This working group focuses on the life cycle inventory of semiconductor manufacturing. We look specifically at wafer and photomask manufacturing, front-end and back-end processes for logic and memory devices .
| WG tracks | Reference contact | Priority | Status | In process tasks |
| Front-end manufacturing | Gauthier | High | Validated | Modelisation approach (v1) |
| High | Almost complete | Characterisation of inputs and outputs by process type | ||
| High | In progress | Recovery of missing model parameters | ||
| High | In progress | Recovery of recipe archetype per device | ||
| High | In progress | Contacting manufacturers and laboratories to obtain average data by process type | ||
| Back-end manufacturing | Arthur | High | In progress | To be filled by Arthur |
| Wafer manufacturing | Gauthier | Medium | In queue | To be started by Matthieu |
| Photomask manufacturing | Gauthier | Low | Not started | To be started, dependant of photomask recovery for elemental analysis |
This working group focuses only on four different industrial chains to create an open LCI.
Wafer manufacturing refers to the processes involved in fabricating blank wafers made of ultra-pure silicon ready to be used in front-end manufacturing.
Photomask manufacturing refers to the processes involved in fabricating photomasks engraved with a specific circuit design.
Front-end manufacturing refers to the processes used for fabricating a chip’s design onto a raw wafer to yield a die (SIA). It broadly includes: oxidation, photolithography, doping, thin film deposition, metallization, etching, chemical mechanical planarization and metrology.
Back-end manufacturing turns a fabricated wafer into as many as tens of thousands of ready-to-use chips (SIA). It broadly includes: dicing, die attach, bonding, encapsulation and testing.
Silicon wafers are used to manufacture logic (CPU, GPU, ASIC, …) and memory (NAND, RAM, etc.) devices. These devices power most modern electronic and digital equipement requiring computing power, memory storage and connectivity.
Silicon wafers used for integrated circuits (IC) manufacturing are defined as monocrystalline or single-cristal silicon wafers.
These wafers are used to manufacture different types of devices :
Depending of devices characteristics, material and processes consumption differ. The biggest parameters that will affect overall consumption and processes involved are :
These parameters affect all types of device mentioned earlier.
Wafer manufacturing has been stable for many years and is a stable market where only wafer size differs (200, 300mm). Single-cristal wafers is a generic component representative of the whole market.
Photomask manufacturing has evolved to answer to industry needs regarding device scaling. A photomask composed of a quartz glass and a chromium layer can be representative of most of the market. However, newer devices at a smaller scale (EUV, etc.) need new materials and technologies.
Some devices can be representative of a specific market or a specific generation but no generalization is possible for logic and memory devices. Model parametrization is necessary to provide a representative assessment of front-end and back-end manufacturing.
Each manufacturing category requires different production lines and steps. All the sub-parts regarding manufacturing will be defined in the dedicated pages below.
Other sub-parts relate to specific flows necessary for semiconductor manufacturing such as Ultra-Pure Water (UPW). However, for now we don't intend to address this kind of sub-parts.
We include manufacturing stages for single-cristal wafers, photomasks, front-end and back-end processes for CPU, NAND and RAM devices.
For now, several devices are excluded : GPU, FPGA, SDRAM (HBM).
We exclude research and development, chip design (except for photomasks manufacturing) and later stages after chips exit the final production line.
We also focus on silicon-based chips, excluding de facto germanium, gallium, indium phosphide-based and other substrates that are used for different applications (power electronics, lasers, etc.
The functional unit is yet to be determined.
Historically, the environmental footprint of integrated circuit manufacturing has been defined by the following UF: the production of cm²/die. This functional unit is calculated by determining the manufacturing footprint of m² of wafer and then accounting for various losses associated with the kerf, defect density, die format and other yield parameters.
Other proxies may be used to complement a chip’s manufacturing footprint: the number of manufacturing steps1); the average environmental footprint of each type of process multiplied by the number of steps per process2); the average footprint based on a fab’s throughput3); or a bottom-up approach for analysts with access to primary data from a fab.
Recent works also point out that allocation of impacts could differ depending of binning, i.e. differentiation of a lot according to performance, and then implying a price-based and performance-based allocation4).
The cm²/die ratio depends entirely on knowing the exact die area, which is not automatically provided by manufacturers. This area is not directly accessible or visible as it is covered by the packaging. Consequently, an inaccurate estimate of the die area can cause significant variations in the footprint of a chip and the equipment that houses it. Proske et al provide a database of die-to-package ratios based on direct measurement or radiographic imaging of 189 chips5) to circumvent this issue.
The reference flow is yet to be determined.
The reference flow is historically a m² of manufactured wafer for wafer and front-end manufacturing. Photomask and back-end manufacturing have yet to be defined.
Regarding front-end manufacturing, Pirson et al provide a comprehensive review of historical and actual trends regarding environmental impacts of IC production per technology nodes and based on scientific literature and LCA databases6).
Furthermore, several datasets and models already exist from LCA databases such as:
Several parametric models are also available from industry specialists:
A silicon wafer is made from single-crystal silicon, which is derived from high-purity sand.
The process involves purifying the sand, melting it, and re-crystallizing it into a large, pure silicon ingot (Czochralski process). This ingot is then sliced into thin discs, which are polished to a mirror-like finish to form the wafer.
Once the disks are obtained, the main steps are:
The resolution of each photolithography technique directly impacts the achievable node size, with EUV technology leading the way in current semiconductor advancements.
For logic wafers: Intel, AMD, Qualcomm. They rely on foundires like TSMC and GlobalFoundries. For memory wafers:
The semiconductor industry is comprised of three main types of companies, each playing a distinct role:
Additionally, it is increasingly common for IDMs to outsource part of their production to foundries. Some IDMs, like IBM and Samsung, also offer foundry services to fabless companies.
The Asia Pacific region is the largest semiconductor industry region, with China as the leading single-country market. Despite a 15% decline since 2021, China's market remains critical. The Americas and Europe show resilience, with Europe's market growing by 4% in 2023. Japan experienced a slight decline, reflecting the competitive nature of the market. These regional dynamics underscore the semiconductor industry's global interconnectedness and sensitivity to geopolitical and economic factors. 8)
The global semiconductor manufacturing industry is highly concentrated in East Asia, the United States, and the European Union with these regions housing over 90% of all facilities. East Asia, with 292 sites, represents around 66% of the total, dominating global chip production. Taiwan alone produces 60% of the world's semiconductors and 90% of the most advanced ones, especially logic chips. Additionally, Korean companies hold a commanding 60% share of the memory chip market. This map visualizes the geographic distribution and processing capabilities of semiconductor manufacturing facilities.
Cutting-edge wafer fabrication, particularly at nodes of 7nm or less, is predominantly found in South Korea and Taiwan. Europe, on the other hand, lacks these advanced facilities and as a consequence lags in adopting more mature technology nodes. A tiny share of wafer capacity for nodes between 10nm to 20nm exists in Europe, mainly due to Intel’s fabs in Ireland and Israel, which are currently not available for contract manufacturing. Furthermore, European fabs like those from STMicroelectronics and GlobalFoundriesoperate nodes from 22nm to 40nm, but the majority of Europe’s capacity (almost 50%) consists of older nodes of 180nm or larger, used extensively for automotive and industrial applications. 9)
TBC
Not in the scope
Not in the scope
The main impacts are:
To help you understand the complexity of semiconductor manufacturing processes, we have created a glossary to bring together key terms and acronyms.
TBC