Cosmos Lab’s KRW 22 Billion Funding Round — A Look at the Patent Landscape for Zinc-Halogen Aqueous Batteries

What is in focus: KRW 22 billion for a tankless zinc-halogen cell

Non-lithium battery startup Cosmos Lab announced a KRW 22 billion Series A round on August 12, 2026. Its core is an integrated cell structure that uses a water-based non-flammable electrolyte and zinc/halogen-based materials without an external electrolyte tank or circulation pump. However, decades of prior patents have already accumulated around the very direction of a ‘tankless zinc-halogen cell’, so differentiation in scope of rights is likely to be decided at the level of detailed configurations.

Starting with the industry landscape, according to reports, Smilegate Investment and a number of financial and strategic investors participated in this round. The company said its integrated cell structure using a water-based non-flammable electrolyte and zinc/halogen-based materials improves safety and space efficiency, and announced that it is co-developing ESS (energy storage system) technology for AI data centers with SoftBank, targeting mass production in 2027. Around the same time, it was reported on August 13, 2026 that Ultium Cells, the GM-LG Energy Solution joint venture, would restart its Ohio plant, which had been idle since January 2026 due to declining EV demand, after seven months.

There are two tension factors. One is energy density. Aqueous batteries have clear strengths in safety, but their storage per unit weight and volume is lower than lithium-ion. The other is the patent landscape. Zinc-halogen chemistry is an old field researched since the 1970s, and even the tankless non-flow structure includes areas already turned into rights by overseas companies.

This article answers three questions. How does a zinc-halogen aqueous battery work, what does it gain and lose compared with lithium-ion, and where can a latecomer’s scope of rights stand within the flow of prior patents?

Basic principle: zinc builds up and dissolves, halogen is held at the cathode

A zinc-halogen battery is an aqueous-electrolyte chemistry in which zinc is oxidized and reduced at the anode (Zn ⇌ Zn²⁺ + 2e⁻) and a halogen (bromine, iodine, etc.) is reduced and oxidized at the cathode. On charge, zinc ions in the electrolyte build up as metallic zinc on the anode surface, while on the cathode side halide ions turn into halogen molecules. On discharge, this process runs in reverse and electricity comes out.

By analogy, the anode is a roof where snow builds up and melts, and the cathode is a sponge. On charge, zinc piles up on the anode like snow on a roof, but if it grows unevenly like sharp icicles, it can reach the opposite electrode and cause a short circuit. This is the dendrite (metal crystals growing in branch-like shapes) problem. Meanwhile, if the halogen generated at the cathode dissolves and wanders around, it causes self-discharge, so a device is needed to hold it like a sponge. Flow batteries store it in an external tank and circulate it with pumps, but an integrated cell must solve this inside the cell.

Because water is used as the electrolyte solvent instead of organic solvents, there is no flash point at all, so the ignition path leading to thermal runaway is structurally limited compared with lithium-ion. In exchange, energy density is lower than lithium-ion, making it better suited to stationary ESS and data center backup power than to electric vehicles with tight weight and volume constraints.

Comparison: lithium-ion and zinc-halogen aqueous batteries

The table below keeps the original comparison table and adds figures disclosed by the company and a cell structure row. Cosmos Lab’s figures are values stated in a company interview published in October 2024 and are not third-party verified.

Category	Lithium-ion (NCM/LFP)	Zinc-halogen aqueous
Electrolyte	Organic solvent (flammable)	Water-based (non-flammable)
Energy density	About 150-220Wh/kg	Relatively low due to aqueous nature (Cosmos Lab stated: over 100Wh/kg, over 250Wh/L)
Thermal runaway risk	Structurally present	No flash point
Cell structure	Integrated cell	Flow type (external tank and pump) or integrated non-flow cell
Cycle life	Varies by cell design	Cosmos Lab stated: over 5,000 cycles at room temperature
Main applications	EVs, small ESS	Long-duration ESS, data center backup

What this table says: the company-stated gravimetric energy density of over 100Wh/kg is below the lower end of the lithium-ion range given in the original article (about 150-220Wh/kg). In exchange, the absence of a flash point and long cycle life become differentiators in stationary storage applications.

The ‘cell structure’ row is the core of this funding news. A flow structure makes it easy to increase capacity through tank size but adds pumps and piping. An integrated structure is closer to the lithium-ion practice of stacking cells, which is advantageous for places with standardized space layouts such as data centers.

Patent perspective: two lines of filings, dendrites and halogen sequestration

A representative patent showing the foundational technology flow of zinc-halogen batteries is US patent US4181777A (“Rechargeable zinc halogen battery”). The original assignee is Diamond Shamrock Technologies S.A., the priority and filing date is September 7, 1978, and the grant date is January 1, 1980. According to Google Patents, this patent has already expired at the end of its term on September 7, 1998 (Expired – Lifetime).

Restating the gist of claim 1, it is a configuration in which an aqueous electrolyte containing zinc ions and halogen ions includes 1-50ppm of selected metal ions such as lead, molybdenum and tungsten, together with at least one of a polysaccharide and sorbitol in an amount sufficient to prevent zinc dendrite formation during charge. Dendrite suppression has been the key challenge blocking commercialization of this chemistry, and many subsequently filed zinc-halogen and zinc-bromine patents solve the same problem in different ways through additive compositions or electrode structure modifications.

