Inside the vacuum chamber, it was like another universe. Here, no sound, no air—only near‑absolute nothingness, and that beam of ultimate light, invisible to the naked eye, pursued by countless intellects and sweat. Thirteen‑point‑five nanometers, the wavelength of extreme ultraviolet (EUV)—short enough to carve future chips' neural pathways onto silicon wafers, yet so short that it would be greedily devoured by everything in this world, including air itself. Hence, this beam's birth and journey must occur within such an ultimate‑purity vacuum environment—like a fragile ghost traversing nothingness.
Xiuxiu stood before the main console, body leaning slightly forward, hands supporting on the cold metal surface, knuckles whitened from pressure. She wore a heavy cleanroom suit enveloping her whole body, only revealing a pair of bloodshot yet stubbornly burning eyes. These eyes now stared fixedly at the huge main screen before her, divided into dozens of display areas. On screen, waterfall‑like data streams poured down silently; graphs flickered in real‑time; various parameter indicator lights glowed in different colors—together depicting that violent storm in the microscopic world inside the vacuum chamber.
This was the ultraclean laboratory at "String Light Research Institute" in Shanghai's Zhangjiang, the "EUV‑light‑source assault fortress" where Xiuxiu and her team had fought countless days and nights. Today was their who‑knows‑which attempt to charge that grail‑like target—**250 watts**.
250 watts. For ordinary people, this number might be less than a household incandescent bulb. But in the EUV‑lithography world, it was a life‑death line—the critical **baseline** measuring whether a light‑source system could support large‑scale, high‑efficiency chip mass‑production. **Power** directly determined how many "photon bullets" could be shot at the silicon wafer for exposure. Insufficient power meant exposure time must be extended, causing lithography‑machine **throughput** to plummet—making chip‑manufacturing costs unbearably high, rendering the so‑called "breakthrough" commercially meaningless.
Yet achieving stable 250‑watt EUV light‑source was as difficult as creating and constraining a continuously burning "micro‑sun" at temperatures reaching hundreds of thousands of degrees—on a microscopic scale.
Xiuxiu's gaze focused on the screen's central most‑critical parameters:
**Power**: Real‑time display of current light‑source output power, unit watts (W). This was today's core‑assault target. Previous countless experiments, their best achievement: briefly touching 248 watts, then—due to various reasons: usually **debris** contaminating collector‑mirrors causing reflectivity drop, or **laser‑and‑tin‑droplet synchronization** slight deviation causing plasma‑excitation instability—rapidly falling back to 230‑watt or lower plateau. Stability, harder than instantaneous peak.
**Bandwidth**: Measuring EUV‑light‑source output‑wavelength purity. Ideal light‑source should output only extremely narrow‑band light near 13.5 nm—like an extremely sharp carving knife. If bandwidth too wide, contaminated by stray light of other wavelengths—like a carving knife with burrs—would cause unnecessary exposure and blurring on chips, affecting pattern precision. Maintaining high power simultaneously requires suppressing bandwidth to level of point‑zero‑something percent nanometer (pm)—demanding extremely precise control of laser parameters, tin‑droplet state.
**Collection Efficiency**: An often‑overlooked‑by‑outsiders yet equally lethal parameter. It measures: after laser‑bombardment‑of‑tin‑droplet generates plasma, radiating EUV light—what proportion of photons can be successfully captured by subsequent **collector mirrors** and directed into optical‑path system. This efficiency typically frustratingly low. Vast amounts of EUV light escape instantly upon generation, or get absorbed/scattered by residual gas, produced debris. Enhancing collection efficiency meant optimizing collector‑mirror multilayer‑coating design (employing special molybdenum/silicon multilayers achieving highest reflectivity at 13.5‑nm wavelength), and maximally reducing **debris** contamination of mirror surface. Debris—inevitable by‑product of laser‑bombarding tin‑droplets—tiny tin particles flying like smoke dust, adhering to expensive, precise collector‑mirrors—causing reflectivity to decay over time—this was EUV light‑source lifespan and power‑stability's biggest killer.
Power, bandwidth, collection efficiency—these three parameters formed an **impossible triangle**, mutually constraining, influencing. Pursuing high power might sacrifice bandwidth purity; also might generate more debris due to higher energy, lowering collection efficiency. For high collection efficiency, over‑optimizing collector‑mirror structure or protection schemes might introduce new optical losses, limiting power enhancement. Xiuxiu's team's work over past years—almost dancing on this impossible triangle's blade edge—seeking that optimal, dynamic **balance point**.
"Laser main‑pulse energy, stabilized at 98.7 percent." A young researcher's voice came via internal‑comm system—carrying a barely perceptible tremor.
"Tin‑droplet generator, frequency locked at 50 kHz, diameter deviation less than ±0.1 µm." Another engineer reported—voice taut.
"Vacuum level maintained at 3×10⁻⁸ Torr." The vacuum group's voice—ensuring that microscopic universe's absolute "emptiness."
"Collector‑mirror active‑temperature‑control‑system feedback: mirror‑surface thermal deformation controlled within preset threshold." Key measure handling high‑power‑induced thermal load.
