Tin and zinc sprinkled dream

Thin film solar cells are being touted in the market and achieving relative successes in the industry in terms of price competitiveness and efficiency. Amorphous silicon, cadmium telluride (CdTe), copper-inidum-gallium-selenide or CIGS and dye-sensitized solar cells are gaining momentum to compete with the regular heavyweight of the solar world, crystalline silicon (c-Si). Some very established thin film leaders like First Solar have found success with their CdTe modules. Enter IBM with their announcement that their experiments with CZTS have yielded a record 9.6 percent efficiency. Interest in CZTS, a composition with selenide (Cu2ZnSnSe4) or sulfide (Cu2ZnSnS4) or an alloy of the two, has been alive for over a decade. Mike Scarpulla, Assistant Professor at the University of Utah’s Department of Electrical and Computer Engineering explains, “Research in CZTS for thin film PV has been significantly advanced by a group in Nagaoka, Japan.” Interesting is how these CZTS researches came about. What spurred examination of tin and zinc as substitutes?
“I was interested in new materials for photovoltaics that would give high performance at low cost and allow the industry to scale up to a significant percentage of our energy demand. The challenge is to find metals to pair with sulfur, an abundant and cheap element,” Scarpulla says. Similar interest was also sparked in Susanne Siebentritt, Professor at the University of Luxembourg whose research interest with the Laboratory of Photovoltaics (LPV) focuses on the development of new structures and processes for the preparation of thin film solar cells and the fundamental materials physics of novel semiconductors used as absorbers. Siebentritt answers, “A couple of years ago, there was a story in the middle of the silicon shortage saying that there would be a shortage for other materials too. This made me think about alternative materials. Together with my colleagues, we wanted to start looking at kesterites.” There are scores of research articles that have been steadily released with insights into CZTS but as mentioned before, it was rather low-key. The IBM breakthrough placed CZTS in the spotlight. IBM’s scientists started playing around with a few different materials. Supratik Guha, Director of Physical Sciences at IBM Research explains how the firm picked up CZTS, which was achieving reported efficiencies of 6.8 percent at that time and turned that number into a more exciting 9.6.

Efficiency

Guha tells, “We had a different way of making it via solution process rather than sputtering. Pretty soon, we found that we ended up with cells of 9.75 percent. That exceeded the world record at that time by about 50 percent and we got pretty excited. And when the efficiency starts reaching ten percent, the interest in that material begins to pick up.” True enough. Japanese thin film producer Solar Frontier soon announced a collaboration with IBM to take CZTS development further. Solar Frontier’s spokesperson tells pv magazine, “This joint research will couple with IBM’s research into CZTS materials and devices with Solar Frontier’s experience and know-how in developing CIS-based thin film solar cells to be a cost competitive solar technology that is inexpensive.”
Staying on the topic of efficiency, how much higher can it get for CZTS cells? Can CZTS dethrone the leaders like CIGS or CdTe for that matter? “At this point, it is not that CZTS is more efficient than the other thin film materials. The benefit of CZTS would come in terms of scalability of thin film technology,” says Scarpulla. He adds, “Now demonstrations in the six to seven and nine to ten percent ranges have been made by a few processing routes for CZTS and CZT(S,Se) cells, respectively.” The hype around the new absorber keeps demanding higher efficiencies in a bid to knock out competition that is often classified as being too toxic or too expensive.
CZTS itself has a similar crystal structure to that of chalcopyrites (CIS for example). Siebentritt elaborates, “Imagine copper-indium-selenide or copper-indium-sulfide, and replace half of the indium with zinc and the other half with tin. Then, you get the CZTS crystal structure.” The optoelectronic properties of CZTS do not seem to vary very much from that of CIGS as well, showing similar light absorption properties. However, as Scarpulla differentiates, CZTS does offer a potential benefit in that it has a higher band gap (that is similar to that of CdTe), meaning higher output voltages compared to CIGS or c-Si solar cells. The absorption properties of CZTS also seem to look good as Siebentritt emphasizes. She adds that there are a couple of experimental values and the absorption coefficient is in the same range as CIS cells. Promising prospects!

Cost and supply drivers

Costs need to be driven down in order for photovoltaic technology to become competitive with the other conventional forms. Indium and gallium, the heavyweights are not cheap. Indium demand is estimated to grow in the next few years, ten-fold till 2013 as Indium Corporation in the U.S. predicts. The prices of indium and gallium went up in 2010, not helping the cause to drive costs down in PV. The potential substitutes zinc and tin now hold the promise of doing just that. CZTS has the potential to be cost-competitive. As Solar Frontier says, “CZTS technology utilizes only abundant and cheaper metals and will lead to lower production costs. Performance and production yields for CZTS, in large scale production, can be as competitive as CIS.” The company also adds that the zinc and tin supply are more abundant compared to indium and gallium making the supply situation more stable. Copper and sulfur are relatively abundant. One key step is to replace the selenium in the record efficiency cells (a S-Se alloy) with sulfur as selenium is a trace element.
Scarpulla addresses a caution. He says that there is some intellectual dishonesty in motivating work on CZTS solely because it is ‘cheap.’ “From a standpoint of cost of the raw elements, according to publicly-available information, CZTS is indeed cheaper than either CdTe or CIGS. However, the costs of the thin film materials are small compared to the other costs like the glass, aluminum frames and others,” he asserts. Guha seems to agree with this. He admits that CZTS materials would be cheaper but emphasizes that the main advantage is that CZTS is a material that can be scalable for the future.

