Silicon based multi-junction solar cells are a promising option to overcome the theoretical efficiency limit of a silicon solar cell (29.4%). With III-V semiconductors, high bandgap materials applicable for top cells are available. For the application of such silicon based multi-junction devices, a full integration of all solar cell layers in one 2-terminal device is of great advantage. We realized a triple-junction device by wafer-bonding two III-V-based top cells onto the silicon bottom cell. However, in such a series connected solar cell system, the currents of all sub-cells need to be matched in order to achieve highest efficiencies. To fulfil the current matching condition and maximise the power output, photonic structures were investigated. The reference system without photonic structures, a triple-junction cell with identical GaInP/GaAs top cells, suffered from a current limitation by the weakly absorbing indirect semiconductor silicon bottom cell. Therefore rear side diffraction gratings manufactured by nanoimprint lithography were implemented to trap the infrared light and boost the solar cell current by more than 1 mA/cm<sup>2</sup>. Since planar passivated surfaces with an additional photonic structure (i.e. electrically planar but optically structured) were used, the optical gain could be realized without deterioration of the electrical cell properties, leading to a strong efficiency increase of 1.9% absolute. With this technology, an efficiency of 33.3% could be achieved.
Modeling single junction solar cells composed of III–V semiconductors such as GaAs with the effects of photon recycling yields insight into design and material criteria required for high efficiencies. For a thin-film single junction GaAs cell to reach 28.5% efficiency, simulation results using a recently developed model which accounts for photon recycling indicate that Shockley–Read–Hall (SRH) lifetimes of electrons and holes must be longer than 3 and 1 μs, respectively, in a 2-μm thin active region, and that the native substrate must be removed such that the cell is coupled to a highly reflective rear-side mirror. The model is generalized to account for luminescence coupling in tandem devices, which yields direct insight into the top cell’s nonradiative lifetimes. A heavily current mismatched GaAs/GaAs tandem device is simulated and measured experimentally as a function of concentration between 3 and 100 suns. The luminescence coupling increases from 14% to 33% experimentally, whereas the model requires increasing electron and hole SRH lifetimes to explain these results. This could be an indication of the saturating defects which mediate the SRH process. However, intermediate GaAs layers between the two subcells may also contribute to the luminescence coupling as a function of concentration.
Single junction photovoltaic devices composed of direct bandgap III-V semiconductors such as GaAs can exploit
the effects of photon recycling to achieve record-high open circuit voltages. Modeling such devices yields insight into the design
and material criteria required to achieve high efficiencies. For a GaAs cell to reach 28 % efficiency without a substrate, the
Shockley-Read-Hall (SRH) lifetimes of the electrons and holes must be longer than 3 s and 100 ns respectively in a 2 μm thin
active region coupled to a very high reflective (>99%) rear-side mirror. The model is generalized to account for luminescence
coupling in tandem devices, which yields direct insight into the top cell’s non-radiative lifetimes. A heavily current
mismatched GaAs/GaAs tandem device is simulated and measured experimentally as a function of concentration between 3
and 100 suns. The luminescence coupling increases from 14 % to 33 % experimentally, whereas the model requires an
increasing SRH lifetime for both electrons and holes to explain these experimental results. However, intermediate absorbing
GaAs layers between the two sub-cells may also increasingly contribute to the luminescence coupling as a function of
The hot carrier solar cell (HCSC) offers one route to high efficiency solar energy conversion and has similar
fundamental limiting efficiency to multi-junction (MJ) solar cells however, the HCSC is at a much earlier stage
of development. We discuss the unique features of the HCSC which distinguish it from other PV technologies,
providing motivation for development.
We consider the potential for a low concentration hot-carrier enhanced single-junction solar cell, enabled
by field enhancing cell architectures. To support this we experimentally show that changing sample geometry
to increase carrier density, while keeping phononic and electronic properties constant, substantially reduces
hot-carrier themalization coefficient. Such a scheme might have similar applications to todays high efficiency
single-junction devices while allowing from some intrinsic efficiency enhancement.
We also use spectral data simulated using SMARTS to identify HCSC spectral insensitivity relative to MJ
devices. Spectral insensitivity increases annual energy yield relative to laboratory test efficiency, reducing the
cost of PV power generation. There are also several practical advantages: a single device design will operate
optimally in a variety of locations and solar power stations are less reliant of accurate, long-range atmospheric
simulation to achieve energy yield targets.
Tunnel diodes constitute an essential part of multi-junction concentrator photovoltaics. These tunnel junctions exhibit a
transition from low-resistance tunneling to high-resistance thermal diffusion, commonly at current densities of the order
of 10<sup>2</sup>-10<sup>3</sup> mA/mm<sup>2</sup>. Experimental evidence of a fundamentally new effect is reported and confirmed in distinct cell
architectures: the dependence of the threshold current density on the extent of localized irradiation. It is also shown that
photovoltaic cells with a non-uniform metal grid can possess an additional spatial dependence to the threshold current
density. These new phenomena should be observable in all solar cell tunnel diodes subjected to inhomogeneous
illumination, and are posited to stem from the lateral spreading of excess majority carriers (similar to current spreading
in LEDs). The implications for concentrator solar cells are also addressed.
Conference Committee Involvement (1)
High and Low Concentration for Solar Electric Applications
14 August 2006 | San Diego, California, United States