tag:theconversation.com,2011:/uk/topics/silicon-technology-6011/articlessilicon technology – The Conversation2023-09-10T20:05:59Ztag:theconversation.com,2011:article/2107232023-09-10T20:05:59Z2023-09-10T20:05:59ZSolar panel technology is set to be turbo-charged – but first, a few big roadblocks have to be cleared<figure><img src="https://images.theconversation.com/files/543607/original/file-20230821-255381-6ttzir.jpg?ixlib=rb-1.1.0&rect=9%2C0%2C6434%2C3939&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>Solar panel technology has made enormous progress in the last two decades. In fact, the most advanced silicon solar cells produced today are <a href="https://pubs.acs.org/doi/10.1021/acsenergylett.0c01790">about as good</a> as the technology will get.</p>
<p>So what’s next? Enter “tandem solar cells”, the new generation in solar technology. They can convert a much greater portion of sunlight into electricity than conventional solar cells.</p>
<p>The technology promises to fast-track the global transition away from polluting sources of energy generation such as coal and gas. But there’s a major catch.</p>
<p>As <a href="https://doi.org/10.1039/D3EE00952A">our new research</a> shows, current tandem solar cells must be redesigned if they’re to be manufactured at the scale required to become the climate-saving technology the planet needs. </p>
<h2>The solar story so far</h2>
<p>A solar cell is a device that turns sunlight into electricity. One important measure when it comes to solar cells is their efficiency – the proportion of sunlight they can convert into electricity.</p>
<p>Almost all solar panels we see today are made from “photovoltaic” silicon cells. When light hits the silicon cell, electrons inside it produce an electric current.</p>
<p>The first silicon photovoltaic cell, demonstrated in 1954 in the United States, had an <a href="https://www.science.org.au/curious/technology-future/solar-pv">efficiency of about 5%</a>. That means that for every unit of the Sun’s energy the cell received, 5% was turned into electricity.</p>
<p>But the technology has since developed. At the end of last year, <a href="https://www.longi.com/en/news/propelling-the-transformation/">Chinese solar manufacturer LONGi announced</a> a new world-record efficiency for silicon solar cells of 26.81%.</p>
<p>Silicon solar cells will never be able to convert 100% of the Sun’s energy into electricity. That’s mostly because an individual material can absorb only a limited proportion of the solar spectrum.</p>
<p>To help increase efficiency – and so continue to reduce the cost of solar electricity – new technology is needed. That’s where tandem solar cells come in.</p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1594581819847942144"}"></div></p>
<h2>A promising new leap</h2>
<p>Tandem solar cells use two different materials which absorb energy from the Sun together. In theory, it means the cell can absorb more of the solar spectrum – and so produce more electricity – than if just one material is used (such as silicon alone).</p>
<p>Using this approach, researchers overseas <a href="https://www.pv-magazine.com/2023/05/30/kaust-claims-33-7-efficiency-for-perovskite-silicon-tandem-solar-cell/">recently achieved</a> a tandem solar cell efficiency of 33.7%. They <a href="https://www.science.org/doi/abs/10.1126/science.adi6278">did this by</a> building a thin solar cell with a material called <a href="https://www.unsw.edu.au/engineering/our-schools/photovoltaic-and-renewable-energy-engineering/our-research/research-activities/perovskite-solar-cells">perovskite</a> directly on top of a traditional silicon solar cell. </p>
<p>Traditional silicon solar panels still dominate manufacturing. But leading solar manufacturers <a href="https://www.pv-tech.org/qcells-to-invest-us100-million-in-perovskite-tandem-production-line/">have signalled plans</a> to commercialise the tandem cell technology.</p>
<p>Such is the potential of tandem solar cells, they are <a href="https://www.nature.com/articles/nenergy201515">poised to overtake</a> the conventional technology in coming decades. But the expansion will be thwarted, unless the technology is redesigned with new, more abundant materials.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/is-it-worth-investing-in-a-battery-for-your-rooftop-solar-heres-what-buyers-need-to-know-but-often-cant-find-out-209219">Is it worth investing in a battery for your rooftop solar? Here's what buyers need to know (but often can't find out)</a>
</strong>
</em>
</p>
<hr>
<figure class="align-center ">
<img alt="automated solar cell production line" src="https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=338&fit=crop&dpr=1 600w, https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=338&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=338&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=424&fit=crop&dpr=1 754w, https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=424&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/543609/original/file-20230821-225972-z8sj6l.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=424&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Tandem solar cells cannot overtake existing technology (pictured) unless they are redesigned.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<h2>The problem of materials</h2>
<p>Almost all tandem solar cells involve a design known as “silicon heterojunction”. Solar cells made in this way normally require more silver, and more of the chemical element indium, than other solar cell designs. </p>
<p>But silver and indium are <a href="https://pubs.rsc.org/en/content/articlehtml/2021/ee/d1ee01814k">scarce materials</a>. </p>
<p>Silver is used in thousands of applications, including manufacturing, making it highly sought after. In fact, global demand for silver <a href="https://www.reuters.com/markets/commodities/record-demand-pushes-silver-into-new-era-deficits-silver-institute-says-2023-04-19/">reportedly rose by 18%</a> last year.</p>