Deposition methods

Thin film deposition sees a myriad of possibilities. Instead of sputtering, IBM used solution processing for its CZTS deposition. Thereafter, the company used vacuum deposition technology and recently, with vacuum deposited CZTS with a pure sulfide phase; Guha says that the efficiencies in the lab are up to 7.1 percent. “This is higher than any published results for sputttered or vacuum deposited CZTS,” Guha explains. The technique IBM is trying to pursue is solution processing that does not require the vacuum or pumps and all the pomp and circumstance of sputtering (see pv magazine, 10/2010). From IBM’s standpoint, the biggest advantage of this method is that material recovery would be higher: in the range of 40 to 60 percent, as Guha points out. Most materials via evaporation and sputtering get lost and recycling is a rather difficult process if utilizing these processes. “The other advantage is that the effective deposition rates can be much higher. With vacuum deposition, you are depositing at a rate of less than ten microns an hour. With solution processing, with one sweep, one can have a big fraction of a layer laid down. One would expect the throughput to be higher,” adds Guha.
Of course, solution processing is not the holy grail yet for CZTS. Being at the infant research stages, other parties with interest in the material are going the sputtering way like Scarpulla with the approach of sputtering CZTS layers from binary sulfide targets. He states that this method has the potential in terms of preventing undesired phase formation and promoting good film microstructure.
Siebentritt’s lab in Luxembourg takes the approach two different ways. “One way is co-evaporation but it needs a second annealing step and the other way is that we prepare a precursor by galvanic deposition, electrochemical deposition and then it is annealed in a selenium atmosphere,” explains Siebentritt. When asked if sputtering fits just as well to CZTS deposition she says that it is plausible but that it does not seem to be as tolerant with the new materials.

Greener and leaner

It would be rather easy here to fight the case for CZTS. There are no ‘heavy’ metals in the combination as Siebentritt points out. “Selenium has a certain toxicity. If we do not use the selenide compound but the sulfide compound, then there are no issues of toxicity.” From the supply side as well, this will make sense for cost savings as mentioned before. Her research at the moment uses a cadmium-sulfide buffer but she asserts that they are in the development phase and there will be no cadmium involvement in the future.
Cadmium. An element that has caused a lot of buzz, after all the lobbying to extend the ban of the substance into the PV world. Fair enough, since cadmium is a toxic substance if ingested, inhaled or mishandled. Scarpulla has a distinct take on the matter of toxicity of cadmium telluride. He says that when the component elements are chemically bonded together and made into solar panels, they are in a stable form.
As he tells pv magazine, “The modules are encapsulated in glass and plastic. So, long story short, they are not going anywhere unless they are dissolved by chemicals (i.e. by being leached by acid in a landfill) or burned in a fire. First Solar has set one of the best examples of corporate responsibility I have ever heard of by guaranteeing full recycling of their panels via a return deposit on the panels held in a trust fund – they have planned to keep the CdTe out of landfills even if the company goes under in 20 years.” Scarpulla adds that the place where toxicity can be most worrisome would be in the mining and refining of the thin film solar cells’ constituent elements and in the fabrication of the cells. It is here that CZTS could offer benefits to companies producing panels and to our environment.
Solar Frontier believes in the environmental friendliness of this composition. The company states, “CZTS technology will only use environmentally friendly materials and will not use cadmium or lead, just like in the case of our CIS modules. This makes it suited for collection and recycling in the future.”
Looking at the elements from a purely chemical point of view, cadmium does rank high in the Agency for Toxic Substances and Disease Registry. It takes up seventh position of hazardousness in the directory of 2007, with Zinc placed 74th. Lenntech, a water treatment solutions company, states that tin as single atoms or molecules are not very toxic. Taking into account what Scarpulla and Siebentritt have said, it is heartening to hear that there is an alternative absorber being developed that does not carry as much debatable weight as other elements in a fundamental sense.

Challenges

CZTS has some hurdles to cross. Nevertheless, Guha states that he expects CZTS based solar cells to be competitive within a few years. IBM itself does not have the intention to manufacture the solar cells but they are working together with Solar Frontier and Taiwanese DelSolar to develop practical solutions with high efficiencies. Tokyo Ohka Kogyo or TOK, a Japanese company, is also cooperating with IBM to develop the solution processing tool for CZTS and precursor chemistry. “We are taking this very seriously and aim to turn CZTS into a viable technology,” emphasizes a confident Guha.
Challenges, as exist for every new development, still need to be overcome. Scarpulla mentions that because there are four elements involved in CZTS and semiconductors are in principle sensitive to parts-per-billion composition variations, composition control is, not surprisingly, one of their biggest challenges. He adds, “Getting the mixture of copper, zinc, tin and sulfur correct and keeping it correct during the entire process is not trivial. However, one of the things that continues to amaze me about CZTS is that record and near-record cells have imbalances of some of the elements in the ten to 20 percent range. For a semiconductor, this is just amazing that it works at all.” Siebentritt believes that understanding the semiconductor physics of CZTS is an important aspect apart from developing the solar cells themselves.
“I do not have a crystal ball. If things go well, we could have lab efficiencies of 15 percent in a couple of years,” concludes Siebentritt. The interest is peaking. Scarpulla points out that the 2011 Photovoltaic Specialists Conference has a topical area for new and emerging photovoltaic materials like CZTS. Predictions are being made, cautiously though, so as to avoid sweeping declarations. What is certain is that the upcoming CZTS is now definitely a feature in the solar industry’s radar screen.