---
title: Directional Harmonic Modes Overview
url: https://www.emergentmind.com/topics/directional-harmonic-modes
type: topic
---

# Directional Harmonic Modes Overview

Searching arXiv for the cited papers to ground the article and confirm metadata.
Directional harmonic modes are harmonic responses whose structure depends explicitly on direction, orientation, or directional coupling variables rather than only on frequency or global harmonic order. In the literature, the phrase appears in several domain-specific senses: as steerable second-harmonic emission from sub-wavelength nonlinear antennas, as topologically or supermode-selected harmonic channels in coupled photonic structures, as localized orientation-resolved spherical-harmonic content on \(\mathbb S^2\), as mixed azimuthal-flow sectors in heavy-ion collisions, as directional spectra generated by multi-hinged wavemakers, and as convexity in a prescribed direction for planar harmonic mappings [1806.04869], [1207.5558], [1711.07773], [1703.03593]. Taken together, these uses indicate that “directional” modifies not the existence of harmonics themselves, but the geometry of their generation, representation, or coupling.

## 1. Conceptual scope

Across the cited work, a directional harmonic mode is not a single standardized object. It is instead a family of constructions in which harmonic content is resolved or controlled by an additional directional variable: a Fourier-plane angle, a propagation branch, an orientation parameter on \(\mathrm{SO}(3)\), a wave-vector sector, a flow-sector coupling, or a geometric direction of convexity. This suggests a unifying description in which harmonic order is supplemented by a directional degree of freedom.

| Domain | Harmonic object | Directional variable |
|---|---|---|
| Nonlinear optics | SHG or higher harmonics | Fourier-plane angle, facet orientation, supermode parity, valley branch |
| Harmonic analysis | Spherical-harmonic or lattice-direction content | Window orientation \(\gamma\), direction set \(V\) |
| Heavy-ion flow | Mixed \(v_n\) sectors | \(p_T\) or \(\eta\)-resolved mode mixing |
| Wavemaking | Free, bound, and spurious harmonics | Horizontal wavevector direction \(k_y\) |
| Harmonic mappings | Harmonic shears | Convexity in direction \(\gamma\) |

Two recurring technical themes appear repeatedly. First, directional harmonic modes are usually defined relative to a basis change or structured decomposition: phased dipoles on crystal facets, valley-Hall kink modes, directional SLSHT coefficients \(g(\rho;\ell,m)\), generalized mixed-harmonic correlation matrices, or analytic shears \(h-e^{2i\gamma}g\). Second, the directional variable is operational rather than decorative: changing excitation position, Euler angle, interferometric phase, paddle profile, or mixing coefficient changes the measured harmonic response [1806.04869], [1207.5558], [2502.09586], [1807.11658].

## 2. Nonlinear-optical realizations: facets, topology, and supermodes

In nonlinear optics, directional harmonic modes are most explicitly realized as angle-selective second-harmonic generation. A representative case is the single-crystalline organic mesowire study of crystalline diaminoanthraquinone (DAAQ), where the forward-scattered SHG radiation pattern is tuned by shifting the excitation spot across a faceted sub-wavelength cross-section. The measured SHG maximum moves from about \(+63^\circ\) to about \(-67.8^\circ\), corresponding to a tuning of about \(130.8^\circ\), while the real-plane SHG remains localized at the pump position. The paper attributes this to facet-controlled localization of the pump near field, selective excitation of spatially distributed nonlinear dipoles, and coherent interference of their radiation in the far field. In the same system, two-photon excited fluorescence is isotropic in angle and delocalized in space, propagating along the mesowire and outcoupling from distal ends; the contrast isolates directional harmonic emission as a coherent nonlinear-scattering effect rather than a generic geometric-outcoupling effect [1806.04869].