<p>Likewise, <a href="https://www.sydney.edu.au/news-opinion/news/2021/07/20/touchscreen-alternative-allays-fear-of-world-indium-shortage.html">indium is used</a> to make touchscreens and other smart devices. But it’s extremely rare and only found in tiny traces.</p>
<p>This scarcity isn’t a problem for tandem solar technology yet, because it hasn’t yet been produced in large volumes. But our research shows this scarcity could limit the ability of manufacturers to ramp up production volumes in future.</p>
<p>This may represent a substantial roadblock in tackling climate change. By mid-century, the world must install <a href="https://www.science.org/doi/abs/10.1126/science.adf6957">62 times more solar power capacity</a> than is currently built, to enable the clean energy shift. </p>
<p>Clearly, a major redesign of tandem solar cells is urgently needed to enable this exponential acceleration of solar deployment.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/how-to-maximise-savings-from-your-home-solar-system-and-slash-your-power-bills-197415">How to maximise savings from your home solar system and slash your power bills</a>
</strong>
</em>
</p>
<hr>
<figure class="align-center ">
<img alt="lumps of silver" src="https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=503&fit=crop&dpr=1 754w, https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=503&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/543610/original/file-20230821-218096-jqcbca.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=503&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Silver is a key component in much electronics manufacturing.</span>
<span class="attribution"><span class="source">Shutterstock</span></span>
</figcaption>
</figure>
<h2>Ramping up the transition</h2>
<p>Some silicon solar cells don’t use indium and require only a small amount of silver. Research and development is urgently needed to make these cells compatible with tandem technology. Thankfully, this work has <a href="https://onlinelibrary.wiley.com/doi/full/10.1002/aenm.202200821">already begun</a> – but more is needed.</p>
<p>A scarcity of materials is not the only barrier to overcome. Tandem solar cells must also be made more durable. Solar panels we see everywhere today are <a href="https://www.cleanenergyreviews.info/solar-panel-warranty">generally guaranteed</a> to produce a decent amount of electricity for at least 25 years. Perovskite-on-silicon tandem cells <a href="https://www.nature.com/articles/s41578-022-00521-1">don’t last as long</a>.</p>
<p>Solar power has already shaken up electricity generation in Australia and around the world. But in the race to tackle climate change, this is only the beginning. </p>
<p>Tandem solar cell research is truly global, conducted within a range of countries, including Australia. The technology offers a promising way forward. But the materials used to make them must be urgently reconsidered.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/despairing-about-climate-change-these-4-charts-on-the-unstoppable-growth-of-solar-may-change-your-mind-204901">Despairing about climate change? These 4 charts on the unstoppable growth of solar may change your mind</a>
</strong>
</em>
</p>
<hr>
<img src="https://counter.theconversation.com/content/210723/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Bruno Vicari Stefani receives funding from the CSIRO Research Office. </span></em></p><p class="fine-print"><em><span>Matthew Wright receives funding from UK Research and Innovation. </span></em></p>Tandem solar cells promise to revolutionise the clean energy transition – but a shortage of materials means they must urgently be redesigned.Bruno Vicari Stefani, CERC Fellow, Solar Technologies, CSIROMatthew Wright, Postdoctoral Researcher in Photovoltaic Engineering, University of OxfordLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1231152019-09-12T12:52:23Z2019-09-12T12:52:23ZQuantum computers could arrive sooner if we build them with traditional silicon technology<figure><img src="https://images.theconversation.com/files/292222/original/file-20190912-190002-1qou69y.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-illustration/3d-render-quantum-processor-1015677370?src=kwTkio1WBdTyeFGIocGfCQ-1-2">Amin Van/Shutterstoc</a></span></figcaption></figure><p><a href="https://theconversation.com/explainer-quantum-computation-and-communication-technology-7892">Quantum computers</a> have the potential to revolutionise the way we solve hard computing problems, from creating <a href="https://theconversation.com/quantum-computer-were-planning-to-create-one-that-acts-like-a-brain-108716">advanced artificial intelligence</a> to simulating chemical reactions in order to create the next generation of <a href="https://www.technologyreview.com/s/603794/chemists-are-first-in-line-for-quantum-computings-benefits/">materials or drugs</a>. But actually building such machines is very difficult because they involve <a href="https://www.technologyreview.com/s/612760/quantum-computers-component-shortage/">exotic components</a> and have to be kept in highly controlled environments. And the ones we have so far can’t outperform traditional machines as yet. </p>
<p>But with a team of researchers from the UK and France, <a href="https://www.nature.com/articles/s41928-019-0259-5">we have demonstrated</a> that it may well be possible to build a quantum computer from conventional silicon-based electronic components. This could pave the way for large-scale manufacturing of quantum computers much sooner than might otherwise be possible.</p>
<p>The theoretical superior power of quantum computers derives from the laws of nanoscale or <a href="https://theconversation.com/explainer-quantum-physics-570">“quantum” physics</a>. Unlike conventional computers, which store information in binary bits that can be either “0” or “1”, quantum computers use quantum bits (or qubits) that could be in a combination of “0” and “1” at the same time. This is because quantum physics allows particles to be in different states or places simultaneously.</p>
