Hands Off the Glovebox: An Electrochemical System for Semiautonomous Decontamination at Los Alamos

The radiological glovebox is among the most recognizable pieces of contaminated infrastructure in the nuclear enterprise, and among the most difficult for disposition. Every glovebox used in plutonium operations eventually reaches the end of its service life. It can be dispositioned in its contaminated state, but at considerable cost: disposal as transuranic (TRU) waste is orders of magnitude more expensive than disposal as low-level waste, and it consumes limited capacity at the Waste Isolation Pilot Plant in Carlsbad, N.M., the nation's only deep geologic repository for defense TRU waste. This economic reality makes decontamination to LLW limits the strongly preferred disposition path. Today, that work is typically performed by an operator reaching through glove ports to repeatedly spray and wipe interior surfaces until surveys confirm acceptable contamination levels. The process is slow, physically demanding, and dose-intensive, and the wipes, rags, and spent reagents it consumes become radioactive waste themselves.

A team at Los Alamos National Laboratory has developed a fundamentally different approach.* The Glovebox Decontamination System (GB-Decon) is the flagship module of the laboratory's Modular Electrochemical Nuclear Decontamination System (MENDS): a closed-loop, semiautonomous platform that draws a continuously regenerated cerium-based oxidant across contaminated glovebox surfaces under negative pressure, with no spraying, no wiping, and no open transfer of liquid. Preliminary results show GB-Decon bringing 304 stainless steel surfaces to LLW decontamination thresholds in approximately 20 minutes of spot treatment beneath its sealed cleaning heads (Fig. 1), and 316 stainless steel in roughly 30 minutes, with the operator's role limited to placing a cleaning head and actuating a valve.

The chemistry underlying GB-Decon has already been subjected to a demanding proof-of-concept evaluation. In validation trials using a companion MENDS bath module built on the same core technology, coupons of Haynes 25-a cobalt-based superalloy-were deliberately contaminated to approximately 10 million disintegrations per minute per square centimeter, five orders of magnitude above the release criterion of less than 100 dpm/cm². The traditional nitric/hydrofluoric acid soak-and-wipe method required more than 12 hours to reach the release limit, with coupons removed from the acid and manually wiped approximately every hour throughout processing. The MENDS system reached the same limit in 19 minutes with no direct handling: Coupons were placed in the bath in custom 3D-printed sample holders, left for the targeted time, rinsed briefly in deionized water, and left to dry (Fig. 2).

This article describes the problem GB-Decon was designed to solve, the electrochemical technology that enables it, how the module operates in practice, and what the validation results suggest for glovebox disposition and waste management across the Department of Energy complex and the broader nuclear industry.

The glovebox problem

A common challenge unites every sector of the nuclear enterprise-power generation, medicine, national security, and research: managing the waste generated during the processing and production of nuclear materials. Facilities capable of long-term nuclear waste storage in the United States are extremely limited, and disposal is expensive. Equipment used in nuclear material processing inevitably becomes contaminated. Nothing prevents a facility from disposing of that equipment in its contaminated state, but the cost difference is substantial: high-level and TRU disposal pathways are orders of magnitude more expensive than LLW disposal. Wherever contaminated equipment can practically be decontaminated to LLW limits, doing so is the economically sound choice-which makes the efficiency, safety, and waste burden of the decontamination process itself a matter of genuine consequence.

Gloveboxes concentrate this challenge. They are large, they are ubiquitous at plutonium facilities, and their interiors are, by design, the most contaminated accessible surfaces on-site. The disposition campaigns at Rocky Flats and at Hanford's Plutonium Finishing Plant demonstrated at scale how burdensome glovebox decontamination and dismantlement can be, and the growing volume of decommissioning work across the industry will only increase demand for this capability.

