Acoustoelectric Superlattice Fundamentals

• Acoustoelectric superlattices are periodic electronic potentials generated by acoustic waves, offering tunability distinct from fixed lithographic patterns.
• SAW-induced pseudopotentials create effective static trapping lattices that modulate carrier dynamics and enable reconfigurable miniband engineering.
• Experimental platforms range from GaAs/AlGaAs 2DEGs to BN-encapsulated graphene, demonstrating applications in ultrafast current generation and tunable topological phases.

Searching arXiv for recent and foundational papers on acoustoelectric superlattices and closely related SAW/superlattice platforms. First search: exact topic phrase and closely matching title. An acoustoelectric superlattice is a periodic electronic potential landscape generated, driven, or functionally defined by acoustic waves, most commonly surface acoustic waves (SAWs) or propagating strain fields in piezoelectric or layered semiconductor media. In one widely used construction, counter-propagating SAWs produce an effective time-independent trapping lattice for carriers, directly analogous to the pseudopotential of a Paul trap; in another, a traveling acoustic perturbation propagates through a semiconductor superlattice and dynamically modulates tunneling transport; more recent work extends the concept to externally programmable two-dimensional miniband engineering in quantum materials, where the superlattice geometry, wavelength, phase, and amplitude are controlled in situ by the acoustic drive rather than by lithographic patterning or interlayer registry (Schuetz et al., 2017, Wang et al., 2020, Meril et al., 6 Jul 2025).

1. Conceptual scope and defining characteristics

In the literature considered here, the term denotes a superlattice whose periodicity is imposed by acoustically generated electric or deformation potentials rather than by a permanently fabricated modulation. The defining ingredients are a periodic acoustic field, carrier coupling through piezoelectric or deformation-potential mechanisms, and an electronically relevant length scale set by the acoustic wavelength or by the interference of several acoustic waves (Schuetz et al., 2017, Meril et al., 6 Jul 2025).

For SAW-based realizations in piezoelectric semiconductors, the basic time-dependent potential is

$$V(x,t)=V_{\mathrm{SAW}}\cos(kx)\cos(\omega t), \qquad \omega=v_s k, \qquad \lambda=\frac{2\pi}{k}.$$

When the acoustic drive is sufficiently rapid, the slow carrier dynamics is governed by an effective static lattice, even though the underlying field oscillates in time. This motivates the designation acoustic lattice or acoustic pseudo-lattice, and the resulting electron system can be viewed as an acoustoelectric superlattice (Schuetz et al., 2017).

A distinct but related usage arises in pre-existing semiconductor superlattices, such as weakly coupled GaAs/AlAs structures, where a propagating train of picosecond deformation pulses creates a moving potential inside the layered stack. In that setting, the superlattice is structural, but the acoustoelectric physics lies in the dynamical acoustic control of miniband transport and current generation (Wang et al., 2020). A further extension is the deliberately programmable two-dimensional acoustoelectric superlattice formed by two obliquely propagating SAWs under a two-dimensional quantum material, where the reciprocal lattice is generated by integer combinations of the two SAW wavevectors and the miniband structure can be reconfigured in real time (Meril et al., 6 Jul 2025).

Two properties recur across these realizations. First, the imposed periodicity is externally tunable through the acoustic frequency, since $\lambda=2\pi/k$. Second, the modulation amplitude is controlled by RF power or strain amplitude, so the superlattice is not fixed by crystal growth or twist angle. This sharply distinguishes acoustoelectric superlattices from conventional lithographic superlattices and from moiré systems whose potential amplitude is set by intrinsic interlayer tunneling and lattice relaxation (Schuetz et al., 2017, Meril et al., 6 Jul 2025).

2. Effective-lattice formation from standing surface acoustic waves

The most explicit theoretical formulation of an acoustoelectric superlattice is the SAW pseudopotential construction for electrons in a two-dimensional electron gas or a quasi-one-dimensional wire. Two opposite SAWs launched by interdigital transducers generate the oscillatory potential $V(x,t)$ above. The central figure of merit is the sound kinetic energy

$E_S=\frac{m v_s^2}{2},$

and the corresponding dimensionless stability parameter is

$q=\frac{V_{\mathrm{SAW}}}{E_S}.$

In the classical description, motion near a minimum reduces to a Mathieu-type problem,

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$

with $\tilde{x}=kx$ and $\tau=\omega t/2$. In the small-amplitude regime $q^2\ll 1$, the trajectory separates into slow secular motion at $\omega_0$ and micromotion at $\lambda=2\pi/k$0, with

$\lambda=2\pi/k$1

The effective pseudopotential is then

$\lambda=2\pi/k$2

The quantum treatment reaches the same result through a high-frequency Floquet expansion of

$\lambda=2\pi/k$3

yielding, to leading order,

$\lambda=2\pi/k$4

An equivalent expression,

$\lambda=2\pi/k$5

makes the trap-depth scaling explicit (Schuetz et al., 2017).