<p>Quantum computer development is still in its infancy and several hardware technologies are available without any single one yet dominating. The most advanced prototypes are currently made from either a few dozen <a href="https://theconversation.com/compute-this-the-quantum-future-is-crystal-clear-6671">ions trapped in a vacuum chamber</a> or <a href="https://theconversation.com/nevens-law-why-it-might-be-too-soon-for-a-moores-law-for-quantum-computers-120706">superconducting circuits</a> kept at near-absolute-zero temperature.</p>
<p>The crucial challenge is scaling up these small demonstrators into large interconnected qubit systems that will have enough computing power to perform useful tasks faster than classical supercomputers. To this end, another technology may eventually turn out to be more suitable. Strikingly enough, this could be the very same technology that today enables our digital society, the silicon transistor, the basic unit of information present in all microprocessors and memory chips.</p>
<p>There are two main reasons why making a quantum computer out of silicon has an aura of great interest around it. First, the <a href="https://theconversation.com/moores-law-is-50-years-old-but-will-it-continue-44511">Moore’s Law</a>-led relentless miniaturisation of silicon devices has enabled the manufacturing of transistors that are only a few tens of atoms wide. This is the scale at which the laws of quantum physics start to apply.</p>
<p>This represents a physical limit that has brought any further miniaturisation of silicon transistors to a halt. But it has also promoted new uses of silicon technology, known as <a href="https://spectrum.ieee.org/video/semiconductors/nanotechnology/how-will-we-go-beyond-moores-law-experts-weigh-in">More-than-Moore electronics</a>. Chief among these new directions is the possibility of encoding a <a href="https://theconversation.com/quantum-computing-poised-for-new-silicon-revolution-32800">quantum bit of information in each silicon transistor</a>, and then using them to build large-scale quantum computers.</p>
<p>By reusing the same technology that the microchip industry has handled for the past 60 years, we could also take advantage of previous multi-billion-dollar infrastructural investments and reduce costs. This means that all the clever engineering and processing that went into the development of modern microelectronics could be adapted to build increasingly powerful quantum processors.</p>
<h2>Silicon quantum chip</h2>
<p>The experiments recently carried out by our collaborating teams at Cambridge University, Hitachi R&D, University College London and CEA-LETI in France, and published in <a href="https://www.nature.com/articles/s41928-019-0259-5">Nature Electronics</a> suggest that this marriage between conventional and quantum electronics can be indeed celebrated. We took engineering solutions from conventional silicon circuits and applied them to interconnect different quantum devices on a chip. This has brought the practical realisation of quantum processors one step closer.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/te_tdvKaMYo?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>We have developed a circuit that operates at near-absolute-zero temperature and employs all commercial transistors. Some of these are so small that they can be used as qubits, whereas others are slightly larger and can be used to connect to different qubits. This architecture is remarkably similar to the one used for random access memory (RAM) in today’s laptops and smartphones.</p>
<p>In the past half a century or so, ordinary computers evolved from room-sized cabinets full of vacuum tubes to today’s hand-held microchip-based devices. There is still a long way to go before a fully-fledged quantum computer becomes available, but history may well repeat itself. The current progress of research suggests that initial quantum processors may be realised with some exotic technology first. But now that we have learnt that silicon can be used to efficiently interconnect qubits, the quantum future could be made of silicon.</p><img src="https://counter.theconversation.com/content/123115/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Alessandro Rossi receives funding from the UKRI Industrial Strategy Challenge Fund through the Measurement Fellowship Scheme at the National Physical Laboratory. He also holds a Chancellor's Fellowship at the University of Strathclyde.</span></em></p><p class="fine-print"><em><span>M. Fernando Gonzalez-Zalba receives funding from the European Commission H2020 Programme, the Royal Society and the Winton Programme for the Physics of Sustainability . </span></em></p>Manufacturing quantum computers would be a lot easier with existing technology than the exotic components currently used to build them.Alessandro Rossi, Chancellor's Fellow, Department of Physics, University of Strathclyde M. Fernando Gonzalez-Zalba, Honorary Research Fellow, University of CambridgeLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/689032016-11-29T02:25:35Z2016-11-29T02:25:35ZThe future of electronics is light<figure><img src="https://images.theconversation.com/files/147248/original/image-20161123-19717-hrc3xx.png?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">A basic design of a light-based chip.</span> <span class="attribution"><span class="source">Arnab Hazari</span>, <span class="license">Author provided</span></span></figcaption></figure><p>For the past four decades, the electronics industry has been driven by what is called “<a href="http://www.mooreslaw.org/">Moore’s Law</a>,” which is not a law but more an axiom or observation. Effectively, it suggests that the electronic devices double in speed and capability about every two years. And indeed, every year tech companies come up with new, faster, smarter and better gadgets.</p>
<p>Specifically, Moore’s Law, as articulated by Intel cofounder Gordon Moore, is that “The number of transistors incorporated in a chip will <a href="https://www-ssl.intel.com/content/www/us/en/history/museum-gordon-moore-law.html">approximately double every 24 months</a>.” Transistors, tiny electrical switches, are the fundamental unit that drives all the electronic gadgets we can think of. As they get smaller, they also <a href="http://www.intel.com/content/www/us/en/silicon-innovations/moores-law-technology.html">get faster and consume less electricity</a> to operate.</p>
