Among the gases used in semiconductor manufacturing, tetrafluoromethane (CF₄) is a greenhouse gas over 6,000 times more potent than carbon dioxide. A KAIST research team has developed a technology that removes this gas with high efficiency while extending the usable lifetime of the catalyst that helps break it down by harnessing the ‘power of disorder,’ in which mixing multiple metal atoms together actually stabilizes the catalyst’s structure.
KAIST (President Chung Sik Bae) announced on September 3 that a research team led by Professor Minkee Choi from the Department of Chemical and Biomolecular Engineering, working in collaboration with researchers from Samsung Electronics, has developed a new catalyst capable of removing CF₄, a greenhouse gas used in processes such as the fabrication of fine semiconductor circuits with high efficiency over long periods of use.
CF₄ is used in processes such as dry etching, in which unwanted portions of a semiconductor wafer are selectively removed to create fine circuit patterns. The problem lies in the CF₄ left over after use. Because its carbon and fluorine atoms are bound together extremely tightly, the gas does not easily decompose, and once released into the atmosphere, it can persist for roughly 50,000 years. Its impact on global warming is also more than 6,000 times greater than that of carbon dioxide.
To prevent CF₄ from being released as is, semiconductor manufacturing sites currently decompose it at high temperatures using steam and a catalyst. A catalyst speeds up chemical reactions, much like those used to reduce pollutants in car exhaust.
However, conventional catalysts have suffered from declining performance the longer they are used. This is because hydrogen fluoride (HF), generated as CF₄ decomposes, combines with moisture to create a highly corrosive environment, causing the catalyst’s fine particles to aggregate or its structure to change. When small catalyst particles clump together into larger masses, the surface area in contact with the CF₄ to be treated shrinks, and performance declines accordingly.
The research team solved this problem, paradoxically, by harnessing the ‘power of disorder.’
Mixing multiple atom types creates a complex, disordered structure that resists phase changes and remains stable. This process is called entropy stabilization. In simple terms, it is a principle in which evenly mixing multiple kinds of atoms makes it difficult for a catalyst to clump together or change into another structure.
Using this principle, the research team evenly incorporated multiple metals — aluminum (Al), zinc (Zn), gallium (Ga), nickel (Ni), and cobalt (Co) — into a single aluminate crystal structure. Aluminate is a material in which several metals are bonded around a basic framework of aluminum and oxygen. Through this approach, the team developed an ‘entropy-stabilized aluminate (ESA) catalyst’ that resists aggregation and structural deformation even under the harsh conditions of high temperature, moisture, and fluorine occurring together.
The performance gap was clear. The new catalyst’s intrinsic activity for decomposing CF₄ was approximately 2.3 times higher than that of a conventional alumina catalyst. Notably, in an accelerated test conducted at about 800°C for 150 hours, the CF₄ conversion of the conventional alumina catalyst dropped from 93% to 48%. The new catalyst, by contrast, maintained a high level, declining only from 98% to 92%. This demonstrated that the catalyst can remove CF₄ with high efficiency while sustaining its performance over extended periods.
The researchers also revealed the decomposition mechanism of CF₄. To do this, they used oxygen isotopes, which allow the movement of oxygen atoms to be tracked. In simple terms, this involves attaching a ‘tag’ to oxygen atoms so that where the oxygen comes from and where it moves to during the reaction can be traced.
The results confirmed that the catalyst first uses the oxygen within its own structure to decompose CF₄, and that the reaction continues as surrounding steam replenishes the oxygen that has been depleted. In effect, the catalyst functions as a kind of ‘oxygen refill system,’ in which steam restores the oxygen the catalyst draws upon. Through this, the research team provided the world’s first experimental confirmation of a CF₄ decomposition process that had previously only been proposed in theory.
The significance of this research goes beyond developing a single catalyst that decomposes CF₄ effectively; it presents a new catalyst design strategy capable of achieving both high decomposition performance and a long service life at the same time. The approach is expected to be applicable to the future development of catalysts for treating a range of semiconductor process gases by varying the types and combinations of metals used.
Professor Choi said, “By applying the principle that disorder in nature can actually make a structure more stable to catalyst design, we achieved both high CF₄ decomposition performance and long-term stability at the same time.” He added, “This work is meaningful in that it presents a new materials design strategy that can be extended to catalysts for treating a range of semiconductor process gases by varying the types and combinations of metals used.”
The study was led by Dr. Seunghyuck Chi, a postdoctoral researcher in KAIST’s Department of Chemical and Biomolecular Engineering, who served as first author, with researchers from Samsung Electronics participating as co-authors. The findings were published in June in the international chemistry journal Angewandte Chemie International Edition.
Paper title: Entropy-Stabilized Aluminate Catalysts that Break the Activity–Stability Tradeoff in CF₄ Hydrolysis, DOI: 10.1002/anie.6752036
This research was supported by the National Research Foundation of Korea (RS‐2024‐00333937 and RS‐2024‐00405261).
