Flat electronic bands and moiré magnetism remain theoretical proposals rather than demonstrated oxide states. The strategy begins with epitaxial synthesis. A functional complex oxide is deposited on a substrate, often above a selectively removable sacrificial layer, at conditions designed to establish its crystallinity, orientation and strain state.

Chemical etching of the sacrificial layer then releases the film without sacrificing its single-crystalline structure. This approach differs from ordinary mechanical exfoliation because most epitaxial oxides lack weak cleavage planes; they are three-dimensionally bonded ionic-covalent lattices. Selective chemical lift-off therefore provides the principal route to ultrathin membranes, although etching, drying and transfer can introduce polymer residue, trapped bubbles, wrinkles, cracks, altered surface termination or changes in oxygen stoichiometry.

Other release methods include laser lift-off, controlled spalling and mechanical exfoliation, but each depends on optical, fracture or layered-material properties that are not generally available in conventional complex oxides. Once released, the membranes are no longer constrained by their original substrates. They can be bent, strained, transferred and rotated, while each layer can be grown under different conditions.

In sequential transfer, two independently prepared membranes are aligned and stacked, allowing researchers to select composition, thickness, crystallographic orientation and strain state separately. A tear-and-stack method instead tears part of one parent membrane, picks it up with a supported stamp, rotates it and places it on the remaining film. This same-parent approach has reported angular-control accuracy of approximately 0.1 degrees and can reduce differences between layers arising from separate growth histories.

Both methods remain sensitive to flatness, cleanliness, trapped contaminants and transfer-induced deformation, and post-stacking anneals may be needed to recrystallize an interface that initially contains an amorphous gap. Twist creates a moiré potential because atoms in the two layers alternately approach different registries as one lattice is rotated across the other. In a symmetric homobilayer with no lattice mismatch, the characteristic wavelength grows as the twist angle becomes smaller, approximately following the inverse-square relation between wavelength and angle in radians.

Mismatch can also generate a moiré period without rotation. The local registry is more consequential in oxides than a simple geometric offset: it changes surface coordination, electrostatic boundary conditions, metal-oxygen orbital overlap, crystal-field splitting and lattice relaxation. Surface termination adds another variable because AO- and BO2-terminated perovskite faces expose different charges, dipoles and bonding environments.

Because oxide membranes retain finite thickness and three-dimensional bonding, the affected response may remain concentrated at the interface or extend into a layer depending on the material, thickness, screening and mechanical boundary conditions. Atomic-scale imaging has confirmed that the geometry works. High-resolution transmission electron microscopy revealed a 14.07-angstrom moiré periodicity in a 22.7-degree twisted strontium titanate bilayer, and later studies produced structural moiré patterns in twisted cobaltate and barium titanate membranes.

In the cobaltate series, the moiré-fringe spacing systematically decreased as twist angle increased, supporting a geometric origin rather than strain-driven reconstruction. Multislice electron ptychography reconstructed distinct AA-like and AB-like regions across an oxide interface, where columns remain aligned or are displaced into different atomic neighborhoods. These observations demonstrate coherent structural superlattices, but they do not by themselves establish flat-band formation or a correlated electronic phase.

They instead provide the microscopic template against which electronic and ferroic responses must be compared. Reconstructed ferroic behavior has been observed in both ferroelectric and nominally paraelectric oxide systems. In twisted barium titanate, titanium-displacement maps revealed periodic vortex-antivortex textures at twist angles of 10.4 and 3 degrees.

The vortex size and spacing followed the moiré modulation, indicating that relative rotation can control the lateral organization of polarization. Calculations indicate that spatially varying stacking energetics drive different lattice relaxations, electrostatic conditions and polarization preferences in neighboring regions of the moiré cell. Strain gradients can further couple to polarization through the flexoelectric effect.

Multislice ptychography also resolved vortex-like polarization in nominally paraelectric strontium titanate, showing that the phenomenon is not restricted to materials with a bulk ferroelectric ground state. The stabilization of these textures remains material-specific because it depends on stacking energetics, polarization-gradient energy, dielectric or ferroelectric susceptibility, layer thickness and interfacial conditions. Electronic reconstruction has likewise been measured, although the strongest proposed effects remain theoretical.

