Jessica G. J. Adams

Design of an achievable, all lattice-matched multijunction solar cell using InGaAlAsSb

A design for a realistically achievable, multijunction solar cell based on all lattice-matched materials with >50% projected efficiencies under concentration is presented. Using quaternary materials such as InAlAsSb and InGaAlAs at stochiometries lattice-matched to InP substrates, direct bandgaps ranging from 0.74eV up to ∼1.8eV, ideal for solar energy conversion, can be achieved. In addition, multi-quantum well structures are used to reduce the band-gap further to <0.7 eV. A triple-junction (3J) solar cell using these materials is described, and in-depth modeling results are presented showing realistically achievable efficiencies of AM1.5D 500X of η ∼ 53% and AM0 1 Sun of η∼ 37%.

Modeling and analysis of multijunction solar cells

The modeling of high efficiency, multijunction (MJ) solar cells away from the radiative limit is presented. In the model, we quantify the effect of non-radiative recombination by using radiative efficiency as a figure of merit to extract realistic values of performance under different spectral conditions. This approach represents a deviation from the traditional detailed balance approximation, where losses in the device are assumed to occur purely through radiative recombination. For lattice matched multijunction solar cells, the model predicts efficiency values of 37.1% for AM0 conditions and 52.8% under AM1.5D at 1 sun and 500X, respectively. In addition to the theoretical study, we present an experimental approach to achieving these high efficiencies by implementing a lattice matched triple junction (TJ) solar cell grown on InP substrates. The projected efficiencies of this approach are compared to results for the state of the art inverted-metamorphic (IMM) technology. We account for the effect of metamorphic junctions, essential in IMM technology, by employing reduced radiative efficiencies as derived from recent data. We show that high efficiencies, comparable to current GaAs-based MJ technology, can be accomplished without any relaxed layers for growth on InP, and derive the optimum energy gaps, material alloys, and quantum-well structures necessary to realize them.

Demonstration of Multiple Substrate Reuses for Inverted Metamorphic Solar Cells

The substrate typically accounts for more than 50% of the bill of materials cost for a III–V solar cell wafer. A substantial fraction of this cost can be mitigated via multiple reuses of the substrate. We report on inverted metamorphic solar cells grown on GaAs substrates that have undergone up to five consecutive reuse cycles. The active solar cell layers are removed from the substrate after each growth cycle via epitaxial lift-off, a non-destructive selective etch process, and the substrate is then repolished in preparation for reuse. The solar cell wafer is transferred to a temporary carrier for processing using standard photolithographic techniques. Data from five reuse cycles demonstrate that the use of repolished substrates does not compromise the performance of inverted metamorphic solar cells.

Optimal Bandgap Combinations—Does Material Quality Matter?

The balance of photogeneration and recombination gives rise to an optimum bandgap for any solar cell. The radiative limit represents the lowest permissible level of recombination in a solar cell and, therefore, places an upper limit on the voltage that can be attained. Introducing additional nonradiative recombination results in a loss in voltage that can only be compensated for by moving to higher bandgaps. Consequently, the optimal bandgap for solar energy conversion will rise with increasing nonradiative recombination rate. This balance was recognized by Shockley and Queisser for single-junction solar cells and is here extended to multijunction solar cells. A rise in optimal bandgaps has been observed in simulated single-, double-, and triple-junction devices as nonradiative recombination increases. Optimal bandgaps between excellent and poor diode quality devices are shown to differ by 100s of meV under 1-sun illumination with both terrestrial and extraterrestrial spectra but exhibit no significant change at high concentration due to the dominance of the radiative component in the recombination dynamics.

Extending the 1-D Hovel Model for Coherent and Incoherent Back Reflections in Homojunction Solar Cells

In this paper we extend the analytical drift-diffusion model, or Hovel model, to model the electrical characteristics of solar cells incorporating a back mirror. We use a compact summation approach to derive modified optical generation functions in Homojunction solar cells, considering both coherent and incoherent reflections from the back reflector. These modified generation functions are then used to derive analytical formulae for the current-voltage characteristics of mirrored solar cells. We simulate the quantum efficiency of a simple GaAs np diode with a planar gold back reflector, and compare the results with the standard Hovel model using a generation function given by the Beer-Lambert law. Finally, we use the model to simulate the performance of a real GaAs solar cell device fabricated using an epitaxial-lift-off procedure, demonstrating excellent agreement between the simulated and measured characteristics.

