NIEL Dose Analysis on triple and single junction InGaP/GaAs/Ge solar cells irradiated with electrons, protons and neutrons

NIEL Dose Analysis on triple and single junction

InGaP/GaAs/Ge solar cells irradiated with electrons,

protons and neutrons

Abstract

Triple junction (InGaP/GaAs/Ge) and single junction (SJ) solar cells were irradiated with electrons, protons and neutrons. The degradation of remaining factors was analyzed as func- tion of the induced Displacement Damage Dose (DDD) calculated by means of the SR-NIEL (Screened Relativistic Non Ionizing Energy Loss) approach. In particular, the aim of this work is to analyze the variation of the solar cells remaining factors due to neutron irradiation with respect to those previously obtained with electrons and protons. The current analysis conirms that the degradation of the Pmax electrical parameter is related by means of the usual semi- empirical expression to the displacement dose, independently of type of the incoming particle. Isc and Voc parameters were also measured as a function of the displacement damage dose. Furthermore, a DLTS analysis was carried out on diodes - with the same epitaxial structure as the middle sub-cell - irradiated with neutrons.

1 Introduction

The prediction of solar cell degradation, due to radiation present in space environment, is of primary importance in the preparation of space missions towards harsh radiation orbits. The radiation analysis of solar cells is required to predict the End Of Life (EOL) performances of the solar arrays. Degradation of solar cell electrical performances depends on the energy, the luence and the type of the incident particles (electrons, protons, neutrons, etc.).

Two methodologies are currently adopted by the space actors to perform on-orbit solar cell performance predictions: the Equivalent Fluence method developed by JPL (Jet Propulsion Labo- ratory) [2 ]. Even if theirstone is the mainly used because of its heritage, it has the disadvan- tage of requiring a large number of irradiation tests with diferent particles having various energies and luences. On the contrary, the more recent DDD approach consents to predict the EOL behav- ior of solar cells starting from a reduced number of irradiation tests allowing for a rapid analysis of emerging cells technologies. The key aspect of this method is that it is based on the calculation of the NIEL (Non Ionizing Energy Loss) doses which, in turn, depends on the amount of the permanent displacement damage induced by particle interactions inside the device and resulting

in the degradation of its electrical performance. Following this approach, a single characteristic degradation curve can be obtained regardless of the particle type when the degradation is related to the actual values of the displacement damage dose only.

In the present work, the results of electron, proton and neutron irradiation tests, performed on triple junction solar cells and related isotype sub-cells will be presented. The NIEL doses, which depend on the displacement threshold energy Ed, were obtained by means of the SR-NIEL tool ( [4 ]). The degradation curve for each cell type, is obtained by plotting the remaining factors (the ratio of the end-of-life EOL value to the beginning-of-life BOL value) of a given electrical parameter as a function of the calculated DDD.

2 Description of irradiated samples

InGaP/InGaAs/Ge TJ solar cells and related component cells with AM0 e伍ciency class 30% (CTJ30), have been manufactured as 2 x 2 cm2 solar cells and 0.5 mm diameter diodes (only top and middle sub cell) 1. The TJ

Figure 1: Schema of triple junction (TJ) and single junction (SJ) isotype sub-cells (Top, Middle and Bottom).

solar cell is composed by a germanium bottom junction obtained by difusion into the germanium P-type substrate, a middle junction of (In)GaAs, whose energy gap is around 1.38 eV and a top junction of InGaP with an energy gap of 1.85 eV. Component cells are single-junction (SJ) cells which shall be an electrical and optical representation of the subcells inside the TJ cell. Therefore, to manufacture them, special attention was put to reproduce the optical thicknesses of all the upper layers present in the TJ structure. Top and middle sub cells for DLTS analysis were also manufactured as diodes with 0.5 mm diameter using a mesa etch to remove the edge defects related to cutting.

3 Experimental Irradiation Procedure

Solar cells have been measured in BOL conditions and then after irradiation, allowing a self- annealing duration of about one month for electrons and protons and about two months for neu- trons.