The methods available for this work have changed little over the decades. The predominant technique remains what is commonly known as spray-and-wipe: An operator applies a decontamination agent through the glove ports and physically wipes surfaces down, repeatedly, until surveys show acceptable levels. This approach carries three interrelated costs. The first is worker safety: Operators must directly manipulate contaminated surfaces, an action that increases radiation dose and ergonomic strain while it raises the risk of containment breaches. The second is secondary waste: Every wipe, rag, and spent reagent batch becomes radioactive waste, a process that adds to the disposal burden the decontamination was intended to relieve. The third is time: Hands-on decontamination is slow, labor-intensive work that occupies trained personnel needed elsewhere.

Los Alamos has evaluated numerous conventional alternatives for equipment used in plutonium operations. Commercial decontamination products-including spray-applied chemical agents and strippable gel coatings that are applied wet, cured, and peeled away along with the contamination-were tested on larger equipment, but these products either proved ineffective or were chemically incompatible with glovebox materials, jeopardizing structural integrity. The situation for small components is similarly unsatisfactory: The traditional method is to soak items in concentrated nitric acid combined with hydrofluoric ­acid-among the most hazardous chemical combinations in routine industrial use-with total processing times that can exceed 12 hours for heavily contaminated items.

What was needed was not another chemical formulation to apply by hand, but a different process architecture altogether: one that removes the operator from direct contact with contamination, reuses a fixed volume of solution rather than consuming reagents, and integrates with existing facility infrastructure. That architecture became GB-Decon (Fig. 3).

Established chemistry, new engineering

The chemistry at the heart of GB-Decon is well established. A literature review directed the Los Alamos team to the Catalyzed Electrochemical Plutonium Oxide Dissolution (CEPOD) process developed at Pacific Northwest National Laboratory in the 1980s for dissolving plutonium oxide and leaching plutonium from scrap and wastes. ­CEPOD uses electrochemically generated, highly oxidizing metal ions-species such as Co³⁺ and Ce⁴⁺-to convert solid metal substrates into soluble complexes through oxidative dissolution. In practical terms, an electrochemical cell generates an aggressive oxidant on demand, the oxidant attacks the contaminated surface and carries the contamination into solution, and the spent oxidant returns to the cell for regeneration.

Among the candidate oxidizers, the Los Alamos team built its system around cerium. The cerium(III)/cerium(IV) redox couple is among the most thoroughly studied in electrochemistry and has an established record in nuclear decontamination applications, and it pairs strong oxidizing power with the property that matters most for a recirculating system: The reduced Ce(III) remains stable in solution and is readily reoxidized electrolytically, repeatedly, without being consumed.

This regeneration capability is the central insight, and it is what makes a closed-loop glovebox system possible. Because the cerium oxidant is continuously regenerated electrochemically-Ce(III) reoxidized to Ce(IV) as quickly as the decontamination reaction consumes it-the process consumes electrons rather than chemicals. A small, fixed volume of solution can perform the work of many batches of conventional reagents, without generating comparable volumes of spent acid and contaminated wipes.

CEPOD's limitation was scale. The original configuration was restricted to a reaction volume of approximately 100 milliliters-suitable for laboratory demonstrations, but not for industrial application. Simply enlarging the vessel would not resolve the problem; scaling the vessel introduces handling difficulties, inefficient electrochemical turnover, and incompatibility with modular system integration. The Los Alamos approach instead circulates a small volume of cerium solution from an external reservoir, through an electrochemical flow cell that continuously regenerates the oxidant, and out to a cleaning fixture at the glovebox surface. Developing that flow cell proved to be the pivotal engineering challenge of the program.

Developing the flow cell

The team initially pursued a pragmatic path: procuring a commercial electrolyzer. Commercial off-the-shelf units offered a potential starting point for adapting CEPOD chemistry into a scalable platform. Testing, however, revealed several critical shortcomings. The commercial units capable of the required oxidation performance were physically large and heavy, making integration into a compact, modular system impractical. They demanded high electrical current, introducing unnecessary electrical hazards and complicating system design. Most significantly, their materials of construction-particularly stainless steel components-were chemically incompatible with the cerium(IV) solution. Exposure to Ce(IV) corroded and degraded the stainless steel, compromising both the electrolyzer and the chemical integrity of the solution it was intended to sustain.

