Natural Hyperbolic Materials

Updated 29 November 2025

Natural hyperbolic materials are crystalline systems with anisotropic dielectric tensors that support open hyperbolic isofrequency contours and extreme subwavelength confinement.
They encompass a range of classes—from van der Waals crystals and electrides to organic aggregates—with tunable hyperbolic windows spanning the infrared to ultraviolet while offering lower losses than metamaterials.
Key applications include hyperlensing, negative refraction, and enhanced spontaneous emission, facilitated by tailored electronic, phononic, and excitonic mechanisms.

Natural hyperbolic materials (NHMs) are crystalline or molecular systems in which the principal components of the dielectric tensor change sign relative to one another within certain spectral bands, yielding an "indefinite" permittivity. This property enables the propagation of extraordinary electromagnetic modes with highly directional, open (hyperbolic) isofrequency contours in momentum space, supporting unbounded wavevectors and extreme subwavelength confinement. Traditionally, hyperbolic photonic platforms relied on capacitive metamaterial stack engineering or natural layered materials with strong structural anisotropy. Recent research has established a much broader landscape of NHMs, encompassing van der Waals crystals, electrides, hexagonal boron nitride, two-dimensional talc, organic J-aggregates, and many more, with tunable hyperbolic windows ranging from the infrared to the ultraviolet. These systems exhibit sharply reduced losses compared to artificial hyperbolic metamaterials and enable essential functionalities such as hyperlensing, negative refraction, Purcell-enhanced emission, valley quantum interference, and nanoscale polaritonic circuitry.

1. Permittivity Tensor, Hyperbolicity Criteria, and Dispersion Relations

NHMs are characterized by a frequency-dependent permittivity tensor, typically of the form

$\epsilon(\omega) = \text{diag}[\epsilon_x(\omega),\, \epsilon_y(\omega),\, \epsilon_z(\omega)]$

in the principal-axis basis. Hyperbolic dispersion occurs when the real parts of any two tensor components have opposite sign at a given frequency, i.e.,

$\mathrm{Re}\,\epsilon_i(\omega) \cdot \mathrm{Re}\,\epsilon_j(\omega) < 0 \quad (i \neq j)$

This hyperbolicity can be:

Type I: One component negative, the others positive (e.g., $\epsilon_\perp<0$ , $\epsilon_\parallel>0$ ),
Type II: One component positive, the others negative (e.g., $\epsilon_\perp>0$ , $\epsilon_\parallel<0$ ).

In uniaxial systems, the bulk extraordinary-wave dispersion is given by

$\frac{k_x^2 + k_y^2}{\epsilon_\perp(\omega)} + \frac{k_z^2}{\epsilon_\parallel(\omega)} = \left(\frac{\omega}{c}\right)^2$

and the 2D isofrequency contour in planar systems (for TM modes) is

$\frac{k_x^2}{\epsilon_y(\omega)} + \frac{k_y^2}{\epsilon_x(\omega)} = \left(\frac{\omega}{c}\right)^2$

These contours are closed ellipses for conventional media and open hyperbolas for NHMs, facilitating high-k states and deeply sub-diffractional confinement (Caldwell et al., 2014, Venturi et al., 24 May 2024, Jia et al., 2022).

2. Classes of Natural Hyperbolic Materials: Structural and Electronic Mechanisms

NHMs can be categorized by their crystal structure, electronic bonding, and mechanism of dielectric anisotropy:

Van der Waals Layered Crystals and Biaxial Media: Materials such as $\alpha$ -MoO $_3$ (Zheng et al., 2018, Bapat et al., 2021, Sahoo et al., 2021), MoOCl $_2$ (Venturi et al., 24 May 2024), WTe $_2$ (Wang et al., 2020), hBN (Caldwell et al., 2014, Dai et al., 2015), and 2D talc (Feres et al., 28 Jan 2025). These systems exhibit pronounced in-plane anisotropy, often governed by polarized phonon or plasmon excitations. Biaxial permittivity enables multiple distinct hyperbolic windows, each with different propagation directions and bandwidths.
Electrides and Non-cubic Ionic Solids: Non-cubic electrides, such as hcp Be, sh Mg, hp4 Na, TiH, Sc $_2$ Sb, Be $_2$ Zr, Ba $_3$ LiN, and layered/2D electrides (Sc $_5$ Cl $_8$ , Ca $_2$ Cu, Y $_2$ Cl $_3) [2511.17859], [1702.05602], feature interstitial quasi-atomic (ISQ) electrons spatially localized in lattice voids. This results in exceptionally flat bands (large effective mass), low plasma frequencies ($ \omega_p\sim0.1-2 $eV), and strong directional anisotropy, allowing hyperbolic windows in both metallic and semiconducting regimes.</li> <li><strong>Layered Hexagonal Compounds:</strong> Materials with P6/mmm and P6$ _3 $/mmc symmetry (Li$ _3 $N, Na$ _3 $N, Li$ _2 $KN, etc.), where interband transitions driven by symmetry-selective orbitals generate hyperbolic windows tunable by strain, doping, and alloying (<a href="/papers/2101.05262" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Ebrahimian et al., 2021</a>).</li> <li><strong>Organic J-aggregates and Semiconductors:</strong> Type-II hyperbolic dispersion realized in organic TDBC J-aggregate films via anisotropic exciton resonances; out-of-plane permittivity remains positive while in-plane permittivity becomes strongly negative over the exciton band (<a href="/papers/2506.07718" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Thomas et al., 9 Jun 2025</a>).</li> <li><strong>Mid-IR and Deep-UV Materials:</strong> hBN in the mid-IR Reststrahlen bands is a canonical NHM for volume-confined polaritons (<a href="/papers/1404.0494" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Caldwell et al., 2014</a>, <a href="/papers/1502.04094" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Dai et al., 2015</a>), and even in the deep-UV, hBN supports type-II hyperbolic exciton polaritons induced by giant anisotropic excitonic oscillators (<a href="/papers/2507.13271" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Choi et al., 17 Jul 2025</a>). MoTe$ _2$ and other TMDCs achieve hyperbolicity over broad visible-UV windows due to strong excitonic effects (<a href="/papers/2004.10870" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Edalati-Boostan et al., 2020</a>).</li> </ul> <h2 class='paper-heading' id='hyperbolic-windows-quantitative-spectral-ranges-and-quality-factors'>3. Hyperbolic Windows: Quantitative Spectral Ranges and Quality Factors</h2> <p>First-principles <a href="https://www.emergentmind.com/topics/direct-fine-tuning-dft" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">DFT</a>+GW+BSE calculations and spectroscopic ellipsometry establish the explicit spectral bandwidths for hyperbolic response in each material:</p> <div class='overflow-x-auto max-w-full my-4'><table class='table border-collapse w-full' style='table-layout: fixed'><thead><tr> <th>Material / Class</th> <th>Hyperbolic window (eV)</th> <th>Mechanism</th> <th>Q-factor</th> </tr> </thead><tbody><tr> <td>hBN (mid-IR, DUV)</td> <td>760–825 cm⁻¹ (Type I)<br\>1370–1610 cm⁻¹ (Type II)<br\>6.17–7.56 (Type II/DUV)</td> <td>Phonons / Excitons</td> <td>66–283</td> </tr> <tr> <td>$\alpha $-MoO$ _3 $</td> <td>545–850 cm⁻¹, 820–972 cm⁻¹ (RB1, RB2)</td> <td>Phonon-anisotropy</td> <td>up to 87 (confinement)</td> </tr> <tr> <td>MoOCl$ _2$</td> <td>0.4–2.5 eV (500–3000 nm)</td> <td>Plasmonic Drude-Lorentz</td> <td>>30</td> </tr> <tr> <td>Electr ides</td> <td>0.1–2 eV, up to 7.4 eV (pressure)</td> <td>Interstitial electrons</td> <td>–</td> </tr> <tr> <td>TDBC J-aggregate</td> <td>2.11–2.43 eV (Type II)</td> <td>Exciton resonance</td> <td>–</td> </tr> <tr> <td>Talc (2D)</td> <td>948–1002 cm⁻¹ (Type I)<br\>1011–1041 cm⁻¹ (Type II)</td> <td>Phonon bands</td> <td>3.6–4.5</td> </tr> </tbody></table></div> <p>Q-factors in hBN reach 283, with confinement ratios $\lambda_0/\lambda_z $up to 86 (<a href="/papers/1404.0494" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Caldwell et al., 2014</a>, <a href="/papers/2507.13271" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Choi et al., 17 Jul 2025</a>). Electr ides and layered hexagonal crystals achieve low losses ($ \mathrm{Im}\,\epsilon \lesssim 0.1 $) over broad bands spanning the infrared to the ultraviolet (<a href="/papers/2511.17859" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Hao et al., 22 Nov 2025</a>, <a href="/papers/2101.05262" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Ebrahimian et al., 2021</a>).</p> <h2 class='paper-heading' id='mechanisms-for-hyperbolic-dispersion'>4. Mechanisms for Hyperbolic Dispersion</h2> <p>Fundamental mechanisms producing hyperbolicity include:</p> <ul> <li><strong>Directional Drude Response:</strong> In metallic and semimetallic NHMs (WTe$ _2 $, electrides), effective <a href="https://www.emergentmind.com/topics/multi-agent-systems-mass" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">mass</a> anisotropy and reduced carrier density (flat bands, mobile ISQ electrons) push the plasma frequency to low energies and split along different axes, enabling broad hyperbolic windows (<a href="/papers/2003.02398" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Wang et al., 2020</a>, <a href="/papers/2511.17859" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Hao et al., 22 Nov 2025</a>).</li> <li><strong>Interband Resonances:</strong> Anisotropic interband transitions (semiconducting electrides, hexagonal layered compounds) produce separated absorption peaks, giving rise to sharp changes in$ \mathrm{Re}\,\epsilon $and opening interband-driven hyperbolic gaps (<a href="/papers/2101.05262" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Ebrahimian et al., 2021</a>, <a href="/papers/2511.17859" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Hao et al., 22 Nov 2025</a>).</li> <li><strong>Anisotropic Phonon Modes:</strong> Infrared-active phonon bands localized along distinct crystal axes in hBN,$ \alpha $-MoO$ _3 $, and talc, generate multiple Reststrahlen windows with hyperbolic dispersion accessible through spectroscopic techniques (<a href="/papers/1404.0494" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Caldwell et al., 2014</a>, <a href="/papers/2501.17340" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Feres et al., 28 Jan 2025</a>).