The DAAQ work also makes the angular representation explicit through back-focal-plane imaging. The measured pattern is parameterized by \((\theta,\phi)\) with
\[
k_x/k_0=\sin\theta\cos\phi,\qquad k_y/k_0=\sin\theta\sin\phi,
\]
and with the \(\mathrm{NA}=0.95\) collection objective the maximum captured half-angle is \(71.8^\circ\). The authors model the radiation using a phased-dipole picture, writing
\[
P_{2\omega}(\theta,\varphi)=\frac12\,n_{2\omega}\varepsilon_0 c r^2 \left|E_{2\omega}(\theta,\varphi)\right|^2,
\qquad
E_{2\omega}(\theta,\varphi)=E_{2\omega}^{(0)}(\theta,\varphi)\cdot N\cdot A(\theta,\varphi),
\]
and support the interpretation with \(3\mathrm{D}\) finite-difference time-domain simulations showing that the pump field localizes differently on the upper two facets as the beam is scanned laterally [1806.04869].

A different mechanism appears in topological photonics. In all-dielectric honeycomb photonic crystals, two valley bandgaps are engineered, one around \(\omega_0\) and another around \(2\omega_0\), so that a mirror-symmetric domain wall hosts double valley-Hall kink modes. The second-harmonic field is then generated by \(\chi^{(2)}\)-coupling between topological interface modes, with the relevant mismatch written as
\[
\Delta \bar{k}=(2\bar{k}_f-2)-\bar{k}_s^-.
\]
Because the lower-gap and upper-gap kink branches can have opposite slopes at the same valley, the generated SHG is in general bi-directional, with forward and backward components \(\bar{k}_s^-\) and \(\bar{k}_s^+\). The directional asymmetry is quantified by
\[
\eta_{\rightleftarrows}=\frac{\eta_{\rightarrow}}{\eta_{\leftarrow}},
\qquad
\mathrm{SHG\mbox{-}DD} = \frac{\eta_{\rightarrow}-\eta_{\leftarrow}}{\eta_{\rightarrow}+\eta_{\leftarrow}},
\]
and the reported SHG directional dichroism reaches \(0.97\) at \(\bar{\omega}=0.52\) [2007.04875]. A common misconception is that topological directionality here is strictly one-way; the paper explicitly states the opposite, since time-reversal symmetry leaves harmonic interface states available in both directions.

A third optical realization uses coupled-waveguide supermodes rather than facet dipoles or valley branches. In an integrated nonlinear interferometer composed of an input linear directional coupler, a central nonlinear directional coupler, and an output linear directional coupler, the pump is prepared in the even or odd fundamental supermode,
\[
E_f = \frac{A_f + B_f}{\sqrt{2}} e^{-i C z},
\qquad
O_f = \frac{B_f - A_f}{\sqrt{2}} e^{i C z}.
\]
This supermode choice shifts the SHG phase-matching condition to
\[
\Delta\beta(\lambda_e)-2C=0
\quad\text{or}\quad
\Delta\beta(\lambda_o)+2C=0.
\]
Experimentally, the even-supermode branch appears at \(\lambda_e = 1.5592\,\mu\text{m}\), the single-waveguide branch at \(\lambda_{\text{single}} = 1.5601\,\mu\text{m}\), and the odd-supermode branch at \(\lambda_o = 1.5610\,\mu\text{m}\). This is a directional harmonic mode in modal rather than angular space: the directional label is the coupled-system parity of the pump supermode, and it selects a distinct SHG spectral response [2106.13204].

## 3. Structured electromagnetic harmonic fields beyond conventional SHG

Directional harmonic modes need not be far-field steering phenomena. In isotropic nonlinear media below the threshold for multiphoton ionization, odd harmonics generated by vector polarization beams can form crown-structured optical harmonics: a ring of narrow filaments whose number and sharpness are determined by the pump polarization topology and harmonic order. The local harmonic source law is written as
\[
I^{m}_{HG}(r,\phi) \sim \left(\frac{I(r,\phi)}{I_0}\right)^m \sin^{\,m-1}\{2\chi(r,\phi)\},
\]
and, for suitable vector beams, the angular factor reduces to a \(\cos(l\phi)\)-type dependence that yields \(2l\) equally spaced intensity filaments. For \(l=6\), the beam forms \(12\) spikes; for \(m=31\), the azimuthal FWHM spatial width of each filament is \(0.10\,c\), corresponding to \(1.6\%\) of the ring circumference, and the radial FWHM is \(0.16w\) [1906.04978]. Here directionality is azimuthal segmentation of the transverse mode rather than emission into different longitudinal channels.