<p>In the technology world, one of the biggest questions of the 21st century is: How small can we make transistors? If there is a limit to how tiny they can get, we might reach a point at which we can no longer continue to make smaller, more powerful, more efficient devices. It’s an industry with <a href="https://www.statista.com/statistics/272115/revenue-growth-ce-industry/">more than US$200 billion</a> in annual revenue in the U.S. alone. Might it stop growing?</p>
<h2>Getting close to the limit</h2>
<p>At the present, companies like Intel are mass-producing transistors <a href="https://www-ssl.intel.com/content/www/us/en/silicon-innovations/intel-14nm-technology.html">14 nanometers across</a> – just 14 times wider than <a href="https://dx.doi.org/10.1016/0022-2836(81)90099-1">DNA molecules</a>. They’re made of silicon, the <a href="http://hyperphysics.phy-astr.gsu.edu/hbase/Tables/elabund.html">second-most abundant material</a> on our planet. Silicon’s atomic size is <a href="http://www.extremetech.com/computing/97469-is-14nm-the-end-of-the-road-for-silicon-lithography">about 0.2 nanometers</a>.</p>
<p>Today’s transistors are about 70 silicon atoms wide, so the possibility of making them even smaller is itself shrinking. We’re getting very close to the limit of how small we can make a transistor.</p>
<p>At present, transistors use electrical signals – electrons moving from one place to another – to communicate. But if we could use light, made up of photons, instead of electricity, we could make transistors even faster. My work, on finding ways to integrate light-based processing with existing chips, is part of that nascent effort.</p>
<h2>Putting light inside a chip</h2>
<p>A <a href="https://reibot.org/2011/09/06/a-beginners-guide-to-the-mosfet/">transistor has three parts</a>; think of them as parts of a digital camera. First, information comes into the lens, analogous to a transistor’s source. Then it travels through a channel from the image sensor to the wires inside the camera. And lastly, the information is stored on the camera’s memory card, which is called a transistor’s “drain” – where the information ultimately ends up.</p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=550&fit=crop&dpr=1 600w, https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=550&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=550&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=691&fit=crop&dpr=1 754w, https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=691&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/147791/original/image-20161128-22729-tc0olq.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=691&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Light waves can have different frequencies.</span>
<span class="attribution"><a class="source" href="https://commons.wikimedia.org/wiki/File:VisibleEmrWavelengths.svg">maxhurtz</a></span>
</figcaption>
</figure>
<p>Right now, all of that happens by moving electrons around. To substitute light as the medium, we actually need to move photons instead. Subatomic particles like electrons and photons travel in a wave motion, vibrating up and down even as they move in one direction. The length of each wave depends on what it’s traveling through. </p>
<p>In silicon, the most efficient wavelength for photons is <a href="http://www.its.bldrdoc.gov/fs-1037/dir-040/_5927.htm">1.3 micrometers</a>. This is very small – a human hair is <a href="http://www.nano.gov/nanotech-101/what/nano-size">around 100 micrometers across</a>. But <a href="http://homepages.rpi.edu/%7Esawyes/Models_review.pdf">electrons in silicon</a> are even smaller – with wavelengths <a href="http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/debrog2.html">50 to 1,000 times shorter</a> than photons.</p>
<p>This means the equipment to handle photons needs to be bigger than the electron-handling devices we have today. So it might seem like it would force us to build larger transistors, rather than smaller ones.</p>
<p>However, for two reasons, we could keep chips the same size and deliver more processing power, shrink chips while providing the same power, or, potentially both. First, a <a href="http://www.nature.com/lsa/focus/circuits/index.html">photonic chip</a> needs only a few light sources, generating photons that can then be directed around the chip with very small lenses and mirrors.</p>
<p>And second, light is much faster than electrons. On average photons can travel about <a href="http://education.jlab.org/qa/electron_01.html">20 times faster</a> than electrons in a chip. That means computers that are 20 times faster, a speed increase that would take about 15 years to achieve with current technology.</p>
<p>Scientists have demonstrated <a href="http://www.nature.com/lsa/focus/circuits/index.html">progress toward photonic chips</a> in recent years. A key challenge is making sure the new light-based chips can work with all the existing electronic chips. If we’re able to figure out how to do it – or even to use light-based transistors to enhance electronic ones – we could see significant performance improvement.</p>
<h2>When can I get a light-based laptop or smartphone?</h2>
<p>We still have some way to go before the first consumer device reaches the market, and progress takes time. The first transistor was made in the year 1907 using vacuum tubes, which were <a href="http://www.edisontechcenter.org/VacuumTubes.html">typically between one and six inches tall</a> (on average 100 mm). By 1947, the current type of transistor – the one that’s now just 14 nanometers across – was invented and it was <a href="https://en.wikipedia.org/wiki/History_of_the_transistor#The_first_transistor">40 micrometers long</a> (about 3,000 times longer than the current one). And in 1971 the first commercial microprocessor (the powerhouse of any electronic gadget) was <a href="https://en.wikipedia.org/wiki/Intel_4004">1,000 times bigger</a> than today’s when it was released.</p>
