---
title: Synthetic Rank-2 Electric Field in Cold Atoms
url: https://www.emergentmind.com/topics/synthetic-rank-2-electric-field
type: topic
---

# Synthetic Rank-2 Electric Field in Cold Atoms

Searching arXiv for the cited works and related context on synthetic tensor/rank-2 gauge fields.
A synthetic rank-2 electric field is an engineered tensor-gauge-field response that couples to dipoles rather than to single-particle charge. In the ultracold-atom constructions discussed in "Synthetic tensor gauge fields" and "Dipole condensates in synthetic rank-2 electric fields," the field arises from controlled Peierls phases generated by tilts, quadratic curvature, or spin-dependent forces, and it produces dynamics that are inaccessible to conventional rank-1 vector gauge fields, including dipolar Bloch oscillations, phase twists of dipole condensates, and dipole supercurrents [2306.15663] [2509.14492]. In the hybrid lattice-gauge-theory perspective of the F3 model, an analogous synthetic rank-2 structure appears in the low-energy theory through a double divergence of a vector electric field, showing that a rank-2 sector can emerge even when no independent symmetric tensor electric field survives microscopically [2203.14917].

## 1. Definition and formal characterization

In the 1D cold-atom realization, the synthetic rank-2 gauge potential is identified as
$$
A_{xx}(t)=\Delta' t,
$$
so that, by analogy with Maxwell theory, the rank-2 electric field is
$$
E_{xx}=-\partial_t A_{xx}=-\Delta'.
$$
The defining feature is that this field acts on a dipole formed by a particle-hole pair rather than on an isolated particle [2306.15663].

In the two-component synthetic-dimension construction, the relevant directions are the real lattice direction \(x\) and a synthetic spin direction \(w\). A spin-dependent vector potential,
$$
A_x(\sigma;t)=\Delta_\sigma t,
$$
produces a spin-resolved electric field \(E_x(\sigma)=\Delta_\sigma\), and the discrete difference along the synthetic direction defines a rank-2 field,
$$
E_{xw}=E_x(\uparrow)-E_x(\downarrow)=\Delta_\uparrow-\Delta_\downarrow.
$$
This is the field that couples to particle-hole dipoles living in the \(x\)–\(w\) plane [2509.14492].

The hybrid U(1) lattice-gauge-theory construction introduces a distinct but related notion. There, the low-energy theory contains a rank-1 vector electric field \(E^a\), and the rank-2 sector is encoded by
$$
({\bf DE})(\mathbf r)=\partial_x\partial_y E^z+\partial_y\partial_z E^x+\partial_z\partial_x E^y.
$$
A frequent misconception is that a synthetic rank-2 electric structure must always be represented by an independent symmetric tensor \(E^{ab}\). In the F3 construction, no independent symmetric tensor \(E^{ab}\) survives in the final low-energy theory, yet a built-in rank-2 Gauss law remains through \({\bf DE}\) [2203.14917].

## 2. One-dimensional optical-lattice construction

The basic 1D realization starts from bosons in a deep optical lattice with on-site interaction \(U\), nearest-neighbor hopping \(t_1\), a strong linear tilt \(\Delta\), and a weak quadratic potential \(\Delta'\):
$$
H=-t_{1}\sum_{m}\bigl(b_{m}^{\dagger}b_{m+1} + \mathrm{h.c.}\bigr)
+\sum_{m}\Bigl(\Delta m + \tfrac{\Delta'}{2}m^{2}\Bigr)n_{m}
+\frac{U}{2}\sum_{m}n_{m}(n_{m}-1).
$$
The regime
$$
t_{1}\ll \Delta,U,\qquad \Delta'\ll \Delta
$$
is chosen so that single-particle hopping is Wannier–Stark–localized, double occupancies are suppressed, and only second-order dipole hops survive [2306.15663].

The relevant composite excitation is a nearest-neighbor particle-hole pair described by
$$
D_m^\dagger=b_m^\dagger b_{m-1}.
$$
In second-order perturbation theory, the effective kinetic term becomes
$$
K_L=-t_2\sum_m\bigl(D_m^\dagger D_{m+1}+\mathrm{h.c.}\bigr),
$$
with
$$
t_2\simeq \frac{t_1^2 U}{\Delta^2}.
$$
Direct long-range dipole hops cancel by interference, so the effective dynamics are carried by these nearest-neighbor dipole processes [2306.15663].