TJ solar cells and component cells have been irradiated with protons and electrons at diferent energies and luences (see Table [6, [8 ] using experimental data of activation rates and corresponding cross sections of various elements 10] . The solar cells were irradiated together with monitor samples to get a better than 5 percent accuracy on neutron luence. Spatial characterization was performed by means of Al-Co monitor sample [3, [12 ]) from the stopping power in each absorbing layer 7 ]. In this work, the TJ solar cell was approximated by a GaAs cell single junction cell since the middle cell is the one that mainly afects the overall TJ performances at EOL.

For fast neutrons from nuclear reactors, the displacement damage dose can be computed (see [4 ]) from:

where D(E) is the damage function (also referred to as displacement kerma function) in units of MeV cm2 and φ(E) is the neutron spectral luence in n cm— 2 MeV— 1 . The damage functions

were obtained from the SR-NIEL calculator available in [13 ]. For the present calculation the spectral luence used is the one for the TRIGA reactor (see Section III).

The following semi-empirical equation is used to it the experimental remaining factors (RFpar) of each electrical parameter as a function of DDD:

being A, C and DDDx it parameters. It should also be noted that A is only relevant for bottom cells.

The displacement threshold energies, Ed, were found using a routine which globally minimizes the square root relative diference (SRRD) of the points with respect to the itted curve.

Fig. 4. It is worth to remark that, for neutrons, the calculated NIEL is almost independent on the values of Ed.

By inspection of Fig. 4 , independently of the type of incoming particle.

Figure 2: Optimal it of Pmax degradation curve for TJ, mid cell, top cell and bottom cell.

5 DLTS Analysis

A DLTS investigation regarding deep levels, induced by neutron irradiation, was carried out using samples described in 7 ]. In those articles the experimental procedure and technique were also discussed. In particular, diodes of 0.5 mm in diameter were prepared using the same epitaxial structure of the middle sub-cell. The DLTS spectra - obtained from middle-cell diodes irradiated

with electrons, protons and neutrons - are shown in Fig. 3 , one can observe

Figure 3: Comparison of the DLTS spectra of middle junctions irradiated by protons (3.69 x 1010 MeV g — 1 ), electrons (1.07 x 1010 MeV g — 1 ) and neutrons (9.96 x 1010 MeV g — 1 ), respectively. Emission rate = 46 s — 1 , pulse width=500 μs, reverse voltage Vr = — 1.5 V, pulse voltage V 1 = —0.1 V.

how the ratio of the peak amplitudes E2/E1 is much larger for the proton irradiated sample than for the electron irradiated one 7] . In addition, the peak E1 was detected a) at all doses for samples irradiated with electrons, b) only at the highest doses for those irradiated with protons and, c) inally, it was not untangled from the background in samples irradiated with neutrons. The E1 and E2 deep levels were identiied as majority carrier traps in the p-type bulk bases 7] .

Figure 4: Concentration of E2 traps induced by irradiations with electrons, protons and neutrons in middle sub cell diodes as a function of DDD with Ed = 21.5 eV.

The DLTS peak heights measured allowed for determining the E2 defect concentrations - in- troduced by displacement damage - which are shown in Fig. [6, 2 deserve further investigations.

6 Conclusions

TJ InGaP/GaAs/Ge solar cells and related component cells were irradi- ated with electrons and protons and, inally, with neutrons at the TRIGA reactor in Casaccia. Solar cell electrical performances degradation was analyzed as a function of the induced Displacement Damage Dose (DDD) computed following the SR-NIEL approach.

The experimentally obtained remaining power factors are well represented by a single semi- empirical expression as a function of the DDD. In addition, the concentration of the E2 deep level was found to depend almost linearly on DDD with a slope which, in turn, depends on the particle type.

The current results make the usage of neutron particles a possible candidate for testing solar cell degradation for application in space environment.