Every quantum technology built on atoms or photons starts with a laser, and that laser quietly sets the performance ceiling. Lasers for quantum technology must hit an atomic or telecom wavelength, hold a narrow linewidth, stay frequency-stable for hours and add almost no intensity noise. The numbers differ widely by application: a cold-atom trap tolerates kilohertz-class linewidth, while an optical clock needs about 1 Hz. This guide maps each application to its laser requirements and explains why single-frequency fiber lasers suit systems that must leave the lab.
What a quantum system asks of its laser
A single-frequency laser emits on one longitudinal mode, so all of its power sits in one narrow spectral line. Four properties then decide whether it works in a quantum experiment:
- Wavelength: set by the atom or the fiber band, not by preference. Rubidium cooling uses 780.24 nm; metastable helium uses about 1083 nm.
- Laser linewidth: how sharply the optical frequency is defined. It controls how well the light couples to a narrow transition.
- Frequency stability: drift over seconds to hours. A drifting laser walks off resonance mid-measurement.
- Relative intensity noise (RIN): power fluctuations that turn into measurement noise or shift atomic energy levels.
A fifth figure, coherence length, follows from linewidth at roughly c/(πΔν). A 1 kHz laser stays coherent over about 95 km, and 100 Hz stretches that to around 950 km. Our guide to coherence length shows how to turn a linewidth spec into a practical limit for your setup.
Map of quantum applications to laser requirements
Requirements span several orders of magnitude, from MHz-tolerant atom traps to Hz-class clocks. The table lists reported and typical values; your own error budget should set the final specification.
| Application | Typical wavelength | Linewidth needed | Stability and noise focus |
|---|---|---|---|
| Laser cooling and trapping (Rb MOT) | 780 nm | kHz-class is ample; the Rb D2 line is 6.07 MHz wide | Drift well below a few MHz; steady power |
| Optical atomic clocks (Sr, Yb) | 698 nm (Sr), 578 nm (Yb) | About 1 Hz after cavity lock | Fractional stability near 10⁻¹⁵ over seconds |
| Atom interferometry, gravimetry | 780 nm (Rb) | Hz-class in the most demanding designs | Frequency noise and 0.1%/√Hz intensity stability in large-scale designs |
| Optically pumped magnetometers | 795 nm (Rb D1), 1083 nm (He) | Sub-MHz to tens of kHz, locked to the line | Frequency jitter and RIN set the field-noise floor |
| Quantum communication (CV-QKD) | 1550 nm | About 100 Hz to under 1 kHz reported | RIN below −100 dBc/Hz reported |
| Neutral-atom and ion qubit control | Set by species | Narrow enough to drive pulses cleanly | Frequency and phase noise limit pulse fidelity |
Linewidth and stability requirements, application by application
Laser cooling and trapping atoms
In a magneto-optical trap the cooling light sits one to three natural linewidths below resonance, which is roughly 6 to 18 MHz red-detuned for rubidium. A kilohertz-class linewidth is therefore far narrower than the trap needs. What matters is that the center frequency stays put and the power stays steady, because drift destabilizes trapping and power swings distort atom number readings. A 780 nm single-frequency laser made by doubling a 1560 nm fiber seed offers 1 to 5 kHz linewidth and works with standard saturated-absorption locking. Neutral-atom and ion computers reuse the same cooling and state-control light, so the same requirements carry over.
Atomic clocks
Optical lattice clocks probe extremely narrow transitions. The strontium clock line at 698 nm is only about 10 mHz wide, so the probe laser must be far quieter than a trap laser. Practical clock lasers reach about 1 Hz and roughly 10⁻¹⁵ fractional stability over a few seconds, using a high-finesse reference cavity. A fiber source does not replace that cavity, but a Hz-level starting point gives the lock less noise to remove. If the term is new, see what Hz-level linewidth actually means in a fiber laser.
