Revue des Energies Renouvelables CER’07 Oujda (2007) 157 – 160
157
Degradation in GaInP solar cells : effect of recombination velocity
M. Idali Oumhand 1, M. Zazoui 1*, S. Makham 2, G.C. Sun 2 and J.C. Bourgoin 2
1Laboratoire de Physique de la Matière Condensée, Université Hassan II B.P. 146, Bd Hassan II, F.S.T., Mohammedia, Maroc
2GESEC R&D, Université Pierre et Marie Curie (Paris VI), 140 Rue de Lourmel, 75015 Paris, France
Abstract - The degradation induced by irradiation in space of the open circuit voltage, Voc, and of the short circuit current, Isc, of GaInP solar cells can be computed when the introduction rate k of recombination centers and their capture cross sections σn.p for minority carriers in the emitter and base are known. We have previously determined σn.p and k for electron irradiation in the case of GaInP cells. In this communication, we report on the determination of k for irradiation with protons of GaInP cells and its dependence versus proton energy. We show that the calculated dependences of Isc and Voc versus the fluence of electrons and protons fit the experimental data.
Finally, we illustrates that the degradation is very sensitive to the recombination velocity
1. INTRODUCTION
For years, the prediction of the degradation of solar cells in space relied on data bases providing values of the cell parameters (short-circuit current, Isc, open-circuit voltage, Voc, and maximum power Pmax) obtained from experimental tests for specific fluences of electrons and protons of variable energies and this for each new cell. More recently, with the advent of multijunction (MJ) cells, the size of these data bases became so large that it is now necessary to find ways to reduce them. Thus, for instance, an empirical law has been proposed to describe the radiation-induced degradation of the maximal power Pmax, which contains two empirical parameters [1]:
(
x)
o
max/P 1 Cln 1 D/D
P = − + (1)
where Pmax/Po is the normalized maximum power after exposure to a given fluence D and C and D are empirical constants. x
More recently, a simple method, still empirical, has been proposed [2] consisting in deriving the equivalence between two fluences (ϕe and ϕp) of irradiation at different energies (E and e
E ), for the same irradiating particles or for different ones, electrons and protons: p
( )
e p n1( )
p1 n
eE E = ϕ E E
ϕ (2)
In this expression, En1 is the energy lost by the energetic particle into atomic collisions, which can be often approximated by the NIEL (for non ionizing energy loss [3]). In this way, a degradation curve describing the variation of a cell parameter versus fluence can be deduced from a standard curve, once the En1 values have been calculated for the energies E and e E . p
Actually, the fluence dependences of Isc, Voc and Pmax could be calculated when the minority carrier lifetimes τ associated with the defects produced by the irradiation in the n and p regions of the cell are known. The determination of the characteristics of the recombination
M. Idali Oumhand et al.
158
centers has now been achieved [4, 5] using a combination of techniques (current voltage in dark and under illumination, electroluminescence, Deep Level Transient Spectroscopy (DLTS)) which can be directly applied to the cell itself. Thus, for GaInP, the order of magnitude of kσ, product of the introduction rate of the recombination centers, k , times their cross section, σ, for capturing a minority carrier, have been obtained for 1 MeV electron irradiation. This allowed to deduce τ(ϕ):
ϕ ν σ τ + τ = 1 k 1
0
(3) where τ0 is the initial lifetime prior irradiation, ϕ the fluence and ν the carrier velocity.
A satisfactory prediction of the degradation of Isc and Voc versus fluence has then been achieved for GaInP cells [6] irradiated with 1 MeV electrons.
The aim of this communication is to perform a similar prediction for GaInP in the case of proton irradiations.
2. PREDICTION OF PROTON INDUCED DEGRADATION IN GAINP CELLS In GaInP, previous studies have shown that the recombination centers induced by electron irradiation, corresponding to two traps labelled H2 and H3 by DLTS, are characterized by introduction rates k equal to 0.17 cm-1 for 1 MeV electrons [7-9].
As developed in [6], the prediction of Isc(ϕ) and Voc(ϕ) has been achieved for a 1 MeV electron irradiation.
Based on the fact that k is proportional to En1 [10]:
1
En
k = β (4)
the known value of k for 1 MeV electrons allows to determine β, since the 1 MeV electron energy loss En1 is known. Knowing that k= 0.17 cm-1 [7, 8] and En1= 3.17 × 10-5 MeVcm2/g [11], we obtain β= 5.4 × 103 g/MeVcm3.
