Ennea-Graphene: Novel 2D Carbon Allotrope

• Ennea-graphene is a two-dimensional carbon allotrope composed of 4-, 5-, 6-, and predominantly 9-membered rings, creating a porous network with nonagons as the key motif.
• Density Functional Theory calculations reveal near-zero-gap metallic behavior, dynamic and thermal stability at 300 K, and high in-plane stiffness for the material.
• Sodium decoration at the nonagonal centers enables reversible hydrogen uptake up to 8.80 wt% H2, positioning ennea-graphene as a promising hydrogen-storage substrate.

Searching arXiv for the specified papers to ground the article in current research. Ennea-graphene is a two-dimensional carbon allotrope introduced through Density Functional Theory calculations as an orthorhombic monolayer composed of 4-, 5-, 6-, and mainly 9-membered carbon rings, with nonagons as the dominant motif. In the reported structure, the unit cell contains 36 carbon atoms arranged in a porous network that repeats in the $xy$-plane, and the material exhibits near-zero-gap metallic-like electronic behavior, dynamical and thermal stability at 300 K, and high in-plane stiffness (Huacarpuma et al., 23 Sep 2025). The same study further identifies sodium decoration at the centers of nonagonal rings as energetically favorable, yielding a Na@Ennea-Graphene complex with reversible hydrogen uptake up to $8.80\,\mathrm{wt}\%$ H$_2$, positioning ennea-graphene primarily within the research domain of hydrogen-storage nanomaterials rather than established bulk carbon phases (Huacarpuma et al., 23 Sep 2025).

1. Crystal topology and structural definition

Ennea-graphene is described as a new 2D carbon allotrope with crystal system orthorhombic and layer group P2/m (No. 10) (Huacarpuma et al., 23 Sep 2025). Its lattice parameters are reported as $a = 12.83$ Å, $b = 8.99$ Å, and $c \gg 15$ Å, where the large $c$ value represents vacuum separation along the out-of-plane direction in the simulation cell. The lattice vectors are given in Cartesian form as

$$\mathbf{a}_1 = (a,\,0,\,0), \quad \mathbf{a}_2 = (0,\,b,\,0), \quad \mathbf{a}_3 = (0,\,0,\,c).$$

The basis comprises carbon atoms occupying sites that generate a network of 4-, 5-, 6-, and 9-membered rings. Typical C–C bond lengths range between 1.40 and 1.45 Å, depending on local ring strain (Huacarpuma et al., 23 Sep 2025). The emphasis on nonagons distinguishes ennea-graphene from more familiar graphene-derived allotropes built predominantly from hexagons. In the reported unit cell, all distinct ring motifs are present within a 36-carbon repeat unit.

This ring topology is central to the material’s proposed function. The nonagonal pores create adsorption environments that later become the preferred sites for sodium decoration. A plausible implication is that the geometric openness of the network, combined with the ring-size heterogeneity, governs both the adsorption landscape and the mechanical anisotropy reported for the sheet.

2. Computational framework and energetic definitions

The reported characterization of ennea-graphene was carried out using VASP with PAW potentials, GGA–PBE exchange–correlation, HSE06 for more accurate band gaps, and Grimme DFT-D2 van der Waals corrections (Huacarpuma et al., 23 Sep 2025). The plane-wave cutoff was 550 eV. Brillouin-zone sampling used a $\Gamma$-centered $5\times5\times1$ $8.80\,\mathrm{wt}\%$0-mesh for structural relaxation and a $8.80\,\mathrm{wt}\%$1 $8.80\,\mathrm{wt}\%$2-mesh for electronic structure. Convergence criteria were $8.80\,\mathrm{wt}\%$3 eV and forces below $8.80\,\mathrm{wt}\%$4 eV/Å.

Ab initio molecular dynamics employed the NVT ensemble with a Nosé–Hoover thermostat at $8.80\,\mathrm{wt}\%$5 K, time step $8.80\,\mathrm{wt}\%$6 fs, and total simulation time of 5 ps, with 10 ps used in selected stability runs (Huacarpuma et al., 23 Sep 2025). These methodological details delimit the evidentiary basis of the reported stability and adsorption results.

Two energy definitions structure the adsorption analysis. The sodium binding energy is defined as

$8.80\,\mathrm{wt}\%$7

and the average H$8.80\,\mathrm{wt}\%$8 adsorption energy per molecule is defined as

$8.80\,\mathrm{wt}\%$9

These expressions are used to quantify the thermodynamic favorability of metal anchoring and molecular hydrogen uptake, respectively (Huacarpuma et al., 23 Sep 2025).

