Development of flash lithium evaporators for NSTX-U

This paper presents the design and validation of a new flash lithium evaporator (f-LITER) for the NSTX-U, which builds upon in-vacuo refill and rapid evaporation tests conducted on LTX-$\beta$ to enable fresh lithium delivery to multiple plasma-facing components while minimizing impurity pickup and service complexity.

The Gist

Imagine a fusion reactor as a high-performance race car engine. To make it run efficiently, the inside walls need to be perfectly clean and smooth. If the walls get dirty or "sticky," the fuel (plasma) leaks out or gets contaminated, and the engine sputters.

For years, scientists have used lithium (a soft, silvery metal) to coat these walls. Think of lithium as a specialized "non-stick" spray that soaks up impurities and keeps the fuel flowing smoothly. However, there's a catch: lithium is like fresh paint. It dries out, gets oxidized (rusted by air), and loses its effectiveness within a few hours. To keep the engine running at peak performance, you need to re-apply the lithium constantly.

This paper describes the journey of inventing a better "paint sprayer" for the NSTX-U, a massive fusion experiment at Princeton. Here is the story of how they went from a slow, messy process to a fast, precise "flash" system.

The Old Way: The Slow, Heavy Paint Roller

In the past, the team used a method that was like trying to paint a ceiling with a heavy, soaking wet roller that took hours to dry.

• The Problem: They would load solid chunks of lithium into a container, heat it up, and let it evaporate. But the container was heavy (high thermal mass). Heating it took hours, and cooling it down took hours.
• The "Rust" Issue: While waiting for the machine to heat up or cool down, the fresh lithium was exposed to tiny amounts of air and moisture in the vacuum chamber. This caused the lithium to "rust" (oxidize) before it even touched the walls.
• The Coverage Gap: The old sprayers only pointed downward, like a showerhead. They could only coat the bottom of the reactor. But the new NSTX-U reactor needs to be coated all around—top, bottom, and sides—to work properly.
• The Waste: To protect the machine during operation, they used metal shutters. But these shutters caught half the lithium spray, wasting the expensive material. When the lithium ran out, they had to pull the heavy equipment out, open the vacuum chamber to the air, refill it, and wait a whole day for it to cool down again.

The Evolution: From "Flash" to "Dropper"

The team realized they needed a system that was fast, light, and could reload without opening the chamber. They developed this in stages on a smaller test reactor called LTX-β:

1. Mark-I (The Flash Light): They built a tiny, lightweight heater. Instead of a heavy roller, it was like a flashbulb. It could heat up and evaporate lithium in just a few minutes. This solved the "waiting around" problem.

   • The Flaw: Because it was so small and fast, it couldn't reach the "high walls" of the reactor. It left the top and sides bare, like a flashlight that only shines on the floor.

2. Mark-Ia (Adding a Reflector): They added a shiny mirror (made of tantalum metal) to bounce the lithium vapor around corners, ensuring the "high walls" got coated.

3. Mark-II (The Felt Liner): The faster evaporation caused the lithium to drip like a leaky faucet. They lined the basket with a special metal "felt" (like a dense sponge made of steel fibers). This felt soaked up the molten lithium, holding it in place so it wouldn't drip, while still letting it evaporate evenly.

The Breakthrough: The "In-Vacuo" Dropper

Even with the Mark-II, there was one big problem: Loading.
To refill the basket, they still had to take solid lithium chunks out of a glovebox, carry them to the machine, and drop them in. Every time they did this, the lithium touched a tiny bit of air, picking up impurities (dirt) that ruined the coating. It was like trying to paint a wall while wearing gloves that were slightly dirty.

The Solution: The Liquid Lithium Dropper
The team invented a new tool: a liquid-lithium dropper.

