By: Phil Kernen
When we consider semiconductor chips and the things they power, we usually think of personal computers, mobile phones, and appliances. The silicon chips in those applications and others process information, helping usher in the digital revolution. However, over the past five years a growing portion of the semiconductor industry is earning attention for its ability to manage something other than information.
By the 1970s, traditional semiconductors had been around for a while. Alone, silicon resistance to the flow of electrons was similar to an insulator like plastic or wood. But with proper chemicals, silicon chips allow electrons with electric pulses carrying ones and zeros to flow across the chip when a voltage is applied, precursor to the age of digitization.
Silicon worked fine in the applications of the day, but farsighted engineers recognized limitations if the amounts of power became very large. Silicon resistance, referenced by the term ‘band gap’, was narrow. Small band gaps aren’t a problem for handling the voltage of a computer, or a mobile phone, such as 1.5 volts. But what about applications that manage batteries or systems involving hundreds or thousands of volts? The conducting abilities of silicon would have limitations.
Experiments began to identify semiconductors with wider band gaps to accommodate higher voltages. Silicon carbide (SiC) was among the contenders and tests revealed three advantages. SiC chips can switch on and off faster than silicon, more efficiently controlling the voltage flow. When the wide band gap of SiC chips encounters large enough voltages in the ‘on’ state, it turns into a conductor with a resistance two thousand times less than silicon. More efficiency. Three, in its ‘off’ state, SiC chips can hold back ten times the power in the form of voltage than silicon, managing and accommodating much larger volumes of voltage. SiC was the answer the engineers were seeking.
While a few companies went to work to commercialize these discoveries, SiC was a creation ahead of its time. Niche products were the outcome for the first few decades. The shift to a broader application occurred when Tesla decided to use SiC chips in the Model 3. Traction inverters convert the direct current (DC) from the car battery into alternating current (AC) to drive the motor. The higher band gap of SiC chips allowed for an inverter that could manage a larger battery than those currently operated by inverters supported by silicon chips.
The Model 3 became a best-selling electric vehicle (EV), and Tesla bought SiC chips in big volumes, applying the idea next to the Model S, its existing luxury sedan. Using the same battery pack, incorporating SiC chips boosted the range from 335 miles to 370, half of which was attributable to the chips alone. Compared to competitors, Tesla was providing SiC-powered cars that offered twice the range at half the price, forcing every automaker in the same direction and launching a push for SiC manufacturing capabilities.
The efficiency of SiC chips may reduce the need for battery materials like cobalt, lithium and nickel, each of which present environmental and social challenges when mined. All else being equal, for a target EV battery range, the greater efficiency of SiC chips means fewer materials needed.
The manufacture and scaling of SiC chips are at the same stage as silicon was in the 1980s when Intel determined how to put more transistors on bigger silicon wafers and obtain large economies of scale in the process. We can’t say whether the miniaturization of SiC transistors will progress similarly, putting more power in smaller packages. Still, chip makers like ON Semiconductor and Wolfspeed are among those assembling factories to churn out as many SiC chips as automakers will need, with many others playing supporting roles providing the machines and diagnostic equipment.
EV applications may be just the beginning, but SiC chips could impact anything with a motor. Consider onboard or external charging stations that convert AC power to DC power to suit the battery, or electrical systems from data centers to windfarms to solar arrays and HVAC. Any systems that have to continuously translate, manage, and regulate the flow of energy are a candidate for SiC chips. Investors are taking note.
Disclosure: This is for informational purposes only and any reference to a specific industry does not constitute a recommendation to buy or sell a company in that industry. The reader should not assume that an investment in the securities in the industries identified or described, was or will, be profitable. https://mitchcap.com/disclosure/