Nearly a month had passed since her return from the Netherlands. Xiuxiu gave herself little time to adapt; practically the day after leaving quarantine, she dragged her still jet‑lagged, weary body straight to the R&D park located on the city's outskirts—enclosed by high walls and stringent security. This would be her "main battlefield" for a long time to come—a core R&D base dedicated to conquering high‑end lithography technology, supported by a major national science and technology project.
The atmosphere inside the park contrasted sharply with the open, international style of ASML's headquarters in Eindhoven. Here it was more reserved, more focused; the air carried a blend of machine oil, electronic components, a particular cleaning‑agent smell, and an intangible, heavy sense of urgency. Every factory building, every laboratory seemed to hold its breath, engaged in a silent yet fiercely consequential race toward the future.
Xiuxiu was assigned to a team primarily responsible for the development and performance enhancement of **deep ultraviolet (DUV) lithography machines**. Her specific task was improving the DUV machine's "heart"—the **light‑source system**. This was her familiar territory, and also one of the key bottlenecks currently hindering the advance of domestic lithography technology from "having" to "refining," crucial for improving chip‑manufacturing yield and reliability.
When she first walked into the lab‑cum‑office assigned to her, she could clearly sense the silent, mixed ripples of curiosity and scrutiny in the air. The team members, mostly engineers and technicians around her age or slightly older, were overwhelmingly male. They paused their work, their gazes converging on this "parachuted‑in," ASML‑glowing "returnee."
Those looks were complex and hard to read. There was the natural awe toward a top‑institution background, a barely perceptible doubt about a female engineer (especially in the field of light sources, considered the hard core of hard‑core technology), perhaps mixed with a subtle distance stemming from her overseas experience, even a tinge of worry about "whether she can adapt to our tough conditions here." No one voiced any skepticism openly, yet that silent atmosphere carried more pressure than any words.
Xiuxiu didn't say much. She calmly greeted the project lead, Engineer Li—a silver‑haired veteran with sharp eyes, steeped in optical engineering for over thirty years—and then simply sat down at her workstation. Her desk stood right next to the observation window of the lab; through the thick lead‑glass pane she could see a DUV light‑source prototype running inside, emitting a low hum and that characteristic, deep‑blue glow.
She knew that here any titles or past credentials were meaningless. What could earn respect was only solid professional skill and the ability to solve concrete problems.
The first specific task she took on was analyzing and optimizing the existing DUV light source's **power stability** and **bandwidth**. During prolonged operation, the current source exhibited non‑negligible fluctuations in output power, and the width of its emission spectrum (bandwidth) had not yet met the ideal design specifications—both factors directly affecting the precision and uniformity of lithographic lines.
To tackle this, one had first to deeply understand the physical principles and evolution of DUV light sources. At the team's technical meeting, Xiuxiu for the first time systematically organized and explained this knowledge, a process that served both to clarify the domestic technological landscape and to demonstrate her own professional depth to the team.
"Colleagues," Xiuxiu stood before the projection screen, her voice clear and steady, without the shyness of a newcomer, only the focus of someone immersed in technical details, "the DUV lithography we currently focus on centers on finding a light source that is sufficiently powerful, sufficiently stable, and with a sufficiently short wavelength to achieve higher resolution. In the history of lithography technology, two milestone DUV sources are worth our in‑depth understanding, because they represent different technological paths and physical limits."
The first schematic appeared on the screen. "The earliest deep‑ultraviolet source widely used in lithography was the **mercury lamp**, particularly its g‑line (436nm) and i‑line (365nm) emission lines. Its principle, in essence, resembles that of a household fluorescent tube: through high‑voltage discharge between electrodes, mercury‑vapor atoms are excited, their electrons jumping to higher energy levels; when they fall back, they release photons of specific wavelengths."
She directed a laser pointer to key parameters on the chart. "The mercury lamp's advantages lie in its relatively simple structure, low cost, and mature technology. However, its shortcomings are equally prominent: first, its **wavelength** is limited—the i‑line's 365nm wavelength is already insufficient for more advanced process nodes; second, increasing its **power** is difficult, and its power density is low, energy dispersed, not conducive to efficient exposure; most crucially, its **stability** and **lifetime** face bottlenecks—electrode and tube degradation leads to intensity decay and fluctuations, requiring frequent maintenance and replacement."
She paused, her eyes sweeping the team members present. Many were nodding; clearly they had firsthand experience with mercury lamps' pros and cons. Domestic earlier lithography equipment had made extensive use of such sources.
"As chip processes moved toward finer features, we needed shorter wavelengths and more concentrated energy." Xiuxiu switched slides; a much more complex device schematic appeared. "This leads to the direction we are currently focusing on—**excimer lasers**, especially the **ArF excimer laser** using an argon‑fluoride (ArF) gas mixture, which produces deep‑ultraviolet light at **193nm**."
Her tone rose slightly, carrying a sense of admiration for the exquisite physical process. "Its principle is more 'violent' and precise. In a specific gas chamber, through high‑voltage pulsed discharge or electron‑beam injection, the inert gas argon (Ar) and the halogen gas fluorine (F₂) react, forming short‑lived, excited‑state 'excimers' (ArF*). These excimers are extremely unstable and rapidly dissociate back to ground‑state atoms, simultaneously releasing energy in the form of radiated photons—the wavelength of these photons is 193nm."