The team therefore designed a custom flow cell. The design began as a detailed CAD model incorporating two internal chambers separated by an ion-exchange membrane: a primary chamber housing the anolyte, the cerium-containing solution in which Ce(III) is oxidized to Ce(IV), and a secondary chamber containing the catholyte, an acidic solution that accepts electrons during the oxidation reaction. This two-chamber configuration enabled efficient electron flow, minimized cross-contamination between electrolytes, and maintained chemical stability during extended operation. Flow channels were engineered to promote uniform fluid distribution and consistent electrochemical exposure across the electrode surfaces, and were incorporated into a removable insert designed for manufacturability and modular replacement.

Additive manufacturing made the design practical to produce. The electrolyzer body is fabricated by stereolithography using a rigid, chemically resistant engineering resin selected for the mechanical strength and dimensional stability required for reliable operation under dynamic flow and thermal conditions (Fig. 4).

The electrodes presented a further design challenge. A key driver was maximizing active electrode surface area while minimizing turbulence, which can degrade oxidation efficiency. Using a compact porous media framework that allowed geometry and flow dynamics to be optimized simultaneously, the team selected an octahedral lattice structure-a geometry that significantly increases the effective surface area available for electrochemical reactions, promotes uniform current distribution, and reduces the turbulent eddies that diminish turnover efficiency. Platinum was identified as the ideal electrode material for its conductivity and resistance to oxidation under strongly acidic, highly oxidizing conditions, but manufacturing a full octahedral lattice from solid platinum would have been prohibitively expensive. Instead, the lattice was 3D printed from grade 5 titanium, chosen for its mechanical strength, manufacturability, and compatibility with additive processes, and then electroplated with a thin but durable platinum layer. This approach achieved the desired electrochemical properties at a fraction of the cost of solid platinum, in a geometry that would be infeasible to produce by traditional machining.

Performance testing validated the design. The custom electrolyzer weighs approximately 6 percent as much as the comparable commercial unit and demonstrated a nearly sevenfold improvement in electrochemical turnover rate-achieved at substantially lower applied voltage and current, reducing power consumption and improving operational safety. A second version with nearly twice the internal height has since been fabricated and tested, and it delivers the anticipated proportional increase in electrochemical turnover rate, a result of the expanded electrode surface area and residence time in the oxidation zone. More detailed design elements, including engineering drawings, internal geometries, and performance simulations, cannot yet be disclosed due to pending intellectual property protections.

With the flow cell validated, the glovebox module could be developed around it.

GB-Decon: Semiautonomous decontamination at the glovebox face

Designing a module to decontaminate glovebox surfaces efficiently required satisfying several engineering criteria simultaneously. The decontamination solution had to remain confined within a closed loop, so a predetermined volume of fluid could be reused continuously and additional waste generation minimized. The custom flow cell had to be incorporated directly into that loop, regenerating Ce(IV) as it is consumed during decontamination. A cleaning fixture was needed that could flow, contain, and recapture solution from the glovebox surface without manual intervention. All solution movement had to be driven by negative pressure, so that a system shutdown or power loss could never propel liquid outward. Finally, each subsystem needed to operate within a modular framework allowing independent control and straightforward reconfiguration.

These principles guided the minimum viable product configuration of GB-Decon. The system is built around a 4-liter perfluoroalkoxy alkane reservoir housing the cerium solution, connected to a fluid-handling assembly in the unit's lower structural frame. The upper enclosed frame contains the electrical components and feedthroughs powering the pumps, heaters, sensors, and monitoring equipment-heater cartridges to elevate solution temperature, thermocouples with active feedback control, independent switches for each subsystem, and a compact display showing real-time Ce(III)-to-Ce(IV) conversion data from an inline UV-Vis flow cell. This monitoring capability provides operators with continuous, quantitative insight into the chemical readiness of the decontamination solution during operation-information that manual methods cannot provide. Activating the primary pump initiates recirculation of the solution throughout the loop, including through the electrochemical flow cell where the oxidant is continuously regenerated.