</li> <li><strong>Excitonic Effects:</strong> Strongly bound, anisotropic excitons in organic J-aggregates and TMDCs (MoTe$ _2 $) introduce negative permittivity within narrow bands, realizing type-II hyperbolicity at visible and ultraviolet frequencies (<a href="/papers/2506.07718" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Thomas et al., 9 Jun 2025</a>, <a href="/papers/2004.10870" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Edalati-Boostan et al., 2020</a>, <a href="/papers/2507.13271" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Choi et al., 17 Jul 2025</a>).</li> <li><strong>Shear and Off-diagonal Tensor Modes:</strong> Crystals with skewed or nondiagonal permittivity tensors (e.g., monoclinic$ \beta $-Ga$ _2 $O$ _3 $) produce shear polariton modes with rotated and asymmetric k-space hyperbolic contours (<a href="/papers/2210.12341" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Jia et al., 2022</a>).</li> </ul> <h2 class='paper-heading' id='tunability-and-design-principles'>5. Tunability and Design Principles</h2> <p>NHMs exhibit multifaceted tunability and robust design rules, as elucidated by DFT and experimental studies (<a href="/papers/2511.17859" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Hao et al., 22 Nov 2025</a>, <a href="/papers/1702.05602" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Guan et al., 2017</a>, <a href="/papers/2101.05262" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Ebrahimian et al., 2021</a>):</p> <ul> <li><strong>Structural Anisotropy is Not Necessary:</strong> Charge localization in ISQ sites or layered molecular orbitals generates sufficient anisotropy; cubic symmetry is not required.</li> <li><strong>Electronic Structure Engineering:</strong> Band structure calculations (flat bands, exciton peaks, interband JDOS) identify materials with strong anisotropic optical response. Electron localization function (ELF$ > $0.5) in voids correlates with hyperbolic behavior.</li> <li><strong>Strain, Doping, Alloying:</strong> Uniaxial strain, chemical substitution, or carrier doping can shift the position and breadth of hyperbolic windows, enabling dynamic control and spectral selection (<a href="/papers/2101.05262" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Ebrahimian et al., 2021</a>, <a href="/papers/1702.05602" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Guan et al., 2017</a>).</li> <li><strong>Pressure Tuning:</strong> In high-pressure electrides, plasma-frequency splittings extend hyperbolic response to multi-eV windows (<a href="/papers/2511.17859" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Hao et al., 22 Nov 2025</a>).</li> <li><strong>Dimensionality:</strong> Both bulk and atomically thin (2D) forms can support NHM phases as long as the principal permittivities attain opposite sign. Monolayer, bilayer, and bulk TMDCs display blue-shifted, thickness-dependent windows (<a href="/papers/2004.10870" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Edalati-Boostan et al., 2020</a>).</li> </ul> <h2 class='paper-heading' id='functionalities-photonic-quantum-and-optoelectronic-applications'>6. Functionalities: Photonic, Quantum, and Optoelectronic Applications</h2> <p>NHMs underpin an array of photonic and quantum functionalities demonstrable on chip or in the far field:</p> <ul> <li><strong>Hyperlensing and Super-Resolution Imaging:</strong> Open hyperbolic contours support high-k polariton modes, surpassing diffraction limits in mid-IR (hBN,$ \alpha $-MoO$ _3 $, talc (<a href="/papers/1404.0494" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Caldwell et al., 2014</a>, <a href="/papers/2501.17340" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Feres et al., 28 Jan 2025</a>, <a href="/papers/2210.12341" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Jia et al., 2022</a>)), visible/UV (MoTe$ _2 $, MoOCl$ _2 $(<a href="/papers/2004.10870" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Edalati-Boostan et al., 2020</a>, <a href="/papers/2405.15420" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Venturi et al., 24 May 2024</a>)), and DUV (hBN (<a href="/papers/2507.13271" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Choi et al., 17 Jul 2025</a>)).</li> <li><strong>Negative Refraction and Beam Steering:</strong> NHMs enable all-angle negative refraction (Ca$ _2 $N (<a href="/papers/1702.05602" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Guan et al., 2017</a>)), dynamic buy-in via gating in 2D systems (<a href="/papers/1509.04383" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Nemilentsau et al., 2015</a>), and valley quantum interference control using angle-dependent polariton hybridization (<a href="/papers/2110.07526" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Bapat et al., 2021</a>).</li> <li><strong>Purcell Factor Enhancement and PDOS Amplification:</strong> Extreme subwavelength confinement yields photonic density of states enhancements up to 100×, Purcell factor boosts ($ >10^2 $), and slow group velocity in deep-UV hBN (<a href="/papers/2507.13271" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Choi et al., 17 Jul 2025</a>), visible MoOCl$ _2 $(<a href="/papers/2405.15420" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Venturi et al., 24 May 2024</a>).</li> <li><strong>Thermal Emission and Near-Field Radiative Transfer:</strong> Hyperbolic polariton modes in 2D NHMs provide anomalous thermal transfer rates, exceeding blackbody far beyond classical limits (<a href="/papers/2210.12341" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Jia et al., 2022</a>).</li> <li><strong>Mid-IR Polarizers and Quantum Emitters:</strong> Wafer-scale thin films of$ \alpha $-MoO$ _3 $act as thermally stable, lithography-free polarizers with extinction ratios$ >10 $dB up to 140 °C (<a href="/papers/2108.08510" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Sahoo et al., 2021</a>, <a href="/papers/1809.03432" title="" rel="nofollow" data-turbo="false" class="assistant-link" x-data x-tooltip.raw="">Zheng et al., 2018</a>).</li> <li><strong>Tunable and Topological Photonic Platforms:</strong> Gate-tunable MoOCl$ _2 $, WTe$ _2 $and hybrid van der Waals heterostructures (graphene/hBN, graphene/$ \alpha $-MoO$ _3$) support real-time steering and mode-selective emission enhancement (Dai et al., 2015, Bapat et al., 2021, Nemilentsau et al., 2015).