3. Stability, elasticity, and electronic structure

The phonon dispersion of pristine ennea-graphene shows no imaginary frequencies across the high-symmetry paths, with acoustic modes below 5 THz and optical modes extending up to 54 THz, which is taken as evidence of dynamical stability (Huacarpuma et al., 23 Sep 2025). Thermal stability was examined by AIMD at 300 K for 10 ps, where total-energy fluctuations remained within $_2$0 eV around $_2$1 eV and the topology of the 4-, 5-, 6-, and 9-membered rings remained intact.

The electronic band structure indicates a PBE gap of approximately 0.011 eV and an HSE06 gap of approximately 0.067 eV, leading to the description of the system as metallic-like (Huacarpuma et al., 23 Sep 2025). The projected density of states shows dominant carbon $_2$2-orbital contributions at the Fermi level. This places ennea-graphene in a near-zero-gap regime rather than a wide-gap semiconducting one.

Its mechanical response is reported through elastic constants for a 2D orthorhombic lattice: $_2$3 N/m, $_2$4 N/m, $_2$5 N/m, and $_2$6 N/m, satisfying the Born–Huang criteria (Huacarpuma et al., 23 Sep 2025). The in-plane Young’s modulus $_2$7 varies from 232.7 to 254.9 N/m, with anisotropy ratio approximately 1.10 and average approximately 255 N/m. The shear modulus ranges from 87.7 to 100.1 N/m and Poisson’s ratio from 0.27 to 0.34.

These values indicate a mechanically robust but not perfectly isotropic membrane. This suggests that the mixed-ring topology perturbs the nearly isotropic elastic response associated with ideal graphene while retaining comparatively high in-plane stiffness.

4. Sodium decoration and Na@Ennea-Graphene

Sodium decoration is analyzed by considering hollow sites H1–H4, bridge sites B1–B4, and top sites A1–A3 (Huacarpuma et al., 23 Sep 2025). The preferred adsorption site is H4, identified as the center of a nonagonal ring. At this site, the sodium adsorption energy is reported as approximately $_2$8 eV, the most negative among the tested positions; the other sites lie between $_2$9 and $a = 12.83$0 eV.

In the optimized geometry, Na sits approximately 2.5 Å above the plane, producing minimal distortion of the carbon lattice (Huacarpuma et al., 23 Sep 2025). The electronic analysis indicates that Na $a = 12.83$1- and $a = 12.83$2-orbitals hybridize with carbon $a = 12.83$3 states at the Fermi level, confirming charge transfer from Na to C. Bader analysis, though not detailed in the paper, is reported to indicate partial Na$a = 12.83$4 character.

The combination of strong binding and limited structural distortion is important for the hydrogen-storage mechanism. The nonagonal ring is not merely a geometrical feature; it functions as the primary anchoring environment that stabilizes dispersed sodium atoms against weak binding scenarios typical of less favorable adsorption sites. A plausible implication is that the porous ring architecture simultaneously supports cationic adsorption centers and preserves accessible volume for H$a = 12.83$5 coordination.

5. Hydrogen adsorption, reversibility, and storage metrics

The Na-decorated structure is reported to adsorb up to four H$a = 12.83$6 molecules per Na atom, corresponding to 32 H$a = 12.83$7 molecules per unit cell and a gravimetric capacity of $a = 12.83$8 H$a = 12.83$9 (Huacarpuma et al., 23 Sep 2025). The average adsorption energies are given for sequential loading as follows:

Loading	Average adsorption energy	Gravimetric capacity
8 H$b = 8.99$0	$b = 8.99$1 eV	2.35 wt%
16 H$b = 8.99$2	$b = 8.99$3 eV	4.60 wt%
24 H$b = 8.99$4	$b = 8.99$5 eV	6.75 wt%
32 H$b = 8.99$6	$b = 8.99$7 eV	8.80 wt%

The adsorbed hydrogen remains molecular, with an H–H bond length of approximately 0.76 Å compared with 0.75 Å for free H$b = 8.99$8 (Huacarpuma et al., 23 Sep 2025). This is consistent with physisorption or weak chemisorption rather than dissociative uptake. The reported adsorption-energy window of about $b = 8.99$9 to $c \gg 15$0 eV per H$c \gg 15$1 is presented as compatible with reversible storage.

Using the van ’t Hoff relation

$c \gg 15$2

the desorption temperature at 1 atm is estimated as 197–232 K (Huacarpuma et al., 23 Sep 2025). In AIMD at 300 K for Na@EG+32H$c \gg 15$3 over 10 ps, stepwise desorption events were observed while the carbon framework and Na anchoring were preserved, supporting the claim of reversibility under near-ambient conditions. The thermodynamic map further indicates that at $c \gg 15$4 atm and $c \gg 15$5C, about 31 H$c \gg 15$6 molecules are retained, whereas at $c \gg 15$7 atm and $c \gg 15$8C, nearly full desorption occurs (Huacarpuma et al., 23 Sep 2025).

