Spin-Exchange Dynamical Structure Factor

• Spin-Exchange DSF is the momentum- and energy-resolved response of two-spin bond operators that reveal singlet excitations in quantum magnets.
• It distinguishes itself from conventional DSFs by probing bond energy fluctuations with interference factors, such as the cos²(q/2) modulation that suppresses intensity at q=π.
• Studies demonstrate that in the S=½ Heisenberg chain the DSF is dominated by two-spinon states, while adjacent bond correlators and higher-dimensional systems expose additional multi-spinon dynamics.

Searching arXiv for fresh, relevant papers on spin-exchange dynamical structure factors and related DSF context. The spin-exchange dynamical structure factor is the momentum- and energy-resolved response associated with bond-exchange operators rather than on-site spin operators. In one common form, relevant to indirect K-edge resonant inelastic x-ray scattering (RIXS) on the $S=\tfrac12$ Heisenberg chain, the effective scattering operator is

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$

and the corresponding response is

$$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$

A more general bond-energy form is

$$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$

This object probes fluctuations of the bond exchange energy, is central to indirect RIXS, and is distinct from the conventional single-spin dynamical structure factor measured in inelastic neutron scattering (INS) (Klauser et al., 2011, Ferrari et al., 2019).

1. Definition and operator content

The conventional single-spin dynamical structure factor is built from on-site operators $S_q^a$, whereas the spin-exchange dynamical structure factor is built from two-spin bond operators. In the Heisenberg-chain RIXS formulation, the experimental cross section takes the form

$$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$

with $\omega_{\rm res}$ the resonance energy, $\Gamma$ the inverse core-hole lifetime, and $\eta$ the fractional change of $J$ in the intermediate state (Klauser et al., 2011).

A closely related bond-operator formulation appears in exact studies of adjacent spin operators. For the $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$0 Heisenberg chain,

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$1

so the exchange response can be assembled from dynamical correlators of adjacent bond components (Klauser et al., 2012).

Response	Operator	Symmetry sector
Single-spin DSF	$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$2	$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$3 triplet final states
Spin-exchange DSF	$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$4 or $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$5	$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$6 singlet final states
Longitudinal bond DSF	$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$7	$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$8

This distinction is not cosmetic. The exchange operator is a spin-rotational scalar, so it selects a different part of the Hilbert space from INS, even when both responses are supported by the same underlying fractionalized excitations (Klauser et al., 2011).

2. Symmetry, selection rules, and continuum kinematics

Because $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$9 is a scalar under global SU(2) spin rotations and the ground state is a singlet, only excited states with $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$0 contribute to $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$1. In the exact SU(2) reduction,

$$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$2

so the response carries an explicit interference factor $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$3 and vanishes at $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$4 (Klauser et al., 2011).

The selection rules differ sharply from INS. Single-spin operators create $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$5 states from the SU(2)-singlet ground state, whereas spin-exchange operators create $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$6 states. These sectors are orthogonal. Nevertheless, both responses are supported on the same two-spinon kinematic continuum, with identical lower threshold

$$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$7

and upper threshold

$$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$8

for $$S^{\mathrm{exch}}(q,\omega)=2\pi\sum_\alpha |\langle 0|X_q|\alpha\rangle|^2\,\delta(\omega-\omega_\alpha).$$9, extended periodically to $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$0 (Klauser et al., 2011).

The coincidence of continuum boundaries despite orthogonal final-state symmetry sectors is one of the defining properties of the spin-exchange DSF in the isotropic chain. It shows that the operator primarily reshuffles spectral weight within a common spinon phase space, rather than generating an altogether different kinematic support (Klauser et al., 2011).

3. Fractionalization and spectral-weight distribution in the Heisenberg chain

For the isotropic $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$1 Heisenberg chain, the spin-exchange response is almost entirely exhausted by two-spinon states. Exact Bethe-Ansatz form factors combined with the ABACUS summation algorithm show that two-spinon states “cover all but $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$2” of the total RIXS spectral weight for $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$3, whereas in the INS single-spin DSF the $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$4 spinon sector carries about $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$5 of the weight for the same size (Klauser et al., 2011).

