Deep underground in XianGuang Research Institute, the ultra‑clean laboratory seemed like an isolated singularity undergoing its own internal Big Bang. The towering waves of the outside financial markets, the life‑and‑death capital battle Mozi was facing—all were shut out by the heavy soundproofing walls and layer upon layer of purification systems, unable to penetrate this microscopic world composed of ultimate cleanliness and precision. Here, time flowed with a different density; every second carried the accumulated weight of the engineers' decade‑long technological Long March, pressing down on one's breath almost to suffocation.
Xiuxiu stood before the main console of the High NA EUV prototype, the heavy cleanroom suit enveloping her like an astronaut's extravehicular garment, completely isolating her from the surroundings, leaving only a pair of bloodshot eyes burning with unyielding fire, fixed intently on the huge main screen split into dozens of display areas. On the screen, data streams rushed like torrential rivers, graphs jumped violently, various parameter indicator lights flashed with heart‑stopping brilliance—together depicting the ultimate microscopic‑scale storm inside the vacuum chamber, a storm that would decide final success or failure.
This was the **final test** of the High NA EUV prototype. Not a single‑component validation, not a subsystem performance check, but the entire lithography system undergoing an ultimate trial of full‑power, full‑process, long‑duration stability under extreme conditions simulating future chip‑mass‑production environments. Success meant that "XianGuang" truly held the key to chip processes below 2nm; it meant that Xiuxiu and her team's ten years of effort, Mozi's limitless capital investment, even the nation's strategic layout in high‑end manufacturing, would reach a glorious milestone. Failure meant that all previous breakthroughs might be dismissed as accidental; it meant they might have to grope in the dark much longer; it meant that, at this sensitive moment of external capital encirclement, the cornerstone of "XianGuang"'s technological confidence could be shaken.
Pressure, like an invisible giant hand, gripped the throat of everyone in the lab. No one spoke; only heavy breathing and the low hum of instruments intertwined. Xiuxiu had not left the vicinity of the control console for **48 consecutive hours**. Drowsiness like tidal waves repeatedly tried to submerge her; she fought it off with strong black coffee and sheer willpower. Food brought to her hand was swallowed mechanically, tasteless as wax. Her body protested with fatigue—muscles sore, eyes dry, temples throbbing—but her spirit, wound tight as a spring, was stretched to its limit. This was the **final hurdle** of her technological Long March; she had to witness it with her own eyes, control it with her own hands.
The early phase of the test seemed unusually smooth. The massive High NA optical system operated steadily; the enormous lens assembly executed complex scanning motions with nanoscale‑precision motion control. The laser‑produced‑plasma (LPP) source module, building on earlier 250‑watt success, demonstrated encouraging stability at the higher power demand, outputting extreme ultraviolet light intensity meeting design expectations. Vacuum system, cooling system, transfer system… all subsystems performed their duties faithfully.
A spark of hope quietly ignited in each heart. Perhaps Lady Luck would stand on their side this time?
Yet, just as full‑power operation entered its 36th hour, with system load peaking at design limits and pushing toward longer‑duration stability, crisis—like a leviathan lurking in the deep sea—suddenly bared its ferocious fangs.
On the main screen, a parameter indicator that should have stayed steadily green suddenly began flashing an alarming yellow; then, within seconds, it unhesitatingly jumped to a heart‑stopping red! Immediately afterward, several key performance indicators linked to it—like toppling dominoes—began showing abnormal fluctuations!
"Report! **Source collector mirror** area—thermal‑load sensor reading exceeds limit! Mirror‑surface temperature gradient above safety threshold!" an engineer monitoring the thermal‑management system shouted hoarsely, with incredulous panic.
"EUV output beam‑quality monitoring shows wavefront aberration rapidly increasing! Central intensity attenuation fifteen percent!" a researcher in charge of optical‑quality monitoring followed up, voice trembling.
