The daily chemical giant has quietly emerged as the champion in the chip cleaning industry.

Per Tech reported on August 19 that on August 5, Kao, a leading Japanese daily chemical giant, released its half-year financial report for fiscal year 2026. In the financial results of the company widely known for laundry detergents and facial cleansers, a little-known business line has become the core driving force for growth — semiconductor precision cleaning agents.

The financial report shows that in the back-end process of semiconductors, Kao occupies about 60% of the market share in precision cleaning agents for package substrates; in the front-end process, Kao ranks first in market share for chemicals used in specific NAND manufacturing processes.

In the same period, Kao's chemical business related to semiconductor precision cleaning agents recorded net sales of 247.2 billion yen (approximately 10.44 billion RMB), a year-on-year increase of 9.4%, which is more than twice the growth rate of Kao's consumer goods business, making it the fastest-growing segment in the current period. At present, Kao's cleaning agents have entered the supply chains of leading semiconductor enterprises such as TSMC, Samsung, and Kioxia.

During the earnings call on August 5, Kao's executives stated that the company is shifting its focus to high value-added businesses such as electronic materials and semiconductor-related products, and its front-end chemical business is expanding from NAND to the DRAM and logic semiconductor sectors that are gaining higher popularity currently.

How on earth did a company widely known for detergents and personal care products take the leading position in the niche market of the semiconductor materials sector, which has extremely high entry barriers?

01. Starting from a bar of soap, entering the precision cleaning market in the 1990s

In 1887, Nagase Tomiro, the founder of Kao, opened a grocery store named "Nagase Shoten" in Tokyo, which mainly sold various Western sundries, with soap being one of the key product categories.

At that time, soap in Japan mainly relied on imports and was expensive, while the quality of domestically produced soap was not high. Nagase Tomiro spotted the business opportunity here, and three years later, he developed a facial cleansing soap with reasonable price and excellent quality. This soap was later named "Kao", and the company got its name from it.

Nagase Tomiro, founder of Kao, and Kao's soap products

The soap business soon hit a ceiling: raw materials. The core of soap is grease, and the deep processing technology of grease was in the hands of European and American manufacturers. In order to avoid being constrained by raw material supply, Kao's Tokyo plant was put into operation in 1925, realizing the large-scale production of glycerol; in 1928, Kao developed Japan's first shortening for the food industry, expanding the application of grease from soap to the food sector. The oleochemical business that originally derived from soap making gradually grew into an independent business line.

In 1951, Kao began to sell surfactants, the core component of synthetic detergents, which was also the origin of Kao's chemical business in later years. Surfactants can manipulate all phenomena that occur when liquid comes into contact with solid: wetting, penetration, dispersion, and stripping. After mastering this technology, Kao obtained the ticket to enter almost all "interface problem" related fields.

In the following decades, Kao's chemical business gradually expanded along this line: in the 1960s, its developed polyurethane catalysts and concrete admixtures were applied to the construction of expressways and high-rise buildings; the waste paper deinking agent launched in the 1970s caught up with the recycled paper wave; the copier toner adhesive in the 1980s took the ride of office automation. These products belong to seemingly unrelated industries, but their core is always interface control.

The intersection between Kao and semiconductors began with an industrial reshuffle triggered by environmental regulations. CFCs were the mainstream solvents for cleaning circuit boards and precision components at that time, with good performance, high volatility and non-flammability, and the whole industry formed deep dependence on them. However, with the signing of the Montreal Protocol in 1987, ozone-depleting chlorofluorocarbons (represented by Freon and other substances) entered the countdown to phase-out.

Under the ban, alternatives must meet several nearly contradictory conditions at the same time: capable of removing grease and flux residues, non-corrosive to metals and resins, easy to rinse clean, and the waste liquid can be properly treated. Various alternative solvents that emerged one after another had obvious shortcomings, and the industry once fell into the dilemma of "compliant products are not practical, practical products are non-compliant".

This is exactly Kao's strength. One of the core contradictions in daily chemical product development is the balance between detergency and mildness, and Kao's accumulation in raw material purification and low-residue formulas perfectly meets the purity requirements of electronic cleaning.

