Rooms Inside a Crystal, a Pipeline Without Pipes: Atomis Takes Nobel-Winning MOF Science to Thailand

A ceremony at Thailand Science Park

On April 23, 2026, three signatures met at Thailand Science Park in Pathum Thani. Dr. Sumittra Charojrochkul signed for the National Energy Technology Center, known as ENTEC and housed under Thailand’s National Science and Technology Development Agency. Daisuke Asari signed for Kobe-based Atomis. Kenjiro Yamanaka signed for Yachiyo Engineering, the Tokyo consultancy that has spent years translating infrastructure plans into projects across Asia. Japan’s ambassador to Thailand, Masato Otaka, witnessed the ceremony.

The document was a memorandum of understanding on applying advanced materials to energy technology. ENTEC described gas storage, carbon capture, hydrogen and biomethane. Yachiyo put greater emphasis on CubiTan, Atomis’s next-generation gas container, and on a “Smart Gas Network” that would deliver, meter and manage gas without a city pipeline.

It was an important institutional act. It was also only a beginning. The releases disclosed no purchase order, investment value, Thai deployment site, number of vessels, operating pressure, storage capacity, certification route, lifecycle-carbon calculation or commercial start date. Calling the agreement a launch would move the story ahead of the evidence.

Why Thailand is interested

Yachiyo’s account starts with a physical absence. Thailand’s urban gas-pipeline system does not reach every household, and bottled liquefied petroleum gas remains a common source of cooking heat. A cylinder network can go where buried steel cannot, but it creates its own chain of filling plants, trucks, inventories, deposits, missed deliveries and heavy containers.

Energy security sharpens the issue. LPG prices have required government support, and Asian supply chains remain exposed to shipping and geopolitical shocks. Thailand also has agricultural and municipal feedstocks that can produce biogas. Upgrade that gas to biomethane, store it efficiently and deliver it without building a conventional pipeline, and a waste stream can become local energy.

That is the proposition. It is not yet the result. The origin of the methane, the energy used to purify and compress it, the distance travelled, leakage throughout the chain and the fate of the adsorbent all belong in the final balance sheet.

The gas business is a logistics business

A gas molecule is useful only when it arrives at the right purity, pressure, quantity, place and price. Pipelines solve the transport problem by creating permanent networks. Liquefied gases solve it by chilling molecules into dense liquids, at the cost of refrigeration and specialized handling. High-pressure cylinders solve it by compression, at the cost of weight, shape and pressure.

Atomis is proposing a fourth layer: adsorption. Instead of relying only on the empty volume of a vessel, it fills the vessel with a material whose internal surfaces attract gas molecules. A useful adsorbent can raise the amount stored at a given pressure or achieve a target amount under less severe conditions. The exact gain depends on the MOF, gas, temperature, pressure, packing density and usable working capacity—not simply the most flattering laboratory adsorption number.

A crystal that behaves like a city of rooms

A metal-organic framework is built from metal ions or clusters connected by organic molecules. The metal nodes act like junctions; the organic linkers act like beams. Repeated across a crystal, they form an ordered network of molecular-size rooms and corridors.

The analogy is tempting but incomplete. A room’s usefulness depends on the door, wall chemistry and how readily a guest enters and leaves. Chemists can vary the node, linker, pore size and internal environment, creating frameworks that favor particular molecules. Some are rigid. Some flex when a guest arrives, opening like a lung. That designability separates MOFs from traditional porous materials such as activated carbon and zeolites.

Before the Nobel: wooden balls and an unfashionable void

The official Nobel history begins in 1974 with University of Melbourne chemist Richard Robson preparing wooden ball-and-stick models for students. The drilled holes encoded bonding directions. Robson realized that molecules themselves might be assembled into predictable, extended structures. In 1989 he reported a diamond-like copper network with spacious cavities. It was elegant and fragile.

In Kyoto, Susumu Kitagawa was studying coordination polymers. A 1992 copper complex formed a honeycomb network with organic molecules sitting in its spaces. Where many chemists saw unwanted emptiness, Kitagawa saw usable volume. Grant reviewers often saw neither purpose nor durability. Porous crystals already existed; zeolites and activated carbon worked. Why build a delicate imitation?

The 1997 experiment that changed the argument

Kitagawa’s answer arrived in 1997. His group created three-dimensional frameworks built from cobalt, nickel or zinc ions and 4,4′-bipyridine. After water was removed, open channels remained stable. Methane, nitrogen and oxygen could enter and leave without destroying the framework.