The final element was the interface between the system and the glovebox itself: the cleaning heads. Gloveboxes present a variety of surfaces and geometries, so the team constructed a suite of specialized fixtures using both subtractive and additive manufacturing. Each head carries a custom-profiled foam gasket, allowing the head, when placed against a glovebox surface, to form a sealed but narrow void through which solution can flow evenly across the contaminated area. Flat interior surfaces are addressed with clear machined fixtures in a range of sizes (Fig. 5), while corners, edges, and areas obstructed by internal components are serviced with 3D-printed fixtures tailored to those shapes (Fig. 6).

Operation is straightforward. Valved quick-connect fittings on every head allow rapid fixture exchange. An eductor in the fluid-handling loop generates the negative pressure that both adheres the cleaning head to the glovebox surface and draws solution across the treatment area. The operator places a fixture against the surface and actuates the eductor; the gasket compresses to create a tight seal while preserving the essential flow gap beneath the fixture. Solution is drawn through the head, across the contaminated surface, and back into the recirculation loop until the operator halts the process with an integrated three-way valve. The negative-pressure design also fails safely: If power is lost or the system shuts down, the maximum credible release is the small dead volume of solution already present in the lines. Because there is no positive pressure in the loop, nothing can spray solution outward or continue driving liquid to the release point; the consequence of a worst-case failure is a small, localized release rather than a pressurized spill.

Decontamination times depend on the metal substrate. Preliminary results indicate that 304 stainless steel surfaces reach LLW decontamination thresholds in approximately 20 minutes of spot treatment beneath a sealed cleaning head, while 316 stainless steel requires roughly 30 minutes per treated area. Compared with traditional spray-and-wipe procedures, this largely hands-off approach significantly reduces radiation dose, ergonomic risk, and the potential for containment breaches, while producing a fraction of the secondary waste.

Proof of concept: validation on Haynes 25

Surface decontamination times on stainless steel are one measure of performance; demonstrating that the underlying cerium chemistry can address severe, realistic contamination is another. To subject the MENDS platform to that more demanding test, the team turned to a companion module built on the same flow cell and closed-loop architecture as GB-Decon: the MENDS Recirculator, an enclosed, continuously circulating decontamination bath developed for small or intricately textured items-the parts and tools that conventionally spend upward of 12 hours soaking in nitric/hydrofluoric acid with periodic manual wiping.

For the validation campaign, coupons of Haynes 25 alloy were sectioned, introduced into a radiological glovebox, and deliberately contaminated to varying levels, with initial values measured in disintegrations per minute per square centimeter. The release criterion for the application is less than 100 dpm/cm². Parallel trials then compared the traditional acid soak-and-wipe method against the MENDS system on comparable coupons.

The results were unambiguous. For the most contaminated coupons-approximately 10 million dpm/cm², five orders of magnitude above the release limit-the standard acid bath required more than 12 hours to reach the limit, and it required sustained operator attention throughout: Approximately every hour, coupons had to be removed from the acid and manually wiped, with each cycle adding operator dose, ergonomic strain, and another contaminated wipe to the waste stream. The MENDS Recirculator achieved the same endpoint in 19 minutes with no direct handling. Coupons were placed in the bath in custom 3D-printed sample holders (Fig. 7), left for the targeted time, rinsed briefly in deionized water, and left to dry. More than 12 hours of hands-on processing was reduced to two brief handling steps, and the wipe waste stream was eliminated entirely.

The modular design compounds the throughput advantage. With the 3D-printed sample holders, the Recirculator decontaminates coupons in batches just as effectively as individually, whereas the acid bath process-with its hourly handling and wiping of each item-is effectively limited to processing one item at a time, substantially constraining throughput. The time savings per item is considerable; the gain in facility-level throughput is greater still.

As a proof of concept, the campaign carries significance beyond the bath module that performed it. It demonstrates that the continuously regenerated cerium oxidant-the same solution GB-Decon circulates across glovebox surfaces-can reduce gross contamination on a challenging cobalt-based superalloy to release-criterion levels in minutes rather than hours, without hydrofluoric acid and without an operator handling the workpiece. The results capture every dimension of the platform's value proposition: substantially reduced processing time, batch-scale throughput, elimination of a hazardous acid process, and removal of the operator from direct contact with contaminated material.