<p>The vast research efforts and the consequential evolution seen in the electronics industry are only starting in the photonic industry. As a result, current electronics can perform tasks that are far more complex than the best current photonic devices. But as research proceeds, light’s capability will catch up to, and ultimately surpass, electronics’ speeds. However long it takes to get there, the future of photonics is bright.</p><img src="https://counter.theconversation.com/content/68903/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Arnab Hazari's research group receives funding from the National Science Foundation, under MRSEC program.</span></em></p>As electronic transistors get tinier, they approach a point at which they won’t be able to get smaller. How can we keep shrinking our devices, and making them more powerful at the same time? Light.Arnab Hazari, Ph.D. student in Electrical Engineering, University of MichiganLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/557952016-03-10T11:11:52Z2016-03-10T11:11:52ZBeyond silicon: the search for new semiconductors<figure><img src="https://images.theconversation.com/files/114507/original/image-20160309-13689-2guygk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">A molecular beam epitaxy machine used to create semiconductor samples.</span> <span class="attribution"><span class="source">John C. Bean (University of Virginia) and Tom Vandervelde (Tufts University)</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span></figcaption></figure><p>Our modern world is based on semiconductors. In addition to your computer, cellphones and digital cameras, semiconductors are a critical component of a growing number of devices. Think of the high-efficiency LED lights you are putting in your house, along with everything with a lit display or control circuit: cars, refrigerators, ovens, coffee makers and more. You would be hard-pressed to find a modern device that uses electricity that does not have semiconductor circuits in it.</p>
<p>While most people have heard of silicon and Silicon Valley, they do not realize that this is just one example of a whole class of materials.</p>
<p>But the workhorse silicon – used in all manner of computers and electronic gadgets – has its technical limits, particularly as engineers look to use electronic devices for producing or processing light. The search for new semiconductors is on. Where will these materials innovations come from? </p>
<h2>What’s a semiconductor?</h2>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/114508/original/image-20160309-13689-191zym2.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Probing a newly fabricated device.</span>
<span class="attribution"><span class="source">Corey Shemelya and Tom Vandervelde, Tufts University</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>As the name suggests, semiconductors are materials that conduct electricity at some temperatures but not others – unlike most metals, which are conductive at any temperature, and insulators like glass, plastic and stone, which usually don’t conduct electricity.</p>
<p>However, this is not their most important trait. When constructed properly, these materials can modify the electricity moving through them, including limiting the directions it flows and amplifying a signal.</p>
<p>The combination of these properties is the basis of diodes and transistors which make up all our modern gadgets. These circuit elements perform a multitude of tasks, including converting the electricity from your wall socket to something usable by the devices, and processing information in the form of zeros and ones. </p>
<p>Light can also be absorbed into semiconductors and turned into electrical current and voltage. The process works in reverse as well, allowing for the emission of light. Using this property, we make lasers, LED lights, digital cameras and many other devices.</p>
<h2>The rise of silicon</h2>
<p>While this all seems very modern, the original discoveries of semiconductors date back to the 1830s. By the 1880s, Alexander Graham Bell <a href="http://physics.kenyon.edu/EarlyApparatus/Electricity/Selenium_Cell/Selenium_Cell.html">experimented with using selenium</a> to transmit sound over a beam of light. Selenium was also used to make some of the first solar cells in the 1880s. </p>
<p>A key limitation was the inability to purify the elements being used. Tiny impurities – as small as one in a trillion, or 0.0000000001 percent – could fundamentally change the way a semiconductor behaved. As technology evolved to make purer materials, better semiconductors followed.</p>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=1068&fit=crop&dpr=1 600w, https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=1068&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=1068&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=1342&fit=crop&dpr=1 754w, https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=1342&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/114500/original/image-20160309-13730-qu75nt.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=1342&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A semiconductor chip fabricated in a Tufts University lab.</span>
<span class="attribution"><span class="source">Corey Shemelya and Tom Vandervelde, Tufts University</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>The first semiconducting transistor was made of germanium in 1948, but silicon quickly rose to become the dominant semiconductor material. Silicon is mechanically strong, relatively easy to purify, and has reasonable electrical properties. </p>
<p>It is also incredibly abundant: 28.2 percent of the Earth’s crust is silicon. That makes it literally dirt cheap. This almost-perfect semiconductor worked well for making diodes and transistors and still is the basis of almost every computer chip out there. There was one problem: silicon is very inefficient at converting light into an electrical signal, or turning electricity back into light.</p>
