Find the answers you need before purchase

A stack (HXS-2, HXS-30) is the electrochemical assembly of electrodes, MEAs, bipolar plates and end plates; you design and attach the power, electrolyte circulation, gas separation and safety systems (BoP). A system (HXB-V1) integrates that stack with the BoP into a complete hydrogen generator that runs on power and water alone. Electrodes (HXP-an, HXP-ca) are the component materials that make up a stack.

200 × 300 mm at 300 µm is a thin large-area sheet suited to electrode cutting, MEA fabrication and stack-level evaluation. 100 × 100 mm at 3 mm is a thicker small-area format for lab-scale testing, single-cell evaluation and catalyst screening. HXP-ca is 260 ± 20 µm thick at 200 × 300 mm, with a choice of Pt/C or PtRu/C. Catalyst loading is 0.2 mgPGM/cm²; for PtRu it is based on the combined mass of Pt and Ru. Standard formats ship in approx. 2 weeks; custom thickness or large sizes take up to approx. 2 months because of substrate procurement and synthesis.

The standard build uses SUS316L end plates and nickel-based metallic bipolar plates, and we can build them in titanium or another stainless steel instead. Because it is made to order, we fix three things with you at the enquiry stage: which parts change (end plates, bipolar plates), the requirement behind the change (electrolyte type and concentration, temperature, potential window), and quantity and delivery. Two points up front. This is separate from changing the flow field: HXS-2 and HXS-30 bipolar plates are made with dedicated tooling, so the geometry stays as it is and only the material changes, while a custom flow-field geometry is available on HXS-0. And because it is not the standard build, our published performance and durability data do not carry over unchanged, so we state the validated scope together with the quotation.

We have supplied hardware for non-electrolysis use. Let us be clear about the scope. The anode (HXP-an) and the cell and stack hardware can be applied to other electrochemical reactions, and we build the active area, cell count and flow field for your purpose. The cathode stays in your design scope: HXP-ca is a Pt/C or PtRu/C electrode for the hydrogen evolution reaction, and we hold no performance data for it under CO₂ reduction or chlorine evolution. For strongly corrosive electrolytes or oxidising conditions we start from material compatibility, and we can build the end plates and bipolar plates in a different material where needed. Tell us the reaction, the electrolyte composition and the target current density and potential window, and we will first set out what we can take on and where your scope begins.

The HXS-0 test cell takes an active area of 1–25 cm² set by the gasket window, and configures from a single cell up to a 20-cell short stack. 3 × 3 cm (9 cm²) is inside the standard range and is built as is. The standard flow field is single-serpentine, and the geometry can also be custom-made for the test objective. MEAs are not included in the standard supply; on request we supply it with HXP-an, HXP-ca and a membrane. Fabrication starts once the active area, cell count and flow field are fixed, so send those three with your test objective and we can confirm delivery straight away.

We use HXB-V1 in house as a stack evaluation station, where it handles electrolyte flow, temperature control and data logging. However, its built in power supply is optimized for the operating window of our 2 kW stack, so voltage and current cannot be swept freely. For i-V curves or evaluation of low voltage short stacks and single cells, we recommend pairing it with a separate programmable power supply.

HXB-V1 is an electrolyzer: the power unit pushes current into the stack and it manages an alkaline electrolyte loop, temperature control and data logging. A fuel cell bench instead needs an electronic load that draws current, plus mass flow controllers, humidification and back pressure control, none of which are present. Owning a potentiostat does not substitute for the gas supply, humidification and back pressure hardware. If the goal is electrolysis cell or stack evaluation, we can propose a single cell or short stack setup instead.

HXS-0 uses a single-serpentine flow field as standard, and the flow-field geometry can be custom-made within the 1–25 cm² range for the test objective. HXS-2 and HXS-30 use bipolar plates made with dedicated tooling, so internal flow-field changes are not supported. Tell us your test objective and operating conditions and we will define either the HXS-0 custom scope or what can be evaluated with the production-stack flow field.

MEA is not included in the standard supply. On request it can be supplied with an MEA comprising HXP-an, HXP-ca and a membrane.

