Mercedes Gabas

This paper presents the comparison of two temperature Accelerated Life Tests (ALTs) for space commercial solar cells. The tests are carried out in dark conditions to circumvent the inherent problems of illumination ALTs. In the first ALT the solar cell working conditions were emulated by means of forward bias (that emulates the electric performance under illumination) while in the second ALT, forward bias was used together with reverse bias (that emulates the shadowing at eclipse periods or any other shadowing event such as for example the shadows cast by antennas) in order to reproduce more adequately the real operation conditions of solar cells. Forward to reverse time ratio is 4:1. Furthermore, the high temperatures used in the second ALT (190, 210 and 230 °C) allow a considerable time reduction of the test with respect to the first ALT (150, 165 and 180°C). The evaluation of the Arrhenius activation energy for an illumination power degradation of 10% was 0.88 eV in the forward bias test while 1.06 eV in the forward-reverse test. Reverse bias promotes similar degradation mode than forward one, namely parallel resistance reduction, but probably with a different degradation mechanism that degrades solar cell more intensely, thus reducing reliability which is more than 7 times shorter in the forward-reverse test than in forward one. Despite this reliability reduction, the time for a 90% reliability in the forward-reverse ALT is 32 years of continuous operation assuming a nominal temperature of 80°C which is typical of many GEO missions. The solar cells used in both ALTs were commercial GaInP/Ga(In)As/Ge triple junction and perform very robustly and exhibiting a high reliability for many space applications. It should be noted that these reliability figures are related to the degradation caused exclusively by high temperatures and other stressors like radiation are not considered here.

We are developing flexible and lightweight III-V multijuntion solar cells for high power density applications such as unmanned aerial vehicles (drones), High Altitude Pseudo Satellites (HAPS) and high altitude stratospheric balloons for satellite-like communications, cheap and widespread internet connectivity, etc. We focus our research on two different types of III-V solar cells covering the range from 1 kW/kg to 3 kW/kg (considering the weight of the solar cell itself and not that of the module). Our approach to achieve a goal of 3 kW/kg is based on 3J inverted metamorphic solar cells. We have recently developed GaInP/GaAs/GaInAs 3J IMM solar cells with a power density of 5.7 kW/kg under 1xAM1.5g for small areas. Demonstrating large area devices (tens of square cm) is the next challenge to tackle. In addition, we are developing lower-cost designs, targeting 1 kW/kg, based on GaInP/Ga(In)As/Ge 3J lattice matched solar cells. In this paper we present the advances in the modelling and manufacturing of this last type of solar cell. Modelling indicates that there are multiple efficiency-thickness combinations reaching the 1 kW/kg target. For example in the case of AM1.5g, a solar thickness of 46 microns requires an efficiency of 30% which is a demanding value. However, if the thickness is reduced to 40 microns, a more easily achievable efficiency of 27% would be required. We have demonstrated a process for thinning the Ge substrates reaching thicknesses as low as 14 microns with highly homogeneous thickness profiles. Preliminary results show that the 3J thinned solar cell exhibits an efficiency which is the 91% of the efficiency achieved by the non-thinned cell.

In this work, we demonstrate a new manufacturing process that allows to fabricate thin solar cells avoiding problems related to the brittleness or fragility of thin wafers without needing any carrier nor temporary substrate. This technique has been successfully applied to Ge single and GaInP/Ga(In)As/Ge triple‐junction solar cells grown on Ge substrates, achieving 47.5‐ and 55.5‐μm‐thick substrates, respectively. Although this technique should allow to thin substrates to any desired thickness, a final thickness around 20–30 μm is recommended due to practical concerns such as the uneven substrate thickness throughout the wafer. In any case, a III–V solar cell with such a thickness should be thin enough for most applications and, in particular, for space ones.

A temperature accelerated life test on Upright Metamorphic Ga_0.37In_0.63P/Ga_0.83In_0.17As/Ge triple‐junction solar cells has been carried out. The acceleration has been accomplished by subjecting the solar cells to temperatures (125, 145 and 165°C) significantly higher than the nominal working temperature inside a concentrator (90°C), while the nominal photo‐current (500×) has been emulated by injecting current in darkness. The failure distributions have been fitted to an Arrhenius–Weibull model resulting in an activation energy of 1.39 eV. Accordingly, a 72 years warranty time for those solar cells for a place like Tucson (AZ, USA), was determined. After the ALT, an intense characterization campaign has been carried out in order to determine the failure origin. We have detected that temperature soak alone is enough to degrade the cell performance by increasing the leakage currents, the series resistance, and the recombination currents. When solar cells were also forward biased an increase of series resistance together with a reduction of short circuit current is detected. The failure analysis shows that: a) several metallization sub‐products concentrate in several regions of front metal grid where they poison the silver, resulting in a two times reduction of the metal sheet resistance; b) the metal/cap layer interface is greatly degraded and there is also a deterioration of the cap layer crystalline quality producing a huge increase of the specific front contact resistance, c) the decrease of short circuit current is mainly due to the GaInP top subcell degradation.

Graphene has been intensively studied in photovoltaics focusing on emerging solar cells based on thin films, dye‐sensitized solar cells, quantum dots, nanowires, and so forth. However, the typical efficiency of these solar cells incorporating graphene is below 16%. Therefore, the photovoltaic potential of graphene has not yet been shown. In this work, the use of graphene for concentration applications on III‐V multijunction solar cells, which indeed are the solar cells with the highest efficiency, is demonstrated. First, a wide optoelectronic characterization of graphene layers is carried out. Then, the graphene layer is incorporated onto triple‐junction solar cells, which decreases their series resistance by 35% (relative), leading to an increase in fill factor of 4% (absolute) at concentrations of 1000 suns. Simultaneously, the optical absorption of graphene produces a relative short‐circuit current density decrease in the range of 0% to 1.8%. As a result, an absolute efficiency improvement close to 1% at concentrations of 1000 suns was achieved with respect to triple‐junction solar cells without graphene. The impact of incorporating one and two graphene monolayers is also evaluated.

Space solar cells have been subjected to an innovative temperature Accelerated Life Test (ALT) to get information on their degradation and their activation energy. Our procedure emulates the thermal stress that solar cells suffer working under space conditions. Solar cells were kept in a climatic chamber under nitrogen atmosphere at three different temperatures for few months. After degradation, results have been fitted to a temperature Arrhenius model, in order to obtain the activation energy and acceleration factor of space GaInP/Ga(In)As/Ge triple junction solar cells.