The establishment of the immersion‑technology path acted like a powerful stimulant, briefly reinvigorating the team's fighting spirit. As technical chief, Xiuxiu led each subsystem in pushing forward theoretical calculations, scheme design, and pre‑research on key subsystems. Fluid control, temperature‑field management, defect detection… each challenge was broken down; though progress was slow, at least the direction was clear, each step taken solidly.
Yet just as they were ambitiously preparing to advance into the "deep‑water zone," a more fundamental, more fatal "curse" quietly surfaced. Its origin lay not in the emerging immersion system, but in the lithography machine's ancient yet precise "eye"—the **optical lens**.
To realize the resolution leap promised by immersion technology, not only must the medium between lens and wafer be changed, but the lens itself must meet almost perverse demands. To achieve higher numerical aperture (NA), the objective lens design must employ larger aperture angles and more complex **aspherical lens elements**.
**Aspherical lens elements**, as the name suggests, have surface profiles that are no longer simple spheres (parts of a sphere) or planes, but more complex surfaces described by higher‑order equations. Compared to traditional spherical lens elements, aspherical elements offer incomparable advantages: they can drastically eliminate **spherical aberration**—an optical aberration caused by the inherent properties of spherical surfaces, where rays from different apertures fail to converge at a single point. In a lithography system demanding atomic‑scale imaging quality, any tiny aberration is unacceptable.
But manufacturing such high‑precision, large‑aperture **aspherical lens elements** is a challenge comparable to Mount Everest of optical engineering. Traditional grinding and polishing techniques rely on skilled artisans' experience and feel; even for spherical surfaces they require extreme skill and long cycles, while for complex mathematical‑surface‑defined aspherical elements they are almost powerless.
The domestic top‑tier optical‑component manufacturer cooperating with the team had already exhausted its efforts. They introduced state‑of‑the‑art computer‑numerical‑control (CNC) grinding and magnetorheological polishing equipment, attempting to "sculpt" the theoretically designed aspherical surface via digital control. Yet progress was extremely unsatisfactory.
The yield was despairingly low.
Substrate blanks sent for processing, after weeks or even months of precision machining, failed to meet design requirements nine times out of ten. The problems lay in **surface‑form accuracy** and **surface roughness**.
Xiuxiu stood outside the cleanroom observation window, watching engineers inside testing the latest batch of delivered lens elements with phase‑shifting interferometers. The screen showed not the ideally smooth colored‑contour map representing a perfect surface, but interference fringes full of chaotic, wave‑like patterns. The accompanying data report clearly displayed the RMS (root‑mean‑square) and PV (peak‑to‑valley) values of **wavefront aberration**, far exceeding permitted design ranges.
**Wavefront aberration** is the core metric for optical‑system imaging quality. Ideally, a perfect plane wave passing through an ideal optical system should remain a perfect plane wave or converge to a perfect point. In reality, due to lens‑surface‑form errors, material inhomogeneity, internal stress, etc., the emergent wavefront becomes distorted. This distortion is wavefront aberration.
It acts like a demon‑revealing mirror, magnifying any nanometer‑scale or even sub‑nanometer flaw on the lens surface into a fatal defect capable of ruining the entire lithographic pattern. A tiny, imperceptible dent or bump, a local slope error, would cause an extra optical‑path difference for light passing that point, resulting in a circuit pattern that should be sharp and clear on the final focal plane becoming blurry, distorted, or generating unwanted stray light.
"Still not acceptable." Engineer Lao Zhou, responsible for optical integration, came out, removed his mask, face etched with weariness and dejection. He handed the test report to Xiuxiu. "PV value three times over limit, RMS nearly double. The key is, this aberration‑pattern mode isn't reproducible; each lens element's error distribution is different, like… a random curse."
Xiuxiu took the report; the paper felt a thousand‑pounds heavy in her hands. Randomness—that was the most terrifying situation. If errors were regular, you could adjust process parameters accordingly. But randomness meant the process itself was unstable—perhaps due to micro‑vibrations during processing, temperature fluctuations, uneven polishing‑fluid distribution, even microscopic defects in the material itself.
"We tried adjusting polishing pressure, rotation speed, even changed batches of polishing powder," Lao Zhou's voice was hoarse, "No significant effect. The veteran craftsmen say this already approaches the limit of existing equipment and process capability. Aspherical surface measurement and feedback compensation is itself a world‑class problem."
Xiuxiu nodded silently. She understood the difficulty. Magnetorheological polishing, though advanced, relies on an accurate "removal function" model—knowing how much material the polishing head removes at a given position with given pressure and dwell time. This model itself needs extremely precise calibration; during processing, changes in polishing‑fluid condition, tool‑head wear cause deviations, making actual removal differ from expected. This is a dynamic, nonlinear control process; any tiny deviation leaves an irreparable trace on the lens surface.
"The curse of the lens." The phrase began circulating privately among the team. Unlike the light source that could excite surging energy, unlike the stage that could display astonishing speed, it stood there silent, stubborn, using those nanometer‑scale, seemingly negligible errors to declare the boundary of human precision‑manufacturing capability. A piece of glass obtained by investing huge resources, consuming long time, might ultimately become scrap because of an unpredictable, incomprehensible microscopic defect. This silent attrition was devastating to morale and confidence.
As technical chief, Xiuxiu bore enormous pressure from all sides. Chief Engineer Li still supported her, but worry in his eyes deepened daily. Calls from above pushing for progress grew more frequent. Inside the team, the enthusiasm kindled by establishing immersion technology was gradually worn away by this seemingly endless frustration. Some began privately questioning whether they should fall back, first adopt a spherical‑lens‑combination scheme with somewhat lower performance but higher yield, even if sacrificing some resolution.