<p>When the primary use of semiconductors was in computer processors connected by metal wires, this wasn’t much of a problem. But, as we moved toward using semiconductors in solar panels, camera sensors and other light-related applications, this weakness of silicon became a real obstacle to progress.</p>
<h2>Finding new semiconductors</h2>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=657&fit=crop&dpr=1 600w, https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=657&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=657&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=825&fit=crop&dpr=1 754w, https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=825&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/114342/original/image-20160308-22147-11p0ld1.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=825&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A section of the periodic table of the elements.</span>
<span class="attribution"><a class="source" href="https://commons.wikimedia.org/wiki/File:Periodic_Table_Radioactivity.svg">Armtuk, Alessio Rolleri, Gringer</a>, <a class="license" href="http://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA</a></span>
</figcaption>
</figure>
<p>The search for new semiconductors begins on the periodic table of the elements, a portion of which is in the figure at right. </p>
<p>In the column labeled <em>IV</em>, each element forms bonds by sharing four of its electrons with four neighbors. The strongest of these “group IV” elements bonds is for carbon (C), forming diamonds. Diamonds are good insulators (and transparent) because carbon holds on to these electrons so tightly. Generally, a diamond would burn before you could force an electrical current through it.</p>
<p>The elements at the bottom of the column, tin (Sn) and lead (Pb), are much more metallic. Like most metals, they hold their bonding electrons so loosely that when a small amount of energy is applied the electrons are free to break their bonds and flow through the material.</p>
<p><a href="https://www.researchgate.net/profile/Thomas_Vandervelde/publication/228744777_The_effect_of_two-temperature_capping_on_germaniumsilicon_quantum_dots_and_analysis_of_superlattices_so_composed/links/0fcfd513e2ecb63454000000.pdf">Silicon (Si) and germanium (Ge)</a> are in between and accordingly are semiconductors. Due to a quirk in the way both of them are structured, however, they are inefficient at exchanging electricity with light.</p>
<p>To find materials that work well with light, we have to step to either side of the group IV column. Combining elements from the “group III” and “group V” columns results in materials with semiconducting properties. These “III-V” materials, such as gallium arsenide (GaAs), are used to make lasers, LED lights, photodetectors (as found in cameras) and many other devices. They do what silicon does not do well. </p>
<figure class="align-right ">
<img alt="" src="https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/114502/original/image-20160309-13689-a56jxk.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">A solar cell being tested under the illumination of an artificial sun, in one of our solar simulators.</span>
<span class="attribution"><span class="source">Corey Shemelya and Tom Vandervelde, Tufts University</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>But why is silicon used for solar panels if it is so bad at converting the light into electricity? Cost. Silicon could be refined from a shovel full of dirt scooped up from anywhere on the Earth’s surface; the III-V compounds’ constituent elements are far rarer. </p>
<p>A standard silicon solar panel converts the sunlight with an efficiency of 10 to 15%. A III-V panel can be three times as efficient, but often costs more than three times as much. The III-V materials are also more brittle than silicon, making them hard to work with in wide panels. </p>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/114501/original/image-20160309-13737-m6ykig.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Ghosts in the machine: A scanning electron microscopy image of a damaged gallium-antimonide semiconductor sample.</span>
<span class="attribution"><span class="source">Corey Shemelya and Tom Vandervelde, Tufts University</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>However, the III-V materials’ increased electron speeds enable construction of much faster transistors, with speeds hundreds of times faster than the ones you find in your computers. They may pave the way for wires inside computers to be replaced with beams of light, significantly improving the speed of data flow.</p>
<p>In addition to III-V materials, there are also II-VI materials in use. These materials include some of the sulfides and oxides researched in the 1800s. Combinations of zinc, cadmium, and mercury with tellurium have been used to create infrared cameras as well as solar cells from companies such as <a href="http://www.firstsolar.com/en/Technologies-and-Capabilities/PV-Modules/First-Solar-Series-3-Black-Module/CdTe-Technology.aspx">First Solar</a>. These materials are notoriously brittle and very challenging to fabricate. </p>
<h2>The future of semiconductors</h2>
<p>How might new semiconductor materials be used? </p>
<figure class="align-left zoomable">
<a href="https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/114503/original/image-20160309-13730-11l1sgw.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">A photonic crystal consisting of a periodic metal structure on a semiconductor.</span>
<span class="attribution"><span class="source">Corey Shemelya and Tom Vandervelde, Tufts University</span>, <a class="license" href="http://creativecommons.org/licenses/by-nd/4.0/">CC BY-ND</a></span>
</figcaption>
</figure>