Because the effective potential is $\lambda=2\pi/k$6, the lattice spacing is

$\lambda=2\pi/k$7

For SAW frequencies in the tens of GHz, spacings around $\lambda=2\pi/k$8 are emphasized, placing the acoustoelectric superlattice in a regime of relatively large energy scales for electron systems (Schuetz et al., 2017). The trapping regime of interest is summarized by

$\lambda=2\pi/k$9

(with $V(x,t)$0 in that expression), together with the requirement of at least one bound state per site,

$V(x,t)$1

with $V(x,t)$2 required (Schuetz et al., 2017).

A notable feature is reconfigurability. Since the lattice is generated by driven SAWs, its minima are not lithographically fixed; they can be translated in situ by changing the RF phases applied to the transducers. The same mechanism extends to two dimensions, where orthogonal SAWs produce a square-lattice geometry and more elaborate geometries are explicitly noted as possible (Schuetz et al., 2017). A closely related experimental realization in BN-encapsulated graphene employs counter-propagating SAWs to form a standing surface acoustic wave,

$V(x,t)$3

which creates a spatially fixed but temporally oscillating modulation; under phase matching and equal intensity, the acoustoelectric current vanishes, consistent with standing-wave formation rather than net momentum drag (McSorley et al., 20 Mar 2025).

3. Dynamical semiconductor superlattices under propagating acoustic fields

A second major branch of the subject concerns semiconductor superlattices driven by traveling acoustic perturbations. In a GaAs/AlAs weakly coupled superlattice with period $V(x,t)$4 nm and length $V(x,t)$5 nm, electron transport occurs by quantum tunneling between adjacent wells while an incoming train of picosecond strain pulses from an Al transducer generates a moving deformation potential inside the structure. The time-dependent Schrödinger equation

$V(x,t)$6

is solved for

$V(x,t)$7

with $V(x,t)$8. The simulations, implemented with a Crank–Nicolson method on GPUs, show that a $V(x,t)$9 GHz strain-pulse train can generate current oscillations near $E_S=\frac{m v_s^2}{2},$0 GHz, corresponding to an up-conversion factor of about $E_S=\frac{m v_s^2}{2},$1. The response follows from fast internal electron oscillations inside the moving acoustic potential, not from simple locking to the pulse repetition rate (Wang et al., 2020).

The same work demonstrates that an electromagnetic-style rectification model,

$E_S=\frac{m v_s^2}{2},$2

is inadequate when propagation is neglected: the simplified model gives the wrong quantitative scale and even the wrong qualitative dependence on $E_S=\frac{m v_s^2}{2},$3 and $E_S=\frac{m v_s^2}{2},$4. This is an important corrective to a common oversimplification. In an acoustoelectric superlattice, the strain field is not merely an oscillatory drive; it is a traveling potential landscape whose spatial propagation through the layered stack is essential to the transport physics (Wang et al., 2020).

The dynamical picture is broadened further by studies of miniband semiconductor superlattices driven by a longitudinal acoustic plane wave,

$E_S=\frac{m v_s^2}{2},$5

with first-miniband dispersion

$E_S=\frac{m v_s^2}{2},$6

In the moving frame, the acoustic drive generates fixed points and separatrices in phase space; when the amplitude exceeds critical values

$E_S=\frac{m v_s^2}{2},$7

global bifurcations appear. These are interpreted as normal and anomalous Doppler processes in the acoustic field, accompanied by electron bunch formation, drift suppression, drift reversal, and absolute negative mobility. Near the second critical amplitude $E_S=\frac{m v_s^2}{2},$8, the real part of the dynamical conductivity becomes negative over a broad range, with gain up to about $E_S=\frac{m v_s^2}{2},$9 GHz and a gain coefficient $q=\frac{V_{\mathrm{SAW}}}{E_S}.$0 for GaAs-like parameters (Apostolakis et al., 2022).

An older but conceptually aligned line of work considers the attenuation coefficient $q=\frac{V_{\mathrm{SAW}}}{E_S}.$1 of acoustic phonons in a one-dimensional semiconductor superlattice under a combined field

$q=\frac{V_{\mathrm{SAW}}}{E_S}.$2

in the hypersound regime $q=\frac{V_{\mathrm{SAW}}}{E_S}.$3. After kinetic averaging, the field response is expressed as a Bessel-weighted harmonic sum,

$q=\frac{V_{\mathrm{SAW}}}{E_S}.$4

and an inversion is reported in which amplification far exceeds absorption, with $q=\frac{V_{\mathrm{SAW}}}{E_S}.$5, suggesting terahertz phonon amplification and a hypersound MASER regime (Dompreh et al., 2011).