• How it works: Imagine a high-tech eyedropper filled with molten lithium. It sits outside the reactor. When it's time to refill, the dropper lowers a needle into the evaporator basket and squeezes out a few drops of liquid lithium.
• The Magic: The dropper never leaves the vacuum environment. The lithium goes straight from the dropper to the basket without ever touching air. It's like refilling a pen with ink without ever taking the cap off or exposing the ink to dust.
• The Result: They tested this on LTX-β. The dropper successfully wetted the metal felt, held the liquid without dripping, and evaporated a perfect 100-nanometer layer of fresh lithium in about 5 minutes.

Why This Matters for NSTX-U

The new system, called f-LITER (Flash Lithium Evaporator), is designed specifically for the big NSTX-U machine.

• Full Coverage: It can spray lithium to the top, bottom, and sides of the reactor, not just the bottom.
• Freshness: Because it can reload in minutes without opening the chamber, the lithium stays "fresh" and effective. The team found that if they wait too long between shots, the lithium gets "old" (oxidized), and the plasma performance drops. With f-LITER, they can refresh the coating between every single shot.
• Less Waste: No more shutters catching the spray. The lithium goes exactly where it's needed.
• Easy Maintenance: The "head" of the sprayer can be detached and swapped out easily, so if it breaks, they don't have to pull the whole machine apart.

The Bottom Line

The paper shows that by moving from heavy, slow, solid-lithium loaders to a fast, lightweight system that uses a liquid dropper to reload inside a vacuum, they can keep the reactor walls perfectly coated with fresh lithium. This allows the fusion engine to run hotter, cleaner, and more efficiently, paving the way for future fusion power plants.

Technical Summary

Technical Summary: Development of Flash Lithium Evaporators for NSTX-U

Problem Statement
The paper addresses the engineering challenges associated with delivering fresh lithium coatings to plasma-facing components (PFCs) in the National Spherical Torus Experiment Upgrade (NSTX-U). Previous experiments on the Lithium Tokamak Experiment (LTX) and LTX-β demonstrated that lithium wall conditioning significantly improves plasma performance by reducing recycling, modifying edge profiles, and suppressing impurities. However, these benefits are highly sensitive to the "freshness" and chemical state of the lithium surface; performance degrades as lithium oxidizes to Li₂O and eventually LiOH over timescales of hours.

Legacy systems on NSTX, specifically downward-facing Lithium EvaporaRs (LITERs), suffered from several limitations:

1. Limited Coverage: They targeted only the lower divertor, failing to provide the upper-divertor and center-column coverage required for NSTX-U's double-null operation.
2. Thermal Inertia and Shutter Loss: The high thermal mass of the LITERs required long cool-down times, and the use of shutters to protect diagnostics during plasma shots intercepted approximately 50% of the evaporated lithium, wasting inventory and creating reliability risks.
3. Operational Complexity: Refilling required venting the vessel, removing probes, and performing a multi-day bakeout process.
4. Impurity Pickup: Manual loading of solid lithium in argon gloveboxes and subsequent transfer introduced impurities (oxygen, water) that degraded the coating quality before operation.

The core problem was the need for a system capable of rapid, repeatable, in-vacuo lithium delivery to all relevant PFC regions (upper/lower divertors, center column, outboard) with minimal impurity introduction and without the operational delays associated with solid lithium handling or shutter mechanisms.

Methodology
The authors pursued a design evolution path starting from LTX-β flash evaporators to develop a "Flash Lithium Evaporator" (f-LITER) suitable for NSTX-U. The methodology involved:

