Optics
Kerr-Lens Mode-Locking: How Lasers Make Femtosecond Pulses
Squeeze a laser pulse down to 5 femtoseconds and you have carved light into a flash barely two optical cycles long — a spike of electromagnetic field so brief it would take 200 trillion of them, laid end to end, to fill a single second. The workhorse that makes such pulses is not an exotic switch or shutter but a subtle trick of the light itself: the beam warps the very crystal it passes through, focusing itself more tightly the brighter it gets.
Kerr-lens mode-locking (KLM) is the technique that turns this self-focusing into an ultrafast, effectively instantaneous shutter. A continuous-wave laser is coaxed into locking hundreds of thousands of its longitudinal cavity modes into phase, so that instead of steady light it emits a train of femtosecond pulses. KLM exploits the optical Kerr effect — an intensity-dependent refractive index — inside the gain crystal, most famously titanium-doped sapphire, to preferentially favor the pulsed state over continuous-wave operation.
- TypePassive mode-locking via optical Kerr effect
- Discovered1991, Spence, Kean & Sibbett (St Andrews)
- Key equationn(I) = n0 + n2·I
- Gain mediumTi:sapphire (Ti:Al2O3), ~800 nm
- Typical pulse5–100 fs, 70–90 MHz repetition rate
- n2 of sapphire≈ 3 × 10⁻¹⁶ cm²/W
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
What Kerr-Lens Mode-Locking Actually Is
A laser cavity supports many longitudinal modes — standing-wave frequencies spaced by the cavity free spectral range, Δν = c/2L. A Ti:sapphire oscillator with a 2-metre round-trip path has modes every ~150 MHz, and its gain bandwidth (~650–1100 nm) can span hundreds of thousands of them. Left alone, these modes oscillate with random relative phases and the output is noisy continuous-wave (CW) light.
Mode-locking forces all those modes into a fixed phase relationship. When phases align, the modes interfere constructively at one instant per round trip, producing a single sharp pulse that bounces back and forth. The more modes locked, the shorter the pulse — a direct consequence of the time–bandwidth product Δν·Δt ≈ 0.315 (for a sech² pulse).
- The setup: a Ti:sapphire crystal pumped by a green laser (~532 nm), folded-cavity mirrors, and a pair of prisms or chirped mirrors for dispersion control.
- The 'shutter': no physical modulator — the Kerr effect in the crystal itself creates a self-amplitude modulation that favors the high-intensity pulsed state.
The Mechanism: Self-Focusing as an Instantaneous Shutter
The engine of KLM is the optical Kerr effect: the refractive index of the crystal depends on the local light intensity,
n(I) = n0 + n2·I,
where n0 is the ordinary index (~1.76 for sapphire), n2 ≈ 3 × 10⁻¹⁶ cm²/W is the nonlinear index, and I is the instantaneous intensity. A laser beam has a Gaussian transverse profile — brightest in the center — so it experiences a higher index at the axis than at the edges. That radial index gradient acts exactly like a converging lens: the beam self-focuses.
Here is the crucial point. This Kerr lens is instantaneous (electronic response, sub-femtosecond) and intensity-dependent. A short, high-peak-power pulse self-focuses strongly; weak CW light barely focuses at all. If the cavity contains an aperture — a real slit (hard aperture) or the finite overlap with the pump beam (soft aperture) — the tightly self-focused pulse suffers less loss than the CW background. Higher intensity wins every round trip, so the pulse grows at the expense of CW light. This is self-amplitude modulation: a saturable-absorber-like action with no absorber and no recovery time.
Key Quantities and a Worked Example
Consider a typical Ti:sapphire oscillator: average power P_avg ≈ 500 mW, repetition rate f_rep ≈ 80 MHz, pulse duration τ ≈ 10 fs.
- Pulse energy: E = P_avg / f_rep = 0.5 W / 8×10⁷ Hz ≈ 6.3 nJ.
- Peak power: P_peak ≈ E / τ ≈ 6.3 nJ / 10 fs ≈ 630 kW. Focused to a ~30 µm spot in the crystal, that is an intensity I ≈ 9 × 10¹⁰ W/cm².
- Kerr index change: Δn = n2·I ≈ (3 × 10⁻¹⁶)(9 × 10¹⁰) ≈ 2.7 × 10⁻⁵ — small, but enough to bend the beam waist by tens of microns over the ~3 mm crystal.
Two other numbers govern the design. The critical power for self-focusing, P_cr ≈ 3.77·λ²/(8π·n0·n2), is roughly a few megawatts for sapphire at 800 nm; oscillators run comfortably below it to avoid catastrophic collapse. And the carrier period at 800 nm is T = λ/c ≈ 2.67 fs, so a 5 fs pulse is fewer than two optical cycles — essentially the shortest a Ti:sapphire pulse can be without shifting to a shorter carrier wavelength.
How KLM Is Achieved, Observed, and Applied
KLM is often not self-starting: from steady CW there is no intensity spike to seed the Kerr lens. Practitioners give the cavity a nudge — tapping a mirror, sweeping a prism, or using a fast intracavity modulation — to create a transient noise spike that self-focusing then amplifies into a stable pulse train. Once locked, it runs for hours.
Diagnostics:
- Autocorrelator — the standard tool to measure sub-100 fs durations, since no electronics are fast enough; the pulse measures itself via second-harmonic generation.
- FROG / SPIDER — reconstruct the full amplitude and phase of few-cycle pulses.
- RF spectrum analyzer — a clean single peak at f_rep confirms stable mode-locking.