This mattered because porosity was no longer merely visible in a crystal structure. It performed a repeatable function. In 1998 Kitagawa argued for an even stranger possibility: porous frameworks need not be rigid. They could be soft, changing form in response to guest molecules while remaining reversible. Omar Yaghi’s highly stable, exceptionally spacious MOF-5 followed in 1999. Together with Robson’s originating architecture, those advances made a field.

The 2025 prize—and what it did not say

The Royal Swedish Academy of Sciences awarded the 2025 chemistry prize to Robson, Kitagawa and Yaghi “for the development of metal-organic frameworks.” By then chemists had created tens of thousands of MOFs for gas storage, carbon capture, water harvesting, separation, catalysis and other uses.

Kitagawa is Atomis’s scientific adviser, and the company is explicitly rooted in his research. That makes “Nobel-linked” accurate. “Nobel-approved product” would not be. The committee evaluated foundational discoveries. It did not inspect Atomis’s composite vessel, production batch records, sensors, refill station, Thai safety case or business model.

From MaSaKa-NeXT to Atomis

Masakazu Higuchi, a Kyoto University MOF researcher, incorporated MaSaKa-NeXT on February 10, 2015. Kyoto University now describes that company as the predecessor of Atomis. The name changed in 2017, when Daisuke Asari—trained in coordination chemistry at Kyoto and seasoned in pharmaceutical R&D and new-business development at Aventis, Sanofi and Nitto Denko—took the enterprise toward commercialization.

The company’s location tells its own story. Atomis began as a Kyoto University startup. It raised ¥350 million in Series A financing in 2019 to build pilot production and accelerate CubiTan. A ¥1.2 billion Series B followed in 2021. In 2022 the company started constructing a headquarters, R&D center and pilot plant on Kobe’s Port Island; by 2023 it said the plant could supply as much as 20 tonnes of MOF a year.

The valley between a gram and a tonne

New materials fail commercially for reasons that rarely appear in Nobel lectures. A framework may adsorb brilliantly as a pristine powder but lose performance when shaped into pellets. A binder may block pores. Water, sulfur compounds or repeated thermal cycles may degrade the structure. A synthesis that works in a flask may require too much solvent, heat or purification at industrial scale. A high-capacity material may simply cost too much.

Atomis’s response has three layers. Its POROS database, the company says, covers more than 100,000 porous crystalline materials and adds practical variables such as cost and stability to published properties. Its production work seeks lower-cost, lower-environmental-burden synthesis than conventional solvothermal routes. Its impact business tries to sell systems and services rather than powder alone.

What CubiTan is meant to be

CubiTan combines four ideas. A reinforced-plastic pressure vessel reduces mass. A cube-like outer form makes the package easier to stack than a round steel cylinder. MOF adsorbent inside provides additional internal surface for methane. An IoT module connects each unit to a software-managed delivery system.

The partners’ dedicated CubiTan site says a vessel weighs 12 kilograms and markets reductions of 90% in size and 75% in weight versus “traditional alternatives.” Those are company claims on a project website; the comparison basis and third-party validation are not supplied there. The Thailand releases themselves provide no numeric CubiTan specification. A responsible pilot must publish the baseline vessel, gas quantity, pressure, temperature, cycle count and usable delivered energy behind any comparison.

Adsorption is not magic compression

Gas molecules naturally accumulate on a surface when molecular attractions make that state favorable. A porous adsorbent multiplies the available surface, allowing more molecules to gather. MOFs make the surface tunable. For methane, the goal is a framework that takes up useful quantities during filling and releases most of them at the customer’s operating pressure.

The difference between total adsorption and working capacity is decisive. Gas that remains trapped at the end of discharge adds impressive laboratory capacity but does not cook a meal. Packing loose crystals into a real vessel reduces idealized volumetric performance. Heat released during filling and absorbed during discharge can also change capacity and flow. CubiTan therefore has to be evaluated as a thermodynamic system, not as a powder plus a box.

The digital cylinder

Atomis and Yachiyo describe a supporting family of products. CubiBase is the filling or charging stand and data relay. CubiLink carries data during vehicle transport. CubiApp supports hand-carried transfers. CubiLoop is the web platform intended to integrate shipment, gas, safety, container and payment management.

This layer may be as commercially important as the MOF. Conventional cylinder systems often suffer from poor visibility: a supplier knows how many containers it owns but not exactly where each one sits or how much gas remains. Remote inventory and predicted replacement can reduce emergency deliveries, unused stock and lost assets. It can also create cybersecurity, connectivity, privacy and vendor-lock-in questions that a materials laboratory cannot solve alone.