Quantum sensing
Optically pumped magnetometers turn laser frequency errors directly into field noise. In one cesium magnetometer study, stabilizing frequency jitter from 28 MHz to 0.12 MHz brought closed-loop noise down to 0.012 pT. Helium sensors pump the 1083 nm line, which is why a 1083 nm magnetic detection laser must be frequency-stabilized and mode-hop-free. Atom interferometers add a phase requirement: the MAGIS-100 design lists frequency noise below 10 Hz and 0.1%/√Hz intensity stabilization. For these builds, a single-frequency seed laser designed as a laser for quantum sensing removes the source as a noise contributor.
Quantum communication
Continuous-variable QKD demonstrations over 100 km of fiber have used 1550 nm lasers with linewidths near 100 Hz, and others under 1 kHz with RIN below −100 dBc/Hz. The local-local-oscillator approach needs two independent lasers, one at each end, so each must be narrow and quiet on its own. RIN decides whether detection reaches the shot-noise limit; see how low RIN needs to be for the measurement background.
Why fiber lasers beat bulk lasers for fieldable quantum systems
External-cavity diodes and Ti:sapphire lasers remain excellent on an optical table. Outside the lab, free-space alignment drifts with temperature and vibration, and a quantum sensor on a vehicle, aircraft or survey platform cannot be realigned by hand. Fiber lasers solve this structurally:
- No alignment to lose: light is guided inside the fiber. An all-fiber laser system built for rubidium cold-atom interferometry was reported as portable, easy to align and insensitive to temperature and vibration.
- Telecom heritage: 1.5 µm components are mature and reliable. The fibered laser system for the MIGA atom interferometer was built on C-band telecom technology, aimed at mobile, field-deployable quantum sensors.
- Frequency conversion keeps quality: SHG from a 1560 nm seed passes its linewidth and phase noise to 780 nm. Direct 780 nm diodes are sensitive to optical feedback and usually need active locking.
- Mode-hop-free in a small package: a centimeter-scale DBR cavity avoids mode hops and burst noise. Our 1083 nm module measures 205 × 150 × 28 mm, runs on 5 VDC and operates from −20 to 50 °C with PM980 output.
- Lower size, weight and power: field platforms are limited by size, weight, power and cost, and compact fiber modules help on all four.
The trade-off is tuning range. Bulk lasers tune across wide bands, while fiber sources cover fixed bands and reach other colors by conversion. For fixed atomic lines such as 780 nm, 795 nm and 1083 nm, that is rarely a limitation. For help choosing a band, the fiber laser wavelength guide compares 1.0, 1.5 and 2.0 µm.
Matching a source to your quantum setup
Once the application sets the wavelength and the noise budget, the pairing is usually straightforward:
- Rubidium D1 optical pumping (795 nm): the 795 nm frequency-converted laser offers 2 to 5 kHz linewidth for SERF magnetometers.
- Telecom-band qubits and QKD (1550 nm): the 1.5 µm high-reliability seed source delivers sub-1 kHz linewidth, mode-hop-free.
- Externally locked references: the 1.5 µm frequency-stabilized seed provides an error-signal port for PDH-style locking, with drift below ±500 kHz RMS over two hours.
- Clock and interferometer front ends: the Hz-level fiber laser reaches below 25 Hz linewidth at 1550 nm.
Central wavelengths are customizable on the converted lasers, and the engineering team can review your linewidth, RIN and package constraints before you order. Contact us with your target transition and noise budget.
FAQ
What linewidth do lasers for quantum technology need?
It depends on the application. Atom cooling works with kilohertz-class linewidth, quantum communication demonstrations report about 100 Hz to under 1 kHz, and optical clock lasers reach about 1 Hz after cavity stabilization.
Why is 780 nm used for rubidium experiments?
It matches the rubidium D2 transition at 780.24 nm, the standard line for laser cooling, magneto-optical trapping and atom interferometry.
Is a fiber laser better than a diode laser for quantum sensing?
For fieldable systems, usually yes. Fiber lasers need no free-space alignment and tolerate temperature and vibration better, though diode and bulk lasers tune across wider bands on a lab bench.