A platform, not a point solution

Because GB-Decon's enabling technology-the compact, chemically resistant flow cell and its closed recirculation loop-is modular by design, it has proven readily adaptable to decontamination challenges beyond the glovebox. The Recirculator described above is one example. Another adaptation addresses the nuclear energy sector. Spent fuel tubes are too long for standard disposal pathways and must be cut down to size, but cutting a contaminated tube risks dispersing radioactive particles, creating environmental and worker safety hazards. Decontaminating the tube interior first removes that risk. The MENDS tube fixture serves this purpose, using the same circulation and oxidation infrastructure as GB-Decon to treat tube interiors before size reduction.

The tube fixture consists of two machined end caps that fit securely over the tube's outer diameter. Negative pressure generated by the eductor draws oxidant into the top end cap, where a hollow-cone spray nozzle directs the solution outward toward the tube walls; the bottom cap is plumbed to the suction port, drawing used solution back into the closed loop for reoxidation (Fig. 8). Testing has identified refinements for future iterations-the nozzle's 120-degree cone angle leaves a narrow untreated zone at the top of the tube, which can be addressed by inverting the tube or increasing the nozzle standoff distance, and a dedicated suction pump would provide stronger, more stable flow than the eductor for some tube sizes-but the system effectively demonstrates proof-of-concept capability for an application with substantial potential impact on nuclear industry waste management.

The common thread across all three modules illustrates the platform's central strength: a single enabling technology, packaged into fixtures tailored to gloveboxes, baths, and tubes, with additional configurations in various stages of development and testing. MENDS is designed to integrate with existing facility infrastructure rather than requiring facilities to adapt to it.

What it means for the waste management community

GB-Decon's benefits align directly with the waste management community's most persistent pressures. Decontaminating gloveboxes and equipment from TRU levels to LLW limits reduces the volume of waste requiring disposal at WIPP, preserving capacity at the nation's only deep geologic repository for defense TRU waste and avoiding disposal costs that run orders of magnitude higher than the LLW alternative. Reusing a small, fixed volume of continuously regenerated solution minimizes the secondary waste streams-spent acids, contaminated wipes, and rinse water-that legacy methods generate in quantity. And removing operators from direct, repetitive contact with contaminated surfaces simultaneously reduces radiation dose, chemical exposure, ergonomic injury, and containment breach risk.

Development of GB-Decon is continuing on several fronts, including refinements to the tube fixture, development of additional module configurations, and integration of the second-generation flow cell, which has demonstrated the anticipated increase in oxidant-regeneration capacity. Even at this stage, the results are substantial. The platform has achieved nearly seven times the electrochemical turnover of commercial electrolyzers, decontaminated stainless steel surfaces in 20 to 30 minutes beneath sealed cleaning heads, and reduced a labor-intensive 12-hour acid-soak process to 19 minutes for heavily contaminated Haynes 25 coupons-without direct handling and with the added advantage of batch processing rather than treating one item at a time. As the nuclear industry confronts increasing decommissioning and dismantlement demands, constrained disposal capacity, and the continuing need to protect workers, GB-Decon offers a path toward glovebox decontamination that is faster, safer, cleaner, and substantially less reliant on manual labor at the glovebox face.

Acknowledgments

The author gratefully acknowledges the sponsors who have enabled and supported this work, with special thanks to the Department of Energy, the National Nuclear Security Administration, the Los Alamos Laboratory Directed Research and Development program, and the Material Recycling and Recovery Program (NA-191). This article is based on a technical paper presented at the 2026 Waste Management Symposium, March 8-12, 2026, in Phoenix (LA-UR-25-28708). A video about the technology is available at tinyurl.com/p22pm8xu.

Rami J. Batrice is a scientist in Los Alamos National Laboratory's Biochemistry and Biotechnology group.

Technical inquiries regarding the MENDS platform and its applications may be directed to Batrice at [email protected]