A Japanese commercial foothold—and a crucial distinction

On April 13, 2026, Atomis announced that a domestic CubiTan service had begun. The company said it had approval for four gases: nitrogen, argon, helium and liquefied carbon dioxide. It targeted R&D users and industrial settings that handle relatively small quantities of multiple gases, offering the compact stackable vessel with digital monitoring of pressure, temperature, impact and location.

That is stronger evidence than a prototype display. It shows that Atomis has moved a version of the container-and-IoT platform into a Japanese service. It does not show that Thailand’s proposed energy configuration is already approved or commercial. The four Japanese gases are not methane; liquefied CO₂ uses different storage physics; and Atomis’s service notice does not say those domestic products depend on MOF adsorption. The Thai claim must therefore stand on methane-specific vessel, adsorbent and network evidence.

A virtual pipeline, node by node

The phrase “virtual pipeline” captures the proposal. Biomethane could be produced near farms or waste facilities, purified, placed in containers, carried by truck and delivered to users beyond the pipe grid. Software would decide when to refill and where to route the next unit. The network would expand by adding containers and hubs rather than trenching streets.

That modularity is attractive for islands, remote communities, disaster response and fast-growing districts. It may be less attractive where demand is dense and steady enough to justify a pipeline. Trucks still need fuel, drivers and roads. Containers still require inspection and recertification. The winning infrastructure will depend on local geography and throughput, not on novelty alone.

Indonesia was the proving ground

Thailand is not Atomis and Yachiyo’s first Southeast Asian test. In 2024 they publicized an Indonesian demonstration aimed at transporting natural gas and biomethane. Japan’s JETRO Asia DX program supported early work, and a later Ministry of Economy, Trade and Industry initiative funded feasibility studies across Indonesia, Malaysia and Thailand.

The CubiTan project site said in 2026 that proof-of-concept testing had “just started” in Indonesia and that permits would follow the results, with commercialization targeted for 2027. That language places the project before regulatory approval and broad sales. The Indonesian experience can inform Thailand; it cannot substitute for Thai testing, gas composition, codes or customer economics.

Why the Thai partnership has three parties

The division of labor reflects a truth about deep technology. A new material rarely reaches society through its inventor alone. It needs a local laboratory, a system integrator, regulators, financiers, gas suppliers, truck operators and customers who value reliability more than scientific elegance.

“Clean energy” needs a boundary

The partners use the language of clean energy and decarbonization. That description may prove justified for biomethane made from wastes that would otherwise release methane, particularly if the system displaces fossil LPG and captures leakage. It is not automatic for fossil natural gas. Burning methane still produces carbon dioxide, and leakage can erase part of its advantage.

The International Energy Agency says methane remains in the atmosphere for about 12 years but traps far more heat while present; it attributes nearly 35 million tonnes of 2024 methane emissions to natural-gas operations. Its biomethane outlook estimates current plants can leak between 2% and 5.5% of output. For CubiTan, leak detection is therefore not a decorative “smart” feature. It belongs at the center of the climate claim.

Biomethane’s conditional opportunity

Biogas emerges when microorganisms break down manure, food waste, sewage or crops without oxygen. Remove carbon dioxide, water and contaminants and the methane-rich product can approach natural-gas quality. Capturing manure or landfill methane can prevent direct atmospheric release; using the resulting energy can displace another fuel.

But feedstock matters. Waste-derived biomethane and purpose-grown energy crops do not share the same land, fertilizer or carbon effects. Upgrading consumes energy. Digestate must be managed. Leaks can occur at the digester, upgrader, filling station, valve and stove. A Thai demonstration should measure the entire chain rather than attaching a green label at the cylinder.

The economic test is delivered heat

A household does not buy adsorption capacity. It buys reliable cooking heat. A fair comparison must divide every capital and operating cost by useful energy delivered: MOF, vessel, valve, sensor, filling equipment, data service, truck, labor, maintenance, inspection, financing and end-of-life treatment.

CubiTan could improve that equation through lower handling weight, stackability, fewer unnecessary trips and better asset utilization. It could worsen it if MOF replacement, specialized filling, connectivity or small production volume adds too much cost. The MoU states no target tariff, customer price or subsidy assumption. Those omissions are normal at this stage and decisive later.

Safety moves from pressure alone to a system

Lower operating pressure at equal delivered gas could reduce stored mechanical energy, but the public Thailand materials do not disclose operating pressure or comparative test data. Composite vessels introduce their own inspection, impact, fire and aging requirements. MOF compatibility with gas impurities and repeated cycling must be established. Valves, relief devices and end-user appliances remain critical.