4. Two-dimensional programmable acoustoelectric superlattices and miniband topology

The most explicit modern generalization of the concept is the externally programmable two-dimensional acoustoelectric superlattice formed by two SAWs with wavevectors $q=\frac{V_{\mathrm{SAW}}}{E_S}.$6 and $q=\frac{V_{\mathrm{SAW}}}{E_S}.$7 underneath a two-dimensional quantum material on a piezoelectric substrate. The imposed potential is

$q=\frac{V_{\mathrm{SAW}}}{E_S}.$8

The reciprocal lattice is generated by

$q=\frac{V_{\mathrm{SAW}}}{E_S}.$9

and the wavelength $$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$0 sets the superlattice period. The acoustic amplitude follows

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$1

or, equivalently, from

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$2

which yields

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$3

This architecture places acoustoelectric superlattices at an intermediate length scale between moiré superlattices and optical lattices, while allowing geometry and amplitude to be tuned independently (Meril et al., 6 Jul 2025).

Massive monolayer graphene provides a canonical example. Near valley $$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$4, the low-energy Hamiltonian is

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$5

The SAW potential couples momentum states differing by $$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$6 or $$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$7, so the folded band structure is obtained in an extended-zone “K-lattice” representation. A moving-frame transformation

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$8

renders the problem time-independent through

$$\frac{d^2\tilde{x}}{d\tau^2}+2q\sin(\tilde{x})\cos(2\tau)=0,$$9

This formulation remains applicable even when $\tilde{x}=kx$0 and $\tilde{x}=kx$1 are incommensurate (Meril et al., 6 Jul 2025).

For $\tilde{x}=kx$2 meV and $\tilde{x}=kx$3 W/m, the highest valence miniband undergoes a first gap closing near $\tilde{x}=kx$4 nm at $\tilde{x}=kx$5, after which it acquires $\tilde{x}=kx$6, followed by a second gap closing near $\tilde{x}=kx$7 nm at the $\tilde{x}=kx$8 point, returning it to $\tilde{x}=kx$9. The first Dirac-point closing is captured by a minimal ten-band model with critical coupling

$\tau=\omega t/2$0

For large $\tau=\omega t/2$1, the folded spectrum can produce highly atomic-like flat bands with bandwidth below $\tau=\omega t/2$2 meV in the examples shown. The Berry curvature is strongly concentrated near $\tau=\omega t/2$3 and the mini-Brillouin-zone boundary; near $\tau=\omega t/2$4 nm, roughly $\tau=\omega t/2$5 of the valley Chern number comes from only about $\tau=\omega t/2$6 of the mini-Brillouin-zone area. At $\tau=\omega t/2$7 W/m, the topological phase diagram contains broad regions where the highest valence miniband has $\tau=\omega t/2$8, and a positive indirect gap identifies valley-Hall phases (Meril et al., 6 Jul 2025).

A more transport-oriented realization appears in BN-encapsulated graphene on $\tau=\omega t/2$9 Y-cut LiNbO$q^2\ll 1$0. There, intense SAWs with $q^2\ll 1$1 and $q^2\ll 1$2 give a resonance near $q^2\ll 1$3, and the acoustoelectric current density

$q^2\ll 1$4

crosses over from a linear regime to a nonlinear regime in which the carrier drift velocity saturates near $q^2\ll 1$5, slightly below $q^2\ll 1$6. The carrier density is described as

$q^2\ll 1$7

and, when $q^2\ll 1$8 becomes comparable to the background density, periodic electron, hole, or mixed electron-hole stripes appear. This supplies a real-space picture of a dynamic acoustoelectric superlattice, with standing-wave configurations acting as reconfigurable gates rather than static electrodes (McSorley et al., 20 Mar 2025).

5. Experimental realizations, platforms, and observables

The platform space is heterogeneous. For effective SAW trapping and pseudolattice formation, candidate systems include GaAs/AlGaAs $q^2\ll 1$9DEGs, AlN/diamond heterostructures, Si/SiGe, TMDCs such as $\omega_0$0 and $\omega_0$1, heavy-hole systems, and trions. A central materials criterion is the enhancement of

$\omega_0$2

which favors high-sound-velocity substrates such as diamond-based heterostructures because larger $\omega_0$3 increases the achievable trap depth (Schuetz et al., 2017).

Graphene-based implementations on LiNbO$\omega_0$4 emphasize piezoelectric coupling and direct electrical readout. In the derivative-sensitive probing modality, an hBN-encapsulated monolayer graphene Hall bar is integrated with delay-line interdigital transducers on $\omega_0$5-Y-cut LiNbO$\omega_0$6, with a SAW center frequency of approximately $\omega_0$7 MHz. The SAW-induced current obeys

$\omega_0$8

and the rectified d.c. component is

$\omega_0$9

so that

$\lambda=2\pi/k$00

This derivative weighting suppresses metallic background and enhances weak spectral features. In graphene/hBN moiré superlattices it enabled fractal Brown-Zak oscillations up to the fifth order and the first acoustoelectric observation of the Hofstadter butterfly, including features at $\lambda=2\pi/k$01 (Song et al., 6 Dec 2025).

