betaFeedback

WorksPhysics and Astronomy

Stable Thin Clathrate Layers

Abstract

We explore elemental Si, Ge and Sn free-standing slabs in diamond, clathrate and other competing structures and evaluate their stability by comparing Kohn-Sham total energies at the same atoms/area coverage ρρ. We find that within 3-6 ML range (in units of diamond monolayer), the surface energy stabilizes clathrates against the bulk-stable diamond structure or other competitors, for all three elements. At~1.3-1.6 ML coverages, Si and Ge stable layers are adatomic decorations of the puckered honeycomb lattice, forming a dumbbell–decorated pattern with rectangular unit cell. Analogical Sn layer is unstable to a web-like net of interconnected Sn9_9 clusters. The zoo of stable thin-layer non-diamond structures is completed by a Sn metallic bilayer at 2.3 ML coverage.

Introduction

For bulk Sn, Ge and Si, cubic diamond (Pearson symbol cF8) phase is the ground state structure (GS), lying 23, 27 and 54 meV/at. [1], respectively, below its energetically closest sp3^{3} adversary – type-II clathrate – according to the basic Density Functional Theory (DFT) setup. The main question of the presented fundamental theoretical DFT study is, whether opening a surface could grant clathrates an overall energetic advantage over diamond, resulting in stable free-standing clathrate thin films.

Sp3^{3} slabs exhibit unsaturated – so called dangling – bonds on the surface, which can be (partially) satisfied by proper reconstruction. The stability of a thin film, measured by internal energy per atom for the given thickness ρ≡N/S\rho\equiv N/S (coverage: number of atoms per area), is thus given by the trade-off between optimal bulk energy and saturation of the surface bonds. The question is, whether the latter effect could significantly favour clathrates.

The inquiry is not restricted to clathrates, but spans wide area of known sp3^{3} Column-XIV allotropes and thanks to long Molecular Dynamics (MD) annealings sped up by Machine Learning Force Fields (MLFF) innumerably many other candidates for the stable films as well, including metallic even-layers of Sn.

Free-standing thin films minimizing per-atom-energy for the given thickness (coverage) ρ≡N/S\rho\equiv N/S can be imposed by structure-neutral substrate such as amorphous or electronically-balanced oxide (ZrO2_2) or graphite wafer. Further stabilization of the most stable – elemental – clathrate slabs could be achieved by lattice-matching substrate, e.g. via coherent clathrate-diamond interface [2].

In the focus here are several-ML (diamond monolayers) thick films, no longer linked to their parent bulk phases by the physical properties. As for the thinnest ones, even interaction between slab’s faces comes into play, hence new – yet unexplored – features are expected.

The paper is organized as follows. In addition to the basic DFT setup applied, the crucial aspects of energy-minimization of a thin film such as the choice of proper termination, surface reconstruction when cut from the mother bulk phase together with MD-related approaches starting from random atomic positions or from the melt are described in § 3. § 4 introducing known parent bulk phases for thin-film production ensues. § 7 reveals realm of stability of clathrate films for each of the elements Sn, Ge, Si and compares the associated terminations and reconstructions of the per-atom-energy minimizing slabs. We go on to discuss the possibility of spontaneous clathrate crystallization from the melt in § 8 and the impact of van der Waals dispersion interaction in its last Subsection 6.4. The most important outputs of our work are summarized in the final § 9. Additional Figures and xyz-files* of all stable films are included in the Supplemental Material (SM) and are referenced as S.X.Y from the main text, where the integers X and Y index particular SM Section and Figure, respectively. Our principal results – internal energy per atom of the most stable films as a function of the thickness – are plotted in Figure 6 and tabulated in the final Section S10 of SM, supplemented with calculated values of electronic band gaps of the slabs.

Stability regions of Sn, Ge, and Si clathrate thin films

Figure 6. Shaded green area: stability region of Sn, Ge, Si clathrate thin films. DFT internal energy with respect to the cF8 bulk phase as a function of coverage is plotted in each case.

State of the Art

We shall start this concise review with honeycomb monolayers of Group-IV (G4) elements. Due to the preferred tetrahedral sp3\mathrm{sp}^{3} hybridization of these elements, the only member creating free-standing hexagonal monolayers is carbon (graphene) [3]. Honeycomb monolayers of Pb, Sn, Ge and Si analogous to graphene – so called Xene materials (X = Si, Ge, Sn and Pb) – do exist, but only on appropriate substrates [4], owing to the Xene lattice stretching and its strong interactions with the wafer.

Notably, high-quality monolayer stanene samples were crystallized as an unusually ultraflat honeycomb lattice on Cu(111) via Molecular Beam Epitaxy (MBE), fairly chemically inert and stable up to 243 K [4]. This observation was indeed unexpected because going down the Group IV sp3\mathrm{sp}^{3} nature of elements becomes more and more pronounced, reflecting in the buckling, adatomic reconstructions or layer-stacking stabilization mechanisms [5, 6, 7]. There are also indications [8] of free-standing few-layer stanene prepared on liquid hexane by non-equilibrium process of laser ablation, but later in the text (consult subsection 4.4) we propose another hexagonal polymorph for this structure.

Supported (e.g. by Cu(111) binding with pzp_z Sn orbitals, sustaining stanene chemical inactivity) or with suitable chemical functionalization, stanene is regarded as an ideal platform to study exotic quantum states of matter; when deposited on BN, it retains its electronic properties and can act as a quantum spin Hall insulator with gap as large as 0.3 eV, according to the theoretical predictions [4, 9, 10]. Its applications range from low-power electronics, nanotransistors and high-performance thermoelectrics to fault-tolerant quantum computation. Extra, superconductivity has been discovered in few-layer (bilayer and more-layer) stanene atop PbTe/Bi2_2Te3_3/Si(111) sandwich structure [11].

N.B.: Unlike preceding works [9, 10], the present paper goes beyond stability assessment limited to atomic & cell relaxations and phonon-frequency evaluation. Our inquiry does not rely on real phonon frequencies (reported also for dumbbell-stanene); all the free-standing thin films were subjected to long Molecular Dynamics (MD) runs (millions of fs-steps) at RT and higher (300 − 700 K), which led to dumbbell-stanene spontaneous transformation to a web of Sn9_9 clusters accompanied by decrease of internal energy per atom in the fixed periodic cell, regardless of reconstruction type, see subsection 4.1 and Figure 2.

Sn9, Sn10, Sn15, Sn50, and Sn100 tin clusters

Figure 2. Sn₉ and Sn₁₀ clusters (from two different view angles) are fundamental units of threadlike tin structures that form spontaneously for low coverages ρ\rho (discontinuous layers with holes) [52, 53]. Sn₁₅ subunit also frequent in Sn clusters comprises two Sn₉ basic units sharing a common triangle. Sn₅₀ and Sn₁₀₀ are examples of our long DFT annealings at room temperatures with 50 and 100 Sn atoms, respectively, enclosed in fixed 24×24×2424 \times24 \times24 ų cubic container. Sn₁₀₀ is composed of 8 Sn₉ clusters in a ring. Blue solid borderline delimits periodic motif.

Reconstructed 3×3\sqrt{3} \times\sqrt{3} honeycomb monolayer of Ge – germanene – was synthesized on the top of 7×77 \times7-Au(111) surface [12]. Similarly, one-atom-thick Si sheets of 3×3\sqrt{3} \times\sqrt{3}-reconstructed silicene were deposited on Ag(111) [13]. No experimentally prepared free-standing germanene and silicene are mentioned in the literature.

Optimal reconstruction of free-standing germanene was sought after in theoretical (DFT) paper [14], where the so called HDP structure was regarded to be energetically the best. We think that their “measure of energetics of the phases” – maximizing cohesive energy per unit area – is not the relevant thin-film stability criterion and we suggest in the present work other way to compare the stability of layers: minimization of per-atom energy, EatomE_{\mathrm{atom}}, for the given coverage ρ=N/S\rho= N/S. Wherefore we recalculated all the samples and plotted the results on the corresponding E=E(ρ)E = E(\rho) graphs.

Energy-minimizing (unreconstructed) bilayer silicene and germanene were theoretically predicted to be composed of exactly planar silicene monolayers in eclipsed (e.g. lonsdaleite) conformation [15, 16, 17]. These bilayers are represented by green pentagons marked as “GS-bilayer” in Figure 6 summarizing our chief results. DFT predictions were confirmed by STM experimental images of bilayer Si on Ag(111) [19] [18]. For both bilayer (few-layer) silicene and germanene, the stacking order of monolayers can induce metal-semiconductor phase transition and can regulate size of the electronic band gap [16, 17, 20]. Variety of stacking patterns – majority of which are not available to the bilayer graphene – are due to the inherent buckling of monolayer Si and Ge [15].

We have not investigated impact of stacking sequence on bi-layer stability in the present paper. The scope of this work is limited to 2×22 \times2, 3×3\sqrt{3} \times\sqrt{3}, c(2×4)c(2 \times4) and c(2×8)c(2 \times8) reconstructions of bilayer (111) bulk cubic diamond cuts with all possible relative placements of adatoms on two mutually interacting slab’s faces.

Comprehensive theoretical (DFT) survey of atomically thin free-standing Sn films was performed by P. Borlido et al. [21], who discovered the ground state of bilayer Sn. This square metallic – still hypothetical – structure is supposed to be prepared experimentally on e.g. Ag(100), Au(100), Pt(100), or Al(100) substrates, lying just 149 meV/at. above the bulk ground state α\alpha-Sn (cF8, cubic diamond).

Surface energy of various (un)reconstructed planar terminations of bulk cubic diamond (cF8) phase of Sn, Ge and Si were discussed in [22, 23, 24, 25, 26, 27, 28]. It was shown by numerous experimental and theoretical groups that (111)-diamond section of Sn, Ge and Si is a cleavage face possessing the lowest surface energy γ\gamma among all the cubic-diamond (Pearson cF8) sections. Note that this statement is valid just after proper surface reconstruction [22, 24]. 5×55 \times5 Dimer-Adatom-Stacking Fault (DAS), 7×77 \times7 DAS and 3×3\sqrt{3} \times\sqrt{3}, 2×22 \times2, c(2×4)c(2 \times4), c(2×8)c(2 \times8) adatomic reconstructions were observed on Si and Ge surfaces [29].

It was revealed recently by means of DFT that despite 5×55 \times5 DAS being the ground-state Si(111) bulk-surface reconstruction, 7×77 \times7 high-temperature DAS reconstruction – most frequently observed in experiments – freezes in a metastable state below RT upon cooling [30, 31]. As for Ge(111) relaxed bulk surface c(2×8)c(2 \times8) adatomic reconstruction is experimentally the most frequent one [24, 29].

R. Zhachuk et al. constructed “strain phase diagrams”: DFT-computed surface energy γ\gamma of various reconstructions of Ge(111) and Si(111)-diamond surfaces as a function of biaxial strain in the range ε∈(−4;4)%\varepsilon\in(-4; 4)\% [29, 32], explaining surface structure transformations during growth of strained Ge(111), Si(111) layers (imposed by the substrate). The calculations were compared with scanning tunneling microscopy data obtained on stepped Si(111) surfaces grown on Ge(111) substrate and on Ge islands grown on a Si(111) [29, 31, 32].

It turns out that DAS reconstructions thanks to the presence of all reconstruction elements (each individually) are intrinsically compressive (ε<0\varepsilon< 0), while purely adatomic reconstructions gain dominance under biaxial tensile strain (ε>0\varepsilon> 0). Namely Si(111) thin films deposited on Ge(111) substrate display c(2×42 \times4) adatomic surface reconstruction above ε≈2.5%\varepsilon\approx2.5\% biaxial tensile strain[29, 32]. Local minima of the surface energy for particular orientations are trade-off between saturation of the dangling bonds and minimization of the surface stress, inherently imposed by the reconstruction units (adatoms, dimers, stacking faults, steps, π\pi-bonded chains) introducing unusual bond lengths and angles[30].

There have been a few experiments leading to atomically-thin clathrate layers, deposited on lattice-matching substrates. Our most recent work reports several-atomic-layer thick Sn decagonal clathrate film deposited on quasicrystalline dd-Al-Ni-Co wafer[33]. The preceding work has shown the versatility of Sn to form quasicrystalline atomic arrangements, growing icosahedral Sn clathrate nanolayer on ii-Al-Pd-Mn[34].

Group of T. Kume deposited atomically-thin Ge type-II clathrate on Ge(111)-diamond surface. The authors hypothesized the existence of thin buffer layer at the interface[35, 36], but we have managed to construct coherent interface between the two[2]. This kind of coherent clathrate/diamond interface had been formerly displayed in several computational investigations exploiting clever empirical force fields[37, 38, 39], but had never been described in detail (in terms of periodicity, unit cell, atomic bonds), probably owing to quite a large lattice misfit (∼11%\sim11\%) between diamond and clathrate part of the same G4 element restricting the size of the coherent patches to relatively small islands. Our paper[2] was thus devoted to minimization of that misfit by optimization of materials on both sides.

However, the resulting Sn and Ge configurations in the aforementioned experiments were to large extent governed by the substrate. No experiment states preparation of free-standing clathrate layer on structure-neutral substrate, either liquid, amorphous or electronically-balanced oxide termination. Therefore in the present work, motivated by these studies, the most stable clathrate layers are sought after, regardless of substrate. One of the questions is, whether a combination of square and triangle clathrate tiles (bearing e.g. quasicrystalline symmetry) provides clathrate films with some energetic advantage over e.g. purely triangular (hexagonal) clathrate tilings.

Thicker clathrate layers (∼μ\sim\mum) can be used in opto-electronics as p-n junctions, Schottky barriers, LEDs and photocells in the visible spectrum, optical and X-ray detectors, thanks to the tunable electronic band gap in the range Eg∈[0.6;2.25]E_g \in[0.6; 2.25] eV by alloying and doping Snx_{x} (Nax_{x}Si136_{136}, Si34−x_{34-x}Gex_x, I8_{8}Si46_{46}, Xe8_{8}Si46_{46}, Nax≤24_{x\leq24}(Si1−y_{1-y}Gey_y)136_{136})[40, 41, 42, 43]. This, however, does not apply to the range of coverages investigated in the present work (ρ∈[1;10]\rho\in[1; 10] ML, Figure 6), since the width of the electronic band gap decreases with the decreasing thickness of the film. Note that all the DFT-computed (slightly underestimated) band gaps of our clathrate slabs are just small fractions of 1 eV, mostly below Eg=0.5E_g = 0.5 eV for all three elements, consult Subsection S10 of SM (outside the absorption rage of visible light, Eg>1.7E_g > 1.7 eV ≅\cong red light). The band gaps of thin clathrate layers are smaller than that of the corresponding DFT-computed bulk clath-II phase, see Figure S8.1 and Figure S8.2 (SM). Hence, their possible applications are as yet unknown.

Electronic density of states of bulk type-II clathrates compared with bulk diamond for Sn, Ge, and Si

Figure S8.1. Electronic Density of States, eDOS, of the bulk type-II clathrates (cF136) compared with bulk diamond (cF8) for Sn, Ge and Si. Fermi energy, EFE_F, is designated by the vertical thick dashed line.

Electronic Density of States of the most stable clathrate slabs compared with competitive diamond slabs

Figure S8.2. Electronic Density of States, eDOS, of the most stable clathrate slabs, compared with that of the competitive diamond slabs of approximately the same thickness ρ\rho. Fermi energy, EFE_F, is designated by vertical thick dashed line. Visit Subsection 5.4 in the main text.

Besides opto-electronic applications of clathrates, they are also famous for thermoelectrics and superconductors, owing to the cage doping, but this paper is aimed solely at elemental clathrate thin films, much harder to be experimentally prepared (especially Sn), their structure and symmetry and its impact on slab’s stability.

Methods & Definitions

Density Functional Theory (DFT)

For atomic relaxations as well as for Molecular Dynamics (MD) simulations and on-the-fly Machine Learning Force Fields (MLFF) Vienna ab initioab\ initio Simulation Package (VASP)[44] implementation of the plane-wave Density Functional Theory was employed powered by the Projector Augmented-Wave potentials (PAW)[45] for core-electron treatment in Perdew-Wang PW91 Generalized Gradient Approximation (GGA)[46] to the exchange-correlation functional describing electron-electron interaction.

