of the Open-circuit voltage versus the junction number for a six-junction tandem structure.

      40

      40

      InxGa1-xN tandem cells have an additional advantage as they can be produced with a simpler technology than the ones used to produce tandem junctions made of different materials. In fact, the InxGa1-xN alloys have

      35 similar properties, which make their deposition in

      successive films easier.

      Efficiency (%)

      Efficiency (%)

      30

      25

      20 5.Acknowledgements

      15

      15

      We would like to acknowledge the use of AMPS-1D program that was developed by Dr. Fonashs group at

      10 Pennsylvania State University (PSU).

      1 2 3 4 5 6

      Number of junctions

      Fig. 4: Variation of the efficiency with the number of junctions included in the tandem cell.

      We notice from Table 2and Fig. 1that the achievable open-circuit voltages decreases as the tandem cell. Contains more layers However, the increase of the short-circuit current density as a function of the number of junctions is almost linear; hence, it compensates the decrease of the open-circuit voltages, which is a function of the inverse of the same variable; this explains the increase of the output maximum power and the cell efficiency with the number of junctions.

      The highest efficiency (40.05%) was reached for the six-junction cell (Fig. 4)with a short-circuit current density of 52.27mA/cm2 and an open circuit voltage of about 0.51V.

      InxGa1-xN tandem cells comprised of two and three junctions have also interesting potentials with efficiency of 23.35% for the two-junction tandem cell and of 24.3% for the three-junction one.

      It is noticeable that the increase of the efficiency is more important when we move from a two-junction to a three-junction InxGa1-xN tandem cell.

4. Conclusion

The theoretical design and performance of InxGa1-xN- based MJ solar cells for high efficiency have been studied by developing a simulatio model. The simulation with AMPS-1d result shows that the InxGa1- xN alloys have interesting performances for tandem cells applications. The efficiency is evaluated from 10.09% for single junction to 40.05% for six junctions.

6.References

1. Wu J, Walukiewicz W, Yu K M, Ager J W III, Haller E E,Lu H, Schaff W J, Saito Y and Nanishi Y 2002Appl. Phys.Lett.803967)

2. Xiao H L, Wang X L, Wang J X, Zhang N H, Liu H X,Zeng Y P, Li J M and Wang Z G 2005J. Cryst. Growtp76401I.

3. S. Jacobs and C. P. Bean, Fine particles, thin films and exchange anisotropy, in Magnetism, vol. III, G. T. Rado and H. Suhl, Eds. New York: Academic, 1963, pp. 271350.

4. Jani O, Honsberg C, Asghar A, Nicol D, Ferguson L,Doolittle A and Kurtz S 2005 31stIEEEPhotovoltaic SpecialistsConf.(Lake Buena Vista, FL,2005)pp 3742.

J. Name Stand. Abbrev., in press

1. Zhang, X. B. et al., Simulation of In0.65Ga0.35N single- junction solar cells, J. of Phys. D: Appl. Phys. 40, 7335-7338 (2007)..

2. Hamzaoui, H., Bouazzi, A. S. and Rezig, B., Theoretical possibilities of InxGa1-xN tandem PV structures, Sol. Energy Mater. Sol. Cells 87, 595-603(2005).

3. Levinshtein, M. E., Rumyantsev, S. L. and Shur, M. S., [Properties of Advanced Semiconductor Materials],Wiley, Chichester,UK,190(2001).

4. Li, N., Simulation and Analysis of GaN-based Photoelectronic Devices, Dissertation (in Chinese), Institute of Semiconductors, Chinese Academy [2] ofSciences, Beijing, 3(2005).

5. Li, N., Simulation and Analysis of GaN-based Photoelectronic Devices, Dissertation (in Chinese), Institute of Semiconductors, Chinese Academy ofSciences, Beijing, 3(2005).