Higher flow is not impossible, but we have not operated and validated it ourselves, so we cannot guarantee performance or lifetime numbers there. The bottleneck is the internal flow field and manifold cross section rather than external piping, and higher flow raises the differential pressure, widening the anode to cathode pressure difference. If high flow is needed for an external process, we recommend separating a stack loop (approx. 4 L/min) from a process loop on a shared electrolyte tank.

The HXP-an electrode figures, such as the 98.7 mV/dec Tafel slope, are three-electrode measurements in 0.3 M KOH at room temperature. The HXS-0 catalogue i-V curve is at 4 cm² active area in 1 M KOH at 80 °C, the representative condition for the stack curves is 0.1 M KOH at 50 °C, and the 88% efficiency figure for HXS-2 is a beginning-of-life value in 0.3 M KOH at 60 °C. Those last two differ because of when they were measured: our current recommended operating condition is 0.1 M KOH at 50 °C, while the 88% figure was taken under the earlier condition. Electrolyte concentration and temperature both shift the cell voltage at a given current density, so the same electrode gives different numbers under different conditions. There is no separate activation step: bring the cell to temperature, run it at rated current, and start measuring once the voltage settles. Electrode orders ship with our recommended operating and measurement conditions. If your result is far off, send us the conditions you used: electrolyte type and concentration, temperature, membrane and ionomer, assembly torque, active area, and the measurement mode including galvanostatic or potentiostatic and the scan rate. Our engineer will compare it against the same condition and work through the cause with you.

A research group at Dongguk University published results using our electrodes in Small (S. Jo et al., Small, 2025, DOI 10.1002/smll.202411883), a study evaluating a cathode catalyst in both PEM and AEM electrolyzers. If you are comparing stack specifications across suppliers, a 2026 review of industrial AEM stacks (Current Opinion in Chemical Engineering 51, 101218) compares nine commercial AEM stack suppliers and includes our HXS-2. The conditions listed for us there, 0.3 M KOH at 60 °C, are the beginning-of-life condition for our published 88% efficiency figure, while the i-V curve and representative operating condition for the same product are 0.1 M KOH at 50 °C. One product carries two conditions, so check which one a number belongs to. Our own data is published in the product datasheets and the long-term data note with the measurement conditions stated, so please check those conditions, since figures taken under different conditions are not comparable. Where unpublished detailed data has to be exchanged both ways, we sign an NDA.

A higher stack temperature carries more water vapour into the hydrogen. Since moisture is the main impurity, removing it with a dryer gave ≥ 99.999% within the operating range we have validated.

Our published data and supply specification are based on 0.1–1 M KOH, with 0.1 M KOH at 50 °C as the representative condition. 0.1 M KOH sits around pH 13, so pH 11 is a hundred times lower in alkali concentration. We have not tested that region and we hold no established lower operating limit. At low concentration the electrolyte conductivity falls, so cell voltage rises at the same current density and the ion-transport conditions change, which means our published performance and durability data do not carry over. For that reason we will not tell you at this stage that it works. Send us the target pH and the reason you need it, such as a water-quality, material or process constraint, and we will come back on whether and how it can be tested.

The HXB-V1 supports both 0 barg and 10 barg operation within one supply specification, and the required pressure remains selectable in the system software after delivery. These are not separate hardware configurations fixed at the time of order.

Electrolysis can begin before the electrolyte reaches its target temperature, so you do not have to wait for warm-up before hydrogen is available.

4 barg. In a fault condition the design routes hydrogen out through the vent line.

Hydrogen permeation through the membrane is set by pressure and temperature and is nearly independent of current, while oxygen production scales with current, so the same permeation is diluted into less oxygen and the ratio rises. Measured on a 23-cell HXS-2: 0.39% at rated load and 0.66% at 40% load at ambient pressure, and 0.98% and 1.75% at the same two points at 6 barg. The multiple matters more than the absolute value. It is 1.25 to 1.43 times at 60% load and 1.67 to 1.94 times at 40% load, and it stays about the same across pressures. How much headroom is left is set by pressure: at ambient there is room to spare at 60% load against the 2% control limit, while at 6 barg and 40% load only 0.25 percentage points remain. Minimum load under pressure is therefore a function of pressure. For deeper turndown we recommend a minimum load limit, depressurising at low load and a hard interlock on the oxygen side. That sweep runs 49 points from ambient to 6 barg; we do not have one at the 10 barg rating, and we will run it to your conditions if you need it.