As a default setup, we applied VASP “Accurate” setting (specifying ENCUT, see below, the grids for Fast Fourier Transform, and the accuracy of the projectors in real space) and default values of energy cutoff of the plane-wave basis set in eV used for electronic wave-function description, ENCUT (Sn: 103.3 eV, Ge: 173.8 eV, Si: 245.4 eV). Energy was converged on K-point mesh density (KPOINTS) determining Bloch vectors for the Brillouin zone sampling to 1 meV/atom precision. Simulated films infinite in two dimensions were separated by at least 15Å vacuum layers in the third dimension to prevent mutual interactions of the periodic copies. s2^2p2^2 electrons were treated as valence for each element. All sites in the unit cell along with the unit cell dimensions were relaxed using a conjugate gradient algorithm to minimize energy with an atomic force tolerance of 0.005 eV/Å and a total energy precision of 10−510^{-5} eV.

Electronic band gaps stated in Subsection S10 of SM were evaluated using the same GGA setup and converged within 0.05 eV-wide window with respect to KPOINTS and smearing width given by parameter SIGMA, which was successively lowered from 0.2 eV up to 0 eV while increasing the K-point mesh density. Note that basic PAW GGA setup is well-known to systematically underestimate a bang-gap width.

It has been demonstrated[21] that despite the spin-orbit coupling does have a measurable effect on the formation energy, the energy shift is nearly constant for all 2D Sn structures (55 – 59 meV/at.) and as such can be neglected. Another DFT extension, van der Waals dispersion interaction (PBE-D3 correction to DFT), is discussed in Subsection 6.4.

Ab Initio Molecular Dynamics (AIMD)

By ab initio Molecular Dynamics (AIMD) we mean long MD runs – (tens of) millions of ∼\sim fs-steps – markedly accelerated by Machine Learning Force Fields (MLFF) as implemented in VASP.

AIMD was used within the present work on three occasions. Firstly, AIMD is the simplest way to amorphous bulk and thin films, described in Subsection 4.5, avoiding ingenious construction. Unfortunately, there is no such an automated ab initio way to clathrates, see § 8, where an attempt has been made to recrystallize clathrate slabs from the melt, either in a free-standing slab form or as a Ge slab on an appropriate InN substrate.

The third instance was the thin-film configurational-space search. Despite the clathrate-slab notation suggested in Table 1, Subsubsection 4.3.1, several-atom-thick layers have actually no relation to any bulk phase: they are unique identities themselves. Long AIMD runs thus attest to the completeness of our thinnest-film enumeration, mainly specific Sn metallic even-layers rediscovered this way in Subsection 4.4. AIMD confirms that there are not any other low-energy thin-film atomic arrangements.

Namely, for the Sn, Ge and Si thinnest-film production in the range of coverages ρ∈[0.85;4.0]\rho\in[0.85; 4.0] ML with the basic step Δρ≈0.5\Delta\rho\approx0.5 ML, 4×1054 \times10^{5} steps 5 fs each in the temperature range T∈[300;500]T \in[300; 500] K were performed, leading to no new phases. As concerns Sn, due to the close competition of square and triangle even-layers in this temperature range and owing to the plethora of Sn phases, Subsection 4.4, we have gradually increased coverage from 2 ML to 2.5 ML via random single-atom addition to the (23.9A˚)3(23.9\mathring{\mathrm{A}})^{3} supercell, witnessing a coexistence of the two phases at approximately 2.29 ML (for 137–138 atoms in the cell).

Extra, we have executed several Replica Exchange (RE) simulations 20-50 thousand 5 fs-step long starting from random atomic configurations with all three elements, in the temperature range from room up to the respective melting point, again, with no new unexpected results. These were either simulations with fixed boundary conditions (coverage) or complemented by energy cost function scaling either linearly with surface (surface energy) or quadratically with the departure of a coverage from its initial value, emulating through the stiffness constant some kind of elastic penalty for the sample thickening, which we would like to artificially suppress.

RE is a particular example of a delicate trade-off between precision and computational cost, since several simulations are running at once, calling for as moderate KPOINTS, ENCUT and POTIM, as possible. In the Sn case, we ended with (3 3 1) KPOINTS, while Si and Ge demanded at least 3 specially chosen KPOINTS and even those may not be enough, since Si slabs melted below 1030 K at default value of ENCUT and 5 fs time-step. It is indeed surprising that even single Ba atom addition to our several-dozen-atom periodic cells of free-standing thin-film samples, does not lead to a solo clathrate cage. We hope that the present work stimulates further research (with enhanced values of all the parameters) towards this goal.

N.B.: All samples obtained by means of AIMD, MLFF and RE were finally DFT-relaxed and only energies after DFT relaxation are stated in this text. Hence accuracy of MLFF enters the results – Figure 6 – in no way.

Atomic Structure, Termination, Thickness, Reconstruction: Definitions

Atomic Structure.

Vast majority of the slabs were carved out from the known bulk phases, described in § 4, by planar sections. The second way to Column-XIV thin films led through the ab initio Molecular Dynamics introduced in Subsection 3.2. Except for Sn – a borderline case between metal and semiconductor – the work deals exclusively with sp3\mathrm{sp}^{3} allotropes where all atoms exhibit more or less tetrahedral environment with quartet of nearest neighbours forming average bond angle around 109.5∘109.5^\circ.

All the known high-pressure allotropes more than 50 meV/at. above the respective cubic diamond (cF8) phase, such as BC8, R8, ST12, BT8 [47], mC16, bc5-t [48], obtained chiefly by (de)compression, were avoided [see Table S1.1, SM, for phases that were taken into account], since the main goal is to predict stable slabs with lower internal energy per atom compared to the corresponding diamond slabs of the same thickness. The only exception will be the amorphous phases, due to the intention to crystallize stable clathrate slabs from the melt and owing to their inherent indifference to the structure of a substrate, in contradistinction to diamond or clathrates.

Config.CSiGeSnPb
fcc457360833175-3.598
bcc43675873418642
graphite-9.231640602570719
hP19713522434596
tI4cF832822846128
cF8131-5.420-4.516-3.863251
lonsdaleite155111816292
clathI239643328279
clathII205542723280
clathIII260814840288
clathIV243693831276*
deca241714135299*
icosa288925949274*
CFS-6241553530286
CFS-5263413936289
T12238453632278
amorph–15812891–101–

Table S1.1. Bulk energies of famous Column-XIV allotropes, that were taken into slab internal-energy competition, see Section 4 in the main text. Internal energies of 17 bulk configurations of G4 elements after cell and atomic relaxations starting from sound initial atomic positions and lattice parameters. cF8 stands for the cubic diamond phase; lonsdaleite (wurtzite) is the hexagonal diamond polytype with eclipsed bond conformations. “deca” stands for “decagonal clathrate approximant”, Pearson symbol oP392. “icosa” means “icosahedral clathrate approximant”, cI920. “clathI” (cP46), “clathII” (cF136), “clathIII” (hP40) and “clathIV” (tP172) denotes type I, II, III and IV clathrate, respectively. “CFS” means Chiral Framework Structures (including T12), see Subsection 4.6 (of the main text). “amorph” denotes sp3sp^3-amorphous phases of Ge and Si, visit Subsection 4.5; for Sn the lower bound corresponds to a-metal, whereas upper bound to an abstract a-sp3sp^3 Sn obtained by a-Ge rescaling to the typical Sn interatomic distances. Kohn-Sham energies in eV/atom of the most stable (GS) structures are highlighted by red. GS energies have been subtracted from the remaining energies and the corresponding differences E−EGSE-E_{\mathrm{GS}} are stated in units of meV/atom for easier comparison of distinct elements. * – amorphized or strongly deformed during relaxation process. Without former annealing or external strain. Converged on K-points with default VASP DFT PAW GGA (PW91) ENCUT-s.

Extra, atomic configurations directly related to diamond, such as hexagonal diamond (lonsdaleite), so called diamond polytypes, were also excluded from the competition, as brand new materials bearing novel properties are in the sights.

Termination.

From now on, termination is a particular orientation and position of a cutting plane with respect to a bulk crystal, producing planar slab of constant thickness:

Thickness.

of a slab is measured by the coverage ρ\rho defined by:

ρ≡N/S(1)\rho\equiv N/S \tag*{(1)}

where NN stands for the number of atoms per surface area SS of a periodic cell. Individual cases of planar terminations suitable for each bulk phase are listed in the next § 4.

For the sake of simple comparison, coverage ρ\rho will be measured in dimensionless “monolayer” (ML) unit, one ML corresponding to ρ\rho value of a single (111) layer of cubic α\alpha-phase (cF8, diamond) of either Si, Ge or Sn.

Reconstruction.

Hereafter, by reconstruction we mean addition of atoms of the same kind, adatoms, (i.e. Sn atoms in the Sn-slab case) to the film’s surface followed by DFT relaxation of both the atomic positions and lattice parameters. We did not attempt any hydrogenation, whereas “bare” surfaces after cell and atomic relaxation will be called anyway “unreconstructed”.

Particular instances of surface reconstruction shall be specified for each parent bulk phase and for every termination thereof in the next § 4.

Slab Stability Assessment

Due to the intrinsic positivity of the surface energy γ\gamma there is a general trend of decreasing per-atom energy with slab thickness. At the same time, flat slabs occur at discrete, characteristic values of coverage ρ\rho and energy per atom. Hence, we seek after energy-minimizing slab structures for a given coverage ρ\rho, corresponding to certain slab thickness.

Let’s consider two directly constructed points E(ρ)E(\rho), e.g. a couple of (111)-diamond slabs with distinct thicknesses ρ1≡N1/S1\rho_1 \equiv N_1/S_1, ρ2≡N2/S2\rho_2 \equiv N_2/S_2 with energies per atom E1E_1, E2E_2, respectively, which may differ in type of surface reconstruction, hence in the value of surface energy γ1≠γ2\gamma_1 \ne\gamma_2. An interpolation between the points has a meaning of a terrace of energy EE such that E1>E>E2E_1 > E > E_2 and coverage ρ\rho such that ρ1<ρ<ρ2\rho_1 < \rho< \rho_2 between the two slabs, whose contribution to the overall energy-per-atom of the whole system tends to zero for periodic boundary conditions (PBC) tending to infinity. Because energy, surface and number of atoms per periodic cell are additive quantities satisfying4:

pi≡Ni/N, i∈{1,2}:p1+p2=1(2)p_i \equiv N_i/N,\ i \in\{1,2\}: \qquad p_1+p_2=1 \tag*{(2)}
S=S1+S2=N1(S1N1)+N2(S2N2),E=p1E1+p2E2(3)S=S_1+S_2=N_1\left(\frac{S_1}{N_1}\right)+N_2\left(\frac{S_2}{N_2}\right), \qquad E=p_1E_1+p_2E_2 \tag*{(3)}

the linear interpolation Eatom(ρ−1)=E(ρ−1)E_{\mathrm{atom}}(\rho^{-1})=E(\rho^{-1}) follows:

SN=p1(S1N1)+p2(S2N2) ⟺ (ρ)−1=p1(ρ1)−1+p2(ρ2)−1(4)\frac{S}{N}=p_1\left(\frac{S_1}{N_1}\right)+p_2\left(\frac{S_2}{N_2}\right)\ \Longleftrightarrow\ (\rho)^{-1}=p_1(\rho_1)^{-1}+p_2(\rho_2)^{-1} \tag*{(4)}

corresponding to this particular interpolation on the graphs Eatom(ρ)E_{\mathrm{atom}}(\rho):

E=E2+ρ2−ρρ2−ρ1ρ1ρ(E1−E2)(5)E=E_2+\frac{\rho_2-\rho}{\rho_2-\rho_1}\frac{\rho_1}{\rho}(E_1-E_2) \tag*{(5)}

The aim of this sort of graphs is to compare thin-film energy per atom, EatomE_{\mathrm{atom}}, of different atomic configurations, terminations and reconstructions for the common intermediate coverage ρ\rho, not associated with any planar slab structure already measured.

In the simplest case of a constant surface energy γ\gamma, hyperbolic trends (5) of the curves Eatom(ρ)E_{\mathrm{atom}}(\rho) for each structural family of slabs, cf. Figure 6, can be derived directly from the definition of the surface energy:

γ≡def.Ecell−EbulkN2S ⟹ Eatom(ρ)=EcellN=2γρ+Ebulk(6)\gamma\mathrel{\mathop{\equiv}^{\mathrm{def.}}}\frac{E_{\mathrm{cell}}-E_{\mathrm{bulk}}N}{2S}\ \Longrightarrow\ E_{\mathrm{atom}}(\rho)=\frac{E_{\mathrm{cell}}}{N}=2\frac{\gamma}{\rho}+E_{\mathrm{bulk}} \tag*{(6)}

where EatomE_{\mathrm{atom}} is energy per slab’s atom, EcellE_{\mathrm{cell}} is overall energy per periodic cell with the total of NN atoms, EbulkE_{\mathrm{bulk}} is energy per one atom of the respective bulk phase and ρ\rho is thin-film coverage (possessing two faces, hence the factor “2” in Equation (6)).

In the case of layered growth mode with constant surface energy γ\gamma of a thin film and/or in the close vicinity of the scattered points in Figure 6, the hyperbolic dependence (6) describes slabs with (non-)integer number of layers. Each point of the hyperbolic convex hull can be thus assigned an (in)complete slab; incompleteness meaning a slab with a terrace.

The free-standing slab setup with PBC (periodic cell extended vertically in the zz direction to include vacuum gap between periodic images of a slab) introduces specific constraints: for systems with lattice parameters up to few tens of Å, it effectively hampers formation of non-planar bulky morphologies (e.g. nanowires, nanoparticles). Importantly, all the slabs in current simulations are only metastable – stabilized by artificial effect of finite PBC – whose increase would progressively lead to spontaneous transformation to nanowires and nanoparticles, respectively, owing to the γ>0\gamma>0 of G4 sp3^3 elements.

While this phenomenon actually helps studying slabs as templates for adlayers that might form on an appropriate substrate by wetting, it have to keep in mind that our simulations do not fairly assess the tendencies to possible island formation [33], once the material is deposited on a substrate. Questions related to dewetting are actually highly relevant for sp3^3-type adlayers [49], e.g. silicon on an insulator.

Competing allotropes

Monolayers

Stabilizing reconstruction evidently depends on the layer thickness. Singular point are monolayers, displaying just one atomic layer to reconstruct. However, there are infinitely many ways to do so.

As for Sn, Ge and Si, clear choice [9, 14, 50] are various reconstructions on (111)-diamond monolayer, so called stanene, germanene and silicene, see Figure 1. Because the basic reconstruction unit is dumbbell, we will call them hereafter db-monolayers [14].

Four reconstructions of MoS2, hP10, hP11, oP10 of (111)-diamond Sn, Ge and Si monolayers

Figure 1. Four reconstructions – MoS2_2, hP10, hP11, oP10 – of (111)-diamond Sn, Ge and Si monolayers: top – front view; middle, bottom – side views. Solid blue line delimits periodic cell excluding vacuum. Red balls designate adatoms. ρada\rho_{\mathrm{ada}} under Pearson symbol stands for the area-density of adatoms expressed through typical length unit LadaL_{\mathrm{ada}}: the distance between nearest neighbours on a sublattice of a bipartite (111)-diamond lattice; for bulk diamond structure Lada(Sn)=4.69L_{\mathrm{ada}}(\mathrm{Sn}) = 4.69 Å, Lada(Ge)=4.08L_{\mathrm{ada}}(\mathrm{Ge}) = 4.08 Å, Lada(Si)=3.86L_{\mathrm{ada}}(\mathrm{Si}) = 3.86 Å. Indicated unit cell sizes in universal units of dd and Lada≈3dL_{\mathrm{ada}} \approx\sqrt{3}d are just for better orientation in various cross sections; here we assume idealized planar projections made of regular hexagons with edge length dd (i.e. we neglect lattice corrugations and distortions due to the chemical bonds, e.g. with red adatoms).

Even though each dumbbell constructed on germanene (silicene) lowers per-atom-energy, MoS2_2 configuration, corresponding to the full coverage, has branches with imaginary frequencies in ab-initio phonon spectra indicating mechanical instability [14]. The ground state of (Sn, Ge, Si) db-monolayer is therefore non-trivial and requires systematic investigation.

For every admissible dumbbell density per unit area in the given supercell – from minimal zero stanene density up to the maximal MoS2_2 dumbbell density – all dumbbell configurations were tested for the resulting internal energy per atom and plotted as a function of coverage ρ\rho, see Subsection 3.4. We have chosen 2×22 \times2 oP10 supercell with originally 32 stanene atoms (without adatoms). Adatoms were placed just on one sublattice of the bipartite stanene lattice followed by the systematic elimination of the symmetrically equivalent adatomic configurations.