<p>High power III-V (gallium-nitride) semiconductor electronics will be the backbone of our electrical grid system, converting power for high voltage transmission and back again. New III-V materials (antimonides and bismuthides) are leading the way for <a href="http://www.mdpi.com/1424-8220/13/4/5054/htm">infrared sensing</a> for medical, military, other civilian uses, as well new telecommunication possibilities. Earth-abundant element combinations are <a href="http://www.intechopen.com/articles/show/title/electrodeposited-copper-oxide-and-zinc-oxide-core-shell-nanowire-photovoltaic-cells">being explored</a> to make <a href="http://iopscience.iop.org/2053-1591/1/2/025002/">new semiconductors</a> for high-efficiency, but inexpensive, solar cells.</p>
<p>And what of the old standby, silicon? Its inability to harness light efficiently does not mean that it is destined for the dust bin of history. Researchers are giving new life to silicon, creating “<a href="http://REAP.ece.tufts.edu">silicon photonics</a>” to better handle light, rather than just shuttling electrons.</p>
<p>One method is the inclusion of small amounts of another group IV element, tin, into silicon or germanium. That changes their properties, allowing them to absorb and emit light more efficiently. </p>
<p>The act of including that tin turns out to be difficult, like many other challenges in material science. But as I tell my students all the time, “if it were easy, then it would not be research.”</p><img src="https://counter.theconversation.com/content/55795/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Thomas Vandervelde receives funding from the Air Force Office of Scientific Research, National Science Foundation, Office of Naval Research, NASA, the Intelligence Community, and Tufts University. </span></em></p>As we reach the limits of what can be done with silicon, the search for new and improved superconductors is on.Thomas Vandervelde, Associate Professor of Electrical and Computer Engineering, Tufts UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/462872015-08-27T05:37:40Z2015-08-27T05:37:40ZWith silicon pushed to its limits, what will power the next electronics revolution?<figure><img src="https://images.theconversation.com/files/93084/original/image-20150826-15393-ge6a27.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><a class="source" href="https://www.flickr.com/photos/rbulmahn/8028777646/in/album-72157631631893436/">rbulmahn</a>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span></figcaption></figure><p>The semiconducting silicon chip launched the revolution of electronics and computerisation that has made life in the opening years of the 21st century scarcely recognisable from the start of the last. Silicon integrated circuits (IC) underpin practically everything we take for granted now in our interconnected, digital world: controlling the systems we use and allowing us to access and share information at will. </p>
<p>The rate of progress since the first silicon transistor in 1947 has been enormous, with the number of transistors on a single chip growing from a few thousand in the earliest integrated circuits to <a href="http://phys.org/news/2015-04-silicon-valley-years-law.html">more than two billion today</a>. Moore’s law – that transistor density will double every two years – <a href="http://www.intel.com/content/www/us/en/silicon-innovations/moores-law-technology.html">still holds true 50 years after it was proposed</a>. </p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=561&fit=crop&dpr=1 600w, https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=561&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=561&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=704&fit=crop&dpr=1 754w, https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=704&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/93082/original/image-20150826-15393-nb90yl.PNG?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=704&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Moore’s law still holds true after 50 years.</span>
<span class="attribution"><a class="source" href="https://commons.wikimedia.org/wiki/File:Moores_law_(1970-2011).PNG">shigeru23</a>, <a class="license" href="http://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA</a></span>
</figcaption>
</figure>
<p>Nevertheless, silicon electronics faces a challenge: the latest circuits <a href="http://www.theguardian.com/technology/2015/jul/09/moores-law-new-chips-ibm-7nm/">measure just 7nm</a> wide – between a red blood cell (7,500nm) and a single strand of DNA (2.5nm). The size of individual silicon atoms (around 0.2nm) would be a hard physical limit (with circuits one atom wide), but its behaviour becomes unstable and difficult to control before then. </p>
<p>Without the ability to shrink ICs further silicon cannot continue producing the gains it has so far. Meeting this challenge may require rethinking how we manufacture devices, or even whether we need an alternative to silicon itself.</p>
<h2>Speed, heat, and light</h2>
<p>To understand the challenge, we must look at why silicon became the material of choice for electronics. While it has many points in its favour – abundant, relatively easy to process, has good physical properties and possesses a stable native oxide (SiO<sub>2</sub>) which happens to be a good insulator – it also has several drawbacks.</p>
<p>For example, a great advantage of combining more and more transistors into a single chip is that it enables an IC to process information faster. But this speed boost depends critically on how easily electrons are able to move within the semiconductor material. This is known as <a href="http://www.electrical4u.com/drift-velocity-drift-current-and-electron-mobility/">electron mobility</a>, and while electrons in silicon are quite mobile, they are much more so in other semiconductor materials such as gallium arsenide, indium arsenide, and indium antimonide. </p>
<p>The useful conductive properties of semiconductors don’t just concern the movement of electrons, however, but also the movement of what are called <a href="http://www.allaboutcircuits.com/textbook/semiconductors/chpt-2/electrons-and-holes/">electron holes</a> – the gaps left behind in the lattice of electrons circling around the nucleus after electrons have been pushed out. </p>