1. Iterative Design on LTX-β: Three generations of evaporators were developed and tested:
   • Mark-I: A low-thermal-mass stainless-steel mesh basket with a solid lithium load. It reduced cycle times to minutes but suffered from geometric shadowing, leaving high-field-side limiters uncoated.
   • Mark-Ia: Modified with a tantalum reflector and removed structural obstructions to improve line-of-sight coverage to the high-field side.
   • Mark-II: Featured a larger, coarser tantalum mesh basket lined with sintered 316L stainless-steel fiber felt to prevent lithium dripping. It increased inventory capacity but still relied on manual solid-lithium loading.
2. In-Vacuo Liquid Lithium Loading: To eliminate impurity pickup during loading, the Mark-II design was modified to accept a liquid-lithium dropper. This device, tested on the Deposition Analysis and Reliability Testbed (DART) and LTX-β, uses a plunger-driven mechanism to dispense molten lithium droplets directly into the evaporator basket while the system remains under vacuum.
3. Performance Characterization:
   • Deposition Rates: Quartz Crystal Microbalance (QCM) measurements were used to quantify evaporation rates and cycle times.
   • Surface Analysis: Temperature-Programmed Desorption (TPD) was performed on witness samples exposed to the evaporators to compare impurity retention between solid-lithium loading and the new liquid-dropper method.
   • Plasma Evolution: Run-day analysis on LTX-β tracked plasma current, density, and particle confinement time ($\tau^*_p$) to correlate performance degradation with the oxidation of the lithium surface over time.

Key Contributions

• Design of the f-LITER: The paper presents the final design for the NSTX-U f-LITER, which integrates a detachable evaporator head, a tantalum mesh basket with a sintered felt liner, and a liquid-lithium dropper for in-vacuo loading.
• Demonstration of In-Vacuo Refill: The authors successfully demonstrated the ability to load molten lithium into an evaporator basket without venting to air or using solid lithium transfer, significantly reducing the exposure of lithium to contaminants.
• Impurity Reduction: TPD results showed that the liquid-dropper loading method resulted in significantly lower release of impurities (O, C, N, H₂O) compared to the traditional solid-lithium loading method.
• Coverage Modeling: The design utilizes a midplane insertion geometry with a reflector to ensure line-of-sight coverage to the upper divertor, lower divertor, and center column, addressing the coverage gaps of legacy NSTX systems.

Results

• Evaporation Cycles: The dropper-loaded evaporator achieved a 100-nm lithium deposition in approximately 5 minutes on LTX-β. Scaling calculations suggest that for NSTX-U, a 25–100 nm coating would require an active deposition time of 5.5–22 minutes, with a total cycle time (including heat-up and cool-down) of roughly 14–31 minutes. This is compatible with between-shot operation.
• Surface Chemistry: TPD analysis confirmed that the liquid-dropper method reduced the co-deposition of impurities. The spectrum for the liquid-loaded sample was dominated by the hydrogen signal (2 amu), whereas the solid-loaded sample showed broad release of oxygen, carbon, nitrogen, and water species.
• Plasma Performance Correlation: Analysis of LTX-β run days showed that plasma performance (current, density, confinement) degraded over the course of a day as the lithium surface oxidized. Re-evaporation restored performance, confirming that "fresh" lithium is critical and motivating the need for rapid, repeatable delivery systems.
• Dripping Prevention: The sintered fiber felt liner successfully held molten lithium within the basket during heating and evaporation, preventing the dripping issues observed in earlier coarse-mesh designs.

Significance and Claims
The paper claims that the developed f-LITER design provides a practical engineering solution for the NSTX-U program to achieve reproducible, high-performance double-null operation. The significance lies in the translation of LTX-β operational experience into a system that:

1. Enables Freshness: Allows for lithium refreshment on the timescale of the shot cycle (minutes), preventing the performance degradation associated with surface oxidation.
2. Expands Coverage: Provides necessary coverage to upper-divertor and center-column regions, which legacy downward-facing systems could not reach.
3. Reduces Complexity and Waste: Eliminates the need for vessel venting during refilling, removes the lithium-loss mechanism of shutters, and simplifies maintenance through a detachable head design.
4. Minimizes Impurities: The in-vacuo liquid loading method directly addresses the issue of impurity pickup, ensuring a higher quality lithium coating.

The authors conclude that while further qualification regarding deposition footprints and integration with the NSTX-U shot cycle is underway, the f-LITER architecture represents a validated path toward enabling high-performance lithium conditioning in next-generation spherical tokamaks. The paper also notes that a similar architecture is planned for deployment on the ST40 spherical tokamak.