For both germane and silicene, oP10 minimizes internal energy per slab’s atom, cf. Figure 6. On the contrary, all the candidate db-monolayers, succumbed in the course of AIMD simulations, starting from random atomic positions and from hP10 db-stanene as well, to a web of interconnected Sn9_9 clusters, Sn9_9-web, (Figure 2) in the Sn case for the given coverage, Figure 6, accompanied by an internal energy decrease. This is our new theoretical prediction, which has not been corroborated by experiment yet. For example, hP10 db-stanene (ρ=1.35ML\rho= 1.35\mathrm{ML}), 160 atoms per periodic cell, transformed into a web of Sn9_9 clusters at relatively high temperatures between 559–600 K, lowering the internal energy by 30 meV/at. in the fixed cell.

As it shall be explicitly illustrated by an example of bilayers, Subsection 4.4, Sn due to its metallic temper favours 6-coordinated atoms resulting in genuine G4 bond-angles equal to 60° (Figure 4, typical of a-Ge and a-Si surface[51]) and 109.5° dominant in a Sn9_9-web to purely sp3^3-like coordination of 4, accompanied by “factitious” bond angles 54°, 66°, 72°, 129°, enforced by the arrangements depicted in Figure 1.

Typical surface of a-Si

(a)

Side view of an a-Si slab

(b)

Bond-angle distributions of amorphous silicon and germanium

(c) Top: Typical surface of a-Si: most of the atoms are 4-coordinated yellow balls, blue balls are sp2\mathrm{sp}^{2}-bonded atoms, green balls are p3\mathrm{p}^{3}-bonded and magenta atoms possess two bond angles ≈100∘\approx100^\circ and one ≈120∘\approx120^\circ. There are two atoms with even sharper pyramidal bonding geometry of θave<80∘\theta_{\mathrm{ave}} < 80^\circ. Periodic cell without vacuum is delimited by blue line. Size of a ball is proportional to its distance from the observer. Middle: Side view of the corresponding a-Si 4.3ML-thick slab. On the top panel, just 7Å-thick layer from the surface is depicted. Bottom: Bond-angle distributions of the best bulk a-Si and a-Ge without defects, produced by DFT on-the-fly trained VASP MLFF, fitted by gaussians, compared with thick slabs of the same material.

Figure 4.

Diamond

As a reference for elemental Sn, Ge and Si diamond slabs (no mention of carbon in this work!), we have chosen (111)-terminated ones. It has been shown by numerous experimental and theoretical groups that (111)-diamond section of Sn[28], Ge and Si [22, 23, 24, 25, 26, 27] is a cleavage face possessing the lowest surface energy γ\gamma among all the cubic-diamond (cF8) bulk sections. Note that this statement is valid just after proper bulk surface reconstruction[22, 24].

Several reconstructions of (111)-diamond surface have been compared: 3×3\sqrt{3} \times\sqrt{3}, 2×22 \times2, c(2×4)c(2 \times4), c(2×8)c(2 \times8) [29, 32], 3×33 \times3 DAS, 7×77 \times7 DAS[31], but the variation of the per-atom energies turned out to be negligible with respect to the clathrate-diamond differences above 3 ML slab’s thickness, cf. Figure 6. Mind that below, say, 10 ML surface states or surface atoms cannot be distinguished since practically all atoms belong to the surface[24], so the notion of the surface energy γ\gamma is not relevant here.

Clathrates

Clathrate Terminations

We have investigated all planar sections of clathrate types I-IV, leading to equilateral-/isosceles-triangle, square and rectangle in-plane tilings and combinations thereof, Table 1. To every such orientation of a section plane with respect to a bulk clathrate there are four distinct terminations[33], e.g.:

Th→Tc′→Th′→Tc(7)T_h \to T'_c \to T'_h \to T_c \tag*{(7)}

schematically depicts the 4-layer Triangular-tiling period in the “z” direction perpendicular to the section plane (shared by both isosceles and equilateral triangle, likewise for the square and rectangle tiles) of puckered terminal planes with characteristic “holes” – ThT_h, Th′T'_h – alternating with flat terminal planes displaying “caps” around the tile vertices: Tc′T'_c, TcT_c, consult Figures S2.1 and S3.1 (SM) for decagonal (iso-tri, rectangle) and dodecagonal tiles (equi-tri, square), respectively.

Note that the only exception from the rule (7) comprises trio of (111)-terminations of type-II clathrate where TcT_c planes are missing. These very missing flat planes of clathrate-III 6-caps, Figure S2.1 in SM, act as a bridge between type-II and type-III clathrates, representing the only difference between the two.

Reconstruction of decagonal Sn, Ge, Si clathrate tiles

Figure S2.1. Reconstruction of decagonal Sn, Ge, Si clathrate tiles: isosceles triangle (T) and rectangle (R) terminated by cage “caps” (c) and “holes” (h). Adatoms (beneath the original atoms, inside the film) highlighted by red rings. 3-coord. atoms before the reconstruction marked with central blue dots, the rest are 4-coord. atoms. Side lengths distinguished by blue (longer) and yellow (shorter) color. Sides that fit together marked with green crosses, rectangles.

Explicit clathrate construction, i.e. atomic decorations of clathrate tiles mentioned above, was achieved via symmetry-preserving dual transformation[54, 55] of the respective Frank-Kasper (FK) phases. Equally, orientation of section planes was inspired by the well-known FK atomic planes, linked by dual transformation to the planes of clathrate-cage centers. It turns out that these are the section orientations yielding planar clathrate slabs containing at least 3-coordinated surface atoms, completely without terraces or cages sticking out from the surface.

This, however, is not the whole story about sensible clathrate construction. terminations, since each clathrate cage itself represents an object with at least 3-coordinated surface atoms, even when entirely isolated. Adding 5-caps or 6-caps onto puckered ThT_h and Th′T'_h planes does not change the surface density of unsaturated (so called dangling) bonds. Therefore, we expect that the growth mode of a clathrate will be either pure island or mixed island-layer type, depending on the energy scales related to the processes of (1) completion of a surface cage and (2) coalescing cages together [33].

Clathrate Nomenclature

We shall designate clathrate slabs according to the element, the associated clathrate tile corresponding to particular orientation of a section plane with respect to a bulk clathrate phase and the thickness measured by number of (111)-diamond monolayers NN, Table 1. For the sake of concise and uncluttered notation, unless otherwise stated, the respective NN will be rounded to the nearest integer value except for half-integers and optimal surface terminations with energy-minimizing reconstruction, detailed in § 7, for the given coverage ρ\rho are meant.

typePearsonsect.tilenotation
IcP46(100)squareSi.square-I-N
IIcF136(110)iso-triSi.iso-II-N
IIcF136(111)equi-triSi.equi-II-N
IIIhP40(001)equi-triSi.equi-III-N
IIIhP40(100)rectangleSi.rect-III-N
IVtP172(001)equi-tri + squareSi.dodeca-IV-N
–oP148–iso-tri + rect.Si.deca-N

Table 1. designation of an NN-monolayer-thick planar slab cut from the parent bulk clathrate, for which Pearson symbol is stated, by planar section with the respective clathrate tile information. Decagonal R2T4 approximant[33] (oP148) does not belong to any canonical clathrate I–IV type

Clathrate Reconstruction

Here we describe, for the first time, the optimal surface-energy minimizing reconstructions of planar clathrate sections listed in Table 1 for each of the four terminations from Equation (7).

The basic reconstruction element is again dumbbell, but unlike (111)-diamond surface termination, surface adatom S is located beneath some group of 3-coordinated surface atoms, whose dangling bonds it should saturate.

There are three basic positions[33] of a clathrate-surface adatom S (Figures S2.1, S3.1 in SM):

  1. Under cage faces aka “caps”:

    1a. Under 5-gonal cage facets: TcT_c, RcR_c, Tc′T'_c, Rc′R'_c.

    1b. Under 6-gonal cage facets: TcT_c, Tc′T'_c, ScS_c.

  1. Dumbbell under the 3-coordinated atom with:

    2a. all neighbours 4-coordinated: Tc′T'_c, Rc′R'_c, RcR_c, ScS_c.

    2b. two neighbours 3-coordinated: RhR_h, Rh′R'_h, ShS_h.

    2c. three neighbours 3-coordinated: Th′T'_h.

  1. Midedge position: under the bond of 3-coordinated atoms: ThT_h, RhR_h, ShS_h.

Tin-Specific Phases

Unlike Ge and Si, tin owing to its position at the bottom of Column-XIV lies on the borderline between semiconductors (C, Ge, Si) and metals (Pb), clearly exhibiting metallic properties at ambient conditions. At the level of (111)-diamond bilayer (ρ≃2\rho\simeq2) square and hexagonal metallic Sn films are competing with one another, Figure Figure 3. The couple of metallic phases are designated as “squares” and “triangles” in result Figure Figure 6.

Two Sn thin-film metallic phases: square and hexagonal

Figure 3. Two Sn thin-film metallic phases: square and hexagonal. Upper panel a.: Square Sn bilayer: front view and side view, firstly published in [21]. Lower panel b.: Hexagonal Sn bilayer: front and side view, yet unpublished, discovered via long MD DFT room-temperature runs in our work. Bottom panel: histograms of number of atoms per unit cell (delimited by blue line, excluding vacuum) per given zz coordinate perpendicular to the bilayer plane. Both phases have the front atoms exactly above the bottom atoms of a bilayer in the zz direction. Direction of view is indicated by arrows.

Whereas square Sn-I phase constitutes the ground state of Sn bilayer[21], hexagonal counterpart emerges during long MD or Replica Exchange (RE) annealings at room temperature. These observations suggest that hexagonal phase is favoured by (vibrational) entropy, corroborated by numerous defects of hexagonal polymorphs compared to the relatively perfect square one. All the metallic atomically-thin Sn films showed strong tendency towards the formation of even-layers.

Interestingly, both types of metallic bilayers can be viewed as two planar sections of γ\gamma-Sn phase[56, 57] in mutually perpendicular directions after symmetry breaking (slight displacement of a couple of atoms). Besides Sn9-web, Figure Figure 2, hexagonal Sn even-layers, are the second discovery of our DFT RE runs.

Despite being presented as stanene 4-layer, in the central Figure Figure 1a presumably is depicted triangular Sn 4-layer with hexagon centers filled with Sn atoms, experimentally prepared by non-equilibrium process: pulsed laser ablation of β\beta-Sn target floating on liquid hexane.

Amorphous Phases

Amorphous Tin.

As opposed to Ge and Si, in the case of bulk Sn there is no such a local minimum of the potential energy landscape deserving the title amorphous. More precisely, the aforementioned minima - even more than one mechanically stable structure - can be readily visited in the DFT universe; the real obstacle, however, resides in their experimental accessibility.

So as to observe a bulk-like amorphous Sn layer by quenched condensation on a cold substrate, it requires at least 8 atomic % of some other metal, e.g. Cu, and even then the resulting alloy crystallizes above approximately 20 K[58, 59, 60]. Another method – quenched simultaneous condensation of Sn and H on a substrate on nitrogen temperature – produces bulk-like amorphous semiconducting Sn, stable up to 180 K[61]. The other procedures include for example ion implantation of radioactive Sn isotopes [62], or intense electrolytic chemical reaction able to extract selected elements from the common precursor, such as sulphur from SnS2_2 crystal [63].

Possible realization through decompression of high-pressure intermediate obtained by compression of type-II clathrate[64] or α\alpha-tin[48, 65] seems impracticable for tin, compared to Si and Ge, since analogically DFT-generated metallic Sn disordered liquid (90 meV/at. above α\alpha-Sn) possesses by 10 meV/at. lower internal energy than sp3^3-amorphous Sn gained by a-Ge rescaling to the typical Sn-Sn distances, Table S1.1, SM. This metallic liquid then readily transforms to various metallic crystalline phases of Sn. The phase space of Sn is much more complicated than that of Si and Ge, and may well hide yet unexplored and even unexpected allotropes [66], though.

Amorphous Silicon.

Surprisingly, bulk a-Si can be prepared completely faultless without any coordination defects showing perfect gaussian bond-angle distribution [67], Figure 4, around (almost ideal) tetrahedral bond-angle 108.6° with the standard deviation from the narrow interval 9 − 11° [68, 69, 70, 71, 72, 73]. This was also the case of our samples generated using VASP MLFF suite trained on-the-fly during ab initio heating from 300 K to 900 K of the quenched melt (10410^4 5 fs-steps at 3000 K) followed by annealing at 900 K[74, 75, 76] (2×1052 \times10^5 MLFF 5 fs-steps) and slow cooling (3×1063 \times10^6 MLFF 5 fs-steps) back to room temperature within the standard DFT setup, Subsection 3.1. The density of thus prepared bulk a-Si, Figure 4, was 98.1% of α\alpha-Si (cF8) value, its internal energy being 158 meV/at. above Si.cF8, Table S1.1 (SM), falling into the interval 0.09 − 0.17 eV/at. stated in the literature[70, 77, 78, 79].

The effect of thorough ab initio annealing and relaxation of a-Si surface has been validated in solo paper[51]. It has been proven that DFT treatment leads to the significant relaxation, namely to the p3^3 and sp2^2 hybridizations of 3-coordinated surface atoms, with average bond angles θave\theta_{\mathrm{ave}} of 99.4° and 119.1°, respectively, quite distinct from 108° θave\theta_{\mathrm{ave}} of 3-fold bulk atoms. This effect is sizable, since 3-fold atoms comprise ∼22%\sim22\% of the surface; the rest consisting of ∼74%\sim74\% sp3^3- and 4% 5-coordination. Besides characteristic hybridizations, another intrinsic surface feature are 3-member rings accompanied by small 50° − 60° bond-angle peak, not found in the bulk a-Si distribution [51], approved by our DFT results, Figure 4.

The a-Si surface energy is supposed to be γ=1.05±0.14\gamma= 1.05 \pm0.14 J/m2^2[51], or 1.34 J/m2^2[75]. The latter paper states somewhat large γ\gamma – larger than quoted experimental values 1.14 J/m2^2[80], 1.23 J/m2^2[81], 1.24 J/m2^2[82], of ideal (111) diamond termination – but without proper annealing after a tearing test. (1 erg/cm2^2 = mJ/m2^2 and 1 eV/Å2^2 = 16.02176634 J/m2^2) According to Materials Project [22], the surface energy of reconstructed (111) Si.cF8 surface is 1.30 J/m2^{2}.

Our DFT computed (111) Si.cF8 unreconstructed surface has energy 1.57 J/m2^{2}, its c(2×4)\mathrm{c}(2 \times4) reconstruction[32] has surface energy γ=1.27\gamma= 1.27 J/m2^{2}, in good agreement with literature. Our a-Si slabs with coverages ranging from 1.75 − 4.98 ML (for unit conversion: 1 ML = 0.15469 at./Å2^{2} is the coverage of one Si diamond monolayer, 1 ML) possess energies from the interval 0.77−0.87 J/m2^{2} or 48−54 meV/Å2^{2}: as expected lower than (111)-cF8 termination; γ\gamma of partially crystallized (to cF8) a-Si samples attacking lower bound. Mind that computed value of a-Si γ\gamma strongly depends on the bulk a-Si energy via Equation (6) (EbulkE_{\mathrm{bulk}} term), which, according to the literature, is from the relatively broad interval, see above. For further comparison, surface energy of the best reconstructed Si clath-II slabs, Figure 7, is 0.865 J/m2^{2}, on the a-Si γ\gamma upper limit.

Structure of the most stable free-standing Sn, Ge, and Si clathrate type-II thin films

Figure 7. The structure of the most stable free-standing Sn (left panels), Ge (middle panels), Si (right panels) clathrate type-II thin films: their termination and reconstruction as a function of a chemical element. Shaded areas correspond to elementary-triangle clathrate tiles (isosceles for Si, equilateral for Ge and Sn). Periodic cell without vacuum is highlighted by solid blue line. Surface adatoms are bigger red balls. The integer in the sample’s label designates number of (111)-diamond ML in thickness (coverage ρ\rho). Top: front view; middle, bottom: side views of the same slab. View directions are provided in parentheses.

Amorphous Germanium

Despite structural properties of bulk a-Ge being the same as those of a-Si, Figure 4, and despite bulk a-Ge contrasted directly via a-Si rescaling to the characteristic Ge-Ge distances succeeded by atomic relaxation, we have observed for the first time that a-Ge surface features are completely different, Figure 5. Note that elemental Ge thin films were in fashion in 1970s and 1980s, when DFT was still in its infancy, hence all we know about structural units of a-Ge (the Ge thin film) surface comes from DFT simulations of our own.