Software adds detection but also failure modes. A sensor may drift. A network may go offline during a flood. A false reading may trigger an unnecessary trip; a missed reading may leave a user without gas. Cybersecurity matters when billing, inventory and delivery instructions share one platform. Safe design needs useful behavior when every digital layer fails.

The humidity problem—and the impurity problem

Real gas is not an ideal single molecule. Biomethane can carry water, carbon dioxide, hydrogen sulfide, siloxanes and other trace compounds depending on its source and upgrading. Some MOFs tolerate moisture; others lose structure or active capacity. A strongly adsorbed contaminant can occupy the very pores meant for methane.

Thailand’s heat and humidity make this more than a laboratory footnote. ENTEC’s role should include representative local gas streams and environmental conditions. Long-cycle tests must show not only initial uptake but capacity retention, release behavior and safe handling after months or years.

Manufacturing is part of the invention

Atomis says its 2023 Kobe pilot plant can produce up to 20 tonnes of MOF annually. That is a meaningful bridge between gram-scale chemistry and commercial supply, but “up to” is capacity, not utilization, qualified output or cost. The company has not publicly broken out how much production is allocated to CubiTan.

For a cylinder network, every batch must meet specifications that matter at device level: adsorption isotherm, density, particle strength, water content, impurity profile and cycle durability. Quality control must connect those measurements to a safe usable gas quantity. A Nobel-winning structure synthesized inconsistently is not infrastructure.

Another branch: carbon capture

Atomis also develops MOFs for separation and carbon capture. In April 2026 it said a collaboration with Kobe Steel and Nagase had completed a 30-kilogram-per-day CO₂ capture demonstration and was examining tonne-scale work. That project supports the broader case that Atomis can formulate materials and work with industrial partners.

It does not validate CubiTan. Capturing CO₂ from a process stream and repeatedly storing deliverable methane in a pressure vessel involve different materials, devices, hazards and economics. The distinction is a useful measure of the white paper’s central theme: one scientific platform can create many routes to society, each of which must cross its own evidence gap.

The white paper’s case study

Japan’s 2026 Science, Technology and Innovation White Paper places Atomis immediately after its account of Kitagawa’s decades of work. It explains that MOFs can store, separate and potentially transform molecules, notes government-backed carbon-capture research, and presents Atomis as a startup connecting university science with corporate needs.

The placement is deliberate. The chapter is about the route from knowledge to innovation after Japan’s two 2025 Nobel Prizes. Atomis is evidence that a startup can form around foundational research. It is not evidence that commercialization is complete. The white paper itself closes by acknowledging barriers in R&D and business formation.

A ledger of what is known

From laboratory to Thailand: a timeline

The milestones that would turn a pact into infrastructure

What investors and policymakers should ask

The first question is not “How much gas can one gram of MOF hold?” but “How much usable gas does one certified container deliver after hundreds of realistic cycles?” The second is whether the answer remains superior after moisture, contaminants, packing density and thermal effects are included.

Then come the network questions. Who owns each vessel? Who fills it? What happens when a sensor fails? What is the customer price without a subsidy? Is the methane fossil, waste-derived or crop-derived? How much escapes? Can the adsorbent be recovered? Does a trucked virtual pipeline outperform LPG, electrification or a real pipeline for the selected community?

The promise in an ordinary delivery

Deep technology becomes real when it becomes boring. A driver arrives on time. A cylinder is light enough to handle. A valve fits. A stove lights. The bill is predictable. No methane escapes. The customer does not need to understand coordination chemistry.

That ordinariness is still ahead for Atomis in Thailand. Yet the April pact matters because it assembles the institutions capable of testing it: a Japanese materials startup, an engineering integrator and Thailand’s national energy laboratory. It places Kitagawa’s molecular rooms inside a concrete national question—how to move energy safely and affordably where pipes do not go.

A fair verdict in 2026

Atomis has more than a drawing. It has a decade-old company, financing history, a Kobe pilot plant, prototypes, government-supported development, an Indonesian proof-of-concept path and now an official Thai research partnership. The underlying MOF science is unquestionably important.

But Thailand has not yet received a proven CubiTan network. The MoU is permission to ask harder questions, not permission to skip them. If the partners publish a credible working-capacity comparison, safety record, lifecycle analysis and delivered-cost case, the small cube could become a genuine piece of distributed energy infrastructure. Until then, its most honest description is also its most interesting: Nobel-winning science at the threshold of the real world.