Bright Picosecond Squeezed Light

• Bright picosecond pulsed squeezed light is generated through nonlinear processes (OPA in χ² and SFWM in χ³) to reduce quantum noise in ultrashort optical pulses.
• Integrated platforms like TFLN, PP-KTP, and Si₃N₄ microresonators enable high brightness, nearly single-mode operation, and measurable squeezing up to several dB.
• Challenges include loss management, multi-pair emissions, and modal purity tradeoffs, motivating advances in detector efficiency and mode-engineering.

Bright picosecond pulsed squeezed light refers to quantum optical states generated via nonlinear interactions, resulting in the reduction (squeezing) of quantum noise in specific field quadratures of ultrashort (picosecond-scale) optical pulses with high photon flux. Such states are key enablers in continuous-variable quantum information science, quantum-enhanced measurement, and nonlinear optical technologies. Recent advances have realized bright, nearly single-mode pulsed squeezed light in integrated waveguides, microresonators, and periodically-poled nonlinear crystals at telecommunication and visible wavelengths, with measured squeezing levels up to several dB and corrected internal levels exceeding –15 dB (Peace et al., 2022, Eckstein et al., 2010, Brusaschi et al., 5 Oct 2025, Terrasson et al., 22 Jan 2026).

1. Physical Principles and Theoretical Description

Bright picosecond pulsed squeezed light is most commonly produced using optical parametric amplification (OPA) in $\chi^{(2)}$ nonlinear media or spontaneous four-wave mixing (SFWM) in $\chi^{(3)}$ platforms. Under the undepleted pump approximation, the relevant interaction Hamiltonians for single- or two-mode squeezing are:

• OPA ($\chi^{(2)}$):

$$H = i\hbar\,(\kappa\,\hat{a}^{\dagger 2} - \kappa^*\,\hat{a}^2)$$

For the single-mode case, and

$$H_\text{PDC} = i\hbar\,\chi\,(\hat{a}_s^\dagger\,\hat{a}_i^\dagger - \hat{a}_s\,\hat{a}_i)$$

for two-mode squeezing (e.g., type-II PDC) (Eckstein et al., 2010, Terrasson et al., 22 Jan 2026).

The squeezing parameter $r$ is the effective interaction strength for a pulse of peak power $P_\text{peak}$ and nonlinear medium of length $L$: $r = \gamma\,P_\text{peak}\,L$ with the nonlinear constant $\gamma$ dependent on material, mode area, and effective nonlinear coefficient. The quadrature squeezing (dB) is $S_{-} = -10\log_{10}(e^{-2r})$ (Peace et al., 2022). For a degenerate OPA, the quadrature variances scale as $\Delta X_{\pm}^2 = (\hbar/2)\,e^{\mp2r}$ (Terrasson et al., 22 Jan 2026). In the two-mode regime, the EPR noise reduction is $S = 10\log_{10}(e^{2r})$ (Eckstein et al., 2010).

2. Generation Schemes and Platform Characteristics

A. Integrated $\chi^{(2)}$ Waveguides

Thin-film lithium niobate (TFLN) strip-loaded waveguides offer high nonlinearity and low loss with engineered quasi-phase matching (QPM) (Peace et al., 2022). Typical device parameters are:

• Waveguide length: $L=4.7$ mm
• Effective mode area: $A_\text{eff}\sim5\,\mu\mathrm{m}^2$
• Nonlinear coefficient: $d_\text{eff}\approx20$ pm/V
• Refractive index: $n\approx2.2$
• $\gamma\sim8$ W$^{-1}$m$^{-1}$ For sub-0.3 W on-chip pump power, $r\sim0.094$ yields on-chip squeezing –1.7 dB (inferred) over a 230 GHz phase-matching bandwidth (Peace et al., 2022).

B. Waveguide-Based PDC in PP-KTP

Type-II PDC in periodically-poled potassium titanyl phosphate (PP-KTP) waveguides facilitates ultrafast two-mode squeezing in the telecom band. With effective nonlinear lengths $L_\text{eff}\sim8$ mm and optimized pump and phase-matching bandwidths, single-mode EPR states are produced, achieving up to $\langle n\rangle = 2.5$ photons per pulse and squeezing levels of 11 dB (Eckstein et al., 2010).