Typical a-Ge surface

(a)

Side view of an a-Ge slab

(b)

Figure 5. Top: Typical a-Ge surface obtained by VASP on-the-fly MLFF from the melt by repeated annealing at 646 K and slow quench to 300 K. Periodic cell without vacuum is delimited by blue line. Most of the surface atoms are (approx. sp3\mathrm{sp}^{3}-) 4-coordinated: yellow balls. Red balls: over-coordinated adatoms. Blue balls: dangling bonds, mainly associated with 5- and 6-caps. Bottom: 4.2ML-thick slab, whose surface is plotted above. On the top panel, just 8Å-thick layer from the surface is depicted.

The density of bulk a-Ge comprises 96.68% of the Ge.cF8 value and is by 128 meV/at. unstable at 0 K to diamond, Table S1.1, SM. a-Ge slabs were prepared by long annealing slightly below Tc=650T_{\mathrm{c}} = 650 K, just as bulk a-Ge.

The surface energy of a-Ge slabs with characteristic 5-caps and 6-caps on the surface, Figure 5, lies in the interval (0.45 − 0.55) J/m2^{2} or equivalently (28 − 34) meV/Å2^{2}, approximately one half of unreconstructed (111) diamond, 1.07 J/m2^{2} (Materials Project states the value 1.11 J/m2^{2} [22]). These a-Ge clathrate-surface elements manifest themselves in the bond-angle distribution as a strong peak around 60°, Figure 4, even stronger than that of a-Si surface of different origin.

For further comparison, surface energy of (111) diamond can be lowered to 0.865 J/m2^{2} by 3×3\sqrt{3} \times\sqrt{3} reconstruction, which is still much higher than a-Ge values. Surface energy of the best clathrate reconstructions, 5-caps and 6-caps with one adatom, consult Figure S2.1 and Figure S3.1 (SM) and Figure 7, are 0.734 J/m2^{2} and 0.559 J/m2^{2}, respectively, again attacking upper threshold of a-Ge range, like in the a-Si case (but substantially lower than γ\gamma of reconstructed Ge diamond surface).

Optimal reconstruction of dodecagonal Sn, Ge and Si clathrate tiles

Figure S3.1. Optimal reconstruction of dodecagonal Sn, Ge and Si clathrate tiles: equilateral triangle (T) and square (S) terminated by cage “caps” (c) and “holes” (h). For the explanation of the symbols and color scheme, consult Figure S2.1. Note that placing the “caps” on the top of the “holes” does not change the number of unsaturated 3-coordinated atoms (distinguished by blue central dots); this holds also for the decagonal cases, see Figure S2.1.

Chiral Framework Structures

To the best of our knowledge, the only sp3^{3} G4 allotropes possessing lower internal energy per atom than type-II clathrate not structurally related to diamond are Chiral Framework Structures, CFS, or spirals: CFS-5 [83], CFS-6 [1] or NGS [84], T12 [85], see also Table S1.1, SM.

All three structures are fully sp3^{3}-hybridized, displaying main peak of the bond-angle distribution around 109.5° and do not contain any imaginary frequences supporting dynamical stability of these allotropes at ambient pressure [83, 85].

Lattice Gas Monte Carlo (LGMC) algorithm was deployed to seek after the best planar terminations for thin films. An attractive toy potential was applied that favored two CN=3 surface atoms over a couple of CN=2, CN=4 atoms: thus simulations tend to produce surfaces similar to the unreconstructed (111)-diamond (or clathrate). Simulation itself was a multiply repeated cooling while recording the lowest-energy configurations (by our own program).

LGMC research of CFS-6 ended in two surface terminations normal to (120) and (001) directions [(001) is the axis of the main spiral with hexagonal cross section], subjected to subsequent reconstructions for several Si, Ge and Sn film thicknesses.

For both sections two surface reconstructions were attempted: adatoms were firstly placed below each bond joining a couple of unsaturated atoms – (001)-1, (120)-1 – and secondly to the valley between them – (001)-2, (120)-2. As concerns thicknesses above 3.5 ML (Si), relaxations have just effect on the slab’s surface; the interior remains spiral-like and does not change its phase upon relaxation [86].

Atomic and cell relaxations left (120) (un)reconstructed surface intact, while surface of (001) termination has changed substantially and manifested 3- and 4-coordinated atoms only. As for (001)-1 and (001)-2 reconstructions, adatoms move towards the spiral’s main axis, saturating as many dangling bonds of the surrounding spiral as possible.

As expected from the already probed clathrate-surface reconstructions, Subsection 4.3.3, the only successful attempt was (120)-1 for Ge and Sn, resembling midge reconstruction of clath-II ThT_{\mathrm{h}} tiles, cf. Figure S2.1, S3.1 (SM), which was forbidden for Si, Figure 6. Si CFS-6, which is energetically degenerated with clath-II bulk, can not be cut and reconstructed effectively, therefore only unreconstructed Si.CFS surfaces were included in Figure 6 summarizing the chief results.

(120) termination is clearly better than (001) owing to the smaller area-density of the coordination defects (dangling bonds) and the absence of 2-coordinated atoms, but both are worse compared to the best clath-II planar sections, Figure 7. Clathrates exhibit lower per-atom energy and lower surface energy γ\gamma for all three elements and thicknesses under question.

As for CFS-5, there are two equivalent sections in directions (100) and (010), with just 3- and 4-coordinated atoms. In this case, 3-fold atoms are arranged in 4-atom chains per periodic box, exhibiting highest area-density of coordination defects among all CFS sections already mentioned. Several solo-adatom reconstructions were attempted, none of them being successful at lowering energy-per-atom or even surface energy γ\gamma, so compositions of several such adatoms seem to be also useless (absent in Figure 6).

Finally, natural planar termination of T12 normal to (001) direction with exclusively 3- and 4-coordinated surface atoms is the same as (120) cut of CFS-6, since both structures are stackings of equal 3-atom thick layers, having almost the same bulk energy per atom, wherefore the reconstructions of T12 were not executed.

Results

Our principal results are coverage ranges for which clathrate slabs are the ground state with ΔE(ρ)≡Eclath(ρ)−Eref(ρ)<0\Delta E(\rho) \equiv E_{\mathrm{clath}}(\rho) - E_{\mathrm{ref}}(\rho) < 0, where Eref(ρ)E_{\mathrm{ref}}(\rho) is energy of the closest competing structure, usually diamond. They are summarized in Table 2, Figure 6 (see also Figure S4.1, Figure S4.2 and Figure S4.3 in SM) and the surface reconstruction of per-atom-energy minimizing clathrate slabs corresponding to ρmax⁡\rho_{\max}, where ΔE\Delta E is maximal, are depicted in Figure 7. xyz-files of all stable thin films except for diamond are given in Section S9 (SM). Finally, in Subsection 5.4 a possible mechanism of electronic stabilization of clathrate thin films with respect to diamond ones is advocated.

elementclath-II surfaceρ0\rho_0ρmax⁡\rho_{\max} MLρ1\rho_1ΔEmax⁡\Delta E_{\max} meV/atom
Sn(111)3.23.210.0-35
Ge(111)&(110)2.53.47.4-34
Si(110)2.53.35.3-44

Table 2. Table 2 coverage intervals at which the ground-state slab is a clathrate (always type-II): columns ρ0\rho_0 and ρ1\rho_1 are endpoints of the clathrate stability intervals expressed in (111)-diamond-monolayer units (ML). ΔEmax⁡\Delta E_{\max} is the “maximum energy gain” relative to competitors at ρmax⁡\rho_{\max}

Comparison of per-atom energy difference between the most stable Sn clathrate slabs and Sn (111)-diamond ones for reconstructed and unreconstructed case

Figure S4.1. Comparison of per-atom energy difference between the most stable Sn clathrate slabs and Sn (111)-diamond ones for reconstructed and unreconstructed case.

Comparison of per-atom energy difference between the most stable Ge clathrate slabs and Ge (111)-diamond ones for reconstructed and unreconstructed case

Figure S4.2. Comparison of per-atom energy difference between the most stable Ge clathrate slabs and Ge (111)-diamond ones for reconstructed and unreconstructed case.

Comparison of per-atom energy difference between the most stable Si clathrate slabs and Si (111)-diamond ones for reconstructed and unreconstructed case

Figure S4.3. Comparison of per-atom energy difference between the most stable Si clathrate slabs and Si (111)-diamond ones for reconstructed and unreconstructed case.

Tin

For the lowest coverages below 2 ML, discontinuous structures are formed, depending on the size of a supercell. At ρ≈1.7\rho\approx1.7 ML, thread-like arrangements of Sn9_9, Sn10_{10} and Sn15_{15} units [52, 53] turn into a Sn9_9-web, Figure 2, Figure 6. These small Sn “houses” seem to be the energy-minimizing interface between metallic Sn bilayers with surrounding vacuum.

Around ρ≈2\rho\approx2 ML, uneven match among hexagonal and square metallic bilayers happens, Figure 3, designated as “triangles” and “squares” in Figure 6, respectively, squares being the clear winner by ≈30\approx30 meV/at., Subsection 4.4.

Finally, above ρ≈3\rho\approx3 ML up to ρ≈10\rho\approx10 ML, clathrate slabs get a word in, Table 2, Figure 6. For more than ρ≈10.4\rho\approx10.4 ML, approaching the bulk limit, thin-film clathrates are surpassed by reconstructed thick diamond layers.

Internal-energy-minimizing clathrate films are (111)-cut of type-II bulk showing equilateral-triangle ThT_h tiling in the section plane, Subsubsection 4.3.1, reconstructed equally by adatoms at the midedge positions (in the middle of each tile’s side), samples Sn.equi-II-N, N∈{3;5;7.5;10}N \in\{3; 5; 7.5; 10\}, § 21.1 (SM) and Figure 7. Owing to its metallic behaviour, Sn favours the highest area-density of adatoms from amid Group IV; this particular reconstruction presenting such an example, cf. Figure S3.1 (SM), resulting in 1.5 adatom per triangle tile – the largest admissible clathrate adatomic surface density.

As concerns CFS, Subsection 4.6, the best reconstructed surface, CFS-6 (120)-1, possesses both higher area-density of defects (in the form of 3-coordinated atoms) as well as greater bulk per-atom-energy, thus dropping out of the game, Figure 6.

Germanium

The most stable Si and Ge monolayer is oP10-reconstructed germanene and silicene, Subsection 4.1, Figure 6. Unlike metallic Sn, there is no known (meta)stable Ge or Si bilayer. Above ρ≈2.5\rho\approx2.5 ML clathrates clearly reign stable-thin-film configurational space, starting with Ge.iso-II-2.5 (S9.2.2 in SM) one-cage thick isosceles-triangle (110)-cut of type-II bulk composed of combined Th+Th′T_h + T'_h tiles, Figure S2.1 of SM, reconstructed by solo adatom per ThT_h tile (i.e. without midedge adatoms in Th′T'_h tiles).

Next, in the thickness range ρ∈[2.5;7.4]\rho\in[2.5; 7.4] ML, Table 2, just as in the case of Sn, (111)-cut clath-II films designated as Ge.equi-II-N, N∈{3;3.5;6}N \in\{3; 3.5; 6\}, § 22 (SM), are by far most stable. This time, however, the best termination are markedly deformed Th′T'_h 6-caps, Subsubsection 4.3.1, with one adatom each. The net result of thorough AIMD annealing and subsequent relaxation is transition of 6-cap to effectively 5-cap plus one atom, Figure 7, for all the Ge.equi-II-N samples. The configurational space above ρ≈7\rho\approx7 ML is obviously a bulk limit, where diamond films become dominant, regardless of reconstruction, Figure 6.

As CFS layers are concerned, the concluding remark from the preceding Subsection 5.1 holds, the best CFS-6 (120)-1 moving even further away from the convex hull compared to Sn.CFS-6.

Silicon

The area-density of adatoms gradually decreases in the sequence Sn → Ge → Si. From the G4, Si is most sensitive to adatom addition, unequivocally favouring isosceles-triangle (110)-cut of type-II bulk composed of combined Th+Th′T_h + T'_h tiles, Figure S2.1 of SM, in the whole range of Si clathrate-thin-film stability: ρ∈[2.5;5]\rho\in[2.5; 5] ML, Table 2, resulting in samples labeled Si.iso-II-N at discrete thicknesses N∈{2.5;4;5}N \in\{2.5; 4; 5\}, § 23 (SM). Like in the Ge case, they are reconstructed by solo adatom in the middle of ThT_h tile (i.e. without adatoms at midedge Th′T'_h positions).

According to Figure S4.1, Figure S4.2, Figure S4.3 (SM), clathrate films are already winning by cheaper surface termination before surface reconstruction, but proper reconstructions render them another important advantage over the diamond slabs.

As expected, for Si, the best CFS surface is unreconstructed, Figure 6, almost overlaying high-energy amorphous E(ρ)E(\rho) curve far from the overall convex envelope.

Note that despite the lower surface energy γ\gamma for both a-Ge and a-Si, Subsections 4.5.3 and 4.5.2, respectively, the stability of clathrate (and even diamond) slabs over amorphous ones, Table 6, is secured by their bulk energy, Table S1.1 (SM).

Electronic Stabilization of Clathrate Films

Electronic densities of states (eDOS) of the most stable Sn, Ge and Si clathrate slabs are displayed in Figure S8.2 (SM). Clathrates definitely enhance the tendency of the sp3^3 elements towards bandgap formation – this is clearly visible already in the bulk, see Figure S8.1 (SM).

Trends are evident both with respect to clathrate/diamond competition and as a function of coverage, because the thinnest 2.5 ML-slabs (samples Sn.equi-II-2.5, Ge.iso-II-2.5, Si.iso-II-2.5) exhibit distinct eDOS as well as ∼3.5\sim3.5 ML-thick ones (samples Ge.equi-II-3.5, Si.equi-III-3.5), whereas eDOS of thicker slabs approaches the bulk limit.

Comparing clathrate vs. diamond, ultrathin slabs confirm and enhance clathrate tendency for the gap formation, while the reconstructed diamond slabs tend to fill the electronic states just above or even across the Fermi energy. This trend must be the substantial reason for the enhanced clathrate stability.

Looking at the slab width variation, it is obvious that 2.5 ML-slab eDOS stand out loosing any correspondence with the bulk. 3.5 ML-thick Ge and Si slabs are distinguished by their bandgaps reaching the bulk-bandgap width, and are broader than bandgaps of thicker slabs of the same kind. These results are further substantiated by calculated values of electronic band gaps of the most stable slabs, Subsection S10 of SM, showing that clathrate thin films exhibit wider gaps (of the order of ∼0.1\sim0.1 eV in the thickness range considered, ρ∈[1;10]\rho\in[1; 10] ML) compared to the diamond ones of the same thickness.

Discussion

Here we attempt to crystallize the most stable clathrate thin films from the melt to directly prove the chief results of our work, Figure 6 and Figure 7. The highest expectations are from the a-Ge slabs, owing to the promising spontaneously reconstructed surface – 5caps and 6-caps with one or two adatoms, Figure 5 – that might serve as a condensation centers for clathrate cages.

Crystallization of Free-Standing Molten Si Slab

a-Si slabs, energetically unstable at 0 K with respect to both diamond and clathrate slabs for any coverage ρ\rho, Figure 6, crystallize exothermally into diamond after annealing at Tc∈[900−1100]T_{\mathrm{c}} \in[900-1100] K; the thicker slab the higher annealing temperature (due to the higher cohesive energy; Tc=1100T_{\mathrm{c}} = 1100 K constituting the bulk limit). All of our a-Si slabs were prepared either by long DFT MD (on-the-fly VASP MLFF) annealing slightly below TcT_{\mathrm{c}} or by Replica Exchange method covering the temperature range [400 K;Tc][400\ \mathrm{K}; T_{\mathrm{c}}].

In particular, partial crystallization of 3.2 ML-thick a-Si slab into cF8 was accompanied by 36 meV/at. energy release while annealing at 900 K and cooling down to 300 K. a-Si slabs of coverages ρ=3.9\rho= 3.9 ML and 4.4 ML lowered their internal energies by 28 meV/at. and 30 meV/at. during the same procedure, respectively. Crystallization of 5.0 ML-thick slab required annealing temperature at least 1000 K. The exothermic change of a-Si phase is an evidence that amorphous slabs can not compete with more stable diamond ones at 0 K at any coverage.

Crystallization of Free-Standing Molten Ge Slab

Unlike a-Si bulk surface containing exclusively p3^{3}-, sp2^{2}- and sp3^{3}-bonded atoms, Figure 4, a-Ge surface displays characteristic 5-caps and 6-caps with one or two adatoms, Figure 5, which can be regarded as nuclei of clathrate cages. Therefore we attempted to crystallize one of the most stable free-standing clathrate films – Ge.equi-II-3.5 – from the melt in fixed-volume simulation, Figure S6.1 (SM). Initial atomic configuration before melting can be found in Subsubsection 9.2.4 (SM).

