Third Order Response Functions

• Third order response functions are defined as the cubic terms in the expansion of an observable’s mean value in response to external perturbations, capturing nonlinear effects.
• They are computed using frameworks like Kubo/Matsubara formalism, Liouville-space expansion, and real-time functional derivatives to model quantum and classical dynamics.
• Applications span nonlinear optical phenomena such as third-harmonic generation, nonlinear Hall effects, and coherent current injection in advanced spectroscopic and photonic devices.

A third-order response function is a central theoretical and computational object encoding the leading nonlinear response of a condensed-matter or molecular system to external fields, particularly in nonlinear optics, multidimensional spectroscopy, quantum transport, and nonequilibrium many-body dynamics. These functions quantify the cubic (third-order) terms in the expansion of an observable's expectation value with respect to external perturbations, enabling rigorous interpretation of phenomena such as third-harmonic generation, nonlinear Hall effects, coherent current injection, pump–probe and multidimensional spectroscopies, and nonlinear device functionalities in a variety of quantum materials and molecular aggregates.

1. General Definition and Formal Structure

The third-order response function, denoted generically as $R^{(3)}$, arises in the perturbative expansion of an observable's mean value under a time-dependent field or perturbation: $$\langle Q \rangle_t = \langle Q \rangle_0 + \int dt_1\, R^{(1)}(t-t_1) E(t_1) + \int dt_1dt_2\, R^{(2)}(t-t_1, t-t_2) E(t_1)E(t_2) + \int dt_1dt_2dt_3\, R^{(3)}(t-t_1, t-t_2, t-t_3) E(t_1)E(t_2)E(t_3) + \ldots$$ where $R^{(3)}$ is the third-order response kernel, and $E(t)$ is the applied field. In quantum Liouville space, the retarded third-order response for a general operator $A$ and perturbation operator $V$ is given by a nested commutator: $$R^{(3)}_{A,V,V,V}(\tau_1, \tau_2, \tau_3) = (i\hbar)^{-3} \, \mathrm{Tr} \left[ \rho_0 \left[ \left[ [A(\tau_0), V(\tau_1)], V(\tau_2) \right], V(\tau_3) \right] \right]$$ where $\tau_j = t - t_j$, and $\rho_0$ is the equilibrium density operator (Jain et al., 2024, Jung et al., 2022). Frequency-domain third-order susceptibilities or conductivities, such as $\chi^{(3)}$ or $\sigma^{(3)}$, are obtained by Fourier transform, relating directly to experimentally observed nonlinear signals (Jiang et al., 2017, Cheng et al., 2018).

2. Analytic and Computational Representations

Multiple frameworks exist for calculating and simulating $R^{(3)}$:

a) Kubo/Matsubara Formalism:

Nested commutators as above, or via multi-time correlation functions, either in the time or frequency domain (Jung et al., 2022). Quantum statistical effects (thermal factors, detailed balance) and symmetry can be efficiently included.

b) Pathway Expansion in Liouville Space:

Particularly important in nonlinear spectroscopy, $R^{(3)}$ is split into four fundamental Liouville-pathways corresponding to rephasing/non-rephasing physical processes: ground-state bleaching, stimulated emission, excited-state absorption, and double quantum coherence, with different time- and frequency-dependence. These pathways correspond to distinct Feynman diagrams and field interaction sequences (Begušić et al., 2020, Jang, 1 Jan 2026, Troiani, 2022).

c) Real-Time Functional Derivative Approach:

Nonperturbative extraction of $R^{(3)}$ directly from real-time dynamics or time-dependent simulations, via systematic variation of field amplitudes and functional derivatives with respect to the perturbing fields (Ono, 10 Jul 2025, Krich et al., 17 Apr 2025).

d) Classical and Semiclassical Dynamics:

Approximate $R^{(3)}$ in large or condensed-phase systems by computing symmetrized Kubo-transformed four-time correlation functions via classical molecular dynamics, linearized semiclassical approximations (LSC-IVR), centroid molecular dynamics (CMD), or ring-polymer molecular dynamics (RPMD) (Jung et al., 2022).

e) Diagrammatic and Random Matrix Averages:

In quantum statistical and physical chaos contexts, $R^{(3)}$ can be represented via diagrammatic expansions and averaged over random matrix ensembles, with explicit cluster-counting and large-$N$ scaling (Jain et al., 2024).

3. Physical Processes and Phenomena

The third-order response function characterizes a broad range of nonlinear phenomena:

• Nonlinear Optical Effects:

Third-harmonic generation (THG), four-wave mixing (FWM), sum- and difference-frequency mixing, and nonlinear refractive index modulation, with explicit analytic and gate-tunable expressions derived for systems such as massless Dirac graphene and metallic graphene nanoribbons (Wang et al., 2016, Jiang et al., 2017, Cheng et al., 2018).

• Nonlinear Hall Effects:

Third-order nonlinear Hall effect is sensitive to Berry curvature and band geometric properties; phenomenologically, contributions decompose into Berry-connection polarizability (intrinsic, symmetry-related) and impurity-scattering (extrinsic, Drude-like) terms, as evidenced in type-II Weyl semimetals such as TaIrTe$_4$ (Yang et al., 12 Jun 2025).

• Current-Induced and Divergent Steady-State Processes:

Intraband divergences in $\sigma^{(3)}$ produce giant DC (or near-DC) nonlinear currents—jerk current, coherent current injection (CCI), current-induced second order nonlinearity (CISNL), and degenerate four-wave mixing—governed by double or higher-order poles as one or more summed frequencies vanish (Ventura et al., 2020, Cheng et al., 2018).