C. Si$_3$N$_4$ Microresonators via SFWM ($\chi^{(3)}$)

High-Q silicon nitride microring resonators—FSR 200 GHz, loaded $Q\sim8\times10^5$—enable bright, picosecond pulsed squeezed light via pulsed SFWM. Escape efficiency $p_e\sim0.75$ allows on-chip squeezing up to 5–5.7 dB in the high-gain regime (internal $r\to14$ dB), with single-mode purity preserved by optimal pump detuning to compensate SPM and XPM (Brusaschi et al., 5 Oct 2025).

D. Ridge PPLN—Bright Amplitude Squeezing for Microscopy

Ridge PPLN waveguides (5 mm, AR-coated) with synchronized 5–6 ps pulses at 532/1064 nm achieve up to –3.2 dB bright amplitude squeezing (single detector), with internal (loss-corrected) squeezing reaching –15.4$^{+2.7}_{-8.7}$ dB at moderate average pump powers (20–40 mW) (Terrasson et al., 22 Jan 2026).

3. Temporal and Spectral Mode Structure

High-brightness, single-mode operation is attained by matching the pump bandwidth to the waveguide phase-matching bandwidth, optimizing the joint spectral amplitude $f(\omega_s,\omega_i)$ to approach a factorable, pure Schmidt mode. In the PP-KTP system, Schmidt number $K\simeq1.05$ (via $g^{(2)}(0)\approx1.95$) is achieved at $\Delta\lambda_\text{pump} = 1.95$ nm, signifying nearly ideal single-temporal-mode pulsed output (Eckstein et al., 2010). In TFLN, temporal walk-off between pump and squeezed pulse (1.5 ps over 4.7 mm) is negligible for 12 ps pulses, supporting efficient squeezing over a 230 GHz PM bandwidth (Peace et al., 2022).

In microresonators, time-resolved $g^{(1)}(\tau)$, $g^{(2)}(\tau)$, and joint temporal intensity (JTI) histograms reveal that mode purity and pulsed squeezing degrade at high gain unless pump detuning is optimally set to pre-compensate nonlinear spectral shifts. Four-fold coincidence correction is required to recover true JTI in the presence of multi-pair emission (Brusaschi et al., 5 Oct 2025).

4. Experimental Realizations, Detection, and Loss Budget

A variety of detection schemes are employed:

• Balanced homodyne detection (for vacuum quadrature noise) (Peace et al., 2022, Terrasson et al., 22 Jan 2026)
• Direct detection (for bright amplitude squeezing in displacement mode) (Terrasson et al., 22 Jan 2026)
• Photon counting with correlation measurements ($g^{(2)}(0)$, mode purity) (Eckstein et al., 2010, Brusaschi et al., 5 Oct 2025)

Losses from coupling, propagation, component efficiency, and detector quantum efficiency (QE) limit observable squeezing. For TFLN strip-loaded devices, total detection efficiency $\eta_\text{tot} \sim 22\%$ (see table below), implying measured squeezing –0.33 dB and inferred on-chip squeezing –1.7 dB (Peace et al., 2022). For PPLN ridge, $\eta_\text{total,exp}\sim0.61$ with dominant limitation from photodiode QE (0.75) (Terrasson et al., 22 Jan 2026).

Loss Channel	$\eta_i$	Loss (dB)
Waveguide in/out coupling	45 %	–7.0
Propagation (WG1, TFLN 4.7 mm)	93 %	–0.29
Dichroic+Free-space optics	66 %	–1.8
Filters/Fiber	50 %	–3.0
Homodyne detectors	98 %	–0.09
Electronic SNR	84 %	–0.75
LO overlap	85 %	–0.70
Total	22 %	–6.6

Key TFLN Strip-loaded Detection Efficiencies (Peace et al., 2022);

Loss correction at the level of $r$ employs $$r_{\text{corr}} = \text{arctanh}[(1-L)\,\tanh r_{\text{meas}}]$$, where $L=1-\eta_{\text{tot}}$ (Terrasson et al., 22 Jan 2026).