<p>Modern ICs use a technique called complementary metal-oxide semiconductor (<a href="http://www.engineersgarage.com/articles/what-is-cmos-technology">CMOS</a>) which uses a pair of transistors, one using electrons and the other electron holes. But electron hole mobility in silicon is very poor, and this is a barrier to higher performance – so much so that for several years manufacturers have had to boost it by including germanium with the silicon.</p>
<p>Silicon’s second problem is that performance degrades badly at high temperatures. Modern ICs with billions of transistors generate considerable heat, which is why a lot of effort goes into cooling them – think of the fans and heatsinks strapped to a typical desktop computer processor. Alternative semiconductors such as gallium nitride (GaN) and silicon carbide (SiC) cope much better at higher temperatures, which means they can be run faster and have begun to replace silicon in critical <a href="http://electronicdesign.com/communications/what-s-difference-between-gaas-and-gan-rf-power-amplifiers">high-power applications</a> such as amplifiers.</p>
<p>Lastly, silicon is very poor at transmitting light. While lasers, LEDs and other photonic devices are commonplace today, they use alternative semiconductor compounds to silicon. As a result two distinct industries have evolved, silicon for electronics and compound semiconductors for photonics. This situation has existed for years, but now there is a big push to combine electronics and photonics on a single chip. For the manufacturers, that’s quite a problem.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/93085/original/image-20150826-15407-8exbg2.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Semiconductor lasers, where alternatives to silicon such as germanium have already found a role.</span>
<span class="attribution"><a class="source" href="https://en.wikipedia.org/wiki/File:Laser_module.jpg">彭家杰</a>, <a class="license" href="http://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA</a></span>
</figcaption>
</figure>
<h2>New materials for future</h2>
<p>Of the many materials under investigation as partners for silicon to improve its electronic performance, perhaps three have promise in the short term.</p>
<p>The first concerns silicon’s poor electron hole mobility. A small amount of germanium is already added to improve this, but using large amounts or even a move to all-germanium transistors would be better still. Germanium was the first material used for semiconductor devices, so really this is a “back to the future” move. But re-aligning the established industry around germanium would be quite a problem for manufacturers.</p>
<p>The second concerns metal oxides. Silicon dioxide was used within transistors for many years, but with miniaturisation the layer of silicon dioxide has shrunk to be so thin that it has begun to lose its insulating properties, leading to unreliable transistors. Despite a move to using rare-earth <a href="http://www.rsc.org/chemistryworld/Issues/2007/March/HafniumOxideHelpsMakeChipsSmallerFaster.asp">hafnium dioxide</a> (HfO<sub>2</sub>) as a replacement insulator, the search is on for
alternatives with even better insulating properties.</p>
<p>Most interesting, perhaps, is the use of so-called <a href="http://www.sandia.gov/%7Ejytsao/WCS.pdf">III-V compound semiconductors</a>, particularly those containing indium such as indium arsenide and indium antimonide. These semiconductors have electron mobility <a href="http://www.yokoyama-gnc.jp/english/research/cmos.html">up to 50 times higher</a> than silicon. When combined with germanium-rich transistors, this approach could provide a major speed increase. </p>
<p>Yet all is not as simple as it seems. Silicon, germanium, oxides and the III-V materials are crystalline structures that depend on the integrity of the crystal for their properties. We cannot simply throw them together with silicon and get the best of both. Dealing with this problem, <a href="http://www.semi1source.com/glossary/default.asp?searchterm=lattice+mismatch">crystal lattice mismatch</a>, is the major ongoing technological challenge.</p>
<h2>Different flavours of silicon</h2>
<p>Despite its limitations, silicon electronics has proved adaptable, able to be fashioned into reliable, mass market devices available at minimal cost. So despite headlines about the “end of silicon” or the spectacular (and sometimes rather unrealistic) promise of alternative materials, silicon is still king and, backed by a huge and extremely well-developed global industry, will not be deposed in our lifetime. </p>
<p>Instead progress in electronics will come from improving silicon by integrating other materials. Companies like IBM and Intel and university labs worldwide have poured time and effort into this challenge, and the results are promising: a hybrid approach that <a href="http://www.eeherald.com/section/news/onws20150111001a.html">blends III-V materials, silicon and germanium</a> could reach the market within a few years. Compound semiconductors have already found important uses in lasers, LED lighting/displays and solar panels where silicon simply cannot compete. More advanced compounds will be needed as electronic devices become progressively smaller and lower powered and also for high-power electronics where their characteristics are a significant improvement upon silicon’s capabilities.</p>
<p>The future of electronics is bright, and it’s still going to be largely based on silicon – but now that silicon comes in many different flavours.</p><img src="https://counter.theconversation.com/content/46287/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Mark Hopkinson does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.</span></em></p>Silicon isn’t the perfect semiconductor, it’s just the one we’re using. How can we ensure our electronics keep get getting faster in the face of silicon’s natural physical limits?Mark Hopkinson, Professor of Semiconductor Materials, University of SheffieldLicensed as Creative Commons – attribution, no derivatives.