After the gauge transformation
$$
U(t)=\exp\!{\Bigl[i\,t\sum_{m}(\Delta m+\tfrac{\Delta'}{2}m^{2})\,n_{m}\Bigr]},
$$
single-particle hopping acquires a site- and time-dependent Peierls phase,
$$
b_m^\dagger b_{m+1}\longrightarrow e^{-i[\Delta+\Delta'(m+\tfrac12)]t}\,b_m^\dagger b_{m+1},
$$
and the second-order dipole term correspondingly acquires the rank-2 gauge potential \(A_{xx}(t)=\Delta' t\) [2306.15663].

## 3. Gauge structure and moment conservation

The 1D dipole Hamiltonian is invariant under the scalar-charge rank-2 gauge transformation
$$
A_{xx}\to A_{xx}+\partial_x^2\alpha,\qquad
b_m\to e^{\,i[\alpha(m+1)-\alpha(m)]}b_m.
$$
This identifies the effective field as a rank-2 gauge potential in the sense relevant to scalar-charge tensor gauge theory [2306.15663].

Its physical action differs sharply from that of an ordinary electric field. In the laboratory frame, the quadratic potential \(\tfrac12\Delta' m^2\) acts as a linear potential on a dipole because each advancement of \(D_m^\dagger\) increases the energy by \(\Delta'\). Unlike a conventional rank-1 electric field \(E_x\), which couples to single-particle charge and shifts the dipole moment
$$
P=\sum_m m\,n_m,
$$
the rank-2 field \(E_{xx}\) couples only to dipoles and conserves \(P\) while modulating the quadrupole moment
$$
Q=\sum_m m^2\,n_m.
$$
This separation between dipole-sensitive and charge-sensitive responses is one of the central signatures of the synthetic rank-2 construction [2306.15663].

The same dipole-selective logic reappears in the two-component system. There, the field \(E_{xw}\) is generated by the difference of spin-dependent forces, so it acts on particle-hole dipoles formed across the synthetic spin direction rather than on a single charge sector alone [2509.14492].

## 4. Dipolar Bloch oscillations

Projecting the 1D system onto the single-dipole subspace yields
$$
H_{\rm eff}=
-t_2\sum_m\bigl(D_m^\dagger D_{m+1}+{\rm h.c.}\bigr)
-\tfrac12 E_{xx}\sum_m m\,D_m^\dagger D_m.
$$
This is precisely a tight-binding model with a constant Stark field \(E_{xx}\), and its eigenstates are Wannier–Stark ladders [2306.15663].

A dipole wave packet therefore undergoes a new type of Bloch oscillation with period
$$
T_D=\frac{2\pi}{|E_{xx}|}=\frac{2\pi}{|\Delta'|}.
$$
The defining diagnostic is not a conventional center-of-mass oscillation of charged particles, but an oscillatory modulation of the quadrupole moment accompanied by strict preservation of the dipole moment:
$$
P(t)=P(0).
$$
The quadratic curvature of the trap is thus reinterpreted as a uniform rank-2 electric field acting on the dipole sector [2306.15663].

This dynamical response is distinct from ordinary Bloch oscillations driven by a rank-1 field. Here the motion is that of a lineon formed by a particle-hole pair, and the field addresses the dipole coordinate directly. A plausible implication is that rank-2 electric-field spectroscopy can separate dipolar transport channels from charge transport channels in settings where the two would otherwise be mixed.