This result overturns the naive expectation that a two-spin operator should predominantly excite four spinons. In the exchange DSF, the dominant final states are still two-spinon singlets. The main difference from INS lies not in the continuum boundaries but in the internal weight distribution. The static interference factor $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$6 suppresses intensity at the antiferromagnetic wavevector $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$7 and enhances it near $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$8 and $$I_{\mathrm{ex}}(\mathbf q,\omega)=\sum_n \big|\langle n|\hat O_{\mathrm{ex}}(\mathbf q)|0\rangle\big|^2\delta\!\big(\omega-(E_n-E_0)\big),\qquad \hat O_{\mathrm{ex}}(\mathbf q)=\sum_{\langle i,j\rangle}e^{i\mathbf q\cdot(\mathbf r_i-\mathbf r_j)}\,\mathbf S_i\!\cdot\!\mathbf S_j.$$9. Fixed-$S_q^a$0 cuts show that $S_q^a$1 has a noticeably broader high-$S_q^a$2 shoulder than $S_q^a$3, and at $S_q^a$4 the response is concentrated just above threshold, where the group velocity vanishes (Klauser et al., 2011).

Low-energy field theory does not capture these features well. The exchange operator probes short distances and energies of order $S_q^a$5. In particular, a straightforward Luttinger-liquid or bosonization treatment of four-spin correlators fails to reproduce the observed RIXS signal: the exchange response vanishes at low energies at $S_q^a$6, while standard Luttinger-liquid theory describes the INS singularity at $S_q^a$7 and $S_q^a$8 (Klauser et al., 2011).

For experimental interpretation, this implies a specific RIXS fingerprint in one-dimensional Heisenberg materials such as Sr$S_q^a$9CuO$$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$0: magnetic intensity confined between $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$1 and $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$2, dominant two-spinon singlet character, strong suppression at $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$3, and spectral weight shifted toward higher energies relative to INS (Klauser et al., 2011).

4. Adjacent bond operators, exact form factors, and multi-spinon sectors

A complementary exact route to exchange-sensitive dynamics is through adjacent bond operators. For the XXX chain in a field,

$$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$4

and the corresponding DSFs can be computed over the whole Brillouin zone using determinant representations of Bethe-Ansatz form factors and ABACUS summation (Klauser et al., 2012).

The longitudinal bond DSF,

$$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$5

is directly relevant because $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$6 contains $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$7 as one component. The paper states that the multi-spinon content and spectral support deduced for $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$8 carry directly to $$I(q,\omega)\propto \left|\frac{\omega_{\rm res}\Gamma(\eta J)}{i\Gamma+\omega}\right|^2 S^{\mathrm{exch}}(q,\omega),$$9, which is also a $\omega_{\rm res}$0 bond operator (Klauser et al., 2012).

At zero field, the connected longitudinal bond response $\omega_{\rm res}$1 displays large weight outside the two-spinon continuum. This is a central difference from the RIXS exchange DSF of (Klauser et al., 2011), which remains almost purely two-spinon.

Observable	Dominant $\omega_{\rm res}$2 sectors	Sum-rule saturation
$\omega_{\rm res}$3	$\omega_{\rm res}$4sp $\omega_{\rm res}$5, $\omega_{\rm res}$6sp$\omega_{\rm res}$7s$\omega_{\rm res}$8 $\omega_{\rm res}$9	$\Gamma$0
$\Gamma$1	$\Gamma$2sp$\Gamma$3s$\Gamma$4 $\Gamma$5, $\Gamma$6sp$\Gamma$7 $\Gamma$8, $\Gamma$9sp$\eta$0s$\eta$1 $\eta$2	$\eta$3

Two points are especially significant. First, the $\eta$4sp$\eta$5 contribution to $\eta$6 lies outside the two-spinon spectrum, so longitudinal bond correlations expose multi-spinon physics that is not confined to the usual two-spinon support. Second, at finite field the decomposition reorganizes into particle-hole and string sectors: for $\eta$7, the $\eta$8p$\eta$9h sector carries $J$0, $J$1, and $J$2 of the weight at $J$3, $J$4, and $J$5, respectively, with $J$6-string features appearing as a gapped mode near $J$7 and $J$8 (Klauser et al., 2012).

These exact adjacent-operator results show that “spin-exchange DSF” is not a unique universal lineshape but depends sensitively on the precise bond operator realized by the probe. In the RIXS operator $J$9, two-spinon singlets dominate. In the longitudinal adjacent-bond correlator, four-spinon weight becomes quantitatively large.