The reported capacity exceeds the U.S. Department of Energy 2025 target of $c \gg 15$9 for onboard hydrogen storage materials (Huacarpuma et al., 23 Sep 2025). In the context of the study, ennea-graphene is therefore proposed as a hydrogen-storage substrate whose ring topology, alkali decoration, and molecular adsorption energetics are jointly optimized.

6. Relation to device-oriented graphene research and ennea-graphene nanoribbons

A separate line of research relevant to ennea-graphene concerns contacting strategies for narrow carbon nanostructures using graphene electrodes. In work on 9-atom wide armchair graphene nanoribbons, electron beam lithography-defined graphene electrodes on Si/SiO$c$0 were combined with reactive-ion etching and post-transfer thermal annealing to form field-effect transistor devices with controlled nanogap geometries (Braun et al., 2021). The reported process used two-step EBL, Ar/O$c$1 RIE, and either CSAR or PMMA etch masks, with the best gaps below 15 nm achieved using CSAR. Annealing at $c$2C under high vacuum for 20–30 min improved contact quality by removing physisorbed water and enhancing $c$3–$c$4 coupling.

Electrical transport measurements in those ribbon devices showed highly nonlinear two-terminal characteristics, strong gate dependence, p-type behavior for $c$5, incipient n-type behavior for $c$6, on/off ratios up to $c$7, and on-currents up to 70 nA at $c$8 V and $c$9 V after annealing (Braun et al., 2021). Raman spectroscopy showed preserved G and 2D modes in the graphene electrodes and characteristic graphene nanoribbon modes after transfer, with no evidence of broadening or new defect peaks.

That work explicitly discusses implications for future ennea-graphene device design, stating that controlled EBL-defined graphene electrodes can tailor channel length down to below 15 nm, matching the 5–50 nm lengths of bottom-up ennea-graphene nanoribbons, and that high-throughput fabrication exceeding 1600 devices per chip enables statistical studies (Braun et al., 2021). It also proposes design guidelines for ennea-graphene FETs, including CSAR etch masks, two-step EBL/RIE, a $$\mathbf{a}_1 = (a,\,0,\,0), \quad \mathbf{a}_2 = (0,\,b,\,0), \quad \mathbf{a}_3 = (0,\,0,\,c).$$030 min vacuum anneal at $$\mathbf{a}_1 = (a,\,0,\,0), \quad \mathbf{a}_2 = (0,\,b,\,0), \quad \mathbf{a}_3 = (0,\,0,\,c).$$1C, and Raman plus gate-tuned transport characterization.

These results do not establish experimentally realized ennea-graphene devices, but they define a plausible integration pathway for ennea-graphene nanoribbons if such structures are synthesized. This suggests that the current literature situates ennea-graphene at the intersection of two research programs: computationally designed porous 2D carbon allotropes for gas storage, and all-carbon nanoelectronic contacting schemes for structurally precise ribbon systems.

7. Research significance, limitations, and open directions

The principal significance attributed to ennea-graphene is that it offers a porous 2D carbon network with rare nonagonal ring motifs that stabilize Na decoration and enable high H$$\mathbf{a}_1 = (a,\,0,\,0), \quad \mathbf{a}_2 = (0,\,b,\,0), \quad \mathbf{a}_3 = (0,\,0,\,c).$$2 uptake (Huacarpuma et al., 23 Sep 2025). Within the reported calculations, the material combines several desirable attributes: dynamical stability, thermal stability at 300 K, metallic-like electronic structure, high in-plane stiffness, strong Na anchoring at nonagonal centers, and reversible hydrogen adsorption reaching $$\mathbf{a}_1 = (a,\,0,\,0), \quad \mathbf{a}_2 = (0,\,b,\,0), \quad \mathbf{a}_3 = (0,\,0,\,c).$$3.

The same study also delineates unresolved challenges. These include experimental synthesis of large-area ennea-graphene, uniform Na dispersion, and long-term cycling stability (Huacarpuma et al., 23 Sep 2025). Future directions explicitly identified in the paper are exploration of other alkali or alkaline earth dopants, hybrid composites to tune adsorption energies, experimental validation via chemical vapor deposition or bottom-up assembly of nonagonal motifs, and investigation of multilayer stacking and real-world device integration.

A common potential misconception is to treat ennea-graphene as an experimentally established carbon phase comparable in maturity to graphene. The available description instead presents it as a theoretically introduced allotrope whose properties have been characterized computationally (Huacarpuma et al., 23 Sep 2025). Likewise, the device-oriented literature on graphene nanoribbon contacts provides process guidance relevant to possible ennea-graphene nanoribbon integration, but not direct demonstration of ennea-graphene transistors (Braun et al., 2021). The present state of the topic is therefore best understood as computational materials design with adjacent experimental methodologies that could support subsequent realization.