Xiuxiu didn't respond; her entire spirit immersed in that real‑time curve representing **power**. The curve had oscillated between 235 and 245 watts for nearly an hour—like a dragon trying to break shackles—several upward surges touching 248‑watt edge, then helplessly falling back. Each fall made control‑room air congeal further. Team members—standing or sitting—all held breath, eyes locked on screen or instruments before them. Air permeated with a special odor—mix of machine oil, ozone, human tension.
Time ticked away—each second like a century. Xiuxiu felt her heart pounding heavily in chest, eardrums throbbing blood‑flow roar. She recalled countless failed days‑nights: because of a micron‑level tin‑droplet‑nozzle machining error causing a week's experimental data wasted; because some imported special‑optical‑coating material embargoed, team forced self‑reliance, starting R&D of substitute from scratch—consuming half‑year time; because an unexpected grid‑fluctuation causing laser‑system crash—nearly damaging core vacuum chamber… Those setbacks, anxieties, team members' weary‑disappointed eyes—all now transformed into heavy pressure weighing her shoulders.
She couldn't collapse. She was this team's backbone, technical‑route decision‑maker—bearing most pressure yet must show most confidence. Mozi outside—warding off capital storms for her; Yue'er theoretical world—providing inspiration sparks. And she—must here—lead team, using engineers' hands—transform theoretical sparks into real light.
"Debris‑mitigation‑system hydrogen‑flow rate—increase another five percent." Xiuxiu's voice somewhat hoarse yet unusually stable—breaking control‑room's suffocating silence. A micro‑adjustment based on her latest conception. She speculated the current power‑plateau bottleneck might lie: as power approaches limit, generated debris‑volume slightly increases; though active‑protection‑system operating—minuscule, cumulative contamination slowly eroding collection efficiency—forming a negative feedback. Slightly increasing hydrogen‑gas flow—serving as protective gas—might more effectively "blow away" those tin particles about to adhere to collector‑mirrors.
Instruction swiftly executed. On screen, power‑curve seemingly quivered slightly—still struggling near 245 watts.
Disappointment rippled—subtle ripples—beginning to spread in control‑room. Someone sighed quietly; someone rubbed swollen eyes.
Xiuxiu closed her eyes—not giving up—but rapidly retrospecting in mind all data, all models, all possible variables. She recalled Yue'er once discussing "nonlinear‑system critical‑point" problem; recalled Mozi seeking "key‑driving‑factors" facing market chaos. Perhaps she needed a more radical, balance‑breaking operation?
"Laser pre‑pulse delay time," Xiuxiu suddenly opened eyes—gaze sharp as blade, "reduce 0.5 picoseconds from current baseline."
This instruction made several core engineers pause. Pre‑pulse function: first flatten tin‑droplet, forming "pancake‑like" target more conducive for main‑pulse excitation—generating more EUV light. Delay‑time minute adjustment would directly affect tin‑droplet deformed shape and density distribution—an extremely sensitive parameter; past optimization already fixed it within an interval considered optimal. Reducing 0.5 ps—a very bold, nearly‑risky attempt—might make morphology more optimal, or completely destroy synchronization—causing power plummet.
"Xiuxiu‑zong, this…" The laser‑system‑responsible engineer hesitated.
"Execute." Xiuxiu's tone brooked no doubt. She had a strong intuition: current "optimal" might be just a **local optimum**; they needed jump out of this comfort zone, assault that true **global peak**. This was the technological‑decision‑maker's courage at critical moment; also near‑instinctive "touch" accumulated through countless failures.
Instruction input into system. Everyone's heart in control‑room rose to throat.
On screen, power‑curve first lurched downward sharply—dropped below 240 watts! A suppressed gasp almost escaped lips. But next second—that curve—as if injected powerful vitality—sprang upward with resolute posture!
246 watts!
248 watts!
249 watts!
250 watts!
Number steadily fixed at **250.3 watts**! Moreover, not violently fluctuating as before—but within extremely narrow range around 250 watts—displaying unprecedented, heart‑calming **stability**!
Succeeded?!
Control‑room—dead silent; everyone froze—as if disbelieving eyes. Time congealed about three seconds.
Then—who knows who first emitted an irrepressible, tear‑choked cheer!
Like dam bursting—huge sound‑wave instantly erupted! Cheers, applause, shouts—even someone excitedly pounding control console; tears streaming freely down faces concealed by masks‑goggles. Young researchers hugging each other, jumping, yelling; older engineers—red‑eyed—forcefully slapping each other's shoulders—everything unspoken. Pressure, grievances, sweat accumulated over years—and now immense joy—volcano‑erupted—flooding entire control‑room.
Xiuxiu didn't move.
She still maintained hand‑support‑control‑console posture—but body uncontrollably began slight trembling. Those nerves taut too‑long‑long—at success‑confirmation moment—suddenly relaxing—bringing not instant ecstasy—but intense, nearly‑knock‑over **exhaustion** feeling. Everything before eyes seeming somewhat unreal; clamorous sounds seemingly arriving through a membrane. She felt legs weakening—almost unable stand.
