A raw deal: Will materials shortages, supply chain challenges threaten tech's future?
In the face of rising trade issues and skyrocketing demand, e-waste recycling and digital supply networks and a holistic approach to supply chain sustainability could help in the medium term.
Duncan Stewart
Gillian Crossan
Ariane Bucaille
Christie Simons
You can’t make certain chips, motors, batteries, or a host of other key technologies without certain materials. Deloitte predicts that multiple regions will run short of gallium and possibly germanium as soon as 2024, impacting chipmakers. By 2025, we may start seeing shortages of rare earth elements (REEs) for magnets in electric car motors and more, as well as lithium and cobalt for batteries.1 Actions can be taken, however, that can help avoid shortages in the near, medium, and long term.
How we got here
Some of these raw material shortages were more predictable: Anticipated growth in battery electric vehicles (BEVs) is forecast to drive demand for lithium and cobalt above the predicted production ramp for years, which could affect BEVs, but also consumer and enterprise devices such a laptops, tablets, and smartphones.2 However, geopolitical tensions between China and western economies have recently impacted supply chains, with the potential for future impacts, as well.
First, China is subject to export controls that limit its ability to purchase critical chips, as well as the technologies and software needed to make the chips domestically3 (see gen AI chips prediction). In July, China announced that it would begin to impose its own controls on the export of germanium and gallium.4 Second, some analysts have expressed concerns that China may also restrict the export of the 17 elements that make up REEs (used for electronics, clean energy, aerospace, automotive, and defense) in 2024, leading to possible shortages for western manufacturers that year or the year after.5 Further, some of the alternate sources of tech-important raw materials are in geographies in which political, regulatory, or social factors may decrease their reliability in the long term.6
There have been raw materials shortages before that affected the tech industry. For example, the tantalum shortage of 2000 impacted capacitor supply.7 But what could be unprecedented in 2024 and 2025 is that these shortages could possibly be across dozens of different raw materials at the same time.8 Also unprecedented, the industries enabled by these materials play a much bigger role in the economy than they used to. Such industries are estimated to be worth over US$160 billion annually, or (in the case of semiconductors or electric vehicle (EV) motors) are worth smaller dollar amounts but are still critical for innovation, growth, and national security (figure 1).9
When speaking of these shortages, time frames matter
In the very short term (late 2023 and early 2024) many companies that need germanium and gallium have inventories that are expected to last them into the first half of 2024. After that, gallium may be the bigger challenge to acquire, as there is a large and reliable source of germanium in British Columbia.10
In the longer term, mines and smelters can be built. Gallium is mainly produced as a byproduct of aluminum production from bauxite,11 and there are bauxite deposits in dozens of countries and on every continent.12 Equally, rare earth deposits are often not that rare, and mines are planned for Australia, Angola, Afghanistan, Canada, and the US over the next few years.13
Between now and then there are three big things that could be done to help mitigate supply chain vulnerabilities to existing or probable material shortages.
- E-waste recycling: Globally, the electronics industry throws away a huge volume of valuable elements, collectively worth an estimated US$50+ billion annually.14 Many of those elements are often vulnerable to supply chain disruptions. A rapid increase in recycling could help boost supply, and a push to innovate e-waste recycling technologies could help to accelerate this even more.
- Digital supply networks (DSNs): These allow industries to do more with less: DSNs can anticipate and plan for raw material shortages and make them less severe by getting exactly the right minerals to the right place at the right time with less waste. They are not expected to stop shortages from happening, but they can help make them less severe or protracted. Building DSNs is often synergistic with other sustainable supply chain efforts including, for example, collaboration with and incentivization of supplier networks to reduce other forms of waste, including GHG emissions (see semi sustainability prediction).
- Stockpiling: As an example, the US currently has strategic reserves of petroleum, grain, bullion, helium, and raw materials critical for military defense.15 Would it make sense to create analogous reserves of critical raw materials for EVs and semiconductors?