Spontaneous re-crystallization of free-standing initially symmetrical Ge film

Figure S6.1. Spontaneous re-crystallization of free-standing initially symmetrical Ge film, sample designated as Ge.equi-II-3.5 in the main text, see Subsubsection S9.2.4 for the atomic coordinates.

Slow cooling down of 3000 K-melt to room temperature in 2×1042 \times10^{4} MD Δt=5\Delta t = 5 fs-steps produced (from both faces) symmetrical disordered slab composed mainly of eclipsed hexagons and pentagons, 120 meV/at. above original clathrate Ge.equi-II-3.5 48-atom sample.

Subsequent annealing 520 K → 350 K in 4×1064 \times10^{6} MLFF Δt=5\Delta t = 5 fs-steps resulted in asymmetrical diamond film with dumbbell reconstruction from one side and kind of warts from the other, Figure S6.1 (SM). This diamond slab, not surprisingly, was still 45 meV/at. above initial clathrate Ge.equi-II-3.5 sample.

A couple of long zero-pressure MLFF annealings at 450 K (or 400 K) and subsequent slow cooling to room temperature – ∼106\sim10^{6} steps 5 fs each – of 192-atom sample [2×2×1[2 \times2 \times1 supercell of 48atom Ge.equi-II-3.5 slab] in fixed periodic cell produced several characteristic 5-, 6- and even 7-cap features on the slab’s surface like in Figure 5, but with no vital signs of a clathrate cage. Phase transition to the original clathrate slab (before melting) was prohibited by a kinetic barrier, as the slab got stuck in amorphous configuration. The same holds for 240-atom sample shown in Figure 5, starting from random atomic positions, annealed millions of 1 fs-steps at fixed 650 K temperature (of bulk a-Ge to cF8-Ge transition) and cooled to room temperature during ∼104\sim10^{4} steps (ought to be longer).

There was just one purely MD simulation (before the advent of MLFF in VASP) with 150-atom slab at fixed temperature 400 K, half million of 5 fs-steps long, starting from random atomic positions, that produced single reconstructed clathrate cage.

Since transformation of the-best-clathrate-slab melt into a diamond one (cubic, not lonsdaleite) was well reproducible in many Si and Ge cases of distinct thicknesses, we assume that clathrates are kinetically forbidden. Next we shall try to crystallize them on specially designed clathrate-matching substrates [2].

Crystallization of Molten Ge Slab on InN

Even more interesting is the (re)crystallization of the same molten Ge.equi-II-3.5 layer on coherent InN binary-diamond substrate. The coherency of such an interface was the objective of article [2], where it has been shown that InN wafer gives Ge type-II or type-III clathrate growth fundamental advantage: its bulk-lattice misfit equals 2.3% instead of 12.6% misfit of Ge diamond film on the same InN substrate.

Recrystallization of Ge.equi-II-3.5 on InN led unexpectedly to a thin disordered Ge film connected with InN via a crystalline layer of regular 5-gons. Figure S6.2 (SM) [see Subsection S9.4.1 for initial xyz-file], 54 meV/at. above the original clathrate-diamond interface. The final result is probably due to the very 5-gonal buffer layer, which, as opposed to the coherent clathrate-diamond interface, saturates all the sp3\mathrm{sp}^{3} bonds from the diamond side. As has been already pointed out [2], clathrate is able to saturate at most 6 out of 9 dangling bonds on the diamond surface, Figure S6.3. Even more can be observed: not just all bonds of the substrate are saturated, but all bond angles with vertices at the substrate atoms are around the ideal tetrahedral angle (≈109.5∘\approx109.5^\circ), Figure S6.5, Figure S6.8. Ge melt constitutes an intermediate situation between Ge clathrate and recrystallized Ge layer from the point of view of bond angles; all N bonds are saturated even by the melt, Figure S6.4, Figure S6.7.

Three stages of spontaneous re-crystallization of Ge films on InN, showing energy versus time plots and atomic structures. Plot labels include MLFF learning, MLFF production, N = 5000, N = 10000, N = 500000, and Δt = 7 fs.

Figure S6.2. Spontaneous re-crystallization of Ge films on InN. Starting from the thinnest clathrate film with 6-caps, see Subsubsection S9.4.1 for the atomic coordinates.

Coherent clathrate/diamond interface made up of Ge/InN. The graphic shows a 32-atom interface, Ge and N legends, coordination-number tables, and a plot of P(θ) for Ge-Ge-Ge, Ge-Ge-N, and In-N-Ge.

Figure S6.3. Coherent clathrate/diamond interface made up of Ge/InN.

32-atom interface showing Ge and N atoms, coordination-number tables, and angular-distribution plot

Figure S6.4. Interface between diamond and clathrate melt after N=10000N = 10000 dt = 7 fs-steps at 1240 K, according to roadmap S6.2.

33-atom interface showing Ge and N atoms, coordination-number tables, and angular-distribution plot

Figure S6.5. Interface between diamond and re-crystallized clathrate melt according to Figure S6.2.

Figure S6.7. Interface between diamond and clathrate melt after N=5000N = 5000 dt =7= 7 fs-steps at 1300 K, according to Figure S6.6.

33-atom interface between diamond and re-crystallized clathrate melt, showing Ge and N atom coordination tables, a plot of $P(\theta)$ versus $\theta$ in degrees for Ge-Ge-Ge, Ge-Ge-N, and In-N-Ge, and the atomic interface structure

Figure S6.8. Interface between diamond and re-crystallized clathrate melt according to Figure S6.6.

The statement above holds for the nitrogen termination of InN substrate (from both sides), the more reactive element from the couple (In, N). If the substrate were terminated by In, it would not grant any benefit to an overlaying Ge clathrate at all. The same would be perhaps true for other electronically balanced substrates such as graphite or ZrO2_2, were it not for the covalent bonds e.g. between graphite and Ge-clathrate-film adatoms, already witnessed in DFT relaxations of our own.

When exactly the same experiment was carried out with thicker Ge layer, Ge.equi-II-6, the same 5-gonal buffer monolayer linked InN substrate with adjacent cubic-diamond Ge thin film, Figure S6.6 (SM), with some disordered mess on its outer surface. Thus it can be concluded that it is the very Ge buffer layer playing first fiddle on a N-terminated substrate. Mind, on the other hand, that the final diamond layer is by 38 meV/at. worse compared to the original clathrate-diamond setup, see Subsubsection S9.4.2 for the initial xyz-file, so the reason for diamond Ge output is kinetic, as well.

Spontaneous re-crystallization of Ge films on InN, with MLFF training and production energy plots and atomic configurations. Labels include MLFF training, MLFF production, energy, time, T: 600 → 1250K, T: 1300 → 1300K, T: 600 → 300K, N = 10000, N = 5000, N = 500000, Δt = 7fs, and ΔE = 38 meV/at.

Figure S6.6. Spontaneous re-crystallization of Ge films on InN. Starting from thicker clathrate film with 6-caps, see Subsubsection S9.4.2 for the atomic coordinates.

van der Waals Dispersion Correction

Karttunen et al. demonstrated that for silicon the dispersion interaction neglected in original PBE functional not only adds notable binding contribution of ∼0.3\sim0.3 eV/at. to the bulk systems, but also impacts considerably on energy differences due to the approximately linear dependence of the dispersion energy on the atomic density[87]. The same picture emerges also for germanium[88]. Both cited sources point out effectiveness of the Grimme’s et al. dispersion correction implemented along PBE functional (“PBE-D3”) in VASP[90, 91] [89] in comparison with other ways of treating dispersion within the DFT framework. Thus the dispersion correction practically doubles ΔE\Delta E between bulk clathrate and cubic diamond structure, from 50 to 100 meV/at. for Si, and from 30 to 90 meV/at. for Ge, concerning cubic type-II clathrate (Pearson cF136), owing to the somewhat higher density of the diamond phase[87].

However, the situation is different in the case of tin. According to references [88, 92] binding energy of cubic-diamond α\alpha-Sn is described accurately by PBE without dispersion correction, while PBE-D3 leads to overbinding by a clear margin. Added to this, PBE-D3 overestimates dramatically dispersion energy of the metallic β\beta-Sn structure ΔE\Delta E between α\alpha and β\beta from 42 meV/at. in favour of α\alpha to 43 meV/at. in favour of β\beta(!). Since vibrational free energy privileges β\beta (and we checked that the dispersion correction does not shift the phonon frequencies markedly), PBE-D3 contradicts the experimental fact the stable low-temperature phase being α\alpha up to approximately room temperature.

In Figure S7.1 (SM), the effect of PBE-D3 correction on our results for Si and Ge is displayed (IVDW=12 in VASP), but we refrain from doing the same for Sn.

Change of the diamond-clathrate stability competition in response to the van der Waals dispersion correction

Figure S7.1. Change of the diamond-clathrate stability competition in response to the van der Waals dispersion correction PBE-D3 (ivdw=12), visit Subsection 6.4 in the main text.

Effect of a long-range dispersion interaction in general increases with the thickness of the slab culminating in bulk, so the thinnest films in the focus of the present study are afflicted to much lesser extent. Nevertheless – as expected, the range of clathrate-film stability has been considerably retrenched by dispersion correction for both Ge and Si.

The original interval of Ge clath-II stability, ρ∈(2.5;7.4)\rho\in(2.5; 7.4) ML, Figure 6, reduced to ρ∈(3;5.5)\rho\in(3; 5.5) ML, Figure S7.1, leaving single (out of 4 stable slabs without vdW) directly constructed thin film Ge.equi-II-3.5 of thickness (coverage) ρ≈3.5\rho\approx3.5 ML on the overall convex hull: its energetic advantage over the diamond slab of the same coverage diminished from 34 meV/at. to 24 meV/at. (both below the diamond slab).

The original interval of Si clath-II stability, ρ∈(2.5;5.3)\rho\in(2.5; 5.3) ML, Figure 6, reduced to ρ∈(2.5;4.0)\rho\in(2.5; 4.0) ML, Figure S7.1, leaving solo (out of 3 stable slabs without vdW) directly constructed thin film Si.iso-II-2.5 of thickness (coverage) ρ≈2.5\rho\approx2.5 ML on the overall convex envelope: its energetic benefit over the diamond slab of the same coverage lessened from 44 meV/at. to 23 meV/at. (both with respect to the diamond slab as well).

Conclusions

This work exposes intervals of stability of several-ML thick clathrate layers, namely [2.5; 7.4] ML for Ge, [2.5; 5.3] ML for Si and [3.4; 10.0] ML for Sn, where ML stands for the unit thickness of (111)-diamond monolayer of the given element. Column-XIV clathrate slabs are stabilized by means of the surface energy in two respects: cheaper terminations and higher energy gain by reconstructions compared to the diamond films, meaning that clathrate films possess lower surface density of unsaturated dangling bonds than any known termination of sp3^3 allotrope and can be reconstructed most effectively. Within these thickness ranges lower surface energy reflects in lower internal energy per atom, which is taken as a stability measure.

Natural planar clathrate terminations, as well as the recipe for their reconstructions have been described for the first time, being superior to their diamond counterparts from the point of view of the surface energy.

Combined (do)decagonal rectangle(square)-triangle clathrate tilings were not efficient in lowering internal energy per atom, with respect to the pure triangular ones.

Apart from clathrate thin films, we have discovered several Sn nanolayer phases during long MD runs at room temperature (200 – 400 K), namely Sn9_9-web phase and hexagonal even-layers. In the latter case, the effect of entropy is playing decisive role in the layer stabilization since strongly defective triangular phase emerged at finite temperature and was presumably prepared during non-equilibrium processes[8], clearly far away (30 meV/at.) from the ground-state.

We hope that the present work instigates future experimental preparation of the suggested elemental stable clathrate thin films even in quasi-free-standing form on electronically balanced and/or amorphous substrates, whose novel unique properties would not be governed by the above-mentioned peculiar and often very complex intermetallic wafers (such as ii-Al-Pd-Mn or dd-Al-Ni-Co). Such a substrate should somehow disqualify the main rival - diamond - with respect to clathrates or could at least serve as a structure-neutral medium. The logical consequence would then be an investigation of influence of finite temperature on stability of clathrate thin layers.

Conflicts of interest

The author declares that she has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data supporting this article have been included as part of the Supplementary Information.

Acknowledgements

I acknowledge support from grants VEGA 2/0144/21, APVV-19-0369 and APVV-20-0124. I also acknowledge opportunity to write this article and finish the calculations after the defense of my Dissertation Thesis [86] entitled “Search for Stable Clathrate Films within Elements from Column XIV of the Periodic Table”, under supervisor Dr. Marek Mihalkovič, containing vast majority of material for this article, at the Institute of Materials and Machine Mechanics in Bratislava. Calculations were performed at the Computing Center of the Slovak Academy of Sciences using the supercomputing infrastructure acquired under Projects ITMS 26230120002 and 26210120002 and EuroHPC grant no. 101101903 (supercomputer Devana). Part of the research results was obtained using the computational resources procured in the national project National competence centre for high performance computing (project code: 311070AKF2) funded by European Regional Development Fund, EU Structural Funds Informatization of society, Operational Program Integrated Infrastructure.

Note: Invisible hyperlinks (white places) at the corners of every page of this Supplemental Material (SM) bring you back to this Table of Contents (TOC).

Note: Invisible hyperlinks (white places) at the corners of every page of this Supplemental Material (SM) bring you back to this Table of Contents (TOC).

Comparison of Column-XIV Bulk Energies

Surface Reconstruction of Decagonal Tiles

Surface Reconstruction of Dodecagonal Tiles

Reconstructed vs. Unreconstructed Surfaces

Figure S4.1 and Figure S4.3 below implies that Sn and Si clathrate slabs possess clear energetic advantage over the corresponding (111)-diamond films of the same coverage already before surface reconstruction (addition of Si atoms to the slab’s surface followed by atomic-position DFT relaxation). However, the appropriate surface reconstruction is responsible for further energy decrease of clathrates with respect to the diamond rivals.

Notably, as can be observed in Figure S4.2, the opposite situation happens for Ge, where clathrate’s internal-energy advantage over diamond is markedly more pronounced for the unreconstructed slabs.

In all the graphs below, the region of clathrate film stability is highlighted by green shadow, compare with Figure 6 of the main text. Reconstructed slabs correspond to the unreconstructed structures (“reconstructed” = “unreconstructed” + adatoms) in each case. As for the reconstructed ones, the most stable clathrate and diamond films for each coverage ρ\rho from Figure 6 of the main text were chosen for comparison, confer Results section.

Comparison of Various Types of Clathrate Slabs

Stability comparison plots for Sn, Ge and Si clathrate thin films

Figure S5.1. Stability comparison (DFT energy per atom at 0K) of seven types of Sn, Ge and Si clathrate thin films: clath-I square, clath-II equilateral triangle, clath-II isosceles triangle, clath-III equilateral triangle and rectangle, decagon and clath-IV dodecagonal tiling reconstructed according to Figure S2.1, Figure S3.1. Each cross represents one attempt for clath-tile reconstruction.

Crystallization of Clathrate Slabs from the Melt

van der Waals Dispersion Correction of G4 Slabs

Electronic Density of States: Clathrate Slabs & Bulk

xyz Files

Tin-Slab xyz Files

Sn9-web

ρ=0.1749\rho= 0.1749 Å−2^{-2} =1.667= 1.667 ML, E−Ebulk=216E - E_{\mathrm{bulk}} = 216 meV/at.