Xiuxiu's heart sank violently, as if plunging into an ice cavern. **Source collector mirror**! That was one of the most precise and fragile core components in the EUV source system. Its function was to collect, as efficiently as possible, the precious 13.5nm extreme ultraviolet light radiating in all directions from the plasma generated by laser‑impacted tin droplets, and converge it into a controllable beam for entry into the subsequent optical system. It was typically composed of special multilayer‑mirror surfaces, its figure accuracy required down to the atomic level.
During full‑power operation, the high‑temperature plasma generated by laser‑impacted tin droplets released enormous energy. Despite advanced cooling systems, some heat inevitably transferred to the collector mirror. When the thermal load exceeded a certain critical point, or the cooling system exhibited even the tiniest efficiency deviation causing uneven temperature distribution (thermal gradient) across different mirror regions, it triggered the material's inevitable thermal‑expansion effect. This expansion was extremely minute—perhaps **nanoscale**, even **sub‑nanoscale**—but for EUV light collection demanding atomic‑level precision, this tiny, non‑uniform **thermal deformation** was enough to utterly destroy the mirror's ideal figure!
Like lightly touching an extremely precise magnifying glass with a red‑hot branding iron—even without direct contact, the radiated heat alone could cause microscopic distortion of the lens, severely degrading, even completely nullifying, its focusing ability. Right now, the collector mirror inside the High NA EUV prototype was undergoing a similar process. Overloaded thermal load had led to **microscopic deformation** of the mirror surface, preventing it from perfectly converging EUV light, thereby causing sharp deterioration of output beam quality, increased wavefront aberration, and attenuation of effective light intensity. If not resolved immediately, not only would this final test declare failure; continued operation risked permanent damage to this core collector mirror—priceless and with a long manufacturing cycle!
The control room fell dead silent; despair spread like plague. Several young researchers even reddened their eyes, unable to accept such a fatal blow so close to success. Thermal deformation—this was one of the most intractable, hardest‑to‑cure "terminal illnesses" in lithography‑machine R&D, especially for EUV lithography.
Xiuxiu felt a wave of intense vertigo. The accumulated exhaustion of 48 sleepless hours, combined with this sudden heavy blow, nearly crushed her. Instinctively she steadied herself against the cold console edge, knuckles whitening from the force. In that instant, countless thoughts flashed through her mind: Reduce power? But that would mean High NA performance couldn't be fully unleashed, rendering the test meaningless. Optimize cooling? Distant water couldn't save a nearby fire; besides, it might already be too late. Replace the collector mirror? That would mean months of time loss and enormous cost…
No! Can't give up! This is the final battle; there is no retreat!
In that split second, a technical solution she had long prepared but never verified under such extreme conditions lit up in her mind like a beacon in darkness—**real‑time thermal‑deformation compensation technology**!
She jerked her head up. Those bloodshot eyes blazed anew with hawk‑sharp light; all fatigue seemed forcibly expelled in that moment. Her voice, hoarse from excitement and urgency, carried unquestionable decisiveness, like a sword cleaving the thick despair in the control room:
"Activate the 'Nuwa' system! Highest response level! Target—source collector mirror, real‑time surface‑figure compensation!"
The "Nuwa" system—this was an **active‑optical‑compensation system** secretly developed over several years by Xiuxiu's team to cope with various unavoidable microscopic deformations (whether thermal‑, gravity‑, or stress‑induced) in high‑end optical systems. Its core lay in utilizing **piezoelectric‑ceramic actuators**.
"Piezoelectric ceramics," Xiuxiu explained to the team in rapid‑fire speech while swiftly pulling up the "Nuwa" system's control interface—both an order and a morale booster, "possess a unique 'piezoelectric effect': when voltage is applied across them, they produce extremely minute yet extremely precise deformation! Conversely, when they deform, they generate corresponding electrical signals. We long ago integrated hundreds of such micro‑piezoelectric‑ceramic actuators in a dense array on the back side of the collector mirror!"
