The Stringlight Research Institute, Carbon-based Chip Ultra-Clean Laboratory. The air cleanliness here had reached an astonishing level, far surpassing the areas previously used for material purification and basic research. The air carried a faint, distinctive scent—a mixture of high-purity chemical reagents, ozone, and various special gases—the unique "aroma" inherent to chip fabrication processes. Unlike the atmosphere of silicon-based chip production lines, which after decades of development had become standardized with a slight metallic and plastic feel, everything here carried the raw tension of early exploration: cautious yet unable to conceal excitement.
Xiuxiu, dressed in the highest-grade full-body anti-static cleanroom suit, resembled a white specter standing silently before the observation window separating the core lithography area from subsequent processing zones. Her gaze penetrated the thick specialty glass, closely following the robotic arm as it lifted a special "wafer" bearing their hopes from a specially modified "pre-alignment coater" adapted for carbon nanotube materials, smoothly transferring it to the next process—the electron beam lithography machine.
This "wafer" was not a traditional silicon disc. Its substrate was a specially treated sapphire wafer with an exceptionally smooth surface, upon which a layer of high-purity semiconducting carbon nanotube film had been deposited through a series of precise chemical vapor deposition and self-assembly processes, purified via "selective etching" technology. This film, thin to nearly a single carbon atom layer, carried the dream of transcending the limits of the silicon era.
The manufacturing flow for carbon-based chips, in its macroscopic steps, resembled silicon-based chip production: both relied on core stages like patterning (lithography), doping, etching, deposition, and packaging. Yet delving into each process detail revealed stark differences—these disparities precisely embodied the potential advantages and enormous challenges of carbon-based technology.
One of the most significant differences was **process temperature**.
Traditional silicon-based chip manufacturing, especially front-end high-temperature processes, often required environments of several hundred or even over a thousand degrees Celsius. For instance, thermal oxidation to grow silicon dioxide insulating layers, high-temperature annealing after ion implantation to repair lattice damage and activate dopants—these procedures demanded extreme heat. High temperatures not only meant enormous energy consumption but, more critically, affected material properties, limiting the range of materials integrable with silicon chips. Moreover, as transistor dimensions shrank to nanoscale, issues arising from high-temperature processes—thermal budget problems, impurity diffusion, material stress—became increasingly prominent, emerging as major obstacles to further miniaturization.
In contrast, a core advantage of carbon-based chip manufacturing was **low-temperature processing**.
Carbon nanotubes themselves possessed exceptional thermal and chemical stability. Fabricating carbon nanotube transistors typically did not require the extreme high-temperature steps characteristic of silicon processes.
Now, under Xiuxiu's watchful eyes, that sapphire wafer coated with carbon nanotube film was being loaded into an electron beam lithography machine specially adapted for its environment. The choice of electron beam lithography over more efficient optical lithography stemmed from the first verification wafer's design—feature sizes at the micron scale. For exploring process feasibility, direct-write electron beam lithography offered greater flexibility and precision control, avoiding the initial high costs and lengthy cycles of developing complex photomasks specifically for carbon-based materials. Within the vacuum chamber, the electron beam precisely scanned and irradiated the specialized photoresist covering the wafer surface according to preset circuit patterns, completing the pattern definition. This entire process proceeded at nearly room temperature.
Next came the critical "etching" and "electrode fabrication" stages.
Through dry or wet etching, carbon nanotube regions unprotected by photoresist were removed, forming isolated zones that would become future transistor channels—carbon nanotube arrays. This step too could be accomplished at relatively low temperatures.
Then, using high-precision electron beam evaporation or sputtering equipment in a vacuum environment, metal electrodes (typically gold, palladium, or other metals forming good ohmic contact with carbon nanotubes) were deposited onto predefined source and drain regions. Although the metal deposition process generated some heat inside the equipment due to energy injection, the substrate temperature was strictly maintained **below 200 degrees Celsius**, far lower than the 400–500 degrees Celsius or higher annealing temperatures common in silicon processes.
Finally, atomic layer deposition and similar techniques were needed to grow an ultra-thin, high-quality high-k gate dielectric layer (e.g., hafnium oxide) over the nanotube channel, followed by gate electrode deposition. Atomic layer deposition itself was a classic low-temperature process, typically controllable between 200–300 degrees Celsius.
**The entire carbon nanotube transistor fabrication flow could strictly limit its peak process temperature to within 300 degrees Celsius.** This stood in stark contrast to silicon-based chip manufacturing, which routinely demanded seven to eight hundred degrees or higher.
Low-temperature processes brought revolutionary potential advantages:
**First, extremely low energy consumption.** Dramatically reduced process temperatures directly translated to drastically decreased energy usage during manufacturing—a crucial step toward green, sustainable development for the chip industry, notorious for consuming massive amounts of water and electricity.
**Second, material compatibility.** Low-temperature environments enabled carbon-based chips to integrate with a wider variety of temperature-sensitive materials. For example, flexible, bendable, stretchable carbon-based electronic devices could be fabricated directly on plastic substrates in the future. It would also become easier to heterogeneously integrate optoelectronic devices, sensors, even biological components with carbon-based chips, realizing true "system-on-chip" or even "lab-on-chip" concepts.