Space-time modulated dispersive media provide another directional construction. In a Lorentz medium with traveling modulation, properly engineered material dispersion allows two codirectional Floquet harmonics to couple and form new eigenstates,
\[
E_A = a_0 e^{j(\omega_0 t-\beta_A z)} + a_1 e^{j[(\omega_0+N\omega_m)t-(\beta_A+N\beta_m)z]},
\]
\[
E_S = a_0 e^{j(\omega_0 t-\beta_S z)} - a_1 e^{j[(\omega_0+N\omega_m)t-(\beta_S+N\beta_m)z]},
\]
in the subluminal regime, and
\[
E_G,\ E_D,\qquad \beta_G=\beta+j\alpha,\ \beta_D=\beta-j\alpha,
\]
in the superluminal regime. The crucial point is that these are codirectional harmonic pairs rather than the usual forward/backward Bragg-type states. For \(v_m<v_b\), the result is periodic energy exchange between \(\omega_0\) and \(\omega_0+N\omega_m\); for \(v_m>v_b\), both frequencies are amplified together [1710.01625].

Parabolic GRIN fibers introduce a further shift in meaning. Their weak-guidance Maxwell model yields guided fields built from a \(2\mathrm{D}\) harmonic-oscillator basis, and coherent-state superpositions produce “harmonic motion modes.” Type I accelerating modes have a centroid that traces a circular trajectory, with
\[
\tilde{\delta}(z)=\delta\, e^{i\beta_1(0,0)z},
\qquad
Z_{\mathrm{I}}=\frac{2\pi}{\beta_1(0,0)},
\]
while Type II breathing modes have a propagation-dependent width
\[
w^2(\delta,z) = \cosh|\delta| - \cos\{\arg[\tilde\delta(z)]\}\sinh|\delta|,
\qquad
Z_{\mathrm{II}}=\frac{\pi}{\beta_1(0,0)}.
\]
These are not harmonic-generation modes, but the papers explicitly frame them as fields with selectable handedness, vortex sign, and harmonic transverse dynamics [2312.11734]. This broadens the term from frequency conversion to controlled harmonic motion in structured electromagnetic media.

## 4. Directional harmonic analysis on the sphere and on \(\mathbb Z^2\)

In signal analysis on the sphere, directional harmonic modes are formalized by the directional spatially localized spherical harmonic transform. For \(f\in L^2(\mathbb S^2)\) and an asymmetric window \(h\), the transform is
\[
g(\rho;\ell,m) \triangleq \int_{\mathbb{S}^2} f(\hat{\boldsymbol{x}})\, \big(\mathcal{D}_\rho h\big)(\hat{\boldsymbol{x}})\, \overline{Y_\ell^m(\hat{\boldsymbol{x}})}\,ds(\hat{\boldsymbol{x}}),
\]
where \(\rho=(\alpha,\beta,\gamma)\in\mathrm{SO}(3)\). The additional angle \(\gamma\) is the orientation variable missing from the earlier symmetric-window SLSHT. With band-limits \(L_f\) and \(L_h\), the output is nonzero up to \(L_g=L_f+L_h\), and inversion is exact provided \((h)_0^0\neq 0\):
\[
(f)_\ell^m = \frac{\hat{f}_{\ell m}}{\sqrt{16\pi^3}\,(h)_0^0},
\qquad
\hat{f}_{\ell m} = \int_{\mathrm{SO}(3)} g(\rho;\ell,m)\,d\rho.
\]
The paper interprets \(g(\rho;\ell,m)\) as a family of localized directional harmonic mode amplitudes, jointly resolved in spectral index \((\ell,m)\), spatial position \((\alpha,\beta)\), and orientation \(\gamma\). Computationally, the direct quadrature cost \(O(L_f^4L_h^3)\) is reduced to \(O(L_f^3L_h^2 + L_f^2L_h^4)\) by factoring rotations and using a \(3\mathrm{D}\) FFT [1207.5558].