First, the data provided by 1 MeV electron irradiation are fitted (Figs. 1 and 2) using kσkσ in the emitter and base, the original value determined by electroluminescence [4]. We then obtain the degradation parameters kσn and kσp. Since k is known, this allows deducing the capture cross sections for electrons and holes σn and σp
Second, we fit the available degradation curves of the NRL cell for 3 MeV protons (Figs. 3 and 4). We obtain kσn and kσp. Finally, the results are plotted, together with that obtained for 1 MeV electrons versus En1 in Fig. 5. The slope 1 of this log-log plot indicates that, as expected, the introduction rate of recombination centers is proportionnal to En1. The slope of k(En1)k (Enl) gives βn = 9.139 × 103 g/MeVcm3 and βn = 4.9 × 103 g/MeVcm3. These values are close of the calculated value.
3. EFFECT OF RECOMBINATION VELOCITY
The degradation of the short circuit current of low fluence is limited by that provided by GaInP top cell as shown in Fig 6. The degradation occurs for a higher fluence for triple-junctions (3J) GaInP/GaAs/Ge then for double-junctions (2J) GaInP/GaAs cell. The only way to explain this observation is in a change of the recombination velocity at the surface of the emitter. The case of Fig 6, fits of the experimental data are obtained using the following values of S=1.3 × 106 cm/s, S= 9.5 × 105 cm/s for 2J and 3J cells respectively.
CER’2007: Degradation in GaInP solar cells – effect of recombination velocity 159
Fig. 1: Degradation induced by 1 MeV electrons of the short circuit current of a
GaInP NRL cells and comparison with experimental data of ref. 11
Fig. 2: Degradation induced by 1 MeV electrons of the open circuit voltage of a
GaInP NRL cell and comparison with experimental data of ref. 11
Fig. 3: Degradation induced by 3 MeV protons of the short circuit current of a GaInP NRL cell and comparison with
experimental data of ref. 11
Fig. 4: Degradation induced by 3 MeV protons of the open circuit voltage of a GaInP NRL cell and comparison with
experimental data ref. 11
Fig. 5: Variation of the degradation parameters knσn (!) and kpσp (∀) versus the energy lost by electrons and protons into atomic collisions for NRL GaInP cells. The full lines, of slope 1, indicate a linearity between k and En1.
M. Idali Oumhand et al.
160
This illustrates that the degradation is very sensitive to the recombination velocity.
Fig. 6: Degradation induced by 1 MeV electrons of the short circuit current (-) of a 2J and 3J cells and comparison with experimental data of 2J (•) and of 3J (■)
4. CONCLUSION
We have determined the introduction rates of minority carriers recombination centers, k , as well as their capture cross-sections in the bases and emitters of GaInP cells induced by electron and proton irradiations. We have determined the dependences of k versus the energy of the irradiating particles. We have shown that, using these values for the recombination centers, we can predict the degradation, i.e. the variations versus fluence, of the short circuit current and open circuit voltage induced by electron or proton irradiations of any energy.
REFERENCES
[1] S.R. Messenger, G.P. Summers, E.A. Burke, R.J. Walters and M.A. Xapsos, Proc. in Photovoltaics: Res.
and Appl., 9, 103, 2001.
[2] M. Mbarki, G.C. Sun and J.C. Bourgoin, Semicond. Sci. and Technol., 19, 1081, 2004.
[3] S.R. Messenger, E.A. Burke, M.A. Xapsos, G.P. Summers, R.J. Walters, I. Jun and T. Jordan, IEEE Trans. Nucl. Sci., 50 (6), 1919, 2003.
[4] M. Zazoui, M. MBarki, A. Zin Aldin and J.C. Bourgoin, J. Appl. Phys., 93, 5080, 2003.
[5] J.C. Bourgoin and M. Zazoui, Semicond. Sci. Technol., 17, 453, 2002.
[6] S. Makham, M. Zazoui, G.C. Sun and J.C. Bourgoin, Semicond. Sci. Technol, 20, 699, 2005.
[7] A. Khan, M. Yamaguchi, N. Dhamarasu, J.C. Bourgoin, K. Ando and T. Takamato, Japan J. Appl. Phys., 41, 1241, 2002.
[8] A. Khan, M. Yamaguchi, J.C. Bourgoin and T. Takamato, J. Appl. Phys., 91, 2391, 2002.
[9] A. Khan, M. Yamaguchi, J.C. Bourgoin, N. de Angelis and T. Takamato, Appl. Phys. Lett., 76, 2552, 2000.
[10] J. Jabbour, M. Zazoui, G.C. Sun, J.C. Bourgoin and O. Gilard, J. Appl. Phys., to be published (Feb. 2005.
[11] R.J. Walters, M.A. Xapsos. H.L. Cotal, S.R. Messenger, G.P. Summers, P.R. Sharps and M.L. Timmons, Solid-State Electron, 42, 1747, 1998.