**Finally, friendliness toward three-dimensional stacking.** In the silicon era, three-dimensional stacking technology was advancing to achieve higher integration density, but high-temperature processes between layers could adversely affect each other. Carbon-based chips' low-temperature processes made multilayer, high-density 3D integrated circuit stacking more feasible and straightforward, potentially breaking the integration limits of two-dimensional planes.
Of course, all these advantages rested upon successfully manufacturing functional carbon-based transistors. And at this very moment, Xiuxiu's team was taking that most difficult first step—from zero to one.
The laboratory atmosphere was so tense it felt tangible, as if one could wring water from the air. Every operator focused completely, as if performing delicate surgery. Each robotic arm movement, every vacuum pump start or stop, each droplet of chemical injection tugged at everyone's heartstrings. Although designing the simplest inverter logic unit with micron-scale linewidths, for carbon-based wafer fabrication where no one had yet completed the full process flow, any minor error could undo all previous efforts.
Time passed silently yet tensely, minute by minute. The wafer sequentially underwent electron beam lithography, etching, electrode deposition, gate dielectric deposition, gate electrode formation… Finally, that sapphire wafer—carrying countless efforts and hopes, its surface now covered with fine metal electrode traces and carbon nanotube channel patterns—was carefully loaded into the final rapid thermal annealing furnace for a relatively gentle low-temperature annealing step aimed at optimizing electrode contacts and stabilizing device performance, with temperature set at a silicon-process-dwarfing 350 degrees Celsius.
When the annealing furnace indicator light turned from red to green, atmosphere restored, and the robotic arm once again lifted that seemingly unchanged wafer out, preparing to transfer it to electrical testing, everyone's breath nearly stopped.
Xiuxiu left the observation window, striding quickly toward the testing area. She stood personally behind the semiconductor parameter analyzer connected to the precision probe station. The test engineer, using a micro-manipulator, precisely pressed tungsten-gold probes finer than a hair onto the first inverter unit's input, output, power, and ground pads on the wafer.
The laboratory fell utterly silent, save for the low hum of instrument fans and the thunderous pounding of hearts.
Xiuxiu took a deep breath and nodded to the test engineer.
The engineer pressed the test start button.
On the parameter analyzer screen, curves began to scan. Voltage, current… data points emerged one by one.
Initially, there was chaotic noise, like a haze. But as scanning progressed, a distinct current-voltage curve with clear nonlinear characteristics gradually took shape on the screen! Then, in dynamic characteristic testing, as input voltage shifted from low to high, output voltage clearly exhibited the "inversion" characteristic—high to low!
Although drive current remained small, switching speed slow, noise floor relatively high, performance far inferior to mature silicon-based transistors, even rougher than the team's simulation-predicted worst-case scenario…
But it worked!
A basically functional field-effect transistor based on high-purity semiconducting carbon nanotube arrays had been successfully manufactured on this small wafer! It achieved the most fundamental logic function—inversion!
"We… we did it! We made it!" someone cried out first, voice trembling with tears.
Instantly, long-suppressed elation burst forth like a dam breaking, flooding the entire testing area! Researchers, young and old alike, hugged each other excitedly—some waving fists vigorously, others removing goggles to wipe uncontrollable tears, while more stared at that rough yet precious characteristic curve on the screen, erupting in thunderous cheers!
Dr. Chen tightly grasped Professor Wang's hand—the two veterans who had led purification and process integration, eyes red-rimmed, lips quivering, unable to utter a word, just shaking each other's arms forcefully.
Xiuxiu didn't join the cheering crowd. She remained standing quietly before the parameter analyzer, leaning slightly forward, gaze fixed unwaveringly on the flickering curve on the screen, as if listening to a newborn's first cry.
Then, she signaled the test engineer to place a wafer fragment that had completed preliminary testing under a high-magnification optical microscope.
She leaned toward the eyepiece.
Her field of vision entered a microscopic world, shimmering with metallic luster and the deep hues characteristic of carbon materials. Interlaced electrode lines resembled miniature city roads, while within regions defined by these "roads" lay the carbon nanotubes she had devoted countless efforts to purify and arrange. Now, as a functional whole, they performed the "switching" mission driven by electrical signals. Under specific-angle light, successful transistor areas seemed to flicker with faint glimmers representing current flow—not physical light, but light of hope, light of the future.
A tremendous warmth surged from the deepest recesses of her heart, nearly staggering her, instantly overwhelming the calm and resilience she had long maintained. Her vision blurred; scalding tears silently traced her cheeks, dripping onto the front of her cleanroom suit.
She knew—she had always known—the immense significance of this crude, even somewhat unsightly carbon-based wafer in her hands. It not only proved the feasibility of the "selective etching" purification path, not only verified the preliminary unlocking of the low-temperature process flow for carbon-based wafer manufacturing.
It was also a declaration, a milestone. It announced to the world that the walls of the integrated circuit kingdom, ruled by silicon for over half a century, were not insurmountable. A new, limitless-potential path had revealed its first glimpse on the horizon.
Gazing at those flickering carbon-based transistors under the microscope—carriers of the future—Xiuxiu knew: an era transcending silicon would, no, **must**, begin here. This was merely a tiny, nascent start, yet its inherent power was sufficient to shake the foundations of digital civilization. She, and her team, were precisely the ones knocking on the door of this new age.