On the lattice \(\mathbb Z^2\), directional harmonic analysis takes the form of directional maximal averaging over finite direction sets \(V\). The linear operator is
\[
A_{V,k}f(x):=\sup_{v\in V}\left|\sum_{n\in \mathbb{Z}} f(x-nv)\,\phi_k(n)\right|,
\]
with frequency localization near arithmetic resonant sets \(v\cdot\beta\approx m\). The authors construct direction sets \(V=V_{N,\varepsilon}\subset \mathbb Z^2\) for which
\[
\|A_{V,k}\|_{\ell^2(\mathbb{Z}^2)\to \ell^2(\mathbb{Z}^2)} \le C_\varepsilon N^\varepsilon
\]
whenever \(2^k\ge N^{C_1}\), and analogous bounds for polynomial orbits \(P(n)v\) when \(k\ge C_d\log N\). The mechanism is an arithmetic Kakeya-type incidence estimate controlling overlaps of tube-combs
\[
K^A_{r,s,v} := \left\{ \beta\in\mathbb{T}^2:\ |\beta|\gtrsim A^{-2},\ |v\cdot \beta-m-b/r|\lesssim 2^{-C_1 s} \right\},
\]
so directional harmonic content is determined jointly by geometry and prime-factor arithmetic rather than only by angular separation [1901.06070].

## 5. Nonlinear harmonic coupling in heavy-ion flow and wavemaking

In relativistic heavy-ion collisions, directional harmonic modes refer to azimuthal-flow sectors and their nonlinear couplings. The paper generalizes the usual harmonic correlation matrix to mixed same-total-order sectors such as \(v_2^2\leftrightarrow v_4\), \(v_2v_3\leftrightarrow v_5\), and \(v_2^3\), \(v_3^2\), \(v_6\). The generic mixed correlator is
\[
v\{n_1 \dots n_l ,- m_1 \dots -m_k\}(p_1,p_2)
=
\left\langle Q_{n_1}(p_1)\dots Q_{n_l}(p_1)\,
Q_{m_1}^\star(p_2)\dots Q_{m_k}^\star(p_2) \right\rangle,
\]
with matched total harmonic order. Principal component analysis of these generalized matrices separates dominant mixed modes from factorization breaking. The central conclusion is that for a generalized correlation matrix built from \(n\) harmonic sectors, the first \(n\) eigenvectors mostly encode mode mixing, while higher eigenvectors encode factorization breaking [1711.07773]. In this usage, “directional” means the anisotropic emission pattern carried by each \(v_n\) sector and its momentum-dependent admixtures.

Directional wavemakers introduce a hydrodynamic version of the same theme. In second-order potential-flow theory for multi-hinged directional wavemakers, the first-order paddle motion is a superposition of hinge contributions,
\[
X^{(1)}(y,z,t) = \frac12 \sum_{i=1}^N \sum_n X_{in}^{(1)} \frac{\max(d_i+z,0)}{d_i} e^{i(\omega_n t-k_{y,n}y)},
\]
and second-order interactions generate bound waves, spurious free waves, and correction waves. The decomposition
\[
\phi^{(2)}=\phi^{(21)}+\phi^{(22)}+\phi^{(23)}
\]
separates the bound second-order wave, the spurious free wave, and the second-order correction wave. A central result is that a double-hinged wavemaker can suppress the second-order spurious progressive wave using only single-harmonic hinge motion, without explicit double-harmonic paddle motion; the optimized flap motions are almost always in opposite phase, and the larger draft is often below the water surface [2502.09586]. The paper also shows that eliminating linear evanescent modes with a flexible exponential profile does not eliminate second-order spurious free waves, because nonlinear boundary matching remains necessary.

## 6. Geometric function theory: directional convexity as a harmonic notion

In complex analysis, directional harmonic modes are not spectral or radiative objects. They are harmonic mappings whose image domains are convex in a prescribed direction. For a harmonic map \(f=h+\overline g\), the Clunie–Sheil-Small shear construction states that \(f\) is univalent and convex in direction \(\gamma\) if and only if
\[
h-e^{2i\gamma}g
\]
is analytic, univalent, and convex in direction \(\gamma\). This converts a harmonic geometric problem into an analytic directional-convexity problem [1703.03593].