In 1991, Kao launched the CLEANTHROUGH water-based cleaning agent, using a surfactant compound system to replace Freon, entering the precision cleaning market and accumulating process data and customer relationships for its later entry into the semiconductor cleaning sector.

02. Products cover front-end and back-end semiconductor processes and extend to adjacent links

In the more than 30 years that followed, Kao's CLEANTHROUGH product line continued to expand along the upgrading path of electronic manufacturing. In 2001, Kao put into production polishing agents for mechanical hard disk (HDD) disk manufacturing, extending its products from cleaning to polishing processes; as chip process nodes continued to shrink, the cleaning objects of its products upgraded from circuit boards to wafers, and it gradually built a product system covering silicon wafer rinsing, particle removal, wetting improvement and photoresist stripping, with corresponding products for almost every cleaning process.

Supporting these products is what Kao calls "precision interface control technology". In principle, its core is to achieve uniformity of cleaning effect on nano-scale complex structures, while maintaining selectivity for multiple materials, and meeting the specification requirements of the semiconductor industry in terms of purity, particle control and metal impurity management.

Specifically, in the front-end wafer manufacturing process, cleaning accounts for about 30% of the total processes. Advanced processes have very low tolerance for particles, metal ions and organic residues, and excess residues are very likely to affect product yield. Traditional RCA cleaning relies on multiple cycles of strong chemicals such as hydrogen peroxide, sulfuric acid, ammonia water and hydrofluoric acid, which has problems such as damaging the wafer surface, high water and power consumption, and heavy waste liquid treatment burden.

Kao's entry point is the KS-2200 series of silicon wafer pre-cleaning and rinsing agents: after a 300mm wafer is polished with silica abrasive, the number of particles attached to its surface is estimated by Kao to reach about 7 trillion, and the silicon surface is naturally hydrophobic, making it difficult for the chemical solution to wet uniformly. The KS-2200 series improves the wettability of the silicon wafer surface through super-hydrophilic modification, making the water film spread evenly, and inhibiting particle adhesion from the mechanism, thus reducing the number of cleaning cycles.

Another key indicator of particle removal is to prevent secondary pollution. Kao's KS-3000 series is oriented to scenarios such as post-CMP cleaning, photomasks, and quartz fixtures. It improves the dispersion stability of particles through surfactants, and regulates the Zeta potential on the particle surface to form electrostatic repulsion, inhibiting the stripped silica and cerium oxide particles from re-adhering during the rinsing process.

In terms of photoresist removal, Kao's KS-7000 series adopts the mechanism of swelling stripping rather than chemical dissolution, reducing the risk of damage to the underlying metal lines; semi-aqueous wetting agents such as KS-2010 are used to improve the wettability of chemical solutions in narrow gap structures.

The packaging segment is the field where Kao has the highest market share. The line width and line spacing of high-end ABF substrates have fallen below 10 microns, with deep and narrow micro-pore structures. If drilling glue residues, electroplating residues and fluxes are not completely removed, reliability failures caused by short circuits or ion migration may occur.

The difficulty of this segment lies in selectivity: multiple materials such as copper, resin, nickel and gold coexist on the same substrate. The cleaning agent must not only penetrate into the micro-structure to remove pollutants, but also not corrode any type of base material, while itself must be easy to rinse and have extremely low ion residues.

Kao has made highly segmented products in this segment, and each product corresponds to a specific process pain point. Rosin-based solder paste and flux residues are the most common pollutants in packaging cleaning, and Kao's solution is CLEANTHROUGH 750H. This cleaning agent is designed for high-density packaged parts with OSP organic solderability preservative film, covering mainstream packaging forms such as BGA, CSP, WLCSP, PiP, and PoP, and does not damage the OSP film itself while dissolving residues.

Another pain point appears after cleaning: copper OSP pads are prone to discoloration during the cleaning process, which affects subsequent welding reliability, and 770A is developed specifically for this problem. With the expansion of power semiconductor production, cleaning objects have further extended to sintered materials and high-melting-point solders in power modules, and Kao has also configured a dedicated product line for this purpose.

