Can a laser write single-atom-scale light sources?
Firing laser pulses just above a carefully measured damage threshold made defects smaller than 5 nm in boron nitride, and nearly every spot became a single, bright, stable source of individual photons.
Source
Laser manufacturing of spatial resolution approaching quantum limit
Study at a glance
- Design
- Other — Femtosecond laser writing of defects in exfoliated hBN flakes near a multi-shot-extrapolated threshold, followed by annealing and single-photon characterisation
- N
- No single N: 16-site array for single-emitter yield, 150 emitters for g2(0) statistics, 200 emitters for stability, roughly ten thousand centres for the wavelength histogram
- Population
- Laser-written colour centres in mechanically exfoliated hexagonal boron nitride flakes
- Outcome
- Feature size (TEM), number of emission peaks, single-photon purity g2(0), brightness, and photostability
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The authors varied the number of femtosecond pulses and pulse energy on hBN flakes to find the energy at which damage first appears if infinitely many shots were used (threshold tracking and lock-in). They then wrote defects with single pulses just above that threshold, annealed the flakes at high temperature, and imaged the sites by TEM and photoluminescence. They measured photon antibunching (g2), saturation brightness, polarisation and emission stability, and built a demo device with an hBN flake on a blue laser diode.
What they found
The extrapolated intrinsic threshold was 4.65 nJ, and single pulses near it produced features below 5 nm, consistent with a statistical-uncertainty estimate of about 3 nm. Lowering pulse energy reduced the number of emission peaks per site from more than ten to one, and g2(0) fell from 0.48 to 0.09. In the single-emitter array 15 of 16 sites were single-colour emitters, most emitters were brighter than 5 million counts per second, and 94% of 200 tested emitters showed negligible spectral diffusion or blinking.
The limits
What it doesn't show
The emission wavelength could not be controlled: centres were spread broadly across roughly 560 to 750 nm, and the defect structures behind them were not identified. The g2 values were not background-corrected and vary between emitters, so not every site is a high-purity single-photon source. The link between the 3 nm uncertainty estimate and the observed sub-5 nm features is a rough scaling argument, and TEM morphology varied from spot to spot. Annealing was essential for any emission, so the laser step alone does not create active emitters.
Key terms
- Colour centre
- A point defect in a wide-bandgap crystal that introduces localized energy levels and can emit light like an artificial atom.
- Single-photon emitter
- A light source that emits photons one at a time, needed for quantum communication and computing.
- g2(0) (second-order correlation)
- The probability of detecting two photons at the same time relative to random light; values below 0.5 indicate a single emitter.
- Threshold tracking and lock-in (TTL)
- Using extra pulses to amplify invisible damage and extrapolate the true material damage threshold, then writing just above it.
- Zero-phonon line
- The sharp emission peak from an optical transition that does not involve lattice vibrations.
- Hanbury Brown–Twiss setup
- An interferometer that splits light onto two detectors to measure photon coincidences and hence antibunching.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What intrinsic damage threshold did the multi-shot extrapolation give for hBN?
Common questions
How can a laser make features far smaller than its wavelength?
Damage only happens where intensity exceeds a sharp threshold, so if the pulse energy sits just above threshold only the very tip of the Gaussian focus modifies the material.
What stops features getting arbitrarily small?
Near threshold, the energy of atoms and electrons is statistically distributed, so the ablation site wanders randomly within a few nanometres; the authors estimate this uncertainty limit at about 3 nm for hBN.
Why does a smaller defect site give a purer single-photon source?
A large site contains many defect complexes that emit together; shrinking the site below the roughly 10 nm donor-acceptor scale makes it likely only one active centre is present.
More on Quantum optics