Spatially resolved electron-energy-loss spectroscopy of a 10.4-degree twisted strontium titanate bilayer found that strontium- and titanium-centered coincidence sites retained bulk-like titanium electronic character, while oxygen and non-coincidence sites displayed reduced valence and altered crystal-field splitting. The response was concentrated near the interface rather than extending through the entire bilayer, providing experimental evidence for registry-dependent charge disproportionation. Density-functional calculations predict that these regions can host localized electronic states and a quasi-flat valence band, while twisted barium titanate is predicted to reach a bandwidth of approximately 5.5 meV near a calculated magic angle of 18.92 degrees.

Direct confirmation of those flat bands and associated correlated phases remains outstanding; narrow calculated bands alone do not establish an interaction-driven magnetic, insulating or other collective state. The platform has also crossed beyond oxide-on-oxide contacts. Experiments with freestanding strontium titanate and monolayer tungsten disulfide produced a hybrid moiré superlattice whose period varied systematically with twist angle.

Differential reflectance and photoluminescence revealed additional excitonic resonances below the ordinary tungsten disulfide A exciton, which became more pronounced and shifted to lower energy as the twist angle approached zero. Their temperature robustness, spatial uniformity and valley polarization supported their assignment to K-valley neutral excitons confined within a periodic moiré potential rather than to defect-related states. Continuum modeling estimated a moiré-potential depth of approximately 50 plus or minus 10 meV, while first-principles calculations attributed it mainly to stacking-dependent dipoles at the polar strontium titanate surface, which produced a calculated bandgap modulation of about 70 meV across the moiré supercell.

The hybrid interface also displayed twist-dependent second-harmonic generation with 60-degree periodicity while retaining the sixfold symmetry of tungsten disulfide. Major obstacles remain before oxide twistronics becomes a predictive or scalable technology. Membranes can contain wrinkles, bubbles, spatially variable twist angles and nanometer-scale gaps, so nominally identical interfaces may have very different atomic contact.

Such imperfections can imitate or obscure intrinsic moiré reconstruction by changing strain, coordination, charge transfer and ferroic responses. Imaging also depends strongly on technique: high-angle annular dark-field microscopy emphasizes projected atomic-column intensity, whereas multislice ptychography reconstructs electrostatic potential and can capture contributions from both layers. Functional claims will require correlated structural, spectroscopic, magnetic and transport measurements, together with aligned-bilayer, thickness and twist-angle controls.

Electron microscopy can establish registry, while piezoresponse, electrostatic-force, Kelvin-probe, near-field optical and magnetic probes could test functional responses whose periodicity matches the buried moiré structure. The review’s central takeaway is that freestanding membranes provide a route to quantum materials whose properties are engineered through both composition and geometry. The platform has already produced structural moiré patterns, polar textures, registry-dependent electronic reconstruction and hybrid excitonic minibands.

Predicted flat bands, interfacial d0 magnetism, topological magnetic textures and interaction-driven correlated phases have not yet been directly verified in twisted complex oxides. Small-angle twist may additionally create stacking domains and dislocation-like networks whose strain and defect chemistry could influence transport or magnetic exchange, but such magnetic textures remain unconfirmed. Turning the demonstrated possibilities into robust discoveries will require larger-area crack-free membranes, uniform atomic contact, controlled terminations and oxygen stoichiometry, precise angle control, and probes capable of linking individual atomic registries to the electronic and magnetic states they generate.

Subject of Research: Freestanding complex-oxide membranes for twisted and hybrid moiré quantum materials

Article Title: Designer Moiré Quantum Materials Enabled by Freestanding Oxide Membranes in Twisted and Hybrid Bilayers

Article References: Kaur, P., Rahul, & Yang, J.-C. (2026). Designer Moiré Quantum Materials Enabled by Freestanding Oxide Membranes in Twisted and Hybrid Bilayers. Advanced Science, 13(58), Article e77727. https://doi.org/10.1002/advs.77727

Image Credits: AI Generated

DOI: 10.1002/advs.77727

Keywords: freestanding oxide membranes, oxide twistronics, moiré superlattices, complex oxides, strontium titanate, barium titanate, polar vortices, charge disproportionation, flat bands, exciton minibands, tungsten disulfide, quantum materials