0*fS6KIOUr0e0YzvIk

This article is contributed by Singyuk Hou

Often referred to by chemists as a sibling of carbon, silicon not only serves as the canvas for transistors in microfabrication and the workhorse of solar panels in photovoltaics but also holds incredible potential as an anode material for Li-ion batteries. Despite its long history in development, silicon, the second most abundant element on earth, has only recently started gaining traction in the battery industry as an anode material.

Lithium-silicon alloy on the benches

Li reacts with silicon via intermetallic alloying, in which the structure of silicon is continuously modified to accommodate the Li as if dissolving sugar in water, whereas graphite provides a rigid structure that only houses Li in the space in between the graphite layers. The nature of the alloying reaction allows silicon to store ten times more Li than graphite with the same weight, which translates to at least a 30% increase in the energy density of the batteries. This potential for higher energy density has motivated the study of silicon anodes since day one.

The first use of the Li-silicon alloy in electrochemical cells was reported by General Motors in 1976, seven years earlier than the first use of a graphite anode [1], [2]. The study demonstrated a reversible reaction between Li and silicon, with around 80% material utilization for hundreds of cycles. These results were obtained in molten salt electrolytes at 400–600 °C, a temperature high enough to almost melt the alloy. For the next 10 to 15 years, silicon was primarily considered for use as an anode material for high-temperature batteries [3]. Graphite, on the other hand, was successfully developed as a Li host at ambient temperature in 1990 [4] and almost immediately commercialized between 1991 and 1994, beginning its reign in modern Li-ion batteries and consumer electronics.

Nano-sizing and amorphization

While silicon was being overshadowed by graphite as an anode material, silicon also had its own issues to contend with.

Specifically, silicon expands by nearly 3 times in volume during alloying and the silicon particles simply cannot accommodate the stress brought forth by this volume expansion without breaking into pieces, as described by Huggins from Stanford in 1989 [5].

Things are different at the nanoscale, however. Starting in 1994, Jeff Dahn at Simon Fraser University led several studies to place nano-sized silicon in carbon matrices [6]. Although the silicon occupied a very small portion of the material, these studies demonstrated the viability of using nanosized silicon as an anode at ambient temperatures, thus inspiring a new branch of structural engineering. In 1998, Wang from Zhejiang University was the first to make silicon nanoparticles for use in anodes without a carbon matrix through ball milling [7], and in 1999 Li from the Chinese Academy of Science synthesized silicon nanoparticles using a molecular precursor [8]. Both demonstrated reversible reactions with Li at room temperature. In 2003, researchers from Carnegie Mellon University and Seoul National University independently showed even better cycling performance using amorphous silicon in nanometer-thick, thin-film electrodes [9], [10].

From 2000 to 2010, research on silicon took off. More and more reports from research institutions demonstrated promising lifetime through nanosizing, amorphization, restricting expansion, or a combination of these techniques. Some of the inventions, such as silicon nanowires [11] and nanoparticles, led to start-ups like Amprius and Sila Nanotechnologies that are still seeking to commercialize silicon-based anode technologies.

Back to micron-size: multinary alloys

Although commercialization is seemingly close at hand, several concerns remain. First, nanosized or amorphous silicon is often made from expensive silicon-containing precursors rather than cheap and widely available silica. Also, nanosizing reduces the packing density of silicon, and the corresponding advantages in energy density.

Another strategy, multinary alloying, has yielded promising outcomes for micron-sized silicon-based materials. Adding other inert atoms alters the mechanical and electrical properties of the silicon alloys and can reduce the volume expansion. Several silicon-based anode materials developed by the battery industry have followed this strategy, including a transition metal-doped silicon from 3M Company patented in 2014 [12], a series of metal-doped silicon and silicon oxides from LG Chem patented between 2002 and 2021 [13], and the silicon monoxide from BTR New Material Group patented between 2013 and 2022 [14].

Can binder and electrolyte bring metallurgical-grade and micron-sized silicon back to the game?

Tesla has long been a proponent of the silicon anode. At the 2020 Battery Day, CEO Elon Musk put metallurgical-grade (MG) silicon on the company’s production timeline [15]. MG silicon, a product of silica reduction, is 98–99% pure, far below the grade for microfabrication but at a price of only a couple of dollars per kilogram and with an annual production of several million tons. This technology, therefore, has the potential to be disruptive in reducing the cost of the silicon anode. A patent entitled “Large-format battery anodes comprising silicon particles” was transferred from Colorado-based startup SiLion to Tesla in October 2021 and hints at the utilization of a conductive polymer coating to stabilize the silicon [16].

Figure 1. The major IP players in different segments of batteries with silicon-based anodes [17].