5. Performance Metrics, Bandwidths, and Optimization

Key metrics for bright picosecond squeezed light include squeezing level (dB), bandwidth, photon number per pulse, and spectral purity. A comparative table from recent literature is shown below:

Platform	Squeezing (dB)	Bandwidth	Notes
Cavity OPO (bulk PPLN, CW)	–15	10 MHz	Vahlbruch et al. (2016)
PPLN WG (single-pass CW)	–6	2.5 THz	Kashiwazaki et al. (2020)
TFLN ridge (fsec pulses)	–4.2	25 THz	Nehra et al. (2022)
TFLN strip-loaded (ps pulses)	–1.7*	0.23 THz	Measured here (on-chip, single-pass, telecom)
Ridge PPLN (ps pulses)	–15.4$^\dag$	$\lesssim$1 THz	Loss-corrected internal quadrature (Terrasson et al., 22 Jan 2026)

$^\ast$Inferred on-chip; $^\dag$Loss-corrected waveguide value

Optimizing performance requires phase-matching and dispersion engineering, maximizing waveguide escape efficiency in microresonators, robust phase stabilization, and matching the pump temporal profile to the device response. Detuning the pump to pre-compensate SPM/XPM, as in Si$_3$N$_4$ microresonators, sustains spectral purity and squeezing at high pump energies (Brusaschi et al., 5 Oct 2025). High-QE detectors and improved modal overlap further enhance detected squeezing (Terrasson et al., 22 Jan 2026).

6. Applications and Technological Implications

Bright, picosecond squeezed light underpins a range of continuous-variable and quantum-enhanced technologies:

• Quantum information processing: broadband squeezed states for CV quantum optics and integrated photonic quantum processors (Peace et al., 2022, Eckstein et al., 2010).
• Quantum key distribution and teleportation: single-mode EPR pairs at telecom wavelengths support high-rate, long-distance CV-QKD and entanglement swapping (Eckstein et al., 2010).
• Quantum-enhanced microscopy: bright amplitude squeezed light at ps durations (e.g., –3.2 dB measured, –15.4 dB corrected) enables sub-shot-noise nonlinear imaging while keeping sample-averaged power low, mitigating photodamage (Terrasson et al., 22 Jan 2026).
• Quantum metrology and sensing: reduced quadrature fluctuations enhance interferometric sensitivity and facilitate quantum state engineering (heralded Fock states, photon subtraction/addition) (Eckstein et al., 2010).

A plausible implication is that the scalable, integrated implementation of such sources—offered by TFLN, PPLN, and Si$_3$N$_4$ platforms—will catalyze the transition to photonic quantum processors and metrology devices with on-chip squeezing bandwidths $\gg$100 GHz (Peace et al., 2022, Brusaschi et al., 5 Oct 2025).

7. Challenges, Limitations, and Future Directions

Losses from coupling, propagation, and limited photodiode QE are the primary bottlenecks for detected squeezing levels, driving efforts to enhance component efficiencies (Peace et al., 2022, Terrasson et al., 22 Jan 2026). In the high-gain regime, multi-pair emission and nonlinear phase shifts (SPM/XPM) induce mode mixing and degrade purity, requiring adaptive pump detuning, fine temporal control, and higher-order correlation corrections in data analysis (Brusaschi et al., 5 Oct 2025).

A key technical limitation remains the balance between achievable squeezing, brightness, and experimental complexity. Cavity-enhanced OPOs can yield squeezing $>10$ dB but are narrowband and difficult to integrate. Single-pass waveguide approaches offer much broader bandwidth, passive stability, and scalability, but with generally lower observable squeezing per dB photon loss, and strong sensitivity to loss and modal purity.

Future work is focused on improving detector efficiency, further mode-engineering to suppress multimode excitation, minimizing photorefractive and thermal effects, and integrating squeezing sources with on-chip homodyne receivers and CV photonic circuits for large-scale quantum computation and measurement (Peace et al., 2022, Brusaschi et al., 5 Oct 2025, Terrasson et al., 22 Jan 2026).