Not just elementary, my dear Watson
Thus far, we have entirely discussed elements. But tech and semi supply chains are not just about raw materials. They are also about:
- Refined, processed, and purified materials. Low purity neon or silicon are plentiful. But the gas necessary for lasers in semi manufacturing and the silicon for ingots need to be ultrapure, and sources of the ultrapure materials are highly concentrated: Ukraine is the source of about 50% of semiconductor grade neon,16 while China is the source of 80% of global polysilicon (used for both semi manufacturing and solar power) supply.17
- Manufactured specialty compounds. It might be an epoxy or resin, a special cleaning fluid or gas, or type of plastic.18 Any geopolitical uncertainty, or an earthquake, typhoon, hurricane, fire, flood, drought, or pandemic could cause significant and lasting supply chain interruption.
Strategies such as recycling, DSNs, and in general more sustainable semi manufacturing (see semi sustainability prediction), plus stockpiling could help. While building a new mine takes five to 10 years, building a new factory, for example, takes approximately two to three years.19
The bottom line
The world produced 57.4 million metric tons of e-waste in 2022,20 of which only 17% was formally collected and recycled,21 although there is significant regional variation, with the United States recycling about 15% in 2019 and the EU about 42.5% as of 2022.22 On current trends, the e-waste number is predicted to rise to 75 million tons by 2030.23 There are a variety of e-waste sources: Large appliances alone make up more than half of e-waste in the European Union, but IT, telecom, and consumer electronics make up almost 30% as of 2020.24 The International Telecommunication Union set a goal for e-waste recycling to reach 30% globally by 2023, but that goal will likely not be achieved.25
Why isn’t more e-waste being recycled? At a high level, it’s often because it’s unprofitable. Recycling doesn’t make money; it costs money. That’s mainly because each device contains a mix of recyclable elements, often chemically bound to other elements or compounds, and physically bound to other elements or compounds, such as solders, plastics, and ceramics.26 Extracting metals requires work, energy, chemical, and physical processes, and often results in dangerous or harmful waste.27 Much of the current recycling is done outside where the toxic e-waste is produced. For example, e-waste from the developed world is often shipped to and processed in the developing world, especially impacting women and children’s health: between 2.9 and 12.9 million women may be at risk from exposure to toxic e-waste through their work in the informal waste sector, and around 18 million children are employed in waste processing industries.28
Although there’s an estimated US$15 billion in valuable metals in e-waste annually (mainly from printed circuit boards),29 the costs are sometimes higher, resulting in the current, relatively low percentage of e-waste being recycled. However, if there was a significant push for supply chain resilience reasons (rather than purely economic ones) for REEs, lithium, cobalt, and semiconductors, we could see two benefits: The percentage of e-waste being recycled would likely climb, and more of the recycling may need to occur on shore in the European Union and North America, helping reduce environmental harm to other countries.
A number of niche recycling and circular logistics, process, and business model providers have emerged, and represent a foundation upon which such scaling of recycling can build. The US government has recently announced various incentives for extracting lithium and REEs from e-waste.30
Of the elemental challenges, REEs are one of the biggest, as the magnets enabled by REEs are found in EVs, wind turbines, defense systems, and much more.31 An interesting longer-term solution may be found in Vietnam. According to the United States Geological Survey (USGS), Vietnam has the second largest rare earth deposits at 22 million tons, after only China.32 And it has already started to ramp up production, with output of raw REEs in 2022 of 4,300 tons, over ten times higher than the year before, and with a 2030 goal of 2 million tons by 2030.33
How much would a strategic US or EU gallium stockpile cost? Low purity gallium costs about US$280 per kilogram, while 99.99999% pure gallium costs about US$450 per kilogram, and gold is roughly US$66,000 per kg.34 Large users of gallium for electronics purposes use “dozens of tons” per year,35 so a three-year reserve would be approximately $20 million (unpurified) to $30 million (purified.)
Not only are REEs and other important metals, such as gallium and germanium, mined in China, but they are also usually smelted there, too. If trade restrictions between China and the West escalate further, and as the technology industry increasingly uses various hard-to-get elements, building supply chain resilience likely requires reducing source material concentration and building more mines and smelters on shore or near shore in the long term. In the near term, resilience can also be boosted via investments in e-waste recycling, digital supply networks, and stockpiling.
These initiatives will take billions of dollars of investment, but the industries that rely on these materials are worth a hundred times more in annual revenues.