• Multidimensional Spectroscopies:

$R^{(3)}$ underpins two-dimensional infrared (2D-IR), 2D sum-frequency generation (2D-SFG), and various forms of two-dimensional electronic spectroscopy (2DES), encoding cross-peaks, diagonal splittings, and pump–probe–probe processes that reveal electronic-vibrational couplings, coherence, and dynamics (Begušić et al., 2020, Jung et al., 2022, Troiani, 2022).

• Nonlinear Device Applications:

Enhanced $R^{(3)}$ in low-dimensional materials enables efficient THz frequency conversion, ultrafast photonic switching, and electrical gating of optical nonlinearities, with device structures tailored to exploit strong field-induced and Fermi level–dependent responses (Wang et al., 2016, Jiang et al., 2017).

4. Model Systems and Key Mathematical Results

Analytic and numerical models yield explicit closed-form results for $R^{(3)}$ in several contexts:

System/Model	Key Result Type	Notable Features or Expressions
2D graphene, Dirac	Kubo formulas	Logarithmic and step-like $μ$-dependence in $\chi^{(3)}$ for THG, FWM; see Eq. (1) (Jiang et al., 2017)
acGNR (armchair ribbons)	Closed-form Kerr and 3rd harmonic conductances	$g^{(3)} \propto N$ factors, sharp Fermi-level thresholds, $10^3\times$ enhancement vs. 2D (Wang et al., 2016)
Gapped graphene (massive Dirac)	Analytic $\sigma^{(3)}$ with three intraband divergences	Poles at $\omega\rightarrow0$, $\omega_2+\omega_3\to0$, $\omega_1+\omega_2+\omega_3\to0$; regularization by $\Gamma_a$ (Cheng et al., 2018)
Exponential trap model	Frequency-dependent $\chi_3(\omega)$	Peak/hump or monotonic decay in $|\chi_3|$, variable-dependent divergence at glass transition (Diezemann, 2014)
Random matrix systems	Ensemble-averaged $R^{(3)}$	Power law tails, universal dips in spectral density around $\omega=0$; explicit Bessel function expressions (Jain et al., 2024)

The explicit structure of $R^{(3)}$ always reflects the spectrum, relaxation mechanisms, and selection rules of the underlying Hamiltonian, with divergences cut off only by finite scattering rates.

5. Extraction and Measurement Strategies

Experimental isolation of the third-order response function typically leverages field intensity scaling. By measuring the observable response at a minimum of three distinct pump intensities, then expanding the signal in powers of intensity (usually via a Vandermonde matrix inversion), one isolates the pure third-order ($S_{(3)}$) response and reduces contamination from higher orders (Krich et al., 17 Apr 2025). This protocol is widely employed in pump–probe, TA, and 2D spectroscopies, with the intensity points optimized to minimize both random noise and systematic higher-order contamination.

First-principles evaluations of $R^{(3)}$ in molecules and solids use either direct perturbative expansion (time-dependent density matrix or Liouville-space methods (Begušić et al., 2020, Jang, 1 Jan 2026)), semiclassical single-trajectory thawed Gaussian approaches (exact for harmonic but accurate for moderately anharmonic systems (Begušić et al., 2020)), or quantum-classical path-integral methods leveraging Kubo transforms (Jung et al., 2022).

Nonlinear current extraction in quantum materials relies on combined DC and AC field drives, with protocol-specific frequency limits probing distinct types of divergences (jerk, dichromatic, trichromatic probes (Ventura et al., 2020, Cheng et al., 2018)).

6. Fundamental Limits, Scaling, and Open Issues

Enhancement and tunability:

Nanoscale and low-dimensional systems (e.g., acGNR, massless Dirac graphene, Weyl semimetals) exhibit orders-of-magnitude enhancements in $R^{(3)}$ due to band structure engineering, Fermi-level tuning (electrical gating), and dimensional quantization (Wang et al., 2016, Jiang et al., 2017, Yang et al., 12 Jun 2025). Gate-tuning, impurity scattering, and external DC fields provide real-time electrical control over the magnitude and sign of the nonlinear response.

Universality and Divergence:

In high-quality crystals and cold semiconductors, the limiting divergence of $R^{(3)}$ as one or more field frequencies sum to zero emerges generically from intraband acceleration processes combined with interband coherence (Berry connection), with relaxation rates as the only cutoff (Ventura et al., 2020, Cheng et al., 2018).

Interpretational caveats:

In glassy or disordered models (e.g., exponential trap model), the presence or absence of a peak (hump) in the frequency-dependent $|\chi_3|$ is not a reliable indicator of spatial cooperativity or a growing length scale, but arises from mean-field kinetic/energetic effects and the choice of probe variable (Diezemann, 2014).

Methodological generality:

Real-time functional-derivative frameworks allow extraction of $R^{(3)}$ in both noninteracting and interacting many-body systems (e.g., via tensor network or time-evolution block decimation), with only moderate computational scaling at third order (Ono, 10 Jul 2025).

7. Applications and Broader Significance

Third-order response functions underpin the operational principles of advanced THz modulators, frequency converters, low-threshold nonlinear devices, and ultrafast measurement protocols in both fundamental research and applied quantum photonics (Wang et al., 2016, Jiang et al., 2017, Krich et al., 17 Apr 2025, Kumar et al., 2023). Their direct connection to band geometry, relaxation dynamics, and quantum coherence makes them a powerful probe of both material properties and device figures of merit. Future directions include higher-order generalizations, nonequilibrium dynamics under strong driving, and integration with ab initio many-body theory for material discovery and device optimization.