Concept · physics
Two-dimensional materials
8 studies1 discoveryEvidence last moved Sep 27, 2026
Two-dimensional materials are atom-thin crystals such as graphene, hBN and transition-metal dichalcogenides that can be stacked with chosen twist angles. The evidence here covers moiré patterns and flat bands, tools to set the twist angle, twist-dependent optics and excitons, and a graphene photodetector.
The twist angle is a new control knob that changes band structure without changing chemistry, and students often assume the 'magic' only lives in twisted graphene. These papers show the same ideas in excitons, nonlinear optics and even photonic crystals, and show how hard it is to know the real twist angle in a sample.
Studies
8
Findings
6
9 supporting · 0 challenging · 3 qualifying citations
Open tensions
2
Latest change
Concept page published
Two-dimensional materials
Currently
What we know
- Two moirés can combine into a longer one.
- Twist angle can be set in situ, though here irreversibly.
- Flat bands pack many states into a narrow energy, in electrons or light.
- Matching a twist-set resonance to the light boosts optical effects.
- 2D excitons can be confined like quantum dots by electrostatics.
Largest unresolved question
Twist angle is often inferred rather than measured: angles were fitted from transport, read from Raman features, or taken from simulations, so region-to-region comparisons carry uncertainty that the individual papers acknowledge.
Common misconceptions
Flat bands in 2D materials only occur at the graphene magic angle.
A flat band appeared in untwisted bilayer graphene on SiC because of substrate-induced asymmetries, and twisted photonic slabs show flat bands at their own angle.
A flat band means superconductivity has been observed.
The bilayer flat band paper explicitly did not observe superconductivity; the band would need to be gated to the Fermi level first.
Stacking two layers only doubles the signal.
Most twisted bilayer graphene regions gave about 3.5-4 times monolayer third-harmonic signal, but near-critical twist regions exceeded 6 times, and a gated one reached about 60 times.
Related
Claim ledger
What the evidence shows
Drawn from 8 studies in this library. Mix labels say which citation roles are present; they are not a strength score. Supports means evidence for a finding; Challenges means evidence against a stated position; Qualifies marks scope.
Two moirés can combine into a longer one.
Aligning graphene with hBN on both sides creates composite 'super-moiré' periods: transport showed resistance peaks matching periods near 35 nm, beyond the 14.0 and 15.3 nm single moirés, confirmed by Brown-Zak oscillations, and a mechanical rotation technique produced the same two moiré wavelengths (14.7 and 14.0 nm) plus super-moiré signs.
- What happens when graphene is aligned to two hBN layers at once?
- Can you twist layers after a 2D stack is built?
Study Role Design N Population Outcome What happens when graphene is aligned to two hBN layers at once? Supports OtherFabricated hBN/graphene/hBN stacks with both hBN layers aligned, characterised by AFM and Raman, and measured by magnetotransport (resistance peaks, Landau fans, Brown-Zak oscillations). Several devices; the angle-dependence plot uses four samples. Double-aligned graphene/hexagonal boron nitride heterostructure devices Positions of resistance peaks (secondary Dirac points), inferred moiré periodicities, Brown-Zak oscillations, Raman 2D peak width Can you twist layers after a 2D stack is built? Supports OtherLab experiment: a PMMA patch patterned on the top hBN flake was pushed with a PDMS hemisphere to rotate layers of assembled heterostructures, followed by Raman mapping, AFM and low-temperature transport measurements in magnetic field. No sample count; main results come from two heterostructures (sample 1 and sample 2), with transport focused on the bilayer region of sample 1. Encapsulated hBN/graphene/hBN van der Waals heterostructures (monolayer and bilayer graphene regions) Ability to rotate and lock layers into alignment, Raman 2D and G peak changes, and transport signatures of moiré superlattices (secondary Dirac points, Hofstadter spectra, Brown-Zak oscillations) Twist angle can be set in situ, though here irreversibly.
