Decades after his death, the name of Ernst Werner von Siemens, the 19th century German inventor, is still widely referred to. Companies in the solar photovoltaic (PV) supply chain depend on him all the time when they take quartz to make silicon in a manufacturing procedure known as the Siemens process.

But as the solar PV supply chain looks to address costs, if silicon producers could alter or cut down the Siemens process in some way, they could save themselves large sums of money.

One quartz mining company in Australia thinks it may have found a way of doing just that. Down in the far South, about 80 miles Northwest of Melbourne, the company claims it may be able to build a plant to produce silicon from the waste taken from a gold mine. Its managers say that not only would the silicon produced be finer than most alternative sources, but it could clear a waste mountain at the same time.

“It [the quartz] has advantages other products can't match in terms of [low] boron and phosphorus content,” asserts Chris Karamountzos, the CEO of Creswick Quartz. It is also low in lithium content. He says that tests from a project completed in March 2010 have proved the quartz to have extraordinary characteristics: “Some of the product has 60 ppm in impurities, which is unheard of. A lot of other sources wouldn't be able to dream of our purity,” he says.

The raw material for the process would be the waste tailings from a 19th century gold mine. The mine, based near Smeaton in the state of Victoria, covers a licensed area of 325 km² although not all this area is in use. Gold is still available in the mine, but is difficult to extract, and gold mining ceased over a century ago. At the moment the waste tailings generate a supply of quartz (SiO₂) – a source for silicon – for the construction sector, and Creswick Quartz has been active in this area for the last 15 years.

Following tests however, the company says it is considering converting some of it to an unusually pure grade feedstock (at least 99.995% pure silicon, approaching the 99.999% purity needed for solar-grade silicon production) required by solar PV manufacturers. This is higher than the more commonly talked about metallurgical-grade silicon which has a purity of 98-99.5% largely because, in turn, the purity of the quartz feedstock is low.

Using the process it is developing, Creswick Quartz says it could supply high-quality, fine-particle zed (down to 10-20 microns) and lump (20-80 mm) quartz suitable for both ultra-high metallurgical and solar-grade silicon production.

Is a purer form of silicon viable?

The less pure forms of silicon are purified using the Siemens process. This means reacting the silicon with hydrochloric acid to produce a gas which then has to be purified. It is then decomposed in a 1000°C furnace, producing silicon deposits. The process can last up to a week and is expensive, energy intensive and difficult.

Scientists and managers involved in the Creswick project reason that, since Creswick Quartz's product is purer than most silicon that currently needs to be purified by the industry, it may be able to bypass this process and save energy and money at the same time.

If the company can bypass the Siemens process and also avoid the mining costs incurred by many other quartz suppliers, the potential benefits are clear: a reduction in mining waste from the area, in addition to a new supply of high quality raw material to the solar industry, generating lower costs and possibly lower carbon emissions than many competing products.

Dr Hal Aral, a scientist at the Australian Commonwealth Scientific and Industrial Research Organisation (CSIRO), the national government body for scientific research in Australia, has been working on an environmentally friendly quartz cleaning procedure which would be the first step in the company's new process. He explains: “removal of these tails puts the landscape [back] to its original form. This is because the removal process does not impact upon the surface by way of extraction at all.”

The quartz in the tails is processed by applying water wash which he says generates high purity quartz that can be below 100 ppm total impurities; this he describes as “a very green cleansing process.” For higher purity applications this can then involve caustic and acid leach treatment. The waste streams of caustic and acid treatment neutralise each other when mixed, and this means the process is environmentally more acceptable than direct acid leach. The neutralisation process, he asserts, does not generate much solid waste because the Creswick quartz contains a very small amount of impurities to start with.

However, Karamountzos cautiously states that “there are quite a number of different treatment options that we are evaluating.”

What do the experts think?

Jenny Chase, Manager of Solar Insight at Bloomberg New Energy Finance, expressed doubts about the project, pointing out examples of discontinued silicon production schemes of one kind or another: “Many companies have tried to develop ‘upgraded metallurgical’ silicon processes to turn metallurgical silicon (silica heated up with carbon so it reduces, adding a lot of carbon impurity in the process) as an alternative to the Siemens process.”

However, those companies, which include Timminco in Canada and Elkem in Norway, gave up when the silicon price fell in 2008 because conventional silicon is cheaper than it was, so there is less of a price differential: “To use the lower purity upgraded metallurgical silicon instead of Siemens silicon, manufacturers of wafers for solar cells had to adapt their own processing technologies, and now silicon can be had for US$55/kg, and they prefer to use the standard silicon and get more efficient cells,” she explains.

No decisions have been made by Creswick management as to whether to continue with the project, but they appear to be serious. “The issue is how to go about our business. Do we target the market for the highest possible end user or see if we have an abundant resource below ground?” muses Karamountzos.

A million tonnes of tailings are available, and the company is about to arrive at a turning point. If it continues to use the tailings for the broader category of quartz, it can sell to Asian customers in the construction sector, but estimates that it will run out of waste tailings above ground in a shorter space of time, perhaps three or four years. It will then need to drill some of the material underground and this will mean a major shift in strategy and higher costs. So it hopes to build a processing plant to produce ultra pure silicon as an alternative. This, says Karamountzos, would probably have a 30-year life span.

There would probably be plenty of demand if the project became viable, despite the recent lull in the business. The global capacity of silicon production is 200,000 mt, according to Solar Insight. However, Jenny Chase remarks: “a lot of it [50,000 mt] is three or four tier factories which might go out of business.” Demand for silicon, on the other, hand, is about 100,000 mt per annum, of which about a fifth comes from the semiconductor sector.

Nevertheless, the solar PV sector is growing quickly. Chase says she's expecting 14 GW of new build PV modules this year in comparison to 7.6 GW in 2009. “It's difficult to see the PV industry has any way to go but up…some new entrants are Chinese companies who have jumped on the bandwagon and some are Western companies that have gone bankrupt. I don't believe they've made a lot of silicon yet.”

Commenting on the potential of the Creswick project, Professor John Monhemius, a Mining Engineer at Imperial College, London University, drew attention to the abundance of quartz on this planet. However, he added: “This quartz is pretty pure so that's unusual. Ultrapure quartz has a certain market value.” He also pointed out: “pure quartz is not likely to be causing environmental problems and is pretty inert anyway. It's more about removing an eyesore.”

Other types of pure quartz are available, for example in Brazil. The big advantage, he indicated, is that the company does not have to mine it, thus lowering its costs. However, although Creswick would avoid the costs of mining the quartz, this may not be as much of a benefit as it first seems, because the mining cost is only a small proportion of the total silicon production costs.

It would also be likely to incur unusually high transportation costs due to its location in relation to most markets. Monhemius also pointed out the unique nature of the scheme, which may mean it could not be copied elsewhere. “Most mining operations are very specific to their own project. It may cause people to look at their mining waste again, but it's not likely to cause a revolution in the business.”

Renewable Energy Focus
Volume 11, Issue 5, September-October 2010, Pages 32-35