Company history

The company begins in someone else’s laboratory. Takamine Jokichi, the chemist who was a close friend of Leo Baekeland, obtained a licence to work Baekeland’s patent, and in 1911 trial production of Bakelite started at the Shinagawa plant of Sankyo — the pharmaceutical house he was tied to, and a forerunner of today’s Daiichi Sankyo. Commercial manufacture and sales followed in 1914, a dedicated plant at Mukojima in 1919, and in January 1932 Sankyo carved the phenolic-resin business out as a separate company, Nippon Bakelite.

What the new company inherited was less a technology than a use. Baekeland’s 1907 synthetic resin was prized for electrical insulation and heat resistance exactly when Japan’s electrical and automotive industries were expanding — but by 1932 his patent was close to expiry, the feedstock was cheap, and any chemical maker could produce the same resin. Selling the material itself meant being dragged into price collapse. So Nippon Bakelite went instead after applications, industry by industry: switchboards, telephone sets, insulating bases, moulding compounds, laminates. Earning on the form the resin was worked into, not on the resin, is the habit that runs from here all the way to semiconductor encapsulant and medical tubing.

The Sumitomo connection predates the merger. In August 1938 Nippon Bakelite and Sumitomo Chemical jointly founded a synthetic-resin works at Amagasaki; that operation later became Sumitomo Chemical Kogyozai. After the war the company listed its shares in March 1949 and came under Sumitomo Chemical’s wing — so that by the early 1950s the two resin businesses shared a parent, a capital line and a feedstock supply.

In March 1955 Nippon Bakelite merged with Sumitomo Chemical Kogyozai and took the name Sumitomo Bakelite. This was not a union of equals so much as a tidying-up inside a keiretsu: both companies already answered to Sumitomo Chemical, which held the feedstock and the basic chemistry and now consolidated the downstream moulding and fabrication into one entity. That division of labour — Sumitomo Chemical upstream, Sumitomo Bakelite downstream in functional materials — is the structure the company has operated under ever since.

It also fixed the company’s size. Next to its parent, Sumitomo Bakelite remained a mid-tier chemical maker. Phenolic resin moved with the volumes of steel and electrical machinery and its price was pushed around by feedstock markets, so the way out was depth in particular uses rather than scale. The strongest position of these years was Decora, a melamine decorative laminate that the company commercialized earlier than rivals and defended on technique — the same pattern of loading value onto a use rather than a compound.

The deliberate turn came in 1962, with a central research laboratory in January and the Shizuoka plant in October. Together they marked the shift from selling general-purpose resin by the tonne to inventing application-specific functional materials in the laboratory — novolac resins, moulding compounds, laminates — and laid the technical ground for the electronics, semiconductor and automotive materials that would follow.

From the 1970s the company poured its resources into one product: epoxy moulding compound (EME), the material that encases a silicon die, shielding it from moisture and shock while preserving its electrical behaviour. As ICs and LSIs went into mass production worldwide, EME demand grew with them, and the Utsunomiya plant opened in November 1984 as the dedicated volume site — the physical form of that concentration.

Winning the world lead in encapsulant, though, was as much a question of geography as of chemistry. The unit price per package is low, so freight and inventory quickly eat the margin: the material has to be made near the customer’s assembly plant and delivered short. Sumitomo Bakelite therefore built alongside Asia’s packaging clusters — SumiDurez Singapore and a Singapore plant in 1989, Suzhou in 1995, Taiwan in 1998 — and turned location itself into the barrier a later entrant cannot easily reproduce at the same quality and lead time. In parallel it bought resin businesses in the West: Occidental Chemical’s phenolic operations in 2000 (now Durez), Goodrich’s electronic-materials research arm in 2001 (now Promerus), Fers Resins of Spain in 2003, and Vyncolit in Belgium and North America in 2005, with a North American holding company set up in 2002 to run them.

The result was the top share of the world encapsulant market, with the company’s material designed into the lines of the major chipmakers. The cost of that concentration showed at the bottom of the cycle. Consolidated revenue fell from ¥241.0bn in the year to March 2006 to ¥170.8bn in the year to March 2010, and the year to March 2009 produced an operating loss of ¥1.6bn and a net loss of $84.5M (¥8bn). Encapsulant volume tracks semiconductor output one-for-one, so when capacity had been added ahead of a demand collapse the fixed costs landed all at once. Being first in the world did not save a company that earned from a single material.

The 2009 loss set the agenda for the next fifteen years. Hayashi Shigeru, president from June 2010, pushed into emerging markets — “Japan’s chemical makers have led the world in high-value-added, leading-edge fields; the task now is Asia’s emerging markets, China above all,” he told 日刊工業新聞 in November 2010 — but the structural answer was a third earnings pillar beside encapsulant. The company bought it. Vaupell Holdings, a US moulder of engineering plastics for aircraft interiors and medical devices, was acquired in June 2014 for about $256M (¥27bn); a 20% stake in the catheter and dialysis-circuit maker Kawasumi Laboratories followed in March 2019 for roughly $32.1M (¥4bn), full ownership by tender offer in October 2020, and in October 2021 the medical business was carved out into SB Kawasumi.

Fujiwara Kazuhiko, president from June 2017, was a career insider from the bio and high-performance-plastics side who had run the HPP division. Under him the policy was stated plainly: reinvest what encapsulant earns into medical devices — catheters, dialysis circuits — and into mobility materials such as thermoset resins and components for electrification, moving the weight of the portfolio from commodity phenolics toward functional and medical materials.

From 2024 generative AI pulled the encapsulant business back up hard, through advanced packaging for training silicon. Revenue reached a record $2.0B (¥305bn) in the year to March 2025, though operating profit slipped to ¥24.8bn from ¥27.2bn on other charges, and the medium-term plan targets ¥60bn of operating profit by March 2027 with capital spending concentrated on the Asian semiconductor sites. The unresolved question is the one the 2009 loss posed. SB Kawasumi is still mid-sized by global medical-device standards and Vaupell’s contribution remains limited, so the third leg does not yet damp the semiconductor cycle — and the shape of the business after the AI peak has still to be designed.