Layers can be rotated after assembly: a PMMA patch pushed by a PDMS tip rotated hBN smoothly while misaligned and locked once aligned, raising the Raman 2D width from 17 to 55 cm^-1.
Flat bands pack many states into a narrow energy, in electrons or light.
Flat bands can arise from stacking and substrate asymmetries: ARPES on bilayer graphene on SiC found a band varying by no more than 2 meV over ±0.017 1/Å, and simulations of twisted photonic crystal slabs found zero group velocity at a 1.89° twist.
- Can twisted photonic crystals mimic magic-angle graphene for light?— Simulation only.
- Can bilayer graphene host a band with almost no dispersion?— Band sits below the Fermi level; essentially one sample.
Study Role Design N Population Outcome Can bilayer graphene host a band with almost no dispersion? Supports OtherSynchrotron ARPES measurements of epitaxial graphene on 6H-SiC, combined with density functional theory and a 4x4 tight-binding Hamiltonian fitted to the data. No participant count; results come mainly from one sample measured at several stages, with the authors stating other samples gave the same results. Monolayer and bilayer graphene grown on nitrogen-doped 6H-SiC(0001) Band dispersion and photoemission intensity near the K point; calculated density of states and band shape Can twisted photonic crystals mimic magic-angle graphene for light? Supports Computational / modelling3D finite-element (COMSOL) band-structure, eigenmode and Q-factor simulations of twisted bilayer photonic crystal slabs, compared with a plane-wave continuum model No sample; simulations at a set of commensurate twist angles Simulated silicon honeycomb photonic crystal slabs coupled through a low-index tunnelling membrane Photonic band structure, group velocity at the K point, mode localization and quality factors versus twist angle Matching a twist-set resonance to the light boosts optical effects.
Twist angle tunes optical responses: twisted bilayer graphene regions near about 12° gave third-harmonic signals above 6 times monolayer (versus about 3.5-4 elsewhere), and twisted TMD trilayers trapped moiré excitons more deeply than bilayers, with 3-5 times brighter, narrower (3.2 vs 12.5 meV) peaks lasting to about 50 K instead of 30 K.
- Does twisting two graphene layers boost frequency tripling of light?
- Does adding a third twisted layer trap excitons more tightly?
Study Role Design N Population Outcome Does twisting two graphene layers boost frequency tripling of light? Supports OtherLab optical measurements (Raman, reflection contrast, THG mapping, ion-gel gating) on CVD-grown twisted bilayer graphene plus continuum-model band calculations. N=63 · 63 twisted bilayer graphene regions characterised by Raman and THG; the gating experiment used one device with several regions. CVD-grown monolayer and twisted bilayer graphene on SiO2/Si substrates THG intensity relative to monolayer graphene, as a function of twist angle and gate voltage Does adding a third twisted layer trap excitons more tightly? Supports OtherMicro-photoluminescence spectroscopy (temperature, power, polarization and magnetic-field dependence) on hBN-encapsulated twisted TMD bilayer and trilayer stacks, with DFT support Individual devices with twist angles of about 3 and 1.5 degrees; no sample count hBN-encapsulated WSe2/WS2 heterobilayers and WSe2/WS2/WSe2 heterotrilayers Moiré exciton peak linewidth, intensity, thermal quenching temperature, and power dependence 2D excitons can be confined like quantum dots by electrostatics.
Gate electrodes can sculpt exciton traps in a monolayer semiconductor: patterned holes and bow ties in MoSe2 devices gave discrete narrow levels that shifted with bias, and three dots were tuned into energy degeneracy.
Graphene's gapless bands give very wide spectral coverage.
Graphene devices already work as fast broadband photodetectors: a gold-patched graphene stripe detector gave 0.6 to 11.5 A/W from 0.8 to 20 µm with no roll-off to 50 GHz.
- Can one graphene detector be sensitive, fast and broadband at once?— Single device design; variability not reported.
