China's impossible semiconductor
How China's breakthrough reveals why the best technology rarely wins
When Chinese scientists announced their "golden semiconductor" breakthrough last week, something curious happened in global markets. Indium prices jumped 8% within hours—not because traders believed silicon's days were numbered, but because they grasped something the breathless headlines missed entirely.
They knew that scarcity transforms technological promise into economic impossibility.
The breakthrough itself is genuinely remarkable. Researchers at Peking University and Renmin University have produced 2-inch wafers of indium selenide with unprecedented quality, achieving results published in Science that demonstrate electron mobility 5-10 times higher than silicon, atomic-scale thickness, tunable bandgap properties, and superior switching characteristics. By every meaningful technical measure, indium selenide outperforms the foundation of modern computing.
Yet this achievement illuminates a fundamental paradox: the better the technology, the more certain its commercial failure.
The abundance illusion
Silicon dominates semiconductors not through superior performance but through geological fortune. Comprising 28% of Earth's crust, silicon enables commodity-scale production that makes modern electronics affordable. Your smartphone contains silicon chips because silicon is abundant, not because it's technically optimal.
Indium tells a completely different story. This silvery metal ranks among Earth's scarcest elements, with crustal abundance similar to silver but distributed in concentrations that make extraction far more challenging. No dedicated mines exist—it appears only as byproduct from zinc processing, recovered in quantities so tiny that global reserves total merely 356,000 tonnes.
Current pricing ranges $240-771 per kilogram for semiconductor-grade purity. Compare this to silicon's commodity pricing, and the economic gulf becomes stark.
Professor Liu Kaihui, who led the breakthrough, celebrates producing 5-centimetre diameter wafers as enabling "mass production." Modern silicon fabs process wafers 300mm in diameter—nearly six times larger, over 35 times the surface area. A single contemporary wafer accommodates thousands more individual chips, creating orders of magnitude difference in unit economics.
The terminology reveals priorities. What academic papers herald as "mass production" represents laboratory-scale prototyping by semiconductor industry standards.
The pattern of broken promises
Materials science exhibits a systematic pattern: breakthrough announcements followed by commercial disappointment. Gallium arsenide demonstrated superior electron mobility in the 1980s, earning predictions of silicon replacement that never materialised. Today, GaAs serves only specialised applications—mobile phone power amplifiers, satellite communications, military radar.
Gallium nitride revolutionised power electronics and LED lighting, yet silicon retains computing dominance. Graphene earned Nobel Prizes whilst remaining largely laboratory-bound fifteen years later. Carbon nanotubes, molybdenum disulfide, and black phosphorus followed identical trajectories—spectacular demonstrations yielding minimal commercial impact.
The pattern persists because academic incentives reward publications demonstrating superior performance, not analyses of supply chain constraints or manufacturing scalability. We keep discovering materials that outperform silicon in laboratory conditions, then wonder why industry doesn't adopt them. The answer isn't technical—it's thermodynamic. Most superior materials depend on elements that stellar nucleosynthesis produced in tiny quantities.
The economics of impossibility
Indium selenide faces perverse feedback dynamics that worsen with success. Unlike materials with expandable supply chains, geological scarcity creates permanent bottlenecks. Any significant adoption would drive demand beyond existing production capacity, triggering price spirals that eliminate economic advantages over silicon.
Current consumption centres on LCD screens and LED applications where small quantities justify premium pricing whilst operating at scales compatible with supply constraints. Expanding into semiconductor processing would immediately stress global supply chains beyond breaking point.
The mathematics are sobering. Modern CPU production requires wafers containing hundreds of billions of transistors. Replacing even a fraction of silicon production with indium selenide would consume global reserves within months whilst pricing the material beyond commercial viability.
Beijing's announcement strategically emphasises applications in "artificial intelligence, autonomous driving and smart terminals" rather than personal computing—language suggesting recognition that indium selenide cannot compete through cost reduction but might create value through performance exclusivity that scarcity maintains.
The strategic dimension
China's indium selenide research reveals sophisticated strategic thinking that international coverage misses. Beijing controls 66% of global indium production—not through mining prowess but through zinc processing dominance. Research into indium applications creates technological dependencies on Chinese-controlled materials whilst competitors pursue technologies China can restrict.
The timing proves telling. Publishing this breakthrough amid escalating US semiconductor restrictions provides international credibility whilst avoiding commitments to scaled production that would reveal supply chain vulnerabilities.
Rather than failed silicon replacement, this represents successful strategic positioning around scarce materials with specialised applications. The breakthrough advances legitimate research whilst creating optionality around supply chains Beijing controls.
Military and aerospace markets represent the most realistic commercial pathways. These sectors value capability over cost whilst operating at scales compatible with geological constraints—radar systems achieving unprecedented sensitivity, satellite communications requiring radiation resistance, sensors operating in conditions that defeat conventional semiconductors.
The specialisation inevitability
Understanding indium selenide's true potential requires abandoning silicon replacement fantasies. Its future lies where extreme performance justifies extreme costs—applications that create value through scarcity rather than abundance. Exclusive access to superior performance becomes competitive advantage rather than economic burden.
The Peking University breakthrough enables such applications by achieving quality standards necessary for reliable device fabrication. Two-inch wafers support prototype development and small-scale production suitable for specialised markets—genuine progress towards commercial viability within appropriate economic constraints.
The innovation paradox
Indium selenide illuminates a broader truth about technological progress that breakthrough announcements systematically obscure. Superior performance often guarantees commercial failure through economic constraints that laboratory conditions cannot replicate.
The pattern extends beyond semiconductors to any innovation dependent on scarce materials—battery technologies requiring lithium, rare earth magnets needing neodymium, catalysts consuming platinum. All face identical constraints where geological scarcity limits scalability regardless of performance advantages.
Yet each breakthrough generates media attention proportional to replacement implications whilst burying economic analysis revealing niche realities. This creates systematic misunderstanding that persists across innovation cycles.
The Chinese researchers have achieved something genuinely valuable by recognising economic constraints rather than ignoring them. Their work succeeds by positioning indium selenide for markets where scarcity creates rather than destroys value.
The golden semiconductor remains trapped not by technical limitations but by the very abundance that made silicon ubiquitous. Sometimes the most advanced technology is destined for the most specialised applications—where being better than silicon means accepting you'll never replace it.