Energy and material estimates in semiconductor processing and manufacturing

The intent of this joint project between Stanford Graduate Business School (Burgelman, Cogan),  the School of Engineering (Shankar), and SLAC National Laboratory (CEECS) is to study materials and energy use in semiconductor technology, one of the most critical components of the modern economy that drives all aspects of electronics and computing from internet to automation and artificial intelligence.  Our analysis attempts to cover the area in two phases: In the first phase (the current part of the research), the focus is on understanding the scientific and engineering analysis of materials and resources, while in the second phase the focus is on the strategic and business analysis based on the first phase analysis.  The first phase research covered the following components: 1) Identify the key material and electronic components in semiconductor technology including critical materials as relevant to the semiconductor industry; 2) Quantify resource usage that includes materials, water, and electricity in processing and manufacturing; 3) Trace the supply chain of these materials and components needed in the industry.  As the semiconductor supply chain is extremely complex, dynamic, interconnected, and quickly evolving, the research has attempted to qualify and quantify material and resource usage based on available data. In addition, the analysis also indicates that mining and extraction of materials is a multi-faceted problem with long term consequences based on policies, technologies, and business decisions. However, the analysis illustrates that the supply chain problem can be addressed by using a circular approach in which the materials and water can be recycled with efficient linkages between different aspects of manufacturing and consumer use.  This is illustrated by comparing energy requirements for material recycling with that needed to mine and extract materials. For higher efficacy, it is important that the circular ecosystem be relatively proximate to the geographical area of manufacturing and/or customer use, depending on the ease of collection, transportation facilities, and technologies needed for further processing.  The research also indicates that a rethinking of materials and resource processing will be of strategic importance to economies such as the US, where a renewed focus on efficient and intelligent reuse of materials across the different aspects of semiconductor manufacturing locally can mitgate the dependencies on external supply chains while catalyzing large parts of the economy with new innovations in technologies for materials collection and its processing.  The analysis provides pointers to the possibility of a large economic boost to the local economies with optimized manufacturing at every processing step to newly designed supply chain landscape that can potentially be a significant part of the US economy even beyond the semiconductor industry. In addition, this may have implications to the security of the economies which are dependent on advanced computing and its ubiquitous applications.

Business Strategy Challenges

In 2022, global e‑waste volumes hit a record 62 million tons in 2022, yet only 22.3% entered formal recycling streams. Although semiconductor devices constitute only an estimated 0.12% of e-waste by weight, they are perhaps its most economically dense (and therefore strategically important) element. Devices like CPU’s and SSD’s embed critical minerals such as antimony, gallium, yttrium, and titanium for which the United States remains > 95 % import‑reliant. 

Recent geopolitical events such as China’s ban on rare-earth exports to the US has heightened interest in the development of US-based critical minerals supply chains. Reliable recovery of minerals from semiconductor e-waste at scale could unlock both economic efficiencies for chipmakers (due to the lower energy costs associated with their recycling vs. their production from virgin stocks) and further both ESG and geopolitical objectives (related to the recovery of critical materials such as rare earths). Yet, the recycling rate for e-waste as a whole remains relatively low: in 2022, the Global E-waste Statistics Partnerships estimated a 56% collection rate for US e-waste. 

This project synthesizes lessons from adjacent industries with exemplary recycling rates (US lead‑acid batteries: 99%; global aerospace parts: 85–90%), then develops a framework to apply these lessons to the semiconductor industry with the goal of unlocking strategic resources – in particular, the critical mineral antimony – from post-consumer e-waste. We recommend several novel initiatives based on these findings, then project their impact.

We conclude that, while projected recovery of critical minerals in the United States is possible and a domestic “urban‑mining” services market is appealing, the amount of materials recovered is strategically negligible: if all 7188kt of the US’ entire 2022 e-waste stream were to be captured and 100% recovered, the recovery effort would result in an estimated +0.37kt antimony, +0.12kt germanium, and +0.5kt gallium. These amounts all represent less than 0.00001% of annual US imports of each material, rendering this specific initiative – the recovery of critical minerals from post-consumer semiconductor devices – strategically insignificant from a materials perspective, before factoring in costs of recovery and processing.

Based on these findings, some suggestions of future strategic priorities are offered including higher efficiency on-site during manufacturing and individual processing steps.

The project was funded by the Stanford BGS fund and the DOE AMMTO EES2 funding of the SLAC National Laboratory.