Facts Only
* Tetrafluoromethane ($\text{CF}4$) is a greenhouse gas over 6,000 times more potent than carbon dioxide.
* $\text{CF}4$ decomposition in semiconductor processes leaves residual gas that can persist for roughly 50,000 years.
* Semiconductor manufacturing sites currently decompose $\text{CF}4$ at high temperatures using steam and a catalyst.
* Conventional catalysts degrade because hydrogen fluoride ($\text{HF}$) combines with moisture, causing particle aggregation and structural changes.
* The research team mixed aluminum ($\text{Al}$), zinc ($\text{Zn}$), gallium ($\text{Ga}$), nickel ($\text{Ni}$), and cobalt ($\text{Co}$) into an aluminate crystal structure.
* This resulted in the development of an 'entropy-stabilized aluminate (ESA) catalyst'.
* The ESA catalyst resists aggregation and structural deformation under high temperature, moisture, and fluorine conditions.
* The new catalyst had an intrinsic activity for $\text{CF}4$ decomposition approximately 2.3 times higher than a conventional alumina catalyst.
* In accelerated testing at $800^\circ\text{C}$ for 150 hours, the conventional catalyst's $\text{CF}4$ conversion dropped from 93% to 48%.
* The ESA catalyst maintained high performance, declining only from 98% to 92% during the same test.
* The decomposition mechanism involves the catalyst using oxygen from its structure, replenished by surrounding steam (an 'oxygen refill system').
* The findings were published in *Angewandte Chemie International Edition*.
Executive Summary
A research team from KAIST, in collaboration with Samsung Electronics researchers, developed an entropy-stabilized aluminate (ESA) catalyst to efficiently remove tetrafluoromethane ($\text{CF}4$) used in semiconductor manufacturing. $\text{CF}4$ is a potent greenhouse gas and persists in the atmosphere for approximately 50,000 years, posing a significant global warming impact. Currently, semiconductor sites decompose $\text{CF}4$ using high-temperature steam and conventional catalysts. The challenge with conventional catalysts is that hydrogen fluoride ($\text{HF}$), generated during decomposition, reacts with moisture to form corrosive environments, causing the catalyst particles to aggregate, which reduces surface area and performance over time.
The new catalyst design addresses this issue by harnessing the 'power of disorder' through entropy stabilization. This was achieved by mixing multiple metal atoms—aluminum ($\text{Al}$), zinc ($\text{Zn}$), gallium ($\text{Ga}$), nickel ($\text{Ni}$), and cobalt ($\text{Co}$)—into an aluminate crystal structure. This approach creates a complex, disordered structure that resists aggregation under harsh conditions of high temperature, moisture, and fluorine exposure. Experimental testing showed the new ESA catalyst exhibited higher intrinsic activity for $\text{CF}4$ decomposition, sustaining performance longer than conventional catalysts which suffered significant loss in efficiency over extended use.
The decomposition mechanism was also investigated using oxygen isotopes, confirming that the reaction functions as an 'oxygen refill system,' where steam replenishes oxygen depleted during the reaction. This research proposes a new material design strategy for catalysts that can achieve high performance while maintaining long-term stability by designing inherent structural resistance to deactivation.
Full Take
The core innovation lies in applying principles derived from condensed matter physics—specifically entropy stabilization—to solve a materials science and chemical engineering endurance problem. The pattern observed is the successful transmutation of a known thermodynamic principle (disorder stabilizing structure) into a tangible engineering solution for catalysis. Conventional catalyst degradation stems from kinetic instability induced by environmental reactions ($\text{HF}$ corrosion), leading to loss of surface area, which is an empirical, reactive limitation. The research successfully shifts this paradigm toward designing intrinsic stability through multi-elemental mixing, suggesting that structural complexity can be leveraged as a protective barrier against functional decay.
The demonstration of the 'oxygen refill system' decomposition mechanism adds a layer of confirmation by validating the reaction pathway; it moves the understanding from merely observing performance changes to understanding the underlying physical chemistry during the process. This implies that optimizing catalytic performance is not just about increasing reaction rates, but about managing the dynamic equilibrium between chemical reactivity and physical stability across time and environmental stress. The potential implication is a general design strategy: moving away from monolithic, single-component catalysts toward complex, disordered multi-metallic systems for industrial applications where longevity is paramount.
What follows is the necessary inquiry into scaling this principle. If entropy stabilization can effectively mitigate aggregation under severe chemical assault, what are the constraints on incorporating more elements or altering the specific metal combinations? Furthermore, does this material design strategy offer broader applicability beyond $\text{CF}4$ removal to other challenging semiconductor gases? How scalable and cost-effective is synthesizing these complex aluminate structures compared to conventional methods?
Sentinel — Human
This text appears to be a factual summary of specific scientific research, characterized by detailed technical terminology and clear attribution, suggesting human curation or translation of primary sources.