7. Future Directions and Challenges

Emergent research on NHMs addresses key challenges and prospects:

Spectral Range Expansion: Current NHMs operate from mid-IR to DUV; targeted synthesis and computational screening may enable THz, telecom, and multi-octave hyperbolic regions (Jia et al., 2022, Choi et al., 17 Jul 2025).
Monolayer vs Bulk Hyperbolicity: Thickness dependence, edge states, and substrate effects on hyperbolicity remain under active investigation (e.g., TMDCs, ZrSiSe) (Jia et al., 2022, Wang et al., 2020).
Active Modulation: Combining electrostatic gating, selective chemical intercalation, phase-change layers, and nanostructure engineering enables dynamic control over mode directionality, bandwidth, and coherence (Bapat et al., 2021, Nemilentsau et al., 2015).
Large-Scale Integration: CVD and MBE growth of NHMs at wafer scale is a prerequisite for device applications; interface effects (band bending, Schottky barriers) require systematic paper (Jia et al., 2022, Sahoo et al., 2021).
Unconventional Crystal Symmetries: Exploration of monoclinic, triclinic, and quasicrystalline NHMs may yield novel shear, magnetoelectric, and nonreciprocal polaritonic phenomena (Jia et al., 2022).
Topological Photonics: Experimentally confirmed phase singularities (optical vortices, winding number $\pm2\pi$ ) in HSEP and HSPhP modes reveal unique topological invariants in NHMs, ripe for development in singular-phase sensors and robust on-chip photonic circuits (Thomas et al., 9 Jun 2025).