## 5. Synthetic dimensions and dipole condensates

The two-component Bose-Hubbard realization uses
$$
\hat H_B=\hat K_B+\hat U_B+\hat V_B,
$$
with
$$
\hat K_B=-t_1\sum_{m,\sigma}\bigl(\hat b^\dag_{m,\sigma}\hat b_{m+1,\sigma}+\text{h.c.}\bigr),
$$
$$
\hat U_B=
\frac{U}{2}\sum_{m,\sigma}\hat n_{m,\sigma}(\hat n_{m,\sigma}-1)
+U_{\uparrow\downarrow}\sum_m\hat n_{m,\uparrow}\hat n_{m,\downarrow},
$$
$$
\hat V_B=\sum_{m,\sigma}m\,\Delta_\sigma\,\hat n_{m,\sigma}.
$$
Treating \(\sigma=\uparrow,\downarrow\) as a synthetic coordinate \(w\), a spin-dependent tilt generates the Peierls phases
$$
\hat{\tilde K}_B=-t_1\sum_{m,\sigma}\Bigl(e^{-i\,\Delta_\sigma t/\hbar}\,\hat b^\dag_{m,\sigma}\,\hat b_{m+1,\sigma}+\text{h.c.}\Bigr),
$$
from which the synthetic rank-2 field \(E_{xw}\) follows directly [2509.14492].

This setting supports two experimentally emphasized electrodynamic responses. First, with \(\Delta_\uparrow\neq0\) and \(\Delta_\downarrow=0\), the center of mass of each spin component,
$$
X_\sigma(t)=\sum_m m\,\langle \hat n_{m,\sigma}(t)\rangle,
$$
exhibits the atomic analogue of perfect Coulomb drag: for large inter-spin repulsion \(U_{\uparrow\downarrow}\gg t_1\),
$$
X_\uparrow(t)=-X_\downarrow(t).
$$
Equivalently, the relative dipole displacement satisfies \(\langle X_r(t)\rangle\to0\) and \(\mathrm{Var}(X_r)\to0\) as \(U_{\uparrow\downarrow}\to\infty\) [2509.14492].

Second, in the deep Mott or dipole regime \(t_1\ll U,U_{\uparrow\downarrow}\), projection onto
$$
\hat D^\dag_m=\hat b^\dag_{m,\uparrow}\hat b_{m,\downarrow}
$$
gives an effective dipole Hamiltonian with a rank-2 Peierls phase,
$$
\hat H_E=-t_2\sum_m\Bigl(e^{-i\,\Delta t/\hbar}\,\hat D^\dag_m\hat D_{m+1}+\text{h.c.}\Bigr)+\cdots,
$$
where \(\Delta=\Delta_\uparrow-\Delta_\downarrow\). A short pulse of duration \(\tau\) imprints the dipole phase twist
$$
\phi_D=\frac{\Delta\tau}{\hbar},
$$
and a uniform twist generates a dipole supercurrent obeying the dipolar Josephson relation
$$
J_D=2\,t_2\,\rho_s^D\sin(\phi_D).
$$
In the full two-component model,
$$
J_D=2\,t_2\,\mathrm{Im}\langle\hat D^\dag_m\hat D_{m+1}\rangle.
$$
For hardcore dipoles, the analytical result
$$
\rho_s^D=\rho_D\,\frac{\sin(\pi\rho_D)}{\pi}
$$
is reported to be in excellent agreement with TEBD simulation [2509.14492].

The same work also identifies a dynamical probe in momentum space: the dipole momentum distribution \(n_k^D\) undergoes Bloch oscillations with period \(T=2\pi\hbar/|\Delta|\sim100~\mu{\rm s}\)–ms for \(\Delta/h\) up to a few kHz [2509.14492].

## 6. Higher-dimensional tensor fields, planons, and topological phases

In two dimensions, synthetic tensor gauge fields can be generated through ring-exchange interactions,
$$
K_P=-t_2'\sum_{\mathbf m}
\bigl(b_{\mathbf m}^{\dagger}b_{\mathbf m+\hat x}^{\dagger}
b_{\mathbf m+\hat x+\hat y}
b_{\mathbf m+\hat y}
+{\rm h.c.}\bigr).
$$
This process moves a dipole oriented along \(\hat x\) by one step in \(\hat y\). By imprinting a site-dependent phase on the underlying one-body hopping, for example via laser-assisted tunneling in a linear tilt, one induces a rank-2 gauge potential \(A_{xy}\) in the effective dipole motion [2306.15663].