The second line is halogen sequestration technology that makes a tankless structure possible. US10892524B2 (“Electrolyte for rechargeable electrochemical cell”, priority date March 29, 2016, filing date March 29, 2017, grant date January 12, 2021, current assignee Eos Energy Technology Holdings, Active, anticipated expiration November 1, 2036), filed by US-based Eos Energy Storage, claims an electrolyte for static (non-flow) zinc-halide batteries. Claim 1 is an aqueous electrolyte containing about 30-40wt% ZnBr₂, about 5-15wt% KBr, about 5-15wt% KCl and about 0.5-10wt% of quaternary ammonium agents. The specification describes quaternary ammonium salts as halogen sequestration agents and states that halogen species generated during charge are reversibly sequestered in the cathode assembly.

Insight: ‘no tank’ is a starting line, not a differentiator. The tankless, pumpless integrated structure that Cosmos Lab highlights is a clear investment message, but from a patent perspective it may be an area where novelty is hard to claim on its own, given that electrolytes for non-flow zinc-halide batteries have already been turned into rights. Therefore, a latecomer’s scope of rights is likely to be formed in detailed configurations, such as which materials and structures hold the halogen and what process makes the electrodes. In an October 2024 interview, the company said it had developed a solvent-free dry electrode process and held about 40 domestic patents, some of which were filed internationally under the PCT (Patent Cooperation Treaty), but individual application numbers could not be confirmed in this review.

Avoidance points: Claim 1 of US10892524B2 requires all of the content ranges for ZnBr₂, KBr, KCl and quaternary ammonium agents. Therefore, switching the halogen sequestration agent to a family other than quaternary ammonium, using an iodide-centered composition instead of bromide, or designing component contents outside the ranges may leave room to distance oneself from literal infringement. US4181777A has already expired, so its composition can be regarded as a free-to-practice area.

Design-around scenario (counterargument): However, if only the sequestration agent family is changed while achieving the same function (holding the halogen inside the cathode) in the same way, it may be subject to a doctrine-of-equivalents assessment. In addition, a separate Eos patent with the same title and the same priority date (March 29, 2016), US11942606B2 (granted March 26, 2024, Active), exists, so it is hard to consider FTO secured by avoiding a single patent. The claim scope of this family requires individual comparison.

Limitations and outlook

Aqueous zinc-halogen batteries have clear strengths in safety, but their low energy density means they require more site area. In applications where safety and long-duration characteristics take priority over site area, such as AI data centers and ESS, this tradeoff can instead act as a strength. The application split, with the lithium-ion camp focusing on EVs and high-energy-density uses and the non-lithium aqueous camp specializing in stationary storage, is confirmed in this case as well.

The optimistic outlook is that co-development with SoftBank leads to mass production in 2027 and absorbs the safety standard demands of data center ESS. There is also a counter-scenario. In a situation where lithium-ion facilities stop and restart with demand swings, as in the Ultium Cells case at the same time, if idle lithium-ion capacity shifts to LFP-based ESS and price competition intensifies, the cost advantage of aqueous batteries may narrow.

The strategic alternative is to build, together with customers, a framework that evaluates the value of aqueous batteries not by energy density or price alone but by ‘total installed cost including fire-response equipment’. If this framework takes hold, non-flammability can be converted into numbers and the basis of comparison may change.

Practical action: Planning and IP staff considering adoption of or collaboration on zinc-halogen ESS should first check the counterpart’s halogen sequestration method (sequestration agent family and content), and prepare an element-by-element comparison chart against the claims of US10892524B2 and the same-titled Eos patent family.

Items requiring verification

• Individual application and registration numbers and claim scope of Cosmos Lab’s patents (KIPRIS search required)
• Test conditions behind the company-stated energy density (over 100Wh/kg, over 250Wh/L) and cycle life
• The type of halogen (bromine, iodine, etc.) and the sequestration method used by Cosmos Lab
• The family relationship and claim differences of same-titled Eos patents (US11942606B2, etc.)
• The reference source for the lithium-ion energy density range (about 150-220Wh/kg) in the original comparison table

Sources

1. Money Today, “A ‘non-burning battery’ for data centers and ESS… Cosmos Lab raises KRW 22 billion” (2026-08-12)
2. Asia Economy, “Cosmos Lab raises KRW 22 billion Series A… accelerating mass production of ‘non-flammable’ non-lithium batteries” (2026-08-12)
3. THE ELEC, “[Y Insight] A battery with no fire worries! Cosmos Lab’s safe water battery” (2024-10-24)
4. News1, “GM-LG Energy Solution Ohio battery plant restarts after seven months” (2026-08-13)
5. Google Patents, US4181777A – Rechargeable zinc halogen battery (Diamond Shamrock Technologies S.A.)
6. Google Patents, US10892524B2 – Electrolyte for rechargeable electrochemical cell (Eos Energy Storage LLC)

Related Reading — All-Solid-State Is at 2027 Pilot Production, LFP Is Already at 63% · ESS 물량 따내면 끝 아니다 — LFP 배터리 생태계로 이어질까 · LG엔솔·삼성SDI, 현대차 휴머노이드 ‘아틀라스’ 배터리 수주전 — 셀 폼팩터 관점