Mobile Solar Power

The militarys need to reduce both fuel and battery resupply is a real-time requirement for increasing combat effectiveness and decreasing vulnerability. Mobile photovoltaics (PV) is a technology that can address these needs by leveraging emerging, flexible space PV technology. In this project, the development and production of a semirigid, lightweight, efficient solar blanket with the ability to mount on, or stow in, a backpack and recharge a high-capacity rechargeable lithium-ion battery was undertaken. The 19% efficient blanket consists of a 10 × 3 solar array of 20 cm2 and single-junction epitaxial lift-off solar cells, which have an efficiency of ∼22% under AM1.5G illumination. A power-conditioning module was also developed to interface the solar panel to the battery. Thirteen systems were outfitted during a Limited Objective Experiment-1 in February 2012, and based on the results, a second version of the system is in development.

Controlling radiative loss in quantum well solar cells

The inclusion of quantum well layers in a solar cell provides a means for extending the absorption and therefore increasing the photocurrent of the cell. In 2009, a single-junction GaAsP/InGaAs quantum well solar cell attained a peak efficiency of 28.3% under solar concentration. Since then InGaP/MQW/Ge quantum well devices have attained efficiencies in excess of 40% under concentration and over 30% under AM0. The principle motivation for incorporating a quantum well stack into a multi-junction solar cell is to increase the photocurrent delivered by the middle junction over the conventional In0.01GaAs bulk junction. This enables additional current to flow through the top and middle cells, resulting in a sharp rise in efficiency. However, quantum wells also provide some freedom to manipulate the radiative recombination in the quantum well solar cell. We show that under radiatively dominated, anisotropic emission, strong radiative coupling between sub-cells takes place, resulting in a multi-junction solar cell that is tolerant to daily and seasonal changes to the solar spectrum.

Higher limiting efficiencies for nanostructured solar cells

It is possible to tailor the band gap of the strain-balanced quantum well solar cell to match the local solar spectral conditions by altering the quantum well depth. This has led to a recent single-junction world-record efficiency of 28.3%, as well as giving advantages for current matching in multi-junction solar cells. Radiative recombination is the dominant loss mechanism for the strain-balanced quantum well solar cell, so practical improvements focus on techniques for light management in the cell, such as enhancing the optical path length with epitaxial mirrors. Furthermore, the compressive strain in the quantum wells suppresses emission into TM-propagating modes, reducing the overall optical loss and increasing the cell efficiency. As biaxial strain can only be engineered into a cell on the nanoscale, quantum well solar cells are seen to have a fundamental efficiency advantage over bulk semiconductor cells.

Towards high efficiency multi-junction solar cells grown on InP Substrates

Progress toward the development of multi-junction solar cells grown on InP substrates is presented. In this material system, the optimal bandgaps for solar energy conversion are attained while the multi-junction structure is realized under lattice matched conditions. In this work, results for the characterization of material and devices of the individual sub cells are shown. For the top cell, InAlAsSb quaternary material is being developed. For the middle, InGaAsP and InGaAlAs are studied, and for the bottom, InGaAs will provide the possibility of adding multiple quantum wells for fine bandgap tunability. In addition, we will discuss electrical characterization of the tunnel diodes.

Physics of quantum well solar cells

Incorporating quantum wells into multi-junction III-V solar cells provides a means of adjusting the absorption edge of the component junctions. Further, by using alternating compressive and tensile materials, a strain-balanced stack of quantum well and barrier layers can be grown, defect free, providing absorption-edge / lattice parameter combinations that are inaccessible using bulk materials. Incomplete absorption in the quantum wells has been addressed using a distributed Bragg reflector, extending the optical path length through the cell and enabling photon recycling to take place. State of the art single-junction quantum well solar cells have now reached an efficiency of 27.3% under 500X solar concentration and are projected to reach 34% in a double junction configuration.