Apart from cleaning, Kao's product portfolio also extends to adjacent processes. The dry film photoresist needs to be stripped in the substrate circuit patterning and solder bump forming processes, and Kao provides strippers including amine-free formulas, the latter leaving more room for customers' wastewater treatment; in 2.5D/3D packaging, stubborn residues left after temporary bonding materials are treated at high temperature or by laser were once a bottleneck for mass production, and the matching cleaning agents are also included in Kao's product list.

03. Directly participate in customer process development, seize the AI-driven capacity expansion boom

Kao has gained a firm foothold in the semiconductor industry not only by relying on products, but also by a business system built around customer production lines.

The business model of semiconductor materials has a feature: once a product passes customer verification and enters the mass production process, the cost of replacing the supplier is extremely high. Replacing the cleaning agent means re-conducting a full set of reliability verification, which has a long cycle and yield risk, so the share of leading suppliers usually has strong stickiness.

In order to make full use of this industry characteristic, Kao has set up Fine Cleaning Centers in Wakayama, Japan and Hsinchu, Taiwan, China, which open the test analysis of cleaning and stripping processes to customers, and participate in customers' process development rather than only selling chemicals.

The Hsinchu Center in Taiwan, China, which was put into use in December 2025, is Kao's first overseas cleaning center. The officially disclosed service objects include PCB, IC substrates, automotive and communication substrates, as well as advanced packaging production lines such as CoWoS, CoWoP, and FOPLP.

Kao has equipped the Hsinchu Precision Cleaning Center with high-specification equipment

Hsinchu is one of the regions with the highest density of global semiconductor manufacturing enterprises. TSMC's headquarters and its most advanced process R&D and mass production bases are all located here. Wafer fabs such as UMC, Powerchip, Macronix and Winbond are also rooted here. In the packaging and testing segment, manufacturers such as King Yuan Electronics and Chipbond Technology are gathered, and in the substrate field, Unimicron and Kinsus have also formed a complete supply cluster in northern Taiwan.

Kao chose to locate its first overseas cleaning center in Hsinchu at this point in time to seize this AI-driven capacity expansion boom. Nowadays, AI chips from companies such as NVIDIA and AMD generally adopt 2.5D/3D packaging solutions such as CoWoS, and TSMC has continued to expand its advanced packaging production capacity in the past two years, which still cannot meet the market demand.

The more packaging layers and the more complex the structure, the higher the number and difficulty of cleaning processes. As a consumable used in every key process, the demand for cleaning agents grows synchronously with the expansion of advanced packaging production capacity.

At the half-year financial report conference call for fiscal year 2026, Kao's executives revealed that they plan to continue to maintain and expand their share in the back-end processes, continue to penetrate in existing advantageous fields such as packaging substrate cleaning, and in addition expand from NAND to DRAM, logic semiconductors and other fields.

Against the background that businesses such as oleochemicals are continuously under the pressure of raw material price fluctuations, the company has clearly proposed to tilt its development focus to electronic materials and semiconductor-related products.

04. Conclusion: Cross-border players in the semiconductor sector all win for the same reason

In semiconductor manufacturing, the proportion of cleaning processes in total cost is far less than that of equipment and main materials, but its impact on yield is direct and rigid. As advanced packaging promotes the continuous increase in the number of cleaning processes, the strategic value of this type of "low-profile, high-stickiness" material business is being re-evaluated.

Kao's case is not an isolated example. Ajinomoto's ABF film, TOTO's electrostatic chuck, and Nitto Denko's dicing tape are all representative products. The main businesses of these Japanese companies are monosodium glutamate, bathroom products and tapes respectively, but each of them occupies the leading position in a niche category of semiconductor materials.

Their common point is that the underlying technologies all come from long-term accumulation in their main businesses. The competition of semiconductor materials is, to a large extent, the competition of basic chemistry background, which is exactly the part that cannot be achieved quickly.

This article is from WeChat official account "Chongxi", author: Chen Junda, editor: Xinyuan, authorized by 36Kr for release.