Succeeded… really… succeeded? She subconsciously glanced at auxiliary parameters.
**Bandwidth**: stabilized at 0.35 pm—better than design specification.
**Collection efficiency**: real‑time calculation displayed—due to power enhancement and debris‑control proper—effective collection efficiency also slight improvement.
Power, bandwidth, collection efficiency—this impossible triangle—at this moment—pushed by them with wisdom, persistence, and a bit luck—toward a brand‑new, historic‑worthy **balance point**! This meant: EUV light‑source not only reached 250‑watt power baseline—its comprehensive performance fully satisfied, even exceeded commercial mass‑production's **throughput** requirements! Lithography‑machine throughput bottleneck—at light‑source this most‑critical link—thoroughly breached by them!
An immense, ineffable **joy and liberation**—like warm current—belatedly surged throughout her body—dispelling that exhaustion. She slowly, deeply inhaled—as if branding this success‑air deep into lungs. She straightened, turned—facing boiling team.
She didn't speak—just removed her own goggles and mask—revealing that exhausted‑to‑extreme yet radiating astonishing‑brilliance face. She looked at these comrades before her—also removed protection—faces bearing tears and smiles—eyes rapidly accumulating moisture. She forcefully pursed lips—not wanting tears fall—but that glistening liquid uncontrollably overflowed eye‑corners—sliding down cheeks.
She bowed deeply—toward team.
No words—but all gratitude, all affirmation, all shared glory—resided in this deep bow.
Team cheers intensified—mixed with "Xiuxiu‑zong!" "We succeeded!" shouts.
Xiuxiu straightened—wiping tears roughly with back of hand—revealing a brilliant, tear‑sparkling smile. She immediately told assistant: "Quick! Connect Mozi‑and‑Yue'er video! Now! Immediately!"
She couldn't wait to share this news with the two most important people. It was Mozi's unconditional capital‑layer support—withstanding countless external pressures and doubts—providing most solid backing—letting her focus wholeheartedly on R&D. It was Yue'er's theoretical‑layer inspiration and exchange—especially during bottlenecks providing cross‑disciplinary perspective—like lighthouse in darkness—gave her breakthrough‑direction crucial spark.
Assistant swiftly operated; control‑room's main screen split two video windows.
One window displayed Mozi's countenance. He seemed in some meeting‑room—background simple modern‑style decoration. He appeared somewhat weary—yet eyes still keen—clearly also handling heavy affairs. When seeing screen's other end—Xiuxiu's red‑rimmed eyes yet radiant‑smiling face—and behind her still cheering‑leaping team background—Mozi first slightly startled—then—that always carrying trace of coolness‑detachment face—revealed an extremely rare, heartfelt, warm‑and‑content smile. He didn't ask "Succeeded?" such nonsense; he just looked deeply at Xiuxiu—then gently nodded, thumb‑up. Everything unspoken. His silent support—at this moment—received most resounding echo.
Other window displayed Yue'er's figure. She seemed in some quiet study or library at String Light Research Institute—behind her ceiling‑high bookshelves. She still held an electronic pen—clearly just calculating something. When seeing Xiuxiu and scene atmosphere—first surprised blinked those clear‑wise eyes—then face blossomed pure‑happy‑child‑like smile—even happily lightly clapped hands.
"Xiuxiu! You succeeded! Right? 250 watts?" Yue'er's voice came via speaker—carrying her unique softness‑eagerness.
"Yes! Succeeded! Yue'er‑jie! Stable above 250 watts!" Xiuxiu forcefully nodded—voice choked with emotion. "Thank you! And Mozi! Thank you both!"
Video—Mozi smiling shook head—indicating no need thank. Yue'er happily said: "Brilliant! Xiuxiu! This is historic breakthrough! I knew you'd make it!"
Three screens—connecting three different worlds: one—engineering sanctuary creating miracles between vacuum‑photons; one—financial hub strategizing within capital‑rules; one—mathematical sanctuary exploring truth within symbols‑logic. Now—because this common, hard‑won victory—tightly linked.
Xiuxiu looked at screen‑two‑close‑friends‑comrades smiles; looked behind still‑immersed‑ecstatic team—heart filled ineffable fullness‑happiness. All hardships, all efforts—at this moment—became immensely worthwhile.
EUV‑light‑source 250‑watt breakthrough—not merely a technological milestone—but a powerful beam—illuminating road ahead—proving their route's correctness‑resilience. She knew—this just a new starting point—ahead High‑NA, more‑precise optical systems, more‑complex integration challenges. But now—she allowed herself—and her team, and her comrades—fully enjoy this hard‑won success joy.
She faced team members—also facing screen‑Mozi‑Yue'er—loudly declared—voice clear‑forceful:
"We—did it!"
Control‑room—again erupted thunderous, prolonged applause‑cheers. This sound—piercing ultraclean‑laboratory's multiple soundproofing layers—seemingly announcing to whole world know. And that beam—stably burning inside vacuum chamber—250‑watt extreme‑ultraviolet light—now silently—for next‑era's chip legend—carving first glorious imprint.