HXS-2 is a compression assembled stack, so it can be opened, disassembled and reassembled, but the bolting protocol compresses to a 5 ton gauge reading and then bolts in two torque steps of 5 Nm and 7 Nm while the compressed length is measured on four sides. You need the press, an aluminium press block, a torque wrench and a vernier caliper. Tell us before you open the stack if you do not have a press and we will work out an approach with you. Gaskets are the most common failure: the O-ring can ride out of its seat and the MEA gasket shrinks inward, so both need seating after clamping. We supply a reassembly guide, and performance and leak-tightness conditions after reassembly are agreed case by case.

KOH is not consumed during operation; only water is, and that make-up is automatic. So replacement is decided from condition and performance trend. Consider replacing when any of the following appears: cell voltage rising steadily at the same current compared with before, a widening spread between individual cell voltages, or a change in electrolyte colour or clarity or visible sediment. If hydrogen in oxygen (H₂-in-O₂) exceeds 2%, stop and inspect. The interval varies widely with operating conditions and gets shorter at higher temperature, higher concentration and higher current. Tell us your operating conditions if you need a recommended range for them.

Five stacked units reach roughly 10 kW, about 5 kg per day. Beyond that we propose a custom design around the HXS-30 stack. Tell us your required flow rate and daily operating hours and we will suggest a configuration.

Units run in parallel and their outlets join one shared hydrogen pipe, with the dryer and storage downstream shared between them. Wiring is one LAN cable from each unit to a network hub and one cable from the hub to a laptop. The multi-unit screen in the desktop app lists every connected unit with its status, and both a command to all units and control of a single unit happen on that same screen. The structure is worth stating plainly: no unit becomes a hardware master, every HXB-V1 is a complete generator, coordination is done in software, and the operator decides how many units run at what output. The software design limit is 50 units, which is a design limit rather than a number we have run.

Electrolyte is 0.1 M KOH in deionised water, nominal flow 4 L/min with 1 to 10 L/min allowable, temperature 50 °C, and both the anode loop and the hydrogen outlet at or near atmospheric pressure with a free vent. The loop order is reservoir, filter, flooded-suction variable-speed pump, heat exchanger, stack-inlet temperature sensor, flowmeter, stack, then outlet temperature and anode pressure back to the reservoir. The stack itself drops about 0.11 bar at 4 L/min, so before adding fittings and pipe losses a variable-speed pump with at least 0.3 bar available head at 4 L/min is a reasonable sizing basis. One point to be explicit about: the 10 barg figure for HXS-2 is the hydrogen-side maximum operating pressure, not a 10 bar allowable differential pressure. A bare stack has no onboard PLC or standard communications interface, so your controller handles flow, both pressures, stack current and voltage, temperatures, leak detection and emergency stop. We do not yet have an approved differential-pressure limit for a bare stack covering both steady and transient states. The loop schematic, port specifications, heat load and the minimum instrumentation list are in the stack integration guide. It is published on the site and also available as a Korean or English PDF.

There is no minimum order quantity, so a standard format can be ordered from a single sheet. Custom substrate procurement can take up to approx. 2 months, so it is faster to start with the standard 200 × 300 mm (300 µm) anode, which uses the same NiFe-LDH coating, and evaluate coating behaviour and stability first. Note that a standard format suits evaluation of the coating and OER performance rather than of the substrate itself. A custom quotation can be prepared in parallel. Where detailed performance data or test results need to be exchanged, we sign an NDA. We do not run a free sample programme.

The anode uses a non-precious-metal catalyst based on NiFe-LDH, while the cathode uses a Pt- or PtRu-based catalyst at 0.2 mgPGM/cm². We use several membrane and cathode-catalyst brands rather than being tied to one supplier, and disclose the configuration used in the delivered stack. HXP-an is synthesized in house, so replacements are supplied directly. You can later source parts yourself or order our electrode and MEA sets. Synthesis recipes and ionomer composition remain confidential.

STEP files for the stack and the 19 inch system enclosure are available, sent by link when the files are large. We can include dimensional data for installation planning as well.