Moiré, Low-Dimensional, and Engineered Quantum Matter

An electron in a thin crystal can behave differently from one in a bulk solid: confinement changes its allowed motion, nearby layers change its electric environment, and twisting a second layer can create a much longer spatial pattern. This chapter explains how these controls change quantum states and their measurable properties.

Start with the material or control that interests you. The first pages develop dimensionality and constituent layers; the later pages explain moiré bands, collective phases, interfaces and artificial lattices. A narrow band can make interactions relatively important, but the resulting phase also depends on filling, symmetry, temperature and disorder.

Helpful background. Band Theory Overview distinguishes dispersion, eigenstates and filling. Tight-Binding Models identifies the sites, orbitals and couplings retained by an effective lattice. How Quantum Matter Is Measured distinguishes a signal from the observable inferred through a forward model. Each branch supplies its own additional prerequisites.

Enter this chapter

Start from the control or the inference that is uncertain. A thin-layer magnet, a superconducting interface and a photonic lattice need not involve moiré physics.

Your question	Start here
Which dimensions remain dynamically active?	Low-Dimensional Quantum Matter
What changes when a crystal is atomically thin?	Two-Dimensional Materials
Which perturbation changes a monolayer Dirac system?	Graphene
Which spin, valley and exciton labels apply?	Transition-Metal Dichalcogenides
What does assembling layers actually couple?	van der Waals Heterostructures
When does a registry pattern become a quantum lattice?	Moiré Superlattices
Which continuum model describes a twisted graphene bilayer?	Twisted Bilayer Graphene
What survives when dispersion is small?	Flat Bands
What causes an insulating feature at a calibrated filling?	Correlated Insulators in Moiré Systems
What establishes pairing and coherence?	Moiré Superconductivity
Is the claim about bands or an interacting topological state?	Moiré Topology
Which symmetry permits thin-layer order?	2D Magnets and Ferroelectrics
What is transferred across an interface?	Engineered Heterostructures
Does an array realize the intended dynamics and state?	Artificial Lattices and Designer Matter
How does a target response become a reproducible material?	Quantum Materials by Design

Keep geometry, Hamiltonian, state and measurement separate

A gate changes charge density, but may also change screening or the electric field between layers. Pressure can change both hopping and crystal structure. To interpret a control, first identify the term it changes in the Hamiltonian, then ask which state is occupied and which observable the experiment measures. A band diagram alone specifies neither the occupation nor the temperature.

For example, a tunneling gap concerns charged excitations, while an optical line can create a bound electron–hole pair. Ugeda and collaborators combined tunneling and optical measurements to separate these scales. Band Gaps and Excitons in Materials explain the distinction used in Exercise 2.

Dimension, constituent and interface

Begin with Low-Dimensional Quantum Matter if confinement or crossover is unfamiliar. Then choose the constituent material and its environment. The graphene page in this chapter explains substrate, strain and gate controls; Graphene and Dirac Materials supplies the honeycomb and Landau-level derivations.

Geim and Grigorieva describe assembling van der Waals layers. Stacking introduces new questions about alignment, charge transfer and hybridization. The heterostructure guide connects those questions to the properties of the individual layers.

From a moiré pattern to a phase question

Moiré Superlattices connects registry, cell area, minibands and filling. Bistritzer and MacDonald show how a continuum model produces graphene-bilayer moiré bands; Balents extends the description to smooth deformations. A periodic continuum approximation need not make the full atomic stack exactly commensurate.

Next use Flat Bands to examine band isolation and interaction matrix elements. The insulating, superconducting and topological pages then address different possible states. None follows from bandwidth alone. In particular, a fractional Hall response is not a fractional Chern number of an ordinary isolated Bloch band; Topological Order develops the many-body distinction.

Worked route: nominal integer filling in a nonuniform device

Consider an invented triangular moiré structure with a nominal physical period $L=14\,\mathrm{nm}$. A calibrated electron excess number density, relative to a stated neutrality reference, is $n=1.18\times10^{12}\,\mathrm{cm}^{-2}$. The geometric filling relation supplied by Moiré Superlattices is

$$A_{\mathrm M}=\frac{\sqrt3}{2}L^2, \qquad \nu=nA_{\mathrm M}.$$

Here $\nu$ counts excess electrons per physical moiré cell; it does not assume a spin or valley degeneracy. Since $1\,\mathrm{cm}^2=10^{14}\,\mathrm{nm}^2$, the density is $0.0118\,\mathrm{nm}^{-2}$. The nominal cell area is $169.74\,\mathrm{nm}^2$ and $\nu\simeq2.003$.

Suppose spatial metrology instead resolves slowly varying patches with periods between $13$ and $15\,\mathrm{nm}$. To isolate the geometric effect, assume the physical carrier density is approximately uniform and each patch admits a local triangular-cell description. The same conversion gives

$$\begin{aligned} L=13\,\mathrm{nm}:&\quad \nu\simeq1.727,\\ L=15\,\mathrm{nm}:&\quad \nu\simeq2.299. \end{aligned}$$

A resistance peak near the nominal filling two is therefore compatible with quite different local fillings. The example does not predict the actual density profile: electrostatic redistribution and reconstruction may change it. They require independent spatial and device information.

