SHINE Explained: From Origins to Now

Why it’s suddenly everywhere

1. 01

   On October 21, 2026, SHINE produced free-electron laser light for the first time: a 4 GeV continuous-wave superconducting linac drove an electron beam through 18 undulators, each 4 meters long with a 55 mm magnetic period, according to Xinhua.

2. 02

   Only two machines of this class exist on Earth — high-repetition-rate hard X-ray lasers. The other one, SLAC’s LCLS-II High Energy upgrade, is still being built in the US; Fermilab shipped its final cryomodule contribution in April 2026.

3. 03

   SHINE is the most expensive national big-science infrastructure China has ever built — nearly 10 billion yuan — buried about 30 meters under Shanghai’s Zhangjiang Science City in a 3.1 km tunnel.

4. 04

   The milestone lands inside a full photon-science cluster: the Shanghai Synchrotron Radiation Facility (running since 2009) and the SXFEL soft X-ray user facility sit in the same district.

The timeline

1. 1971–1976The free-electron laser is invented

   John Madey at Stanford proposes making relativistic electrons lase by wiggling them through an alternating magnet array — an undulator. His team demonstrates the first FEL in the infrared in 1976.

2. 2009-04LCLS switches on

   SLAC’s Linac Coherent Light Source becomes the world’s first hard X-ray free-electron laser, roughly ten orders of magnitude brighter than any synchrotron X-ray source before it.

3. 2011–2017The XFEL club forms

   SACLA starts up in Japan (2011); the European XFEL in Hamburg (2017) adds MHz-class burst repetition. China’s Dalian Coherent Light Source lases in the vacuum ultraviolet in 2017 — the country’s first FEL light.

4. 2018-04-27SHINE breaks ground

   Construction starts in Zhangjiang, Shanghai: a 3.1 km tunnel about 30 meters underground for an 8 GeV superconducting linac, three undulator lines and ten experimental stations. ShanghaiTech University is the legal owner-builder.

5. 2022-03Tunnel breakthrough

   The beamline tunnel linking the experiment hall to the accelerator’s Shaft 3 is holed through — civil engineering moves toward the fit-out phase.

6. 2025-11Superconducting linac milestone

   The L1/BC1 segment of the linac accelerates beam past 330 MeV using 1.3 GHz and 3.9 GHz superconducting cryomodules — the machine’s core technology proving itself in the tunnel.

7. 2026-10-21First light

   A 4 GeV continuous-wave beam is threaded through the first undulator line and SHINE lases — the decisive step on the way to full completion.

8. 2027Target: full operation

   Planned specs at handover: 8 GeV electrons, photon energies of 0.2–15 keV on the first two undulator lines, femtosecond pulses at up to 1 MHz — one million X-ray flashes per second.

The backstory

A laser with no lasing medium

Every laser you have ever met — pointers, fiber optics, DVD players — works the same way: electrons inside a material get pumped up and drop back down in lockstep, releasing identical photons. The material is the gain medium, and its atomic energy levels dictate which colors of laser light are even possible. That is why, for most of the 20th century, there was no such thing as an X-ray laser: no convenient material has energy levels in the right place.

Madey’s insight was to throw the material away. Send a beam of nearly light-speed electrons through an undulator — a row of magnets with alternating north-south poles — and the electrons are forced to wiggle. Wiggling electrons radiate. If the wiggle is tuned just right, the light from the front of the electron bunch interacts with the back of the bunch, which bunches tighter, which radiates harder: a runaway feedback called self-amplified spontaneous emission, SASE. The “gain medium” is the electron beam itself — free electrons, hence the name. And because the wavelength is set by the electron energy and the magnet spacing rather than by atoms, you can dial it down to X-rays, with peak brightness billions of times beyond any synchrotron.

Why now: the repetition-rate race

First-generation XFELs were single-shot marvels. The original LCLS fired 120 pulses a second; each experiment was a precious, rationed flash. The frontier since the 2010s has been repetition rate — firing a million pulses a second instead of a hundred — because most real science is statistics: weak signals, rare events, movies stitched from millions of frames. Getting there required replacing warm copper accelerators with superconducting radio-frequency cavities that can run continuously without melting.

