| Technology Update|
Simulating Solar Cells with TCAD
Ric Borges, Product Marketing Manager, Synopsys, reviews a range of photovoltaic technologies and explains how TCAD simulation can help engineers produce more efficient solar cells.
History of TCAD in the Solar Arena
There is nothing new about using TCAD simulation software for solar cell development. As soon as it became available in the mid-1980s, solar cell engineers recognized its potential for optimizing their designs and fabrication sequences. There was a synergy between those initial tools, which naturally were optimized for silicon, and the solar cells of the time, many of which were silicon-based.
The development of solar cells progressed fairly well through the 1990s, providing the only way to power satellites and finding use in some terrestrial applications too. Solar cells even attracted a residential market: despite the cost, some people felt compelled to have green technology powering their homes.
As the technology advanced, the industry developed and tested various alternatives to expensive crystalline silicon cells. These new ‘thin-film’ materials, which are ‘amorphous’ rather than crystalline, include cadmium telluride-based materials, CIGS (copper indium gallium diselenide)-based materials and amorphous silicon.
In recent years, many new and established companies have been stepping up their research into solar technology in the light of widespread discussion of climate change, helped by government tax credits in some countries, which has also helped to increase popular awareness of solar technology.
Solar Cells: Where Are We Now?
These companies aim to address two key market needs. One is conversion efficiency; the other is cost. As one might expect, bringing cells to market involves a trade-off: the most efficient cells are also the most expensive, and the cells that are cheapest to mass-produce are the also the least efficient. So, how can we use TCAD tools to address the issues raised by each competing technology?
At the higher-efficiency end of the market is the ‘multi-junction’ solar cell, an expensive, specialized type of cell used for satellites. A multi-junction solar cell is arranged like a sandwich, comprising a stack of 'sub-cells', each of which is effective in capturing a particular range of the solar spectrum.
This type of cell now has over 40% efficiency, and by considering alternative ways to construct a sandwich, there is scope to improve this further still. TCAD helps designers to understand exactly what thicknesses and materials to use, how they heat up, and many other aspects of device physics and surface interactions (Figure 1), in order to achieve the best possible performance.
Figure 1. Design considerations for a generic solar cell
Crystalline silicon solar cells have been around for a long time. They are expensive because they rely on the same silicon substrates as the rest of the silicon industry, which sometimes are in short supply. Their fabrication depends on a lot of the same complex techniques as other silicon-based semiconductor devices.
On the positive side, crystalline silicon solar cells have the second-highest efficiency of any class of solar cells: around 21-22%. Even a half-percentage point efficiency gain is worthwhile, so companies working on these cells are using TCAD to gain insight into the internal workings of their cell designs: how the light is collected and converted into electrical current, how the electrical current begins to flow in the cell and how it is pulled out of the terminals.
This insight then helps to inform ways to refine the structure further. The solution might be to change the composition of some of the layers, for example, or to reduce cost by using less silicon. Since traditional cells have one contact on the top and another on the bottom, one of which partially blocks the light, another way to improve a cell is to find ways to contact the cell from one side only – for example, by drilling holes through the cells with lasers. This allows more light into the cell and also makes the assembly process easier.
Multi-crystalline silicon solar cells are related to crystalline solar cells. The difference is that instead of just using crystalline silicon, the starting material is crushed silicon: grains of crystalline silicon embedded in an ingot and then sliced – almost like making a meat loaf. Although these solar cells aren’t as efficient as the fully crystalline cells, they do have a cost advantage.
Companies manufacturing multi-crystalline cells can use TCAD simulation in much the same way as they would with crystalline cells: many of the same types of operations for optimizing the cells apply, although there are some additional complications where the crystals touch each other or become discontinuous.
One of the materials used for 'thin-film' cells is amorphous silicon, which is electrically very different from crystalline silicon. It can be deposited on very large substrates, which means you can use manufacturing technology to yield large-area cells.