In three dimensions, if that phase varies linearly along \(z\), the construction realizes a magnetic component
$$
B_{xyz}=-\partial_z A_{xy}.
$$
When restricted to dipoles along \(\hat x\) moving in the \(\hat y\)–\(\hat z\) plane, the result is the dipolar Harper–Hofstadter Hamiltonian
$$
H_{\rm dHH}
=\tfrac12\sum_{\mathbf m}
\Bigl[
J_y\,e^{-i\,m_z\phi}\,
D_{\mathbf m,\hat x}^{\dagger}D_{\mathbf m+\hat y,\hat x}
+
J_z\,
D_{\mathbf m,\hat x}^{\dagger}D_{\mathbf m+\hat z,\hat x}
+\mathrm{h.c.}
\Bigr],
$$
with flux per plaquette \(\phi=B_{xyz}a\) [2306.15663].

In its gapped regimes, this model supports dipolar Chern-insulator phases whose edge modes carry quantized dipole currents in the absence of any net charge current. The same framework thereby connects synthetic rank-2 electric and magnetic fields to topological transport of dipoles and to planon dynamics in higher dimensions [2306.15663].

## 7. Gauge-theory embedding and experimental realization

The hybrid rank-1 and rank-2 U(1) lattice-gauge theory of the F3 model provides a continuum framework in which the same vector electric field participates in both conventional and rank-2 constraints. The rank-1 Gauss law is
$$
\nabla\!\cdot\!E=-\frac{2\pi}{p}\rho^e,
$$
while the rank-2 fracton-sector Gauss law is
$$
{\bf DE}=-\rho^f.
$$
These appear in the continuum Lagrangian
$$
\mathcal{L}_{\rm F3}
= \frac{p}{2\pi}E^a\partial_t A^a - A^a j_a^m - E^a j_a^e
-\frac{p}{2\pi}E^0\Bigl(\partial_a A^a+\tfrac{2\pi}{p}\rho^e\Bigr)
- A_a^0\Bigl(\epsilon_{abc}\partial_b E^c+\rho_a^m\Bigr)
- A^0\Bigl(\tfrac{p}{2\pi}{\bf DE}+\rho^f\Bigr).
$$
The synthetic rank-2 sector is therefore not merely a cold-atom engineering trick; it also appears as a structural element of a hybrid continuum gauge theory obtained from Higgsing and continuum reduction [2203.14917].

On the experimental side, the 1D optical-lattice proposal specifies lattice depths \(\gtrsim10E_r\), tunneling \(t_1/h\sim50\)–\(200\) Hz, on-site interaction \(U/h\sim500\)–\(1000\) Hz, linear tilt \(\Delta/h\sim5\)–\(15\) kHz, quadratic tilt \(\Delta'/h\sim1\)–\(10\) Hz, effective dipole tunneling \(t_2\sim\) a few Hz, and a Bloch-oscillation period \(T_D\sim0.1\)–\(1\) s, well within typical trap lifetimes. In 2D or 3D, Rydberg or dipolar molecules with nearest-neighbor interaction \(V\gtrsim t_1\) give ring-exchange scales \(t_2'\sim10\)–\(100\) Hz, and laser-assisted tunneling phases of order unity access dipolar Harper–Hofstadter physics [2306.15663].

The two-component realization emphasizes complementary diagnostics. Center-of-mass displacements \(X_\sigma(t)\) are directly imaged via in situ density; perfect drag is identified by \(X_\uparrow(t)=-X_\downarrow(t)\). The dipole supercurrent can be obtained either from the time derivative of the net dipole moment or from spin-spin correlations,
$$
J_D=2\,t_2\,\mathrm{Im}\bigl\langle \hat S_m^+ \hat S_{m+1}^- \bigr\rangle,
\qquad
\hat S_m^+=\hat b^\dag_{m,\uparrow}\hat b_{m,\downarrow},
$$
and the required correlators are measured by appropriate global or site-selective \(\pi/2\) pulses and density readout [2509.14492].

Taken together, these constructions establish a consistent notion of synthetic rank-2 electric field across several settings: a quadratic-curvature-induced field \(E_{xx}\) for lineons in 1D, a spin-gradient-induced field \(E_{xw}\) for dipole condensates in a synthetic dimension, and a continuum hybrid gauge-theory sector encoded by \({\bf DE}\). The common principle is that engineered phases and constraints are arranged so that the field couples exclusively to dipolar degrees of freedom, producing electrodynamics that preserve or diagnose dipole structure rather than ordinary charge transport.

Source: https://www.emergentmind.com/topics/synthetic-rank-2-electric-field