Xiuxiu knew that meant compromise, meant admitting defeat before even starting. Yet looking at those unqualified test reports, seeing team members' bloodshot eyes and increasingly silent expressions, an unprecedented sense of helplessness and self‑doubt began gnawing at her heart.
Had she been too radical? Leading the team onto this difficult path—was it a mistake? Had she overestimated domestic foundational‑industry levels, overestimated her own problem‑solving ability? That version of her who had witnessed peak technology at ASML—had she subconsciously taken that industrial ecosystem and support system for granted, overlooking the extreme difficulty of climbing from scratch on this land?
For many nights in a row, she lay sleepless. In bed, what floated before her eyes wasn't family faces, nor serene Dutch scenery, but those wavefront‑aberration maps full of bizarre interference fringes. They twisted, rotated, like a silent mockery, tightly entangling her.
One night, she woke again at three a.m., heart pounding, sweat soaking her nightclothes. Immense pressure almost suffocated her. She walked to the living room, didn't turn on lights, curled on the sofa in darkness, feeling she might collapse under the invisible burden. Loneliness, anxiety, and a hint of fear‑of‑failure responsibility surged like tidewater, nearly drowning her.
Just then, her phone rang abruptly in the silence. The name flashing on screen was "Mozi."
She froze a long moment, almost thinking it hallucination. So late, why would he…
She took a deep breath, tried to make her voice sound normal, pressed answer.
"Hello?" Her voice carried a trace of unavoidable exhaustion and hoarseness.
A pause on the other end, then Mozi's steady voice came through, background quiet, as if he too was working. "Xiuxiu? Still up? Or… did I disturb you?" His acuity, as ever.
"No… it's fine." Xiuxiu reflexively denied, yet her taut nerves inexplicably relaxed a little hearing this familiar voice.
"You don't sound well." Mozi's tone held undeniable certainty. "Hit a tough problem? About the lens?"
Xiuxiu was stunned. How could he know? She hadn't mentioned specific progress, especially difficulties, in emails or calls.
Seeming to guess her puzzlement, Mozi explained calmly. "I've been following some industry dynamics and supply‑chain information. High‑precision aspherical lens elements are a recognized bottleneck for lithography‑machine breakthroughs. I guessed you'd likely get stuck here too."
His words held no show‑off or clairvoyant pretense, just a cool inference based on information and logic. But this ability to accurately deduce her predicament from afar, using mere traces, and this sudden call at her most vulnerable moment, instantly breached Xiuxiu's built‑up defenses.
She could no longer maintain surface strength; her voice choked. "…Yield's too low, wavefront aberration… can't be controlled. Feel… no hope." She haltingly poured out the difficulties and inner pressure. It was the first time she'd shown such fragility and helplessness before anyone.
Mozi listened quietly, didn't interrupt, didn't offer facile advice, didn't utter empty phrases like "hang in there."
Only after she finished did his low, clear voice return.
"I know that feeling. Like in the market, when all your models fail, all logic seems broken, only noise and incomprehensible volatility around—that sense of isolation and helplessness."
He paused, as if organizing thoughts, then continued.
"But Xiuxiu, remember, you're not facing an incomprehensible 'curse.' However complex 'wavefront aberration' is, it follows physical laws. That it looks like random noise now just means you haven't found the hidden, decisive 'parameter' or 'feedback loop.'"
His voice carried a peculiar, calming power. "It's like solving an extremely complex equation—too many variables, nonlinear relations. You might need to introduce new measurement methods, build more refined models, even need some luck to discover that key variable. But that doesn't mean the equation has no solution."
"Your value, and your team's, lies in finding this solution. The domestic foundation is weak, the process backward—that's precisely why you exist: to fill those gaps, tackle those 'hard bones' nobody wants to touch." His tone grew exceptionally firm. "I trust your judgment. Choosing immersion was the correct strategic direction. And a correct strategic direction often means encountering the cruelest resistance tactically. The current difficulty precisely proves you're walking the hardest, yet most correct, path."
"Don't be intimidated by temporary failure and low yield. Each failed lens element's test data isn't garbage; it's valuable clues pointing to the answer. Analyze them, compare them, find the commonalities underlying those 'random' errors. This takes time, patience, and—most crucially—your ability under pressure to maintain that calm that sees the problem's essence."
Mozi's words were like a cool beam piercing the fog and panic in Xiuxiu's heart. He didn't offer specific technical solutions; he gave her something more important—empathy based on deep understanding, reaffirmation of strategic direction, and an iron‑clad rational spirit that believes in logic and regularity even in extremity.
Xiuxiu held the phone, tears falling silently. But this time, not from despair—from immense solace in being understood and supported. She wasn't groping alone in darkness; on a distant other battlefield, someone also wrestling with complexity understood her situation and believed she could overcome.
"Thank… thank you, Mozi." She choked out, voice steadier than before.
"You're welcome." Mozi's voice softened slightly. "It's late. Get some rest. Tomorrow, with a fresh perspective, look at those data again. The answer must be hidden there."
After hanging up, Xiuxiu sat in darkness a long while. The city outside still slept, but her heart seemed to have undergone a baptism. Pressure remained, difficulties hadn't diminished at all, yet that near‑suffocating feeling of collapse vanished.
She stood up, turned on the desk lamp, took out those lens‑test reports that had made her feel hopeless before. This time her gaze wasn't trapped by those glaring out‑of‑spec numbers and chaotic fringes; she began focusing on finding patterns, contemplating Mozi's words—finding that hidden "key variable."
The "curse" of the lens remained severe; the road ahead still thorny. But in this deep night, a call bridging physical distance infused her with an intangible yet vital force—spiritual support. She knew that after dawn, she would carry this re‑gathered calm and courage back to that difficult battlefield, to continue, with her team, challenging that nanometer‑scale, destiny‑deciding precision limit.