23.9138  0.00000  0.00000
0.00000  23.9138  0.00000
0.00000  0.00000  23.9138
100 atoms
0.96713  0.89507  0.02979  50  1  1
0.18163  0.30665  0.95156  50  2  2
0.00243  0.36734  0.04387  50  3  3
0.11868  0.91534  0.92111  50  4  4
0.18669  0.44073  0.96766  50  5  5
0.12692  0.38930  0.86865  50  6  6
0.81752  0.94640  0.91512  50  7  7
0.22004  0.78823  0.05703  50  8  8
0.88308  0.80335  0.06302  50  9  9
0.94704  0.92053  0.90656  50  10  10
0.08813  0.54847  0.12274  50  11  11
0.30671  0.70521  0.98685  50  12  12
0.04241  0.44174  0.94857  50  13  13
0.19810  0.35715  0.06952  50  14  14
0.86364  0.84115  0.85282  50  15  15
0.19066  0.17224  0.99467  50  16  16
0.42224  0.74919  0.03300  50  17  17
0.70438  0.93414  0.08492  50  18  18
0.54387  0.91304  0.98796  50  19  19
0.90148  0.00005  0.99533  50  20  20
0.44552  0.17554  0.93494  50  21  21
0.89188  0.65382  0.08673  50  22  22
0.08124  0.15403  0.07424  50  23  23
0.64025  0.85720  0.16452  50  24  24
0.93849  0.79543  0.94609  50  25  25
0.68329  0.64525  0.96937  50  26  26
0.98708  0.10118  0.00128  50  27  27
0.88829  0.53527  0.14043  50  28  28
0.07818  0.01873  0.03478  50  29  29
0.42503  0.29041  0.08684  50  30  30
0.24033  0.95059  0.90392  50  31  31
0.62870  0.00321  0.00457  50  32  32
0.18179  0.07626  0.08287  50  33  33
0.36438  0.36431  0.89509  50  34  34
0.98208  0.56142  0.06184  50  35  35
0.694230.897190.96087503636
0.502530.387360.05731503737
0.457730.043320.91417503838
0.825450.916950.04931503939
0.550730.817380.08054504040
0.066420.184670.94560504141
0.349770.856900.04637504242
0.102120.436890.07187504343
0.161420.635100.06777504444
0.801360.823020.96140504545
0.619600.675390.07927504646
0.100480.749350.06073504747
0.355410.104220.87435504848
0.773260.707270.90756504949
0.351390.544930.98922505050
0.444850.628990.97813505151
0.682300.806900.05652505252
0.601030.807390.95993505353
0.833740.509910.93448505454
0.097700.280420.05107505555
0.258040.827220.93791505656
0.230320.913940.03167505757
0.976140.663150.98586505858
0.453750.913360.09431505959
0.858350.632660.95822506060
0.324380.365990.10798506161
0.382380.809740.92592506262
0.311870.423310.99262506363
0.040160.957120.83578506464
0.423340.928290.96173506565
0.061810.055520.90817506666
0.777600.718420.03918506767
0.109630.544410.98974506868
0.231810.602400.95877506969
0.683400.563690.06703507070
0.170410.717220.95411507171
0.529310.102500.00184507272
0.282990.614210.08203507373
0.181440.046560.96071507474
0.094890.890370.04486507575
0.161460.805020.86034507676
0.904540.462660.03091507777
0.575830.952240.11153507878
0.446580.441570.95606507979
0.562000.684180.96535508080
0.959510.541630.93519508181
0.330260.083750.08606508282
0.440120.031000.05198508383
0.807610.555700.04933508484
0.320170.735290.11244508585

Table 3.

0.33400  0.29204  0.99652  50  86  86
0.67111  0.52100  0.93956  50  87  87
0.41257  0.58602  0.10125  50  88  88
0.06569  0.80185  0.94906  50  89  89
0.04873  0.31091  0.93358  50  90  90
0.33098  0.02608  0.97160  50  91  91
0.40518  0.46092  0.08388  50  92  92
0.45882  0.30776  0.96075  50  93  93
0.43056  0.16401  0.06316  50  94  94
0.61008  0.61374  0.87023  50  95  95
0.32467  0.16033  0.98493  50  96  96
0.53509  0.53990  0.93945  50  97  97
0.53490  0.58237  0.06596  50  98  98
0.15668  0.13509  0.87445  50  99  99
0.59489  0.47624  0.03422  50  100  100

triangles

ρ=0.2225 A˚−2=2.121 ML, E−Ebulk=198 meV/at.\rho= 0.2225\ \mathrm{\AA}^{-2} = 2.121\ \mathrm{ML},\ E - E_{\mathrm{bulk}} = 198\ \mathrm{meV/at}.

3.55789  -0.00806  0.00000
-3.56552  4.68674  0.00000
0.00000  0.00000  28.0000
4 atoms
0  0.25  0.07803  50  1  1
0  0.75  0.92197  50  2  2
0  0.25  0.96739  50  3  3
0  0.75  0.03261  50  4  4

squares

ρ=0.2448 A˚−2=2.333 ML, E−Ebulk=158 meV/at.\rho= 0.2448\ \mathrm{\AA}^{-2} = 2.333\ \mathrm{ML},\ E - E_{\mathrm{bulk}} = 158\ \mathrm{meV/at}.

5.72637  0.00000  0.00000
0.00000  5.72637  0.00000
0.00000  0.00000  20.5771
8 atoms
0  0  0.56989  50  1  1
0.5  0.5  0.56989  50  2  2
0  0  0.41700  50  3  3
0.5  0.5  0.41700  50  4  4
0  0.5  0.62798  50  5  5
0.5  0  0.35890  50  6  6
0  0.5  0.47756  50  7  7
0.5  0  0.50932  50  8  8

Sn.equi-II-3

ρ=0.3428 A˚−2=3.268 ML, E−Ebulk=142 meV/at.\rho= 0.3428\ \mathrm{\AA}^{-2} = 3.268\ \mathrm{ML},\ E - E_{\mathrm{bulk}} = 142\ \mathrm{meV/at}.

12.44832   -0.00000   -0.00008
-6.22416   10.78059   -0.00011
-0.00022   -0.00049   34.25425
46 atoms
0.33338   0.05896   0.26941   50   1   1
0.94107   0.27436   0.26942   50   2   2
0.72567   0.66667   0.26941   50   3   3
0.70092   0.03421   0.26532   50   4   4
0.33337   0.29911   0.26533   50   5   5
0.96582   0.66667   0.26533   50   6   6
0.95506   0.28834   0.02648   50   7   7
0.33331   0.04491   0.02648   50   8   8
0.71162   0.66659   0.02648   50   9   9
0.33332   0.28934   0.02653   50   10   10
0.71067   0.04395   0.02652   50   11   11
0.95605   0.66661   0.02653   50   12   12
0.53980   0.46057   0.07974   50   13   13
0.53938   0.07916   0.07972   50   14   14
0.92083   0.46015   0.07974   50   15   15
0.75269   0.87046   0.21963   50   16   16
0.11783   0.24731   0.21964   50   17   17
0.12956   0.88219   0.21964   50   18   18
1.00000   0.99995   0.10916   50   19   19
0.66670   0.33332   0.19192   50   20   20
0.54890   0.46285   0.21964   50   21   21
0.53716   0.08599   0.21963   50   22   22
0.91402   0.45112   0.21965   50   23   23
0.74587   0.87267   0.07972   50   24   24
0.12686   0.25411   0.07975   50   25   25
0.12729   0.87309   0.07974   50   26   26
0.00003   0.99999   0.19192   50   27   27
0.66667   0.33329   0.10916   50   28   28
0.46594   0.53418   0.15057   50   29   29
0.46583   0.93171   0.15056   50   30   30
0.06830   0.53407   0.15058   50   31   31
0.59842   0.79912   0.15056   50   32   32
0.20077   0.40159   0.15058   50   33   33
0.20089   0.79923   0.15057   50   34   34
0.07730   0.41060   0.34526   50   35   35
0.12885   0.66668   0.32593   50   36   36
0.33338   0.46215   0.32593   50   37   37
0.33339   0.92276   0.34525   50   38   38
0.58947   0.66668   0.34525   50   39   39
0.53791   0.87120   0.32592   50   40   40
0.19051   0.52380   0.26067   50   41   41
0.33337   0.80952   0.26066   50   42   42
0.47623   0.66666   0.26066   50   43   43
0.83399   0.66656   0.10590   50   44   44
0.83265   0.16593   0.10590   50   45   45
0.33327   0.16728   0.10590   50   46   46

Sn.equi-II-5

ρ=0.5501 A˚−2=5.244 ML, E−Ebulk=96 meV/at.\rho= 0.5501\ \text{Å}^{-2} = 5.244\ \text{ML},\ E - E_{\text{bulk}} = 96\ \text{meV/at.}

12.46276   0.00004   0.00003
-6.23135  10.79309  -0.00004
 0.00009  -0.00009  41.74180
74 atoms
0.33759  0.05312  0.21120  50  1   1
0.94690  0.28451  0.21120  50  2   2
0.71549  0.66242  0.21120  50  3   3
0.71549  0.05310  0.21120  50  4   4
0.33758  0.28451  0.21120  50  5   5
0.94688  0.66241  0.21120  50  6   6
0.95688  0.28649  0.01775  50  7   7
0.32957  0.04308  0.01775  50  8   8
0.71350  0.67043  0.01775  50  9   9
0.32957  0.28649  0.01775  50  10  10
0.71350  0.04311  0.01775  50  11  11
0.95692  0.67043  0.01775  50  12  12
0.61782  0.61360  0.27509  50  13  13
0.99573  0.38219  0.27509  50  14  14
0.38642  0.00428  0.27509  50  15  15
0.99572  0.61358  0.27509  50  16  16
0.38640  0.38218  0.27509  50  17  17
0.61781  0.00427  0.27509  50  18  18
0.53981  0.46019  0.06000  50  19  19
0.53981  0.07966  0.06000  50  20  20
0.92034  0.46019  0.06000  50  21  21
0.20780  0.79221  0.30980  50  22  22
0.20780  0.41564  0.30981  50  23  23
0.58436  0.79220  0.30981  50  24  24
0.74895  0.87449  0.17649  50  25  25
0.12551  0.25105  0.17649  50  26  26
0.12551  0.87449  0.17649  50  27  27
0.99999  0.00001  0.08775  50  28  28
0.66665  0.33335  0.33446  50  29  29
0.66665  0.33334  0.15183  50  30  30
0.54290  0.45710  0.17347  50  31  31
0.54290  0.08584  0.17347  50  32  32
0.91416  0.45710  0.17347  50  33  33
0.74857  0.87430  0.06230  50  34  34
0.12570  0.25143  0.06230  50  35  35
0.12570  0.87430  0.06230  50  36  36
0.41914  0.20960  0.31282  50  37  37
0.79040  0.58086  0.31282  50  38  38
0.79041  0.20960  0.31282  50  39  39
0.99999  0.00001  0.15577  50  40  40
0.66665  0.33335  0.08417  50  41  41
0.33332  0.66668  0.33051  50  42  42
0.46549  0.53451  0.11781  50  43  43
0.46549  0.93101  0.11781  50  44  44
0.06899  0.53451  0.11781  50  45  45
0.13203  0.86797  0.36675  50  46  46
0.13204  0.26412  0.36675  50  47  47
0.73588  0.86797  0.36675  50  48  48
0.59743  0.79873  0.11954  50  49  49
0.20127  0.40256  0.11954  50  50  50
0.20127  0.79873  0.11955  50  51  51
0.00375  0.38021  0.46855  50  52  52
0.61980  0.62359  0.46855  50  53  53
0.37638  0.99625  0.46855  50  54  54
0.37641  0.38021  0.46855  50  55  55
0.00375  0.62362  0.46855  50  56  56
0.61980  0.99625  0.46855  50  57  57
0.41297  0.20651  0.42629  50  58  58
0.79349  0.58703  0.42629  50  59  59
0.79349  0.20651  0.42630  50  60  60
0.33332  0.66668  0.39854  50  61  61
0.20761  0.79239  0.42399  50  62  62
0.20762  0.41528  0.42399  50  63  63
0.58472  0.79238  0.42399  50  64  64
0.66666  0.33335  0.40212  50  65  65
0.26433  0.13219  0.36849  50  66  66
0.86781  0.73567  0.36849  50  67  67
0.86781  0.13219  0.36849  50  68  68
0.33339  0.16671  0.08294  50  69  69
0.83328  0.16672  0.08294  50  70  70
0.83329  0.66661  0.08294  50  71  71
0.50003  0.49997  0.40335  50  72  72
0.99991  0.49998  0.40336  50  73  73
0.50002  0.00010  0.40336  50  74  74

Sn.equi-II-7.5

ρ=0.7981 A˚−2=7.608 ML, E−Ebulk=75 meV/at.\rho= 0.7981\,\mathrm{\AA}^{-2} = 7.608\,\mathrm{ML},\ E - E_{\mathrm{bulk}} = 75\,\mathrm{meV/at.}

12.50059  -0.00001  0.00007
-6.25030  10.82582  -0.00009
0.00033  -0.00026  52.72611
108 atoms
0.33844  0.05386  0.16954  50  1  1
0.94615  0.28458  0.16954  50  2  2
0.71543  0.66156  0.16954  50  3  3
0.29003  0.95257  0.82048  50  4  4
0.66258  0.70993  0.82048  50  5  5
0.047430.337410.820485066
0.662580.952570.820485077
0.047420.709960.820485088
0.290060.337420.820485099
0.715420.053850.16954501010
0.338440.284580.16954501111
0.946140.661560.16954501212
0.951260.279470.02386501313
0.328230.048740.02386501414
0.720540.671770.02386501515
0.328230.279450.02386501616
0.720530.048740.02386501717
0.951250.671770.02386501818
0.671460.720210.97325501919
0.279800.951210.97325502020
0.048790.328540.97325502121
0.617900.613130.22015502222
0.995220.382110.22015502323
0.386890.004780.22015502424
0.995220.613110.22015502525
0.386880.382100.22015502626
0.617890.004780.22015502727
0.048780.720200.97325502828
0.671450.951200.97325502929
0.279780.328540.97325503030
0.541300.458700.05090503131
0.541300.082580.05090503232
0.917420.458700.05090503333
0.082300.541130.94596503434
0.458870.917700.94596503535
0.458870.541130.94596503636
0.207810.792200.24744503737
0.207800.415620.24744503838
0.584390.792200.24744503939
0.749260.874630.14250504040
0.125380.250740.14250504141
0.125380.874620.14250504242
0.873380.126620.85368504343
0.873380.746710.85368504444
0.253280.126620.85368504545
0.000010.999990.07199504646
0.000010.999980.92623504747
0.666660.333340.26717504848
0.666670.333330.12142504949
0.333350.666650.87583505050
0.543610.456390.13955505151
0.543610.087200.13955505252
0.912800.456400.13955505353
0.082460.541210.85570505454
0.458790.917540.85570505555
0.458790.541200.85571505656
0.753880.876930.05385505757
0.123070.246120.05385505858
0.123070.876930.05385505959
0.876450.123550.94334506060
0.876450.752860.94334506161
0.247140.123550.94334506262
0.419520.209770.25007506363
0.790240.580480.25007506464
0.790240.209770.25007506565
0.000011.000000.12575506666
0.000010.999990.87267506767
0.666670.333330.06766506868
0.333350.666650.92969506969
0.333330.666670.26371507070
0.465140.534860.09582507171
0.465140.930250.09582507272
0.069750.534870.09582507373
0.930450.465200.90088507474
0.534800.069550.90088507575
0.534800.465200.90088507676
0.131870.868130.29252507777
0.131870.263760.29252507878
0.736250.868130.29252507979
0.596930.798460.09758508080
0.201540.403070.09758508181
0.201540.798460.09758508282
0.798760.201230.89937508383
0.798760.597480.89937508484
0.402520.201230.89937508585
0.004140.380780.37293508686
0.619230.623370.37292508787
0.376600.995870.37292508888
0.376630.380770.37292508989
0.004140.623400.37293509090
0.619230.995870.37293509191
0.413410.206710.33973509292
0.793300.586600.33973509393
0.793290.206710.33973509494
0.333330.666670.31756509595
0.207890.792120.33769509696
0.207890.415790.33769509797
0.584210.792120.33769509898
0.666670.333340.32073509999
0.264170.132090.2940450100100
0.867910.735840.2940450101101
0.867910.132090.2940550102102
0.166740.333430.8720950103103
0.666560.833250.8720950104104
0.166730.833270.8720950105105

Table 1 (PDF p. 42).

0.49994 0.50006 0.32131 50 106 106 0.00013 0.50007 0.32132 50 107 107 0.49993 0.99988 0.32132 50 108 108

Sn.equi-II-10

ρ=1.047\rho= 1.047 Å−2^{-2} = 9.978 ML, E−Ebulk=65E - E_{\mathrm{bulk}} = 65 meV/at.