The next route depends on the missing premise. Use the stack and moiré articles for structural and filling calibration; Flat Bands for the active-band and flavor count; and Correlated Insulators for the insulating mechanism. Exactly two electrons per nominal cell would require about $1.1783\times10^{12}\,\mathrm{cm}^{-2}$, but improving that average calibration alone would not resolve spatial inhomogeneity. Neither a homogeneous Mott state nor a four-flavor active band follows from the nominal integer.

Interfaces, order and artificial lattices beyond moiré

2D Magnets and Ferroelectrics explains how symmetry, anisotropy and interaction range affect order. Engineered Heterostructures explains induced couplings and accompanying unwanted channels.

For an artificial array, matching a single-particle spectrum does not establish its particle statistics or many-body state. Exercise 3 gives a two-site example. Artificial Lattices and Designer Matter develops that comparison, and Quantum Materials by Design connects predicted responses with stability and fabrication. Dated assessments of candidate phases are collected in the Quantum Matter research guide.

Exercises

1. Equal bandwidth, different projection control

Two proposed active-band models have bandwidth $W=2\,\mathrm{meV}$ and intraband interaction scale $U=6\,\mathrm{meV}$. Their remote-subspace separations are $60\,\mathrm{meV}$ and $5\,\mathrm{meV}$, respectively. A separately estimated matrix-element scale coupling the retained and remote subspaces is $V=3\,\mathrm{meV}$ in each. Does equal $U/W$ make the two projections equally controlled?

Solution

Both ratios $U/W$ equal three. The additional mixing diagnostic is $V/\Delta_{\mathrm{remote}}$, which is $0.05$ for the first model and $0.6$ for the second. The first provides a small parameter for this particular off-block mixing test; the second does not. Equal intraband interaction ratios conceal that difference.

This comparison is not a universal error bound and does not choose either system’s phase. Other off-block processes, occupation, temperature, disorder and the required observable accuracy must also be assessed. Use Flat Bands for the active subspace and Hubbard Physics in Materials for the retained interaction model; the general elimination and its error logic belong to Effective Hamiltonians in Many-Body Systems. Do not substitute the intraband $U$ for a specified off-block matrix element without a model that relates them.

2. An unchanged optical line with changed bands

In a clean, low-excitation direct-gap benchmark, matched quasiparticle continuum thresholds in two dielectric environments are $1.95\,\mathrm{eV}$ and $1.80\,\mathrm{eV}$. The same assigned bound neutral optical transition remains at $1.70\,\mathrm{eV}$. Find the two binding energies. Does the unchanged optical line imply unchanged band edges?

Solution

For the specified matched transition, the binding is the continuum threshold minus the bound-pair energy. The bindings are $0.25\,\mathrm{eV}$ and $0.10\,\mathrm{eV}$. The $0.15\,\mathrm{eV}$ reduction in threshold is accompanied by the same reduction in binding, leaving the optical transition unchanged.

The optical line alone cannot separate these two shifts. The subtraction requires the same valley, momentum and transition assignment, calibrated probes and a justified bound-state interpretation. Use the constituent-material article for those labels, Excitons and Band Gaps for the excitation distinction, and Dielectric Response for screening by the environment. These numbers are synthetic, not data taken from the cited experiments.

3. One-particle agreement does not fix particle statistics

Two isolated two-mode platforms implement the same one-particle Hamiltonian $h=-J\sigma_x$, with $J>0$. At $t=\pi\hbar/(4J)$, a particle initially on site one has equal probabilities on the two sites. Does that calibration establish the same many-particle state? As a check, start with one particle on each site and compare identical noninteracting bosons with spin-polarized noninteracting fermions.

Solution

The calibration determines this one-particle mixing operation. It does not determine the Fock space, particle statistics, interactions or preparation. At the stated time, each creation operator is transformed with diagonal amplitude $1/\sqrt2$ and off-diagonal amplitude $i/\sqrt2$.

For identical bosons in the same internal state, the two amplitudes leading to one particle on each output cancel. The output probabilities are $P(1,1)=0$ and $P(2,0)=P(0,2)=1/2$. For spin-polarized fermions, double occupation of either single mode is forbidden and $P(1,1)=1$. A common one-particle Hamiltonian therefore supports different many-particle predictions.

The general symmetrization treatment supplies the state construction. Artificial Lattices supplies the implementation and calibration questions. A classical optical wave experiment can reproduce the one-particle mode equation without implementing this two-particle counting experiment; distinguish those claims explicitly. Distinguishability, interactions or loss would require a different calculation.

References

• L. Balents, “General continuum model for twisted bilayer graphene and arbitrary smooth deformations,” SciPost Physics 7, 048 (2019), doi:10.21468/SciPostPhys.7.4.048. Continuum descriptions and smooth structural deformations.
• R. Bistritzer and A. H. MacDonald, “Moiré bands in twisted double-layer graphene,” Proceedings of the National Academy of Sciences 108, 12233–12237 (2011), doi:10.1073/pnas.1108174108. The graphene-bilayer continuum miniband construction.
• A. K. Geim and I. V. Grigorieva, “Van der Waals heterostructures,” Nature 499, 419–425 (2013), doi:10.1038/nature12385. The layer-assembly and interface platform.
• M. M. Ugeda et al., “Giant bandgap renormalization and excitonic effects in a monolayer transition metal dichalcogenide semiconductor,” Nature Materials 13, 1091–1095 (2014), doi:10.1038/nmat4061. A joint optical and quasiparticle investigation distinguishing the gap and binding scales.