Applications are enormous: KLM Ti:sapphire oscillators seed the amplifiers behind attosecond science and high-harmonic generation; they anchor optical frequency combs (Hall & Hänsch, 2005 Nobel Prize) that underpin optical atomic clocks; they drive two-photon microscopy, femtosecond micromachining, LASIK eye surgery, and pump-probe spectroscopy that films chemical reactions as they happen.
How KLM Compares to Related Techniques
KLM belongs to the family of passive mode-locking methods, but its response is uniquely fast. A real saturable absorber — the SESAM (semiconductor saturable-absorber mirror) that Keller pioneered in the 1990s — has a recovery time of picoseconds, which sets a floor on how short its pulses can be. SESAMs win on reliability: they self-start almost every time. KLM wins on speed: the Kerr response is essentially instantaneous, which is precisely why KLM holds the record for the shortest pulses from any oscillator.
- Additive-pulse mode-locking (APML): uses the Kerr effect in a coupled fiber cavity to reshape pulses interferometrically — physically KLM's cousin, but needing a stabilized second cavity.
- Colliding-pulse mode-locking (CPM): the pre-1990 dye-laser champion (~27 fs), rendered obsolete by solid-state KLM.
- Active mode-locking: an electronic modulator drives the loss; bandwidth-limited to picoseconds.
In practice, KLM and SESAM are often combined: a SESAM starts and stabilizes the laser while the Kerr lens shortens the pulse to the few-cycle regime — the best of both worlds.
Significance, Records, and Open Questions
KLM was discovered by accident. In 1990–91 at the University of St Andrews, David Spence, Philip Kean, and Wilson Sibbett found their Ti:sapphire laser spontaneously producing 60-femtosecond pulses with no mode-locking element in the cavity — 'self-mode-locking,' later explained as the Kerr-lens mechanism. Their 1991 Optics Letters paper launched the femtosecond revolution in solid-state lasers.
Within a decade, combining KLM with chirped-mirror dispersion compensation, groups (notably Krausz, Keller, and others) pushed Ti:sapphire oscillators to sub-two-cycle, ~5 fs pulses — near the fundamental limit set by the 2.67 fs carrier period at 800 nm.
- Open challenges: reliable self-starting KLM without mechanical perturbation; scaling to higher average power (multi-watt, thin-disk and Yb-based KLM) while keeping few-cycle durations; and stabilizing the carrier-envelope phase, which for few-cycle pulses determines the actual field waveform — essential for attosecond pulse generation.
KLM remains the gateway technology to the attosecond frontier: nearly every attosecond experiment on Earth begins with a self-focusing beam quietly building a pulse inside a sapphire crystal.
| Technique | Effective response time | Shortest pulses | Trade-off |
|---|---|---|---|
| Kerr-lens mode-locking (KLM) | < 1 fs (near-instantaneous) | ≈ 5 fs (Ti:sapphire) | Often not self-starting; needs a cavity nudge |
| SESAM (semiconductor saturable absorber) | ~0.1–10 ps recovery | ~50–100 fs | Reliably self-starting; slower recovery limits pulse floor |
| Additive-pulse mode-locking (APML) | < 1 fs (Kerr, fiber) | ~10–100 fs | Needs interferometrically stabilized coupled cavity |
| Colliding-pulse (dye laser, pre-1990) | ~fs (saturable dye) | ~27 fs | Fragile dye jets; obsolete for solid-state |
| Active (AOM/EOM) mode-locking | ~ns (electronic) | ~1–10 ps | Limited by modulator bandwidth, not the medium |
Frequently asked questions
What is Kerr-lens mode-locking in simple terms?
It is a way of making a laser emit ultrashort pulses instead of a steady beam, using the light's own intensity to focus itself inside the crystal. Bright pulses self-focus more than dim continuous light, so an aperture in the cavity favors the pulsed state and lets it grow. The result is a train of femtosecond pulses with no physical shutter or modulator.
Why is it called 'Kerr-lens' mode-locking?
The 'Kerr' refers to the optical Kerr effect, in which a material's refractive index rises with light intensity, n(I) = n0 + n2·I. Because a laser beam is brightest at its center, this creates a lens-like index profile that focuses the beam — a self-induced 'Kerr lens.' 'Mode-locking' means the technique locks the laser's many longitudinal modes into phase to build a pulse.
How short can Kerr-lens mode-locked pulses be?
Ti:sapphire KLM oscillators routinely produce pulses of 10–100 fs, and with careful dispersion control using chirped mirrors they reach about 5 fs — fewer than two optical cycles. That is near the fundamental limit, since a single optical cycle at 800 nm lasts only ~2.67 fs; shorter pulses require moving to a shorter carrier wavelength.
Why is Ti:sapphire the preferred crystal for KLM?
Titanium-doped sapphire has an enormous gain bandwidth (roughly 650–1100 nm), which is essential because supporting a broad spectrum is what allows an ultrashort pulse. It also has good thermal properties, a useful nonlinear index (n2 ≈ 3 × 10⁻¹⁶ cm²/W) for the Kerr lens, and can be pumped by common green lasers around 532 nm.
Is Kerr-lens mode-locking self-starting?
Usually not. Starting from steady continuous-wave light there is no intensity spike to trigger the self-focusing, so the laser needs a nudge — tapping a mirror, sweeping a prism, or using an intracavity modulator — to create a transient noise fluctuation. Once the Kerr lens amplifies that fluctuation into a pulse, mode-locking is stable and can run for hours.
How is Kerr-lens mode-locking different from a SESAM?
A SESAM (semiconductor saturable-absorber mirror) is a real absorbing device with a picosecond recovery time, so it self-starts reliably but cannot make the very shortest pulses. The Kerr lens responds essentially instantaneously (sub-femtosecond), which is why KLM holds the record for shortest oscillator pulses. The two are often combined: a SESAM starts the laser and the Kerr lens shortens the pulse.