Chemical use in the semiconductor manufacturing industry

ABSTRACT

Background: The semiconductor industry is known to use a number of chemicals, but little is known about the exact chemicals used due to the ingredients being kept as a trade secret.

Objectives: The objective of this study was to analyze chemical use using a safety data sheet (SDS) and chemical inventory provided by a major semiconductor company, which operated two factories (A and B).

Methods: Descriptive statistics were obtained on the number of chemical products and ingredients, photoresists, and carcinogens, classified by the International Agency for Research on Cancer (IARC), as well as trade secret ingredients. The total chemical use per year was estimated from chemical inventories mass (kg).

Results: A total of 428 and 432 chemical products were used in factories A and B, respectively. The number of pure chemical ingredients, after removing both trade secret ingredients and multiple counting, was 189 and 157 in factories A and B, respectively. The number of products containing carcinogens, such as sulfuric acid, catechol, and naphthalene was 47/428 (A) and 28/432 (B). Chemicals used in photolithography were 21% (A) and 26% (B) of all chemical products, and more than 97% among them were chemicals containing trade secret ingredients.

Conclusions: Each year, 4.3 and 8.3 tons of chemicals were used per person in factories A and B, respectively. Because of the high level of commercial secrecy and the use of many unregulated chemicals, more sustainable policies and methods should be implemented to address health and safety issues in the semiconductor industry.

KEYWORDS:

• Semiconductor
• chemical
• SDS
• trade secret ingredients
• photolithography

Acknowledgments

This work was supported by BrainKorea21 (BK21) Plus project of the National Research Foundation (NRF) [Grant number 2580-20180100].

Disclosure statement

All authors have no conflicts of interest to declare.

SUPPLEMENTAL DATA

The supplement data can be assessed here.

Additional information

Funding

This work was supported by the BrainKorea21 (BK21) Plus Project of the National Research Foundation (NRF) [Grant number 5280-20180100].