12.51618 0.00000 0.00000 -6.25809 10.83933 0.00000 0.00000 0.00000 62.18220 142 atoms 0.33865 0.05400 0.14502 50 1 1 0.94600 0.28466 0.14502 50 2 2 0.71534 0.66135 0.14502 50 3 3 0.28466 0.94600 0.85498 50 4 4 0.66135 0.71534 0.85498 50 5 5 0.05400 0.33865 0.85498 50 6 6 0.66135 0.94600 0.85498 50 7 7 0.05400 0.71534 0.85498 50 8 8 0.28466 0.33865 0.85498 50 9 9 0.71534 0.05400 0.14502 50 10 10 0.33865 0.28466 0.14502 50 11 11 0.94600 0.66135 0.14502 50 12 12 0.00492 0.38702 0.81203 50 13 13 0.61298 0.61791 0.81203 50 14 14 0.38209 0.99508 0.81203 50 15 15 0.95141 0.27916 0.02146 50 16 16 0.32776 0.04859 0.02146 50 17 17 0.72084 0.67224 0.02146 50 18 18 0.32776 0.27916 0.02146 50 19 19 0.72084 0.04859 0.02146 50 20 20 0.95141 0.67224 0.02146 50 21 21 0.38209 0.38702 0.81203 50 22 22 0.00492 0.61791 0.81203 50 23 23 0.61298 0.99508 0.81203 50 24 24 0.67224 0.72084 0.97854 50 25 25 0.27916 0.95141 0.97854 50 26 26 0.04859 0.32776 0.97854 50 27 27 0.61791 0.61298 0.18797 50 28 28 0.99508 0.38209 0.18797 50 29 29 0.38702 0.00492 0.18797 50 30 30 0.99508 0.61298 0.18797 50 31 31 0.38702 0.38209 0.18797 50 32 32 0.61791 0.00492 0.18797 50 33 33 0.04859 0.72084 0.97854 50 34 34 0.67224 0.95141 0.97854 50 35 35

0.279160.327760.97854503636
0.541190.458810.04431503737
0.541190.082390.04431503838
0.917610.458810.04431503939
0.082390.541190.95569504040
0.458810.917610.95569504141
0.458810.541190.95569504242
0.207800.792200.21105504343
0.207800.415600.21105504444
0.584400.792200.21105504545
0.749260.874630.12216504646
0.125370.250740.12216504747
0.125370.874630.12216504848
0.874630.125370.87784504949
0.874630.749260.87784505050
0.250740.125370.87784505151
0.415600.207800.78895505252
0.792200.584400.78895505353
0.792200.207800.78895505454
0.000000.000000.06239505555
0.000000.000000.93761505656
0.666670.333330.22778505757
0.666670.333330.10411505858
0.333330.666670.89589505959
0.333330.666670.77222506060
0.543660.456340.11954506161
0.543660.087310.11954506262
0.912680.456340.11954506363
0.087310.543660.88045506464
0.456340.912680.88045506565
0.456340.543660.88045506666
0.209760.790240.78666506767
0.209760.419510.78666506868
0.580490.790240.78666506969
0.754020.877010.04699507070
0.122990.245980.04699507171
0.122990.877010.04699507272
0.877010.122990.95301507373
0.877010.754020.95301507474
0.245980.122990.95301507575
0.419510.209760.21334507676
0.790240.580490.21334507777
0.790240.209760.21334507878
0.000000.000000.10800507979
0.000000.000000.89200508080
0.666670.333330.77518508181
0.666670.333330.05849508282
0.333330.666670.94151508383
0.333330.666670.22482508484
0.465110.534890.08246508585
0.465110.930220.08246508686
0.069780.534890.08246508787
0.930220.465110.91754508888
0.534890.069780.91754508989
0.534890.465110.91754509090
0.131890.868110.24930509191
0.131890.263780.24930509292
0.736220.868110.24930509393
0.596950.798470.08405509494
0.201530.403050.08405509595
0.201530.798470.08405509696
0.798470.201530.91595509797
0.798470.596950.91595509898
0.403050.201530.91595509999
0.263780.131890.7507050100100
0.868110.736220.7507050101101
0.868110.131890.7507050102102
0.004110.380750.3174850103103
0.619250.623350.3174850104104
0.376650.995890.3174850105105
0.376650.380750.3174850106106
0.004110.623350.3174850107107
0.619250.995890.3174850108108
0.413400.206700.2893250109109
0.793300.586600.2893250110110
0.793300.206700.2893250111111
0.333330.666670.2705150112112
0.207930.792070.2876050113113
0.207930.415860.2876050114114
0.584140.792070.2876050115115
0.666670.333330.2732250116116
0.263950.131980.2506050117117
0.868020.736040.2506050118118
0.868020.131980.2506050119119
0.623350.619250.6825250120120
0.995890.376650.6825250121121
0.380750.004110.6825250122122
0.995890.619250.6825250123123
0.380750.376650.6825250124124
0.623350.004110.6825250125125
0.206700.793300.7106850126126
0.206700.413400.7106850127127
0.586600.793300.7106850128128
0.666670.333330.7294950129129
0.415860.207930.7124050130130
0.792070.584140.7124050131131
0.792070.207930.7124050132132
0.333330.666670.7267850133133
0.131980.868020.7494050134134
0.131980.263950.7494050135135
0.736040.868020.7494050136136
0.500050.499950.7262450137137
0.999900.499950.7262450138138
0.500050.000100.7262450139139
0.499950.500050.2737650140140
0.000100.500050.2737650141141
0.499950.999900.2737650142142

Table 4.

Germanium-Slab xyz Files

Ge.db-monolayer (oP10)

ρ=0.1884\rho= 0.1884 Å−2^{-2} =1.359= 1.359 ML, E−Ebulk=308E - E_{\mathrm{bulk}} = 308 meV/at.

7.84310  0.00000  0.00000
0.00000 38.86961  0.00000
0.00000  0.00000  6.76570
10 atoms
0.02034  0.00000  0.81941  32  1  1
0.47966  0.00000  0.81941  32  2  2
0.52034  0.00000  0.18059  32  3  3
0.97966  0.00000  0.18059  32  4  4
0.25000  0.03815  0.67495  32  5  5
0.25000  0.96185  0.67495  32  6  6
0.75000  0.96185  0.32505  32  7  7
0.75000  0.03815  0.32505  32  8  8
0.25000  0.00000  0.37240  32  9  9
0.75000  0.00000  0.62760  32 10 10

Ge.iso-II-2.5

ρ=0.3534\rho= 0.3534 Å−2^{-2} =2.548= 2.548 ML, E−Ebulk=251E - E_{\mathrm{bulk}} = 251 meV/at.

15.67423  0.00000  0.00000
 0.00000 11.19385  0.00000
 0.00000  0.00000 28.11850
62 atoms
0.00001  0.00000  0.00000  32  1  1
0.36467  0.00000  0.10991  32  2  2
0.36467  0.00000  0.89009  32  3  3
0.50130  0.00000  0.15808  32  4  4
0.50130  0.00000  0.84192  32  5  5
0.76010  0.00000  0.04475  32  6  6
0.76010  0.00000  0.95525  32  7  7
0.91225  0.00000  0.07406  32  8  8
0.91225  0.00000  0.92595  32  9  9
0.49329  0.00000  0.05356  32 10 10
0.49329  0.00000  0.94644  32 11 11
0.24297  0.11297  0.00000  32 12 12
0.24297  0.88703  0.00000  32 13 13
0.55267  0.16644  0.10305  32 14 14
0.55267  0.83356  0.89694  32 15 15
0.55267  0.16644  0.89694  32 16 16
0.55267  0.83356  0.10305  32 17 17
0.30975  0.18824  0.07465  32 18 18
0.30975  0.81176  0.92535  32 19 19
0.30975  0.18824  0.92535  32 20 20
0.30975  0.81176  0.07465  32  21  21
0.95601  0.17575  0.12814  32  22  22
0.95601  0.82425  0.87186  32  23  23
0.95601  0.17575  0.87186  32  24  24
0.95601  0.82425  0.12814  32  25  25
0.08999  0.18182  0.00000  32  26  26
0.08999  0.81818  0.00000  32  27  27
0.69977  0.19543  0.07184  32  28  28
0.69977  0.80456  0.92816  32  29  29
0.69977  0.19543  0.92816  32  30  30
0.69977  0.80456  0.07184  32  31  31
0.05137  0.29872  0.07307  32  32  32
0.05137  0.70128  0.92693  32  33  33
0.05137  0.29872  0.92693  32  34  34
0.05137  0.70128  0.07307  32  35  35
0.66295  0.31657  0.00000  32  36  36
0.66295  0.68343  0.00000  32  37  37
0.83205  0.29264  0.10315  32  38  38
0.83205  0.70736  0.89685  32  39  39
0.83205  0.29264  0.89685  32  40  40
0.83205  0.70736  0.10315  32  41  41
0.44360  0.31369  0.07471  32  42  42
0.44360  0.68631  0.92529  32  43  43
0.44360  0.31369  0.92529  32  44  44
0.44360  0.68631  0.07471  32  45  45
0.19963  0.33467  0.10130  32  46  46
0.19963  0.66533  0.89870  32  47  47
0.19963  0.33467  0.89870  32  48  48
0.19963  0.66533  0.10130  32  49  49
0.51078  0.38706  0.00000  32  50  50
0.51078  0.61294  0.00000  32  51  51
0.25094  0.50000  0.15702  32  52  52
0.25094  0.50000  0.84298  32  53  53
0.26308  0.50000  0.05291  32  54  54
0.26308  0.50000  0.94709  32  55  55
0.38975  0.50000  0.11150  32  56  56
0.38975  0.50000  0.88850  32  57  57
0.75200  0.50000  0.00000  32  58  58
0.84553  0.50000  0.07091  32  59  59
0.84553  0.50000  0.92909  32  60  60
0.99777  0.50000  0.04467  32  61  61
0.99777  0.50000  0.95533  32  62  62

Ge.equi-II-3

ρ=0.4037\rho= 0.4037 Å−2^{-2} = 2.911 ML, E−Ebulk=222E - E_{\mathrm{bulk}} = 222 meV/at.

10.88372   -0.00000   0.00204
-5.44186    9.33195  -0.00089
0.00667     0.00048  34.96546
41 atoms
0.33470   0.04609   0.25061   50   1   1
0.95527   0.28723   0.25061   50   2   2
0.72578   0.66666   0.24872   50   3   3
0.71201   0.04620   0.25327   50   4   4
0.33247   0.28712   0.25327   50   5   5
0.96033   0.66666   0.25513   50   6   6
0.92142   0.25204   0.07214   50   7   7
0.33606   0.08131   0.07214   50   8   8
0.74943   0.66668   0.07242   50   9   9
0.33610   0.29972   0.04765   50   10   10
0.70306   0.03364   0.04765   50   11   11
0.96650   0.66668   0.04764   50   12   12
0.53293   0.44755   0.09586   50   13   13
0.55305   0.08189   0.09515   50   14   14
0.91793   0.46807   0.09555   50   15   15
0.75604   0.87970   0.21275   50   16   16
0.12609   0.24659   0.21253   50   17   17
0.12317   0.87502   0.21254   50   18   18
0.00216   0.00060   0.11854   50   19   19
0.66819   0.33376   0.18882   50   20   20
0.54300   0.45362   0.21275   50   21   21
0.54616   0.08674   0.21253   50   22   22
0.91481   0.45831   0.21254   50   23   23
0.75205   0.88580   0.09586   50   24   24
0.13783   0.25146   0.09515   50   25   25
0.11654   0.86528   0.09555   50   26   26
0.00110   0.99957   0.18882   50   27   27
0.66823   0.33274   0.11854   50   28   28
0.46653   0.53371   0.15486   50   29   29
0.46666   0.92995   0.15442   50   30   30
0.06985   0.53440   0.15449   50   31   31
0.59949   0.79963   0.15486   50   32   32
0.20338   0.40338   0.15442   50   33   33
0.20212   0.79894   0.15449   50   34   34
0.12584   0.49295   0.29437   50   35   35
0.05444   0.66665   0.32424   50   36   36
0.32321   0.44254   0.30830   50   37   37
0.29955   0.84037   0.29437   50   38   38
0.55285   0.66666   0.29650   50   39   39
0.54732   0.89077   0.30830   50   40   40
0.36303   0.66666   0.24997   50   41   41

Ge.equi-II-3.5

ρ=0.4781\rho= 0.4781 A˚−2\mathrm{\AA}^{-2} =3.448= 3.448 ML, E−Ebulk=183E - E_{\mathrm{bulk}} = 183 meV/at.

10.71370   -0.01100   -0.00829
-5.49272   9.36400   -0.25389
-0.21007   -0.40637   30.60224
48 atoms
0.03938   0.81651   0.19581   32   1   1
0.04665   0.82625   0.27601   32   2   2
0.11219   0.35595   0.23590   32   3   3
0.13361   0.25074   0.05846   32   4   4
0.21634   0.29959   0.39743   32   5   5
0.18713   0.50189   0.07438   32   6   6
0.16572   0.49868   0.41335   32   7   7
0.15609   0.05849   0.16867   32   8   8
0.17348   0.07731   0.30260   32   9   9
0.16140   0.68985   0.16920   32   10   10
0.16828   0.70342   0.30314   32   11   11
0.24190   0.22156   0.23555   32   12   12
0.24199   0.62007   0.23625   32   13   13
0.36033   0.24629   0.07152   32   14   14
0.39287   0.23103   0.43153   32   15   15
0.35751   0.74388   0.04028   32   16   16
0.39007   0.72863   0.40028   32   17   17
0.36533   0.08255   0.12695   32   18   18
0.38570   0.12853   0.35199   32   19   19
0.36468   0.84638   0.11982   32   20   20
0.38505   0.89237   0.34486   32   21   21
0.50849   0.75334   0.23626   32   22   22
0.50840   0.35482   0.23555   32   23   23
0.98625   0.47286   0.12520   32   24   24
0.00594   0.49970   0.34986   32   25   25
0.98530   0.09568   0.12195   32   26   26
0.00628   0.12315   0.34661   32   27   27
0.95036   0.26790   0.16891   32   28   28
0.96249   0.28829   0.30290   32   29   29
0.78790   0.68665   0.16891   32   30   30
0.80003   0.70702   0.30290   32   31   31
0.74445   0.47524   0.12195   32   32   32
0.76413   0.50205   0.34661   32   33   33
0.74412   0.85179   0.12521   32   34   34
0.76510   0.87925   0.34986   32   35   35
0.70374   0.14867   0.19581   32   36   36
0.71101   0.15841   0.27601   32   37   37
0.63820   0.61894   0.23591   32   38   38
0.58467   0.47626   0.05846   32   39   39
0.56325   0.47301   0.39743   32   40   40
0.53406   0.67535   0.07439   32   41   41
0.61677   0.72416   0.41335   32   42   42
0.58211   0.27149   0.16866   32   43   43
0.58897   0.28505   0.30261   32   44   44
0.57690  0.89760  0.16922  32  45  45
0.59430  0.91642  0.30314  32  46  46
0.36517  0.44353  0.12499  32  47  47
0.38523  0.53138  0.34682  32  48  48

Ge.iso-II-4

ρ=0.5461 A˚−2=3.939 ML\rho= 0.5461\,\mathrm{\AA}^{-2} = 3.939\,\mathrm{ML}, E−Ebulk=176 meV/at.E - E_{\mathrm{bulk}} = 176\,\mathrm{meV/at.}

-15.82209   15.82209   0.00000
-0.00007   -0.00007  15.71212
47.38157   47.38157  -0.00043
192 atoms
0.00000    0.88780    0.12974   32   1   1
0.00000    0.50254    0.93508   32   2   2
0.94334    0.00665    0.08217   32   3   3
0.94333    0.24350    0.00107   32   4   4
0.89604    0.92204    0.95858   32   5   5
0.90596    0.94148    0.05038   32   6   6
0.90724    0.94027    0.11385   32   7   7
0.90850    0.09119    0.00154   32   8   8
0.90851    0.15890    0.08174   32   9   9
0.90729    0.30984    0.96935   32  10  10
0.90595    0.30842    0.03291   32  11  11
0.89610    0.32794    0.12476   32  12  12
0.89967    0.70122    0.97045   32  13  13
0.90120    0.69291    0.03171   32  14  14
0.91681    0.69646    0.12495   32  15  15
0.91682    0.55370    0.95830   32  16  16
0.90118    0.55689    0.05163   32  17  17
0.89971    0.54879    0.11290   32  18  18
0.84765    0.05524    0.97125   32  19  19
0.84821    0.05655    0.03191   32  20  20
0.83387    0.04919    0.12574   32  21  21
0.83387    0.20091    0.95740   32  22  22
0.84819    0.19337    0.05138   32  23  23
0.84772    0.19496    0.11203   32  24  24
0.83869    0.79897    0.94772   32  25  25
0.84356    0.80720    0.05037   32  26  26
0.84406    0.80743    0.11390   32  27  27
0.84113    0.66051    0.00105   32  28  28
0.84112    0.58926    0.08229   32  29  29
0.84409    0.44269    0.96935   32  30  30
0.84354    0.44266    0.03297   32  31  31
0.83878    0.45095    0.13564   32  32  32
0.80668    0.74184    0.08225   32  33  33
0.80669    0.50793    0.00105   32  34  34
0.75000    0.99383    0.98263   32  35  35
0.750000.994880.02032323636
0.750000.998550.14880323737
0.749990.841540.97019323838
0.750000.842070.03112323939
0.750000.860230.12929324040
0.750000.750660.00081324141
0.750000.499090.08258324242
0.750000.251990.93431324343
0.750000.255010.06302324444
0.750000.256130.10069324545
0.750000.390370.95384324646
0.750000.407840.05223324747
0.750000.408350.11322324848
0.693320.741840.08225324949
0.693310.507930.00106325050
0.661290.799000.94773325151
0.656440.807200.05037325252
0.655940.807430.11390325353
0.658880.660510.00106325454
0.658890.589260.08229325555
0.655920.442700.96935325656
0.656470.442660.03297325757
0.661230.450960.13564325858
0.652350.055250.97125325959
0.651790.056550.03191326060
0.666130.049190.12574326161
0.666140.200920.95740326262
0.651810.193370.05138326363
0.652280.194960.11203326464
0.600320.701220.97045326565
0.598800.692910.03172326666
0.583190.696460.12495326767
0.583190.553710.95829326868
0.598830.556890.05163326969
0.600300.548790.11290327070
0.603930.922050.95860327171
0.594040.941480.05038327272
0.592760.940270.11385327373
0.591500.091190.00154327474
0.591490.158900.08174327575
0.592710.309850.96935327676
0.594050.308420.03291327777
0.603900.327940.12476327878
0.556660.006650.08217327979
0.556670.243500.00107328080
0.499980.907910.97172328181
0.500000.906930.03132328282
0.500000.887800.12974328383
0.500000.753590.02030328484
0.500000.001010.00159328585

Table 5.