That is the race SHINE just entered. The European XFEL reaches high rates only in short bursts; SLAC’s LCLS-II switched on its continuous-wave superconducting linac in the 2020s, and its High Energy upgrade — the only peer to SHINE, doubling the beam to 8 GeV — is still under construction, with Fermilab completing its cryomodule contribution in April 2026. SHINE’s October 21 first light, at 4 GeV with the full first undulator chain engaged, means the second of these two machines has crossed the hardest threshold: the superconducting linac and the undulators working together in the tunnel, lasing. Decade-scale bets of this kind are having a moment — weeks earlier, the IceCube Neutrino Observatory took a Nobel Prize 38 years after it was proposed. The gap between a technical milestone and full operation has another current counterpart: Hohhot Shengle Airport is rehearsing an overnight transfer before its proposed November 26 opening.

Zhangjiang’s decade-long bet

SHINE did not appear out of nowhere; it is the third act of a deliberate Shanghai strategy. The Shanghai Synchrotron Radiation Facility opened to users in 2009 and grew into China’s most-used big-science facility. The SXFEL soft X-ray free-electron laser followed as the country’s first X-ray FEL user facility. SHINE is the flagship: nearly 10 billion yuan, 3.1 kilometers of tunnel 30 meters under Zhangjiang Science City, with ShanghaiTech — a university founded only in 2013 — as the owner-builder, and the Chinese Academy of Sciences’ Shanghai institutes responsible for the accelerator, undulators and the extreme-light experimental station.

Burying the machine is engineering, not drama. Thirty meters of earth damps the vibrations and temperature swings that would otherwise smear out an X-ray beam focused on atoms — and it puts a radiation shield between a 8 GeV electron beam and the city above. The same district now hosts the full genealogy of light sources: third-generation synchrotron, soft X-ray FEL, and a hard X-ray FEL that, at completion, will be one of two of its kind in the world.

What a “molecular movie” actually is

The phrase from the official coverage — a camera that shoots molecular movies — is literal enough to be useful. A synchrotron takes still photographs of matter: expose a protein crystal long enough, average over trillions of molecules, and you get a structure. An XFEL flash is so short (femtoseconds, the time scale on which chemical bonds actually vibrate) and so intense that it outruns damage entirely: the pulse diffracts off a single tiny crystal in the instant before the crystal is vaporized — “diffract before destruction.” String a million such frames together and you have a film of a molecule moving: bonds breaking, charge flowing, a photosynthetic reaction center doing its one trick. That is the capability SHINE is being built to deliver, at a million frames per second, to life science, quantum materials and energy research.

Key facts

Investment

~¥10 billion

Tunnel

3.1 km, ~30 m down

Electron energy

8 GeV (target)

Photon range

0.2–15 keV (phase 1)

Repetition rate

up to 1 MHz

Completion target

end of 2027

What people are saying

Common questions

How is a free-electron laser different from a normal laser?

A normal laser amplifies light inside a material, so the available colors are fixed by atomic energy levels — which is why practical X-ray lasers long seemed impossible. A free-electron laser has no gain medium at all: relativistic electrons are forced to wiggle through an alternating magnet array (an undulator) and amplify their own emitted light via the SASE feedback process. The wavelength is set by electron energy and magnet spacing, so it can be tuned deep into the X-ray range.

Why bury it 30 meters underground?

Stability and shielding. The X-ray beam must stay steady to atomic scales, and surface vibrations, traffic and temperature swings would smear it out; thirty meters of earth is a natural damper. The rock and soil also absorb radiation from the 8 GeV electron beam, letting a major accelerator sit inside a city science district.

How is SHINE different from the Shanghai Synchrotron (SSRF)?

SSRF is a third-generation storage-ring source: extremely useful, but its X-rays are relatively dim and its pulses relatively long — still photography. SHINE is a fourth-generation free-electron laser: pulses billions of times brighter at their peak and short enough (femtoseconds) to freeze molecular motion. They complement rather than replace each other, and both sit in the same Zhangjiang photon-science cluster.

When can scientists actually use it?

First light is the start of commissioning, not the finish line. The official target is completion and handover by the end of 2027, with the first phase offering two undulator lines covering 0.2–15 keV and ten experimental stations. Between now and then, the team will raise the beam energy from 4 to 8 GeV and bring the stations online one by one.