Chirped Pulse Amplifier
Stretch the pulse, amplify the stretched pulse, then squeeze it back — and a tabletop laser briefly outshines the Sun.
Chirped pulse amplification (CPA) raises ultrashort (10–100 fs) laser pulses from nanojoules to joules of energy without damaging the gain medium. (1) A mode-locked oscillator produces a broad-bandwidth seed pulse. (2) A grating- or fiber-pair stretcher chirps the pulse — different colors arrive at different times — extending duration by 10⁴–10⁵× and dropping peak intensity by the same factor. (3) The stretched pulse traverses a regenerative or multi-pass amplifier (Ti:sapphire, Yb:fiber, or Nd:glass) gaining energy by stimulated emission. (4) A grating compressor (mirror-image of the stretcher) recompresses the amplified pulse to near-transform-limited duration. Output: terawatt to petawatt peak power, used for LASIK, micromachining, attosecond science, and laser-plasma physics.
Stretching uses the wavelength dependence of grating diffraction angles to give different colors different optical path lengths through a 4-pass grating-pair-and-telescope geometry, adding ~10⁸–10¹⁰ fs² of group-delay dispersion (GDD) and stretching a 30 fs pulse to ~1 ns. The stretched pulse traverses Ti:sapphire — a Ti:Al₂O₃ crystal pumped at 532 nm by a Q-switched Nd:YAG — for 4–10 multi-pass round trips, gaining 5–7 orders of magnitude in energy while peak intensity stays well below the crystal's damage threshold. After amplification, a grating compressor with the opposite GDD (achieved by reversed grating geometry) restores all colors to coincidence, producing a recompressed pulse within a factor of ~2 of the seed's transform limit. Output peak power P_peak = E_pulse / τ_pulse: a 1 J pulse compressed to 30 fs is ~33 TW; ELI-NP achieves 10 PW (30 J in ~25 fs). Plasma mirrors and pre-pulse control raise contrast to >10⁸ for clean target interactions. OPCPA (optical-parametric CPA) uses parametric amplification in BBO or LBO crystals instead of a population-inverted gain medium, supporting broader bandwidths.
In plain terms
Imagine you want to slam a huge amount of water through a small pipe at once — the pressure would burst the pipe. So instead, you stretch the flow over a long time, push it through gently, and then squeeze it back into a single giant burst at the far end. CPA does that to a laser pulse: stretches it, amplifies the stretched (gentle) pulse safely, then squeezes it into a femtosecond burst so bright that, for an instant, the spot it hits gets brighter than the surface of the Sun.
Also called: CPA, ultrafast laser amplifier, Ti:sapphire CPA, OPCPA.
Why it matters
- Made petawatt-scale peak powers practical — the entire field of relativistic laser-plasma physics relies on CPA.
- Enables LASIK eye surgery, where a femtosecond pulse cuts the corneal flap with no thermal damage; millions of procedures per year.
- Foundation of attosecond science (2023 Nobel Prize) — CPA pulses drive high-harmonic generation that produces sub-femtosecond XUV pulses.
- Powers laser micromachining, ultrafast spectroscopy, laser ion acceleration, and inertial-confinement fusion fast-ignition concepts.
Where it lives
- ELI-NP, Romania — 2 × 10 PW Ti:sapphire CPA
- ELI Beamlines, Czech Republic — multi-petawatt facilities
- Apollon, France — 10 PW laser
- SLAC LCLS-II auxiliary CPA lasers
- Commercial: Coherent / Spectra-Physics / Light Conversion
Key numbers
- Invention: 1985, Strickland & Mourou (Rochester)
- Nobel Prize: 2018 Physics (Strickland, Mourou + Ashkin)
- Stretch ratio: 10⁴–10⁵×
- Compressed duration: 20–100 fs
- Peak power achievable: kW lab → 10 PW (ELI-NP)
- Focused intensity record (~2024): ~10²³ W/cm²
Surprising facts
- A 1 PW pulse lasting 25 fs is just 7.5 µm long in free space — you could hold the entire pulse in your palm if it didn't immediately vaporize the air.
- The focused intensity of a petawatt beam (~10²² W/cm²) exceeds the entire Sun's power output concentrated onto a pinhead.
- The 1985 Strickland–Mourou paper used optical fiber as the stretcher; grating-pair stretchers came later and enabled the 10⁴–10⁵× stretch ratios needed for petawatt operation.
Open questions
- B-integral nonlinear distortion: Even with stretching, residual peak intensity in the amplifier accumulates nonlinear B-integral phase. Above ~1 radian, spatial and spectral distortions degrade compressibility. Managing B is a perennial design constraint — large beams, multi-pass geometries, and beam-quality diagnostics are the operational answer.
- Pre-pulse contamination from ASE: Amplified spontaneous emission produces a pedestal preceding the main pulse — for relativistic intensity targets, this pre-pulse can pre-ionize the surface before the main pulse arrives. Plasma mirrors raise contrast above 10⁸, but residual control of pre-pulse timing remains a per-experiment art.
A short history
Donna Strickland and Gérard Mourou published 'Compression of amplified chirped optical pulses' in Optics Communications in 1985, describing fiber-stretcher CPA at the Laboratory for Laser Energetics, Rochester. Subsequent grating-pair stretchers (Maine et al. 1988) enabled large stretch ratios. Through the 1990s and 2000s, Ti:sapphire CPA scaled from terawatt to petawatt. Strickland and Mourou shared the 2018 Nobel Prize in Physics with Arthur Ashkin (optical tweezers). ELI's 10 PW facilities (Romania 2019; Czech Republic) brought multi-petawatt science online; 100 PW concepts (XCELS, SEL) are in planning.
Sources
- Strickland & Mourou, Optics Communications 56, 219 (1985) (https://www.sciencedirect.com/science/article/pii/0030401885901518)
- Nobel Prize 2018 — Strickland, Mourou, Ashkin (https://www.nobelprize.org/prizes/physics/2018/press-release/)
- ELI Beamlines lasers (https://www.eli-beams.eu/facility/lasers/)
- Maine et al., IEEE J. Quantum Electron. 24, 398 (1988) (https://ieeexplore.ieee.org/document/137)