Would you be surprised to learn that the most advanced economy in the world is only forty percent efficient at utilizing its key input? Shockingly, of the 100 Quads of energy consumed in the United States in 2014, only 40 quads performed useful services such as cooling a building or transporting the kids to soccer practice. The remaining was lost as heat. A shameful waste or a tremendous opportunity?
Venture capitalists and entrepreneurs alike share the DNA that sees the glass half full. Some of our portfolio companies such as Switch Materials, Tivra and A2M are developing breakthrough efficiency technologies that will chip away at the waste. And these technologies are having an impact with the energy used per unit of GDP slowly trending down over time. Nevertheless, major energy consuming processes such as power generation, transportation, and industrial processing (i.e. steelmaking) remain guilty of producing a tremendous quantity of waste heat. This is where the ability of thermoelectric devices in converting heat to electricity could be the silver bullet in the global efficiency game.
The thermoelectric effect was discovered almost two centuries ago and global research efforts continue to accelerate, with over 300 new patents filed annually. But the commercial promise has never been realized, as conversion efficiency remains low, costs remain high, and durability requirements are extreme. As usual, materials science is coming to the rescue.
In thermoelectrics jargon, ZT is a metric for the effectiveness of a material converting heat to electricity. Increasing ZT is the goal of most thermoelectric research and is generally an exercise in minimizing thermal conductivity while maximizing electrical conductivity. State-of-the-art materials have a ZT of about 3, while a value of 2 would generally achieve a conversion efficiency of 15-20%, depending on the application. Researchers use all sorts of techniques to improve performance including combinatorial materials discovery, exotic thin film processing, complex superlattice device structures, and engineered nanostructuring. Unfortunately, many of these innovations remain stuck in the lab. The most common commercial devices utilize variations of lead telluride or bismuth telluride based on relics of research from the 1950’s!!! Thin film techniques suffer in particular, as it is difficult to maintain the required temperature gradient that drives the thermoelectric effect if the device is too thin. More generally, many of these alternatives suffer from complex manufacturing, poor scalability, and high cost.
As we have seen in the solar industry, the highest efficiency doesn’t always win the day. Silicon has become a de facto standard in solar primarily because of the ability to drive down costs. An interesting cost target for a thermoelectric system is below $2/watt. However, in many cases, the material costs alone exceed this. Unfortunately, much of the research work over the last decade has been on technologies that are in direct conflict with the concept of low cost. For example, thin film deposition of complicated alloys with intricate superlattice structures is probably better left for an Intel chip rather than an industrial energy device. Of course, many of the commonly used materials such as Tellurium don’t help the cost roadmaps and would be fundamentally limited in supply if a product was to scale to actually make a dent in global energy use. Fortunately, numerous industrial and entrepreneurial groups have focused efforts on low cost materials such as Silicides and Skutterudites that have the potential to be low cost. If simple and scalable manufacturing processes to make these materials can be developed and married with simple nanostructuring techniques to improve ZT, these materials have a real chance of taking off.
At Pangaea we love thinking outside of the box. Perhaps the success of the silicon solar industry was the inspiration for several groups in developing silicon-based thermoelectrics despite their seemingly unsuitable thermal properties.
Leveraging billions of dollars of depreciated semiconductor infrastructure is certainly a great story that would be attractive to VCs, however we expect the challenges in getting these materials into acceptable performing, high durability devices means that this technology is in the early stages of a very long gestation period. Given that long development cycles within VC-backed companies generally fail to result in the financial value creation that VCs are looking for, we figure this approach might be better suited to the lab for now.
Being a thermoelectric device would not be a lot of fun. They are squeezed between two surfaces that need to be as far apart in temperature as possible. Add in temperature fluctuations causing thermal expansion and high compressive forces required for efficient heat transfer and you have a material compatibility and systems engineering challenge of monumental proportions. As a result, we see the companies creating significant value in the business of converting materials to modules and systems. Typical in the industry is a 2-3X value bump from material to module and at least another 2X bump from module to system. Even Pangaea Limited partner BASF has chosen to deviate from its usual “The Chemical Company” model to become a module supplier for automotive thermoelectrics.
At the risk of deviating from the topic, it is worth pointing out that thermoelectrics are not unique in how revenue is captured in the value chain. Materials science is usually the enabler of breakthrough performance, lower cost, and longer life but integrating that into a useful product is not always easy and creates real value when done well. This is typically rewarded with significant margin capture as you move up the chain. It is with this in mind that Pangaea invests in companies that have taken novel advanced materials technologies to create breakthrough products that have the potential to make our world a better place. As for thermoelectrics, we would argue that some of the approaches that are being pursued are full of hot air but within the next few years some new low cost, high performance approaches will really deserve the world giving them some of its wasted heat.