Debates
Tensions and limits
Some items are genuine disagreements on the same question. Others mark different assays, populations, or outcomes.
Twist angle is often inferred rather than measured: angles were fitted from transport, read from Raman features, or taken from simulations, so region-to-region comparisons carry uncertainty that the individual papers acknowledge.
Twist angle is often inferred rather than measured: angles were fitted from transport, read from Raman features, or taken from simulations, so region-to-region comparisons carry uncertainty that the individual papers acknowledge.
- What happens when graphene is aligned to two hBN layers at once?
- Does twisting two graphene layers boost frequency tripling of light?
- Does adding a third twisted layer trap excitons more tightly?
Study Role Design N Population Outcome What happens when graphene is aligned to two hBN layers at once? Supports OtherFabricated hBN/graphene/hBN stacks with both hBN layers aligned, characterised by AFM and Raman, and measured by magnetotransport (resistance peaks, Landau fans, Brown-Zak oscillations). Several devices; the angle-dependence plot uses four samples. Double-aligned graphene/hexagonal boron nitride heterostructure devices Positions of resistance peaks (secondary Dirac points), inferred moiré periodicities, Brown-Zak oscillations, Raman 2D peak width Does twisting two graphene layers boost frequency tripling of light? Supports OtherLab optical measurements (Raman, reflection contrast, THG mapping, ion-gel gating) on CVD-grown twisted bilayer graphene plus continuum-model band calculations. N=63 · 63 twisted bilayer graphene regions characterised by Raman and THG; the gating experiment used one device with several regions. CVD-grown monolayer and twisted bilayer graphene on SiO2/Si substrates THG intensity relative to monolayer graphene, as a function of twist angle and gate voltage Does adding a third twisted layer trap excitons more tightly? Supports OtherMicro-photoluminescence spectroscopy (temperature, power, polarization and magnetic-field dependence) on hBN-encapsulated twisted TMD bilayer and trilayer stacks, with DFT support Individual devices with twist angles of about 3 and 1.5 degrees; no sample count hBN-encapsulated WSe2/WS2 heterobilayers and WSe2/WS2/WSe2 heterotrilayers Moiré exciton peak linewidth, intensity, thermal quenching temperature, and power dependence
PaperFren reads this as a limit on how far one study travels — different assays, populations, or outcomes — not a forced fight between papers.
Graphene analogies stop at band structure: the photonic crystal study reproduces magic-angle flat bands but not electron interactions, and the bilayer flat band sits too far below the Fermi level to test superconductivity.
Graphene analogies stop at band structure: the photonic crystal study reproduces magic-angle flat bands but not electron interactions, and the bilayer flat band sits too far below the Fermi level to test superconductivity.
- Can twisted photonic crystals mimic magic-angle graphene for light?
- Can bilayer graphene host a band with almost no dispersion?
Study Role Design N Population Outcome Can twisted photonic crystals mimic magic-angle graphene for light? Supports Computational / modelling3D finite-element (COMSOL) band-structure, eigenmode and Q-factor simulations of twisted bilayer photonic crystal slabs, compared with a plane-wave continuum model No sample; simulations at a set of commensurate twist angles Simulated silicon honeycomb photonic crystal slabs coupled through a low-index tunnelling membrane Photonic band structure, group velocity at the K point, mode localization and quality factors versus twist angle Can bilayer graphene host a band with almost no dispersion? Supports OtherSynchrotron ARPES measurements of epitaxial graphene on 6H-SiC, combined with density functional theory and a 4x4 tight-binding Hamiltonian fitted to the data. No participant count; results come mainly from one sample measured at several stages, with the authors stating other samples gave the same results. Monolayer and bilayer graphene grown on nitrogen-doped 6H-SiC(0001) Band dispersion and photoemission intensity near the K point; calculated density of states and band shape
PaperFren reads this as a limit on how far one study travels — different assays, populations, or outcomes — not a forced fight between papers.