5. Relation to conventional dynamical structure factors

The spin-exchange DSF should be distinguished from the conventional on-site spin DSF even when both are treated in the same integrable model. For the finite-field XXX chain, exact Bethe-Ansatz and pseudofermion dynamical theory yield threshold dispersions and momentum-dependent exponents for

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$00

with lower thresholds controlled by branch lines of real-rapidity excitations across the full Brillouin zone, up to narrow small-$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$01 windows. In the zero-field limit, both longitudinal and transverse thresholds reduce to the two-spinon lower boundary

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$02

with universal edge exponent $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$03. Near saturation, the transverse structure factor collapses to the exact $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$04-peak

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$05

and $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$06 in the thermodynamic limit (Carmelo et al., 2015).

Further structure appears in field-dependent conventional DSFs through Bethe-string continua. In the thermodynamic limit of the XXX chain at finite field, $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$07 and $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$08 receive substantial contributions from $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$09-string states, predominantly $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$10-strings, generating gapped upper continua above the lower continua from real-rapidity states. By contrast, Bethe-string contributions to $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$11 are small at low spin density and become negligible above $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$12, while $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$13 has negligible Bethe-string weight at any finite field (Carmelo et al., 2020).

In the anisotropic XXZ chain, the exact two-spinon longitudinal DSF in the massive regime has an explicit closed form with exact continuum support, elastic contribution, and square-root edge singularities. In the isotropic limit it reduces to the standard Heisenberg-chain continuum boundaries. These conventional $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$14 benchmarks are not exchange DSFs, but they fix the underlying spinon kinematics against which bond-operator responses can be compared (Castillo, 2020).

The distinction matters methodologically. One finite-field XXX-chain study states explicitly that threshold positions and exponents for bond spin-exchange correlators relevant to RIXS are not derived there, although the same Bethe-Ansatz dressing, phase shifts, and mobile-impurity ideas provide a framework for such analyses (Carmelo et al., 2015).

6. Extensions, higher symmetries, and higher-dimensional expectations

The operator notion of spin-exchange DSF extends beyond the SU(2) chain. For the SU(3) Heisenberg chain, a bond-spin-exchange operator appropriate to RIXS is

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$15

with response

$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$16

The variational Monte Carlo study of the SU(3) chain computes instead the local-generator DSF $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$17, but it shows that the spectrum is organized by two-soliton continua, critical Wess–Zumino–Witten SU(3)$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$18 scaling, and singular peaks at $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$19. This suggests that an SU(3) bond-exchange DSF should likewise reflect multi-soliton continua, with the symmetry and multiplet structure $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$20 controlling the accessible sectors (Vörös et al., 2021).

In frustrated two-dimensional magnets, the situation is presently more indirect. For the triangular-lattice $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$21-$$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$22 Heisenberg model, the calculated quantity is the conventional spin DSF $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$23, specifically $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$24, not the spin-exchange DSF. However, the paper defines the corresponding bond-exchange response and argues that the same underlying excitation content—magnons, deconfined spinons, and gauge-field excitations—should shape both responses. It states that roton-like minima and the softening at $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$25 should correlate with enhanced low-energy bond-energy fluctuations at $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$26 as $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$27 increases, and that the projection-induced low-energy signal at $$X_q=\frac{1}{\sqrt N}\sum_j e^{iq j}\left(\mathbf S_{j-1}\!\cdot\!\mathbf S_j+\mathbf S_j\!\cdot\!\mathbf S_{j+1}\right),$$28 suggests exchange-operator responses may also be sensitive to gauge-field-driven monopole dynamics at the Brillouin-zone corners (Ferrari et al., 2019).

Taken together, these results locate the spin-exchange dynamical structure factor at the intersection of two themes. The first is operator specificity: different bond operators emphasize different sectors, from almost purely two-spinon singlets in indirect RIXS on the Heisenberg chain to large four-spinon weight in adjacent longitudinal bond correlators. The second is excitation universality: whenever the underlying system supports fractionalized quasiparticles, strings, or gauge-field modes, exchange-sensitive probes inherit that kinematic content but redistribute spectral weight according to bond-operator symmetry and interference.