What Does It Cost to Build Quantum Communication Crystals?

A tiny flaw inside a crystal can spit out single bits of light on demand. These bright flaws, known as a colour centre, act as building blocks for future hack-proof networks. Yet making these glowing spots is not simple. A new study has tracked the hidden factory side of the quantum communication footprint. The findings show striking differences across twenty crystal recipes.

Mapping twenty ways to make a flaw

The team paired five crystals with four flaw-making methods to build a single photon chip. To keep the test fair, they used a cradle-to-gate boundary. This approach counts every burden from digging rocks out of the ground up to the factory door. However, the study deliberately left out real-world operating gear like deep freezers and network cables to isolate chip building alone.

The team tracked total power, water use, health risks, and cash costs for each chip. The results varied by a huge margin. The cumulative energy demand measures the total power swallowed across the entire supply chain. This value ran from 76 megajoules for laser-treated hexagonal boron nitride (h-BN) up to 687 megajoules for diamond. Shooting charged atoms into diamond pumped out 77 kilograms of greenhouse gas per chip. In contrast, growing layered h-BN produced just 8 kilograms.

Where the costs hide

How you put a flaw into a crystal changes everything. Blasting targets with fast laser pulses proved to be the most power-thrifty option for all five crystals. Laser writing needed 292 megajoules of energy per chip in diamond. Other diamond methods averaged 600 megajoules.

Even so, power use tells only half the story. Water demand brings other shocks. Growing thin crystal layers through heteroepitaxial growth swallowed six cubic metres of water per chip for gallium nitride (GaN). Cooling hot chambers, scrubbing dirty exhaust gas, and rinsing parts with ultra-pure water drove this high demand.

Cash costs also swung wildly. A simple h-BN laser-cut chip costs under $25 to produce. Meanwhile, a GaN chip topped $200. Pricey starter raw materials accounted for roughly $90 of that GaN sum. Finally, diamond carried the highest human health risks from toxic factory smoke, whereas h-BN ranked lowest.

Why greener choices bring compromises

Knowing where each factory burden comes from lets engineers pick cleaner designs early on. For example, shops could swap out harsh cleaning acids or switch to short laser pulses. Yet low-impact choices bring practical hurdles. Cheap h-BN chips require far less power, but diamond still sets the gold standard for pure quantum performance.

Key questions remain open. Scientists do not yet know if cheaper crystals can hold quantum data as securely as diamond. Furthermore, real factory yields for whole wafers remain untested, and several fresh raw materials still lack complete health hazard data.