Timeline
How understanding moved
Study years are when the paper was published. Evidence edits are dated changes to this page's claims. Explanations are when PaperFren added a Discovery — not a claim that the science happened that day.
2026
- Aluminum a few atoms thick superconducts at about three times its bulk temperature
Concept page published
Two-dimensional materials
Change log
What changed
Dated edits to this page's evidence: studies added or removed from a claim, claims added or withdrawn, and new explanations tagged here. Rewordings are not listed.
- Concept page published
- Aluminum a few atoms thick superconducts at about three times its bulk temperatureEvidence: Preliminary
Papers
8 studies in this library bear on Two-dimensional materials, ordered by citations.
- Can one graphene detector be sensitive, fast and broadband at once?
Narrow graphene stripes flanked by gold patches let one detector respond strongly from visible light to the far infrared while staying extremely fast.
- Can you twist layers after a 2D stack is built?
A small polymer patch on the top flake lets a soft stamp rotate layers inside an already assembled graphene–boron nitride stack until they lock into perfect alignment, producing two coexisting moiré patterns.
- What happens when graphene is aligned to two hBN layers at once?
Aligning graphene with both its top and bottom boron nitride layers creates combined 'super-moiré' patterns with long periods that change the electrons' energy spectrum at low energies.
- Can twisted photonic crystals mimic magic-angle graphene for light?
Twisting two stacked silicon photonic crystal slabs to a specific small angle flattens the light bands, slowing light almost to a stop and trapping it in small regions, just as magic-angle graphene does for electrons.
- Does adding a third twisted layer trap excitons more tightly?
Sandwiching a twisted WS2 sheet between two WSe2 sheets created two overlapping moiré patterns that trapped excitons more tightly, giving sharper, brighter light emission that survived to higher temperatures.
- Can bilayer graphene host a band with almost no dispersion?
Bilayer graphene on silicon carbide contains a band so flat that its energy barely changes with momentum, packing a very high density of electron states into a narrow energy range.
- Does twisting two graphene layers boost frequency tripling of light?
Bilayer graphene twisted to about 12 degrees converts infrared light into triple-frequency green light far more strongly than ordinary graphene, because an electronic resonance lines up with three photons' energy.
- Can electric gates trap excitons in custom shapes in 2D crystals?
Patterned electrodes can trap light-emitting excitons in a one-atom-thick semiconductor into rings and tunable quantum dots, and several dots can be tuned to the same energy.
Compare studies
Select 2–10 studies. Design and N are labels, not a ranking.
Nothing selected yet.
Questions
What is still open
Twist angle is often inferred rather than measured: angles were fitted from transport, read from Raman features, or taken from simulations, so region-to-region comparisons carry uncertainty that the individual papers acknowledge.
Graphene analogies stop at band structure: the photonic crystal study reproduces magic-angle flat bands but not electron interactions, and the bilayer flat band sits too far below the Fermi level to test superconductivity.
Ask PaperFren about Two-dimensional materials
Study this conceptflashcards and short-answer questions
Explain how a super-moiré lattice forms and how experiments detect it.
When graphene is aligned with hBN on both sides, each interface makes its own moiré, with periods near 14-15 nm. Interference between the two gives a longer composite period, such as about 35 nm. Transport detects it as extra resistance peaks at low carrier densities and as Brown-Zak oscillations at high temperature. An in situ rotation experiment observed two moiré wavelengths and signs of the same composite pattern.
Why did twisted TMD trilayers give sharper moiré exciton lines than bilayers, and what remains unproven?
The trilayer peaks were 3-5 times brighter, narrower (about 3.2 meV vs 12.5 meV) and survived to about 50 K, consistent with a deeper trapping potential from overlapping moiré patterns. The deeper-potential explanation is inferred from spectra, not imaged. Only a few hand-made devices were compared, and single-photon emission was not tested with photon correlations.