There is still a lot of research to do in order to improve this type of cell: its conversion efficiencies are currently low – from single digits to low teens – and long-term exposure to light and real-life conditions can cause it to degrade, though significant progress is being made. TCAD tools allow manufacturers to do a lot of useful simulations but their predictive nature is not as great as in the crystalline types or the multi-junction types – because amorphous silicon is structurally more complex, it takes a lot of detailed material characterization to understand its electrical characteristics for the modeling to work well.
Cells based on other amorphous materials – CIGS and cadmium telluride – are currently in production, and there is still a lot of work underway to improve their efficiency. Again, this involves working more at the material level. You can apply TCAD to understand the limitations of the cell designs with respect to the impurities or defects in these materials, and what potential the material would have were it not for these defects.
Organic solar cells are made of polymers. They are at the lower end of the efficiency curve, so they aren’t particularly interesting for residential or industrial applications. However, their low cost of manufacture and flexible form make them suitable for campers: someone might like to unfurl an organic cell when they need to charge a cell phone, for example, or they might wish to use one to cover a tent, giving them a power supply in a remote location.
The polymers are very disordered materials and don’t lend themselves to having their electrical properties analyzed. The role for simulation software is currently limited but new research into the electrical transport mechanisms in these materials offers the prospect of predictive simulation in the future.
All of the types of solar cell mentioned above have something in common: a layer of semiconductor material that absorbs light to create carriers – electrons and holes – that are collected as electrical current, which flows out through terminals.
Now a completely different concept of solar cell is emerging. ‘Third-generation’ cells rely on engineered semiconductor structures that are more complex, such as quantum wells and quantum dots. One design objective is to get cells to absorb more of the visible spectrum by changing the characteristics of the material and the way in which the cell absorbs light (Figure 2). Another is to increase the internal absorption efficiency, so that photons absorbed in the cell will generate multiple electrons and multiple holes instead of just one electron and one hole.
Figure 2. Understanding how optical generation changes with wavelength of light
Simulation is interesting to third-generation cell researchers because it allows them to experiment without taking risks: they can explore designs and see if they will work without having to build anything.
Analyzing Solar Cells
It is useful to look at some specific examples of where TCAD has helped to make cells more efficient and less expensive.
First of all, we have shown in several cases that you can use tools to simulate a processing technique called ‘gettering’, which removes metallic and other impurities from the areas of silicon where the light is collected. Design teams want to do this to make crystalline silicon cells more efficient. When the light comes into a cell, it generates charge carriers; removing impurities increases the chances that these carriers will hang around long enough to get to the terminal and be collected as current, rather than recombining and dissipating. Without the insight that TCAD offers, designers would have to conduct many different experiments and risk never arriving at the optimal solution.
TCAD is also helping design teams to understand the optimal shape for ‘texturing’ the surface of crystalline silicon cells. A cell with a plain surface reflects about 30% of the light that hits it. In order to increase the number of photons going into cells, teams are trying to roughen surfaces, sometimes by etching micro-pyramids into the material.
Simulation is also useful when preparing to mass-produce a lab-proven cell. To achieve good yield, the process developed for a nominal design in a research lab must be controlled and made repeatable. TCAD helps to determine which process parameters need to be controlled very tightly, and which are less critical, which in turn helps the manufacturer to specify the equipment it needs to buy. Statistical analysis allows designers to understand the effect of varying cell thicknesses and other parameters on the performance of the cell, something that it would be costly – and in some cases impossible – to achieve experimentally.
Ric Borges is a product marketing manager for Technology CAD (TCAD) at Synopsys. Ric has 20 years of experience in the semiconductor industry, including engineering and management roles in technology development and circuit design. His expertise ranges across many semiconductor technologies (CMOS, SiGe, GaAs, GaN, SiC) and high frequency circuit design. Prior to joining Synopsys in 2005, Ric was director of device engineering at Nitronex Corporation, a wireless chip company addressing the 3G and Wimax base-station markets.
©2010 Synopsys, Inc. Synopsys and the Synopsys logo are registered trademarks of Synopsys, Inc. All other company and product names mentioned herein may be trademarks or registered trademarks of their respective owners and should be treated as such.