She called up a structural diagram of the mirror back, covered with actuator arrays like a neural network. "The 'Nuwa' system, by real‑time monitoring of the collector‑mirror surface's temperature distribution and indirectly inferred figure errors via wavefront sensors, feeds these data into a high‑speed processing unit. The processing unit, based on our pre‑calibrated, complex model describing the relationship between voltage and actuator displacement, calculates in real time the different control voltages that need to be applied to each actuator!"
The screen began simulating this process: color‑contour maps representing mirror‑surface distortion caused by thermal deformation were being dynamically compared and iterated with maps representing compensation‑voltage distributions computed by the "Nuwa" system.
"Then," Xiuxiu's voice held a miracle‑creating certainty, "these precisely controlled voltages are applied to the corresponding piezoelectric‑ceramic actuators. These actuators, following instructions, produce **nanoscale**, precise elongation or contraction—like countless extremely tiny, invisible hands 'pushing' or 'pulling' that thermally deformed collector mirror from behind!"
"Through this dynamic, active micro‑adjustment, we can to some extent 'offset' the non‑uniform deformation caused by thermal load, restoring the mirror surface as closely as possible to its ideal figure state! This isn't eliminating the heat source; it's 'deceiving' the optical system into 'thinking' the mirror remains perfect!"
This was the engineer's wisdom—a roundabout solution sought while acknowledging the irresistibility of physical laws. Like using precise massage techniques to relax a muscle cramped from tension, restoring its function.
The order was swiftly issued. The "Nuwa" system's control module activated; data streams poured in like floodgates opened. Load‑indicator lights on the high‑speed processing unit flashed wildly, computing those complex control signals for hundreds of channels needing real‑time updates.
Everyone held their breath, eyes locked on the main screen—on that most‑critical performance indicator representing EUV output beam quality: wavefront aberration.
Time passed second by second, each second like a century.
At first, the abnormally fluctuating curve still struggled in the red‑alarm zone, showing no sign of improvement. Disappointment began to breed again.
Xiuxiu clenched her lower lip, almost drawing blood. She believed in her design, believed in the team's preparation—but this was the ultimate test; there were no guarantees.
Suddenly! About three minutes after the "Nuwa" system began full‑power operation, that wavefront‑aberration curve quivered—like a drowning struggler finally grabbing a lifeline—its upward momentum halted!
Then, under everyone's disbelieving gaze, it began a slow yet determined **descent**!
Fifteen percent… fourteen percent… thirteen percent…
Slow though it was, the trend was clear! **The real‑time thermal‑deformation compensation technology was working!**
"It's dropping! It's dropping!" a young researcher couldn't help crying out, voice choked.
In the control room, deathly despair gave way to a suppressed, after‑disaster excitement. No one cheered, for the battle wasn't over yet, but each pair of eyes reignited with the spark of hope.
Xiuxiu let out a long, long exhale—the stagnant breath that had almost suffocated her in her chest. She felt a debilitating weakness wash over her; her body swayed, forcing her to steady herself again on the console. Forty‑eight consecutive hours of vigil, plus those minutes of high‑stakes decision‑making and waiting under extreme tension, had nearly drained all her physical strength.
But she did not fall. Her gaze, through the goggles, remained firmly fixed on that curve on the screen—slowly yet surely returning to normal territory—filled with utmost **determination**.
She knew this wasn't final victory yet. How long could the "Nuwa" system hold? Was the compensation precision sufficient for High NA's extreme demands? What about system stability under prolonged operation? These were all unknowns.
But at least they had withstood this deadliest blow, hadn't fallen before the finish line.
She raised a slightly trembling hand, adjusted her communicator, and spoke to the team in the steadiest voice she could muster: "Keep monitoring; real‑time fine‑tuning of 'Nuwa' system parameters. We… haven't lost yet."
Her voice, transmitted through the face shield, carried a tinge of exhausted hoarseness, yet stood like a sea‑calming needle, steadying everyone's heart.
The final battle for the light source entered its cruelest stalemate phase. And Xiuxiu—commander of this technological Long March—still stood firm on her position, guarding with her nearly depleted energy and that never‑yielding heart, the beam of ultimate light that, deep in the vacuum, burned stubbornly despite all tribulations.