"Compared with the mercury lamp, the ArF excimer laser's advantages are revolutionary." Xiuxiu began the critical comparative analysis. "First, **shorter wavelength**—the 193nm wavelength greatly improves resolution, sufficient to support processes from 90nm down to the most advanced 7nm (when combined with immersion technology); second, **high power**—modern high‑power ArF lasers can deliver tens to over a hundred watts output, meeting the needs of high‑speed scanning exposure, with extremely high power density; third, good directionality and coherence, making it easy for optical systems to shape and control."
"But," she pivoted, pointing to the challenges they currently faced, "it also brings new, thornier difficulties. First is **stability**. Excimer laser operation is pulsed; the energy stability of each pulse (Energy Stability) is crucial—the tiniest fluctuation directly translates into line‑width variation. This involves a series of extremely precise control issues: discharge stability, gas consumption and replenishment, thermal management, and more. Second is **bandwidth**. Although the laser's linewidth is already narrow, for the highest‑precision lithography, its spectral width (narrow bandwidth) must be further compressed to reduce chromatic‑aberration effects in the optical system; this typically requires introducing complex line‑narrowing modules (LNM). Moreover, the laser gas degrades during operation, producing impurities that require continuous purification and replenishment, adding to system complexity and cost."
Linking to specific data curves and fault logs, she indicated the exact defect ranges in pulse‑energy stability of the existing prototype, as well as simulated analyses of how unmet bandwidth targets affected actual lithographic resolution. Her analysis went beyond principles, delving into engineering‑implementation details, pointing out several key components and control loops that could be causing the problems.
The meeting room fell silent, save for Xiuxiu's clear voice and the laser‑pointer dot on the screen. Those previously scrutinizing gazes gradually changed. Doubt receded, replaced by attentive listening and thinking. What she demonstrated was not just theoretical familiarity, but keen insight into real engineering problems and a clear problem‑solving approach.
"Therefore," Xiuxiu concluded, "what we face is not unknown principles, but extreme engineering optimization. We need collaborative breakthroughs in the precision of discharge circuits, real‑time feedback control of gas management, heat‑sink efficiency, and the stability of line‑narrowing modules. This may mean we need to redesign certain feedback‑control algorithms, optimize gas‑circulation pathways, even search for more durable electrode materials."
After the meeting, several senior engineers who had initially held reservations approached her to discuss some of the specific technical points she had raised. The questions were professional and concrete, full of the smoke of actual combat. Xiuxiu responded one by one, not only answering but also proposing several improvement options—feasible alternatives based on her ASML experience.
In the following weeks, Xiuxiu practically made the lab her home. Wearing an anti‑static suit, she stayed with team members beside the bulky light‑source prototype, repeatedly adjusting parameters, recording data, analyzing waveforms. She personally operated expensive diagnostic equipment, measuring pulse‑energy jitter, analyzing subtle spectral changes. When she discovered that a certain key temperature sensor lacked sufficient accuracy, she didn't complain about supply‑chain issues; instead, she spent the night scouring domestic and international supplier materials, found an alternative model with comparable performance but a more reliable procurement channel, and personally wrote a technical‑verification report.
When she proposed modifying a core control algorithm, the young software engineer in charge, Xiao Zhang, looked uneasy, worried that changes might introduce unforeseeable risks. Xiuxiu didn't force an order; she spent an entire afternoon with Xiao Zhang at the whiteboard, step‑by‑step deriving the algorithm's mathematical foundation, explaining how the revised control logic could more effectively suppress energy fluctuations, and promising to conduct thorough simulation tests together, sharing the risk.
Sweat soaked the shirt beneath her anti‑static suit; prolonged staring at oscilloscope screens reddened her eyes, yet Xiuxiu's gaze remained bright and determined. Through concrete action, she slowly dissolved the initial barriers and doubts.
A month later, after countless subtle adjustments and optimizations, the improved control algorithm and optimized gas‑management system were loaded onto the prototype. During a 48‑hour full‑power continuous‑operation test, the core team members waited anxiously before the monitoring screens.
The data curves ran steadily. The fluctuation range of pulse energy was tightly held within an unprecedentedly narrow band, significantly outperforming the design specifications. The spectral analyzer showed the bandwidth had also reached the intended narrowing target.
When the test time ended and all key parameters remained within the green qualification zone, a long‑suppressed cheer erupted in the lab. Engineer Li, who had kept a tense face throughout, finally relaxed into a rare, broad smile. He walked over to Xiuxiu, gave a firm nod, and said just one phrase: "Well done, Engineer Xiu!"
"Engineer Xiu"—that plain, tech‑circle‑flavored address made Xiuxiu pause for an instant; then a warm current surged through her heart. She knew those two simple words meant acceptance, recognition—that she had earned her place in the team through her own professionalism and ability.
Looking around at her colleagues, whose once‑scrutinizing faces now shone with excitement and admiration, and at the steady curve on the monitoring screen, she felt a mix of emotions. This was only a small step in DUV light‑source improvement; the road toward the towering peak of EUV remained long and steep. But this step—taken with her own knowledge and sweat, on the land that had nurtured her—was a solid one.