Acoustoelectric layered heterostructures also support active microwave functionality. A three-layer In$\lambda=2\pi/k$02Ga$\lambda=2\pi/k$03As / LiNbO$\lambda=2\pi/k$04 / Si stack uses carrier drift in a thin semiconductor film to exchange energy with a guided piezoelectric acoustic mode. Gain appears when $\lambda=2\pi/k$05, with the synchronous condition $\lambda=2\pi/k$06 and the required drift field $\lambda=2\pi/k$07. In a $\lambda=2\pi/k$08 device operating at $\lambda=2\pi/k$09 GHz, a terminal gain of $\lambda=2\pi/k$10 dB is reported with $\lambda=2\pi/k$11 mW DC power dissipation; broadband gain extends from $\lambda=2\pi/k$12 to $\lambda=2\pi/k$13 GHz, and the nonreciprocal transmission contrast exceeds $\lambda=2\pi/k$14 dB at $\lambda=2\pi/k$15 GHz. The acoustic noise figure is $\lambda=2\pi/k$16 dB (Hackett et al., 2022). Although this architecture is described as “superlattice-like” rather than as a superlattice in the strict band-structure sense, it exemplifies the broader acoustoelectric strategy of engineering layered media to maximize acoustic-field confinement, carrier overlap, and nonreciprocal transport.

6. Related phenomena, common distinctions, and scientific significance

Several distinctions are essential for precise usage. First, an acoustoelectric superlattice need not be static in the laboratory frame. In SAW pseudolattice schemes, the structure is stationary only in the effective, time-averaged description, while the physical fields are moving waves; in traveling-wave transport through semiconductor superlattices, by contrast, the moving nature of the perturbation is itself the central observable (Schuetz et al., 2017, Wang et al., 2020).

Second, acoustic driving is not generically reducible to electromagnetic rectification. The propagating strain field in a semiconductor superlattice creates a time-dependent potential that drags and modulates the electron wavefunction as it traverses the stack, and omission of this propagation can lead to qualitatively incorrect predictions for bias and strain dependence (Wang et al., 2020). A related misconception is that acoustoelectric response is necessarily ordinary and longitudinal. In chiral superconductors, weak particle-hole asymmetry allows a deformation potential to couple linearly to real and imaginary clapping modes, which generate a transverse alternating current with

$\lambda=2\pi/k$17

The resonance appears below the pair-breaking threshold $\lambda=2\pi/k$18, where the clapping-mode masses are $\lambda=2\pi/k$19, and becomes especially strong when $\lambda=2\pi/k$20 (Matsushita et al., 2021). This is not a superlattice realization, but it clarifies that acoustoelectric phenomena can be governed by collective modes rather than by single-particle drag alone.

Third, acoustoelectric superlattices differ from moiré superlattices in how control parameters are partitioned. In moiré systems, periodicity and coupling are constrained by twist angle, lattice mismatch, and relaxation. In acoustoelectric superlattices, the wavelength is set by SAW frequency, the geometry by wavevector orientation, the spatial origin by phase, and the amplitude by RF power. This separation of controls underlies proposals for in-situ band inversion, flat-band formation, dynamic gate operations, and phase-tunable localization (Meril et al., 6 Jul 2025, McSorley et al., 20 Mar 2025).

The broader significance follows from these control knobs. In SAW trapping architectures, confinement to the lowest band maps the system to an extended Anderson-Hubbard model,

$\lambda=2\pi/k$21

with tunneling estimate

$\lambda=2\pi/k$22

and Coulomb scale roughly

$\lambda=2\pi/k$23

This opens the regime $\lambda=2\pi/k$24 relevant to Mott-like physics and spin exchange $\lambda=2\pi/k$25 (Schuetz et al., 2017). In two-dimensional programmable ASLs, the same external tunability supports valley-Hall minibands and narrow topological bands (Meril et al., 6 Jul 2025). In traveling-wave semiconductor superlattices, the moving acoustic potential enables ultrafast current generation, frequency multiplication, and phonon-mediated gain (Wang et al., 2020, Apostolakis et al., 2022, Dompreh et al., 2011).

Taken together, these results show that the acoustoelectric superlattice is not a single device class but a unifying paradigm: acoustic fields act as externally programmable generators of periodic electronic structure, transport instabilities, and collective electromechanical response across semiconductors, van der Waals heterostructures, moiré materials, and correlated quantum media.