Effective Second-Harmonic Generation Coefficient and C-eigenvalue of Nonlinear Susceptibility Tensors

1. Problem Statement

The paper addresses the challenge of accurately calculating the effective second-harmonic generation (SHG) coefficient ($d_{eff}$) in uniaxial crystals.

• Context: SHG is a fundamental second-order nonlinear optical process where fundamental frequency light is converted to its second harmonic. The efficiency of this conversion depends on the effective SHG coefficient.
• Complexity: The calculation of $d_{eff}$ involves optimizing a trilinear form of the nonlinear susceptibility tensor ($\chi^{(2)}$) subject to geometric constraints imposed by the phase-matching conditions (specifically, the relationship between the ordinary 'o' and extraordinary 'e' light polarization vectors).
• Gap: While the C-eigenvalue of piezoelectric-type tensors (a mathematical concept introduced by Chen, Jákli, and Qi) is known to represent the maximum electromechanical coupling, its direct relationship to the effective SHG coefficient in the specific context of uniaxial crystal phase matching had not been fully explored or simplified for computational efficiency.

2. Methodology

The authors employ a combination of tensor analysis, constrained optimization, and symmetry arguments to derive simplified mathematical models.

• Tensor Representation: The nonlinear susceptibility tensor $\chi^{(2)}$ is treated as a third-order piezoelectric-type tensor (symmetric in its last two indices).
• Geometric Constraints:
  • The problem is set in a uniaxial crystal coordinate system where light propagation is along the $x_3$-axis.
  • The polarization vectors for ordinary light ($\mathbf{a}$) and extraordinary light ($\mathbf{b}$) are defined by angles $\phi$ and $\theta$.
  • These vectors satisfy orthogonality ($\mathbf{a}^T\mathbf{b}=0$) and unit norm constraints.
• Optimization Reformulation:
  • The authors formulate the calculation of the effective SHG coefficient as a constrained maximization problem: $\max \{ \mathbf{T}\mathbf{b}\mathbf{a}\mathbf{a} \}$ or $\max \{ \mathbf{T}\mathbf{a}\mathbf{b}\mathbf{b} \}$.
  • Variable Reduction: By exploiting the specific geometric structure of uniaxial crystals (where $\mathbf{a}$ lies in the $x_1-x_2$ plane and $\mathbf{b}$ is derived from $\mathbf{a}$ and the propagation angle), the authors reduce the number of independent variables in the optimization models from 4 (or more) down to 2.
  • Analytical Derivation: Using Lagrangian multipliers and properties of quadratic forms, they derive closed-form expressions for the maximum values.
• Comparison with C-eigenvalues: The study compares the derived maximum effective SHG coefficients with the largest C-eigenvalue ($\lambda_{max}$) of the susceptibility tensor, which is defined as the global maximum of the tensor contraction over the entire unit sphere without the specific phase-matching constraints.

3. Key Contributions

A. Dimensionality Reduction of Optimization Models

The paper proves that for uniaxial crystals, the complex optimization problem for the effective SHG coefficient can be reduced to a 2-variable problem.

• Type-I Phase Matching ($d_{eff}^1$): The maximum is reduced to finding the maximum of a function involving the square root of the sum of squares of quadratic forms:
  $$d_{eff}^1 = \max_{a_1^2+a_2^2=1} \sqrt{(a_2Q_1(\mathbf{a}) - a_1Q_2(\mathbf{a}))^2 + (Q_3(\mathbf{a}))^2}$$
• Type-II Phase Matching ($d_{eff}^2$): The maximum is reduced to the largest eigenvalue of a $2 \times 2$ matrix constructed from the tensor components:
  $$d_{eff}^2 = \frac{1}{2} \max_{a_1^2+a_2^2=1} \left( R_{11}(\mathbf{a}) + R_{22}(\mathbf{a}) + \sqrt{(R_{11}(\mathbf{a}) - R_{22}(\mathbf{a}))^2 + 4R_{12}(\mathbf{a})^2} \right)$$
  This significantly simplifies numerical computation.

B. Theoretical Inequality (C-eigenvalue vs. Effective SHG)

The authors establish a rigorous inequality relationship:
$$d_{eff} \leq \lambda_{max}(\mathbf{T})$$

• Interpretation: The effective SHG coefficient (constrained by phase-matching geometry) is always less than or equal to the largest C-eigenvalue (the global unconstrained maximum of the tensor).
• Equality Condition: The equality holds only for specific crystal classes where the optimal phase-matching direction coincides with the optimal C-eigenvector direction. In many cases, $d_{eff} < \lambda_{max}$.

C. Verification via Crystal Examples

The theoretical framework is validated through analytical calculations for five specific crystal classes:

1. $\bar{4}2m$ (e.g., KH$_2$PO$_4$): $d_{eff} < \lambda_{max}$.
2. $4mm$ (e.g., LiNbO$_3$): $d_{eff}^1 = |\chi_{15}|$, $d_{eff}^2 = 0$, while $\lambda_{max} = \max(|\chi_{15}|, |\chi_{33}|)$.
3. $4$ (e.g., Urea): $d_{eff} = \lambda_{max}$.
4. $62m$ (e.g., Benitoite): $d_{eff} = \lambda_{max}$.
5. $6$(e.g.,$\alpha$-LiIO$_3$): $d_{eff}^1 = \lambda_{max}$, but $d_{eff}^2 \leq \lambda_{max}$.

4. Results

• Computational Efficiency: The proposed method transforms a high-dimensional constrained optimization problem into a manageable 2-variable maximization, making it easier to compute effective coefficients for design purposes.
• Physical Insight: The study clarifies that while the C-eigenvalue represents the intrinsic maximum nonlinear capability of the material, the effective coefficient achievable in a specific optical setup (uniaxial phase matching) is often lower due to geometric constraints.
• Crystal-Specific Behavior: The examples demonstrate that the relationship between $d_{eff}$ and $\lambda_{max}$ is crystal-class dependent. For some crystals (like Urea and Benitoite), the phase-matching constraints do not limit the maximum achievable efficiency relative to the tensor's intrinsic potential, whereas for others (like KDP), there is a distinct gap.

5. Significance

• Crystal Design: Provides a new theoretical tool for designing and optimizing nonlinear optical crystals. Researchers can now quickly estimate the upper limits of SHG efficiency for specific crystal orientations without resorting to heavy numerical simulations.
• Theoretical Unification: Bridges the gap between tensor analysis (C-eigenvalues) and practical nonlinear optics (phase matching), offering a unified mathematical framework.
• Algorithm Development: The reduction to 2 variables suggests that efficient algorithms can be developed for real-time optimization of device parameters (like crystal cut angles) in laser systems.
• Educational Value: Offers clear analytical expressions for common crystal classes, aiding in the teaching and understanding of nonlinear optical properties.

In summary, this paper successfully simplifies the calculation of effective SHG coefficients in uniaxial crystals by reducing variable dimensions and establishes a critical theoretical bound comparing these coefficients to the tensor's largest C-eigenvalue, verified across multiple crystal symmetries.