One line of work studies convolutions of harmonic shears built from
\[
\phi(z)=\int_0^z \frac{d\xi}{1-2\xi e^{i\mu}\cos\nu+\xi^2 e^{2i\mu}},
\]
and shows that if \(f=h+\overline g\) and \(F=H+\overline G\) satisfy
\[
f_{\mu_1}*F_{\mu_2}=\phi,\qquad f_\alpha:=h+e^{-2i\alpha}g,
\]
then \(f*F\) is univalent and convex in the direction \(-(\mu_1+\mu_2)\), provided it is locally univalent and sense-preserving. For specific dilatations \(\omega(z)=az^n\), local univalence is established when \(n=1,2\) and \(|a|\le 1\), or \(n\ge 3\) and
\[
|a|\le n-1-\sqrt{n^2-2n}.
\]
The direction parameter is carried by the phase \(e^{-2i\alpha}\) in the analytic shear and by the orientation \(\mu\) of the seed function [1703.03593].

A related paper studies combinations
\[
\hat f=\eta f_1+(1-\eta)f_2
\quad\text{and}\quad
\tilde f=\eta h_1+(1-\eta)h_2+\overline{\eta g_1+(1-\eta)g_2}
\]
for harmonic right half-plane and vertical strip mappings. Under explicit hypotheses involving
\[
\psi_{\mu,\nu}(z)=\frac{1}{1-2ze^{-i\mu}\cos\nu+z^2e^{-2i\mu}},
\]
it proves that these combinations are univalent and convex in direction \(\mu\) when the dilatations satisfy \(|\omega_k(z)|<\alpha_k<1\) and
\[
|\eta|\le \frac{(1-\alpha_1)(1-\alpha_2)}{2(\alpha_1+\alpha_2)}.
\]
Here the directional harmonic mode is a directionally convex shear geometry, preserved under controlled mixing of analytic and co-analytic parts [1807.11658].

## 7. Cross-domain mechanisms and interpretive cautions

Several recurrent mechanisms organize the literature. Coherence is decisive in optical settings: the DAAQ mesowire work distinguishes directional SHG from isotropic TPEF precisely through coherent addition of nonlinear dipole fields [1806.04869], while the supermode-based interferometer distinguishes even and odd pump branches through modal phase matching rather than simple single-waveguide conversion [2106.13204]. Basis choice is equally decisive in analysis: directional SLSHT replaces global \(Y_\ell^m\) coefficients by \(g(\rho;\ell,m)\), and heavy-ion PCA replaces isolated \(v_n\) observables by mixed-sector eigenmodes [1207.5558], [1711.07773]. Additional degrees of freedom also recur: sub-wavelength facets, valley branches, interferometric phase, asymmetric windows, multiple hinges, and complex shear coefficients each enlarge the control space.

Several misconceptions are explicitly corrected by the cited work. Directionality does not always mean beam steering in real space; in spherical analysis it means orientation-dependent localized spectral content, and in harmonic mapping theory it means convexity in a prescribed direction. Directionality does not imply nonreciprocity; valley-Hall SHG remains bi-directional because the system is time-reversal symmetric [2007.04875]. Suppressing linear near-field artifacts does not guarantee nonlinear cleanliness; the exponential flexible wavemaker removes linear evanescent modes but still produces spurious second-order waves of magnitude comparable to a single flap [2502.09586]. Nor is every multi-peak harmonic spectrum evidence of modal phase matching; the nonlinear interferometer paper distinguishes genuine even/odd supermode SHG from the three-peak response obtained by single-waveguide excitation [2106.13204].

Taken together, the literature supports a broad but technically precise interpretation: directional harmonic modes are harmonic degrees of freedom whose amplitude, localization, propagation, or geometric admissibility is resolved by direction-dependent structure. The specific mathematics changes—from Fourier-plane polar angles, to \(\mathrm{SO}(3)\) rotations, to mixed-harmonic correlation matrices, to analytic shears—but the common principle is the same: harmonic behavior becomes directionally structured when symmetry, geometry, or coupling introduces a nontrivial directional selector.

Source: https://www.emergentmind.com/topics/directional-harmonic-modes