高斯光束经非线性晶体的变化及泵浦过程中腰斑与发散角等技术问询

Great question—this mixes core Gaussian beam propagation with nonlinear optics specifics, so let’s break it down clearly.

1. Core Changes When Gaussian Beams Pass Through Nonlinear Crystals #

Whether we’re talking about second-harmonic generation (SHG) or spontaneous parametric down-conversion (SPDC), the output light retains Gaussian spatial modes (thanks to the coherent, spatially uniform nature of nonlinear coupling in most crystals), but key properties shift based on the interaction:

• SHG: A fundamental Gaussian beam (wavelength λ₁) generates a second-harmonic beam (λ₂ = λ₁/2) when phase-matched. The output beam stays Gaussian, but its power is concentrated where the pump’s intensity is highest—right at the pump’s waist, since Gaussian beams peak there. Conversion efficiency depends on the pump’s peak intensity, crystal length, and phase matching quality.
• SPDC: A pump Gaussian beam (λₚ) produces entangled signal (λₛ) and idler (λᵢ) beams, satisfying 1/λₚ = 1/λₛ + 1/λᵢ. Both output beams are Gaussian, and their spatial modes are correlated (e.g., their waist positions and sizes track each other) due to momentum conservation (kₚ = kₛ + kᵢ) in the nonlinear process.

2. Waist Size and Divergence: Intuition vs. Reality #

Your gut feeling that the waist stays the same is partially true only in very thin crystals—but in general, both waist size and divergence change, driven by diffraction limits and nonlinear phase matching. Let’s break this down:

First, recall the universal diffraction-limited relationship for Gaussian beams:

w₀ * θ = λ / π

where w₀ is the waist radius, θ is the far-field divergence angle, and λ is the beam wavelength. This rule holds for any Gaussian beam, linear or nonlinear—so if the wavelength changes, either w₀, θ, or both must adjust to satisfy this equation.

• SHG Scenario:
  The second-harmonic wavelength is half the pump’s. If the crystal is extremely thin (much shorter than the pump’s Rayleigh length), the harmonic beam’s waist will be nearly identical to the pump’s (w₀₂ ≈ w₀₁). But since λ₂ = λ₁/2, the divergence angle will drop by half: θ₂ = λ₂/(πw₀₂) ≈ θ₁/2. For longer crystals, phase matching effects can tweak the waist size slightly, but the diffraction limit relationship still holds.
• SPDC Scenario:
  Signal and idler wavelengths are longer than the pump’s (e.g., λₛ = λᵢ = 2λₚ in degenerate SPDC). Again, for thin crystals, the signal waist is close to the pump’s (w₀ₛ ≈ w₀ₚ), but since λₛ is twice λₚ, the divergence angle doubles: θₛ = λₛ/(πw₀ₛ) ≈ 2θₚ. For non-degenerate SPDC, the waist and divergence of signal/idler will differ from each other but still follow the diffraction limit rule based on their individual wavelengths.

In short: Waist size doesn’t stay fixed unless the wavelength stays the same—and since SHG/SPDC always change the wavelength, divergence must adjust (or waist size, if phase matching imposes constraints) to maintain the diffraction-limited relationship.

3. General Methods for Analyzing Nonlinear Gaussian Beams, and ABCD Matrix Applicability #

General Approach #

To model Gaussian beams in nonlinear processes, follow this framework:

• Start with Gaussian beam formalism: Represent all beams (pump, output) using their complex amplitude form, which includes waist size, curvature radius, and Gaussian phase terms as functions of propagation distance.
• Solve nonlinear coupling equations: For SHG, this means solving the coupled wave equation dE₂/dz ∝ E₁² exp(iΔkz) (where Δk is phase mismatch). For SPDC, you’ll work with equations describing the generation of entangled signal/idler pairs from the pump.
• Combine with diffraction effects: If the crystal length is comparable to the beam’s Rayleigh length, you can’t ignore diffraction while the nonlinear interaction happens. Use beam propagation method (BPM) for numerical simulations, or analytic approximations when phase matching is perfect (to simplify the coupling-diffraction interplay).

ABCD Matrix Modifications #

Standard ABCD matrices are designed for linear optical systems—they don’t account for nonlinear power transfer or mode coupling. But they can still be useful with caveats:

• Weak nonlinearity (low conversion efficiency): If less than ~10% of the pump is converted (common in many lab setups), the pump’s beam parameters barely change. You can treat the output beam (harmonic, signal/idler) as a new Gaussian beam generated at the pump’s waist, then use ABCD matrices to calculate its propagation after leaving the crystal (since free-space/linear optics propagation follows ABCD rules).
• Strong nonlinearity: When the pump’s power or beam parameters change significantly, ABCD matrices won’t work. You’ll need to combine coupled mode theory with Gaussian beam parameter evolution to model the interaction fully.

内容的提问来源于stack exchange，提问作者Aks