0.50000  0.24902  0.08172  32  86  86
0.50001  0.36206  0.95334  32  87  87
0.50000  0.34292  0.05197  32  88  88
0.50000  0.49621  0.06306  32  89  89
0.50000  0.49435  0.10075  32  90  90
0.99998  0.90791  0.97172  32  91  91
0.00000  0.90693  0.03132  32  92  92
0.99999  0.75551  0.98260  32  93  93
1.00000  0.75359  0.02030  32  94  94
1.00000  0.74775  0.14810  32  95  95
1.00000  0.00101  0.00159  32  96  96
0.00000  0.24902  0.08172  32  97  97
0.00000  0.36206  0.95334  32  98  98
0.00000  0.34292  0.05197  32  99  99
0.00000  0.34197  0.11162  32  100  100
0.00000  0.49621  0.06306  32  101  101
0.00000  0.49435  0.10075  32  102  102
0.05666  0.00665  0.08217  32  103  103
0.05667  0.24350  0.00107  32  104  104
0.10393  0.92205  0.95860  32  105  105
0.09404  0.94148  0.05038  32  106  106
0.09276  0.94027  0.11385  32  107  107
0.09150  0.09119  0.00154  32  108  108
0.09150  0.15890  0.08174  32  109  109
0.09272  0.30985  0.96935  32  110  110
0.09406  0.30842  0.03291  32  111  111
0.10390  0.32794  0.12476  32  112  112
0.10032  0.70122  0.97045  32  113  113
0.09880  0.69291  0.03172  32  114  114
0.08319  0.69646  0.12495  32  115  115
0.08319  0.55371  0.95829  32  116  116
0.09883  0.55689  0.05163  32  117  117
0.10030  0.54879  0.11290  32  118  118
0.15235  0.05525  0.97125  32  119  119
0.15179  0.05655  0.03191  32  120  120
0.16613  0.04919  0.12574  32  121  121
0.16614  0.20092  0.95740  32  122  122
0.15181  0.19337  0.05138  32  123  123
0.15228  0.19496  0.11203  32  124  124
0.16129  0.79900  0.94773  32  125  125
0.15644  0.80720  0.05037  32  126  126
0.15594  0.80743  0.11390  32  127  127
0.15888  0.66051  0.00106  32  128  128
0.15888  0.58926  0.08229  32  129  129
0.15592  0.44270  0.96935  32  130  130
0.15647  0.44266  0.03297  32  131  131
0.16123  0.45096  0.13564  32  132  132
0.19332  0.74184  0.08225  32  133  133
0.19331  0.50793  0.00106  32  134  134
0.25000  0.99383  0.98263  32  135  135
0.250000.994880.0203232136136
0.250000.998550.1488032137137
0.249990.841540.9701932138138
0.250000.842070.0311232139139
0.250000.860230.1292932140140
0.250000.750660.0008132141141
0.250000.499090.0825832142142
0.250000.251990.9343132143143
0.250000.255010.0630232144144
0.250000.256130.1006932145145
0.250000.390370.9538432146146
0.250000.407840.0522332147147
0.250000.408350.1132232148148
0.306680.741840.0822532149149
0.306690.507930.0010532150150
0.338690.798970.9477232151151
0.343560.807200.0503732152152
0.344060.807430.1139032153153
0.341130.660510.0010532154154
0.341120.589260.0822932155155
0.344090.442690.9693532156156
0.343540.442660.0329732157157
0.338770.450950.1356432158158
0.347650.055240.9712532159159
0.348210.056550.0319132160160
0.333870.049190.1257432161161
0.333870.200910.9574032162162
0.348190.193370.0513832163163
0.347720.194960.1120332164164
0.399670.701220.9704532165165
0.401200.692910.0317132166166
0.416810.696460.1249532167167
0.416820.553700.9583032168168
0.401180.556890.0516332169169
0.399710.548790.1129032170170
0.396040.922040.9585832171171
0.405960.941480.0503832172172
0.407240.940270.1138532173173
0.408490.091190.0015432174174
0.408510.158900.0817432175175
0.407290.309840.9693532176176
0.405950.308420.0329132177177
0.396100.327940.1247632178178
0.443340.006650.0821732179179
0.443330.243500.0010732180180
0.500000.755510.9826032181181
0.500000.747750.1481032182182
0.500000.341970.1116232183183
0.500000.502540.9350832184184

Table 6.

0.00001  0.48841  0.97807  32  185  185
0.75000  0.26178  0.97745  32  186  186
0.25000  0.26178  0.97745  32  187  187
0.50001  0.48841  0.97807  32  188  188
0.75000  0.98864  0.10562  32  189  189
1.00000  0.76126  0.10506  32  190  190
0.25000  0.98864  0.10562  32  191  191
0.50000  0.76126  0.10506  32  192  192

Ge.equi-II-6

ρ=0.8042\rho= 0.8042 Å−2^{-2} = 5.800 ML, E−EbulkE - E_{\mathrm{bulk}} = 120 meV/at.

10.86109  -0.01836  -0.00140
-5.44644  9.39675  -0.00201
-0.00506  -0.01138  38.26388
82 atoms
0.34308  0.99782  0.20027  32  1  1
0.95338  0.22993  0.20030  32  2  2
0.72149  0.60798  0.20028  32  3  3
0.72123  0.99788  0.20033  32  4  4
0.34308  0.22938  0.20035  32  5  5
0.95299  0.60771  0.20028  32  6  6
0.96265  0.23156  0.02246  32  7  7
0.33972  0.98994  0.02231  32  8  8
0.71411  0.61309  0.02088  32  9  9
0.33792  0.23100  0.02467  32  10  10
0.71970  0.99048  0.02487  32  11  11
0.94899  0.61236  0.02657  32  12  12
0.62392  0.55966  0.26098  32  13  13
0.00197  0.32778  0.26100  32  14  14
0.39179  0.94936  0.26101  32  15  15
0.00168  0.55928  0.26101  32  16  16
0.39187  0.32751  0.26096  32  17  17
0.62336  0.94936  0.26094  32  18  18
0.55050  0.40345  0.05898  32  19  19
0.54819  0.03198  0.05903  32  20  20
0.91888  0.39903  0.05877  32  21  21
0.21366  0.73739  0.29415  32  22  22
0.21406  0.36150  0.29395  32  23  23
0.59000  0.73776  0.29395  32  24  24
0.75507  0.81993  0.16734  32  25  25
0.13133  0.19586  0.16734  32  26  26
0.13097  0.81953  0.16713  32  27  27
0.00616  0.94435  0.08239  32  28  28
0.67271  0.27901  0.31721  32  29  29
0.67247  0.27846  0.14409  32  30  30
0.54885  0.40170  0.16467  32  31  31
0.54892  0.03090  0.16469  32  32  32
0.920280.402280.16462323333
0.759750.820360.06012323434
0.127060.189790.06017323535
0.130480.824320.05992323636
0.425060.155370.29660323737
0.796420.526730.29667323838
0.795840.155300.29662323939
0.005780.944930.14701324040
0.672830.278770.08011324141
0.339160.612310.31428324242
0.471420.479720.11194324343
0.470750.874920.11187324444
0.075660.479560.11171324545
0.138240.813660.34849324646
0.137880.210000.34825324747
0.742000.813600.34828324848
0.603330.743520.11303324949
0.206930.347630.11302325050
0.207000.743880.11280325151
0.007880.321200.44041325252
0.626310.569690.43883325353
0.384690.946790.43897325454
0.385210.326750.43642325555
0.007100.556040.43472325656
0.625740.944970.43662325757
0.426580.155100.40225325858
0.793630.525760.40252325959
0.798070.157390.40231326060
0.338860.612990.37890326161
0.218900.737380.40136326262
0.214950.366660.40117326363
0.584390.733960.40112326464
0.673280.279610.38118326565
0.269290.077420.34942326666
0.873920.682340.34958326767
0.874090.078100.34935326868
0.126130.387800.97224326969
0.120480.611340.98264327070
0.373060.435690.98467327171
0.349930.834800.97218327272
0.618790.612120.95792327373
0.548570.786050.98497327474
0.782770.733390.48905327575
0.830590.980270.47662327676
0.006280.727730.47865327777
0.007150.226140.50337327878
0.229770.957240.48911327979
0.180930.155790.47631328080
0.311260.611260.02520328181
0.005960.918310.43609328282

Table 7.

Silicon-Slab xyz Files

Si.db-monolayer (oP10)

ρ=0.2110 A˚−2=1.364 ML, E−Ebulk=430 meV/at.\rho= 0.2110\ \text{Å}^{-2} = 1.364\ \text{ML},\ E - E_{\mathrm{bulk}} = 430\ \text{meV/at.}

7.40937  0.00000  0.00000
0.00000  35.75410  0.00000
0.00000  0.00000  6.39543
10 atoms
0.01913  0.00000  0.81934  14  1  1
0.48087  0.00000  0.81934  14  2  2
0.51913  0.00000  0.18066  14  3  3
0.98087  0.00000  0.18066  14  4  4
0.25000  0.03788  0.67701  14  5  5
0.25000  0.96212  0.67701  14  6  6
0.75000  0.96212  0.32299  14  7  7
0.75000  0.03788  0.32299  14  8  8
0.25000  0.00000  0.37407  14  9  9
0.75000  0.00000  0.62593  14  10  10

Si.iso-II-2.5

ρ=0.3851 A˚−2=2.489 ML, E−Ebulk=322 meV/at.\rho= 0.3851\ \text{Å}^{-2} = 2.489\ \text{ML},\ E - E_{\mathrm{bulk}} = 322\ \text{meV/at.}

15.01360  0.00000  0.00000
0.00000  10.72374  0.00000
0.00000  0.00000  24.81168
62 atoms
0.00530  0.00000  0.00000  14  1  1
0.36693  0.00000  0.11410  14  2  2
0.36693  0.00000  0.88590  14  3  3
0.50034  0.00000  0.16477  14  4  4
0.50034  0.00000  0.83523  14  5  5
0.76091  0.00000  0.04838  14  6  6
0.76091  0.00000  0.95162  14  7  7
0.91223  0.00000  0.07669  14  8  8
0.91223  0.00000  0.92331  14  9  9
0.49346  0.00000  0.05520  14  10  10
0.49346  0.00000  0.94480  14  11  11
0.24599  0.11233  0.00000  14  12  12
0.24599  0.88767  0.00000  14  13  13
0.55251  0.16366  0.10837  14  14  14
0.55251  0.83634  0.89163  14  15  15
0.55251  0.16366  0.89163  14  16  16
0.55251  0.83634  0.10837  14  17  17
0.31244  0.18904  0.07990  14  18  18
0.31244  0.81096  0.92010  14  19  19
0.31244  0.18904  0.92010  14  20  20
0.312440.810960.07990142121
0.941470.195310.11305142222
0.941470.804690.88695142323
0.941470.195310.88695142424
0.941470.804690.11305142525
0.095120.181830.00000142626
0.095120.818170.00000142727
0.697570.192920.07633142828
0.697570.807080.92367142929
0.697570.192920.92367143030
0.697570.807080.07633143131
0.060530.307060.07633143232
0.060530.692940.92367143333
0.060530.307060.92367143434
0.060530.692940.07633143535
0.663010.318160.00000143636
0.663010.681850.00000143737
0.816740.304590.11295143838
0.816740.695410.88705143939
0.816740.304590.88705144040
0.816740.695410.11295144141
0.445680.310960.07990144242
0.445680.689040.92010144343
0.445680.310960.92010144444
0.445680.689040.07990144545
0.205600.336350.10836144646
0.205600.663650.89164144747
0.205600.336350.89164144848
0.205600.663650.10836144949
0.512140.387670.00000145050
0.512140.612330.00000145151
0.257780.500000.16477145252
0.257780.500000.83523145353
0.264670.500000.05519145454
0.264670.500000.94481145555
0.391200.500000.11411145656
0.391200.500000.88589145757
0.752820.500000.00000145858
0.845910.500000.07668145959
0.845910.500000.92332146060
0.997220.500000.04837146161
0.997220.500000.95163146262

Table 8.

Si.equi-II-3

ρ=0.5032\rho= 0.5032 Å−2^{-2} = 3.253 ML, E−Ebulk=280E - E_{\mathrm{bulk}} = 280 meV/at.

10.27433   -0.00034    0.00000
-5.13746    8.89766   -0.00000
0.00001   -0.00000   26.09222
46 atoms
0.33334    0.04309    0.27406   14   1    1
0.95693    0.29027    0.27407   14   2    2
0.70976    0.66667    0.27406   14   3    3
0.71884    0.05218    0.27250   14   4    4
0.33333    0.28116    0.27251   14   5    5
0.94782    0.66666    0.27251   14   6    6
0.95693    0.29027    0.02445   14   7    7
0.33334    0.04309    0.02445   14   8    8
0.70976    0.66667    0.02445   14   9    9
0.33333    0.28116    0.02601   14  10   10
0.71884    0.05218    0.02601   14  11   11
0.94782    0.66666    0.02601   14  12   12
0.54259    0.45614    0.07488   14  13   13
0.54385    0.08644    0.07488   14  14   14
0.91356    0.45741    0.07488   14  15   15
0.75311    0.87719    0.22364   14  16   16
0.12407    0.24689    0.22364   14  17   17
0.12280    0.87592    0.22363   14  18   18
0.00001    0.00000    0.10469   14  19   19
0.66667    0.33334    0.19382   14  20   20
0.54259    0.45614    0.22363   14  21   21
0.54386    0.08644    0.22364   14  22   22
0.91356    0.45741    0.22364   14  23   23
0.75311    0.87718    0.07488   14  24   24
0.12407    0.24689    0.07488   14  25   25
0.12281    0.87593    0.07488   14  26   26
0.00001    0.00000    0.19382   14  27   27
0.66667    0.33334    0.10469   14  28   28
0.46565    0.53450    0.14926   14  29   29
0.46550    0.93115    0.14926   14  30   30
0.06885    0.53434    0.14926   14  31   31
0.59782    0.79884    0.14926   14  32   32
0.20101    0.40218    0.14926   14  33   33
0.20117    0.79898    0.14926   14  34   34
0.10121    0.43454    0.34567   14  35   35
0.13149    0.66665    0.32727   14  36   36
0.33332    0.46483    0.32727   14  37   37
0.33333    0.89878    0.34565   14  38   38
0.56544    0.66667    0.34565   14  39   39
0.53517    0.86850    0.32727   14  40   40
0.10121    0.43454    0.95284   14  41   41
0.13150    0.66665    0.97125   14  42   42
0.33332    0.46483    0.97125   14  43   43
0.33333    0.89878    0.95286   14  44   44
0.56544    0.66667    0.95286   14  45   45
0.53517    0.86850    0.97125   14  46   46

Si.iso-II-4

ρ=0.6056 A˚−2=3.915 ML, E−Ebulk=232 meV/at.\rho= 0.6056\,\text{\AA}^{-2} = 3.915\,\mathrm{ML},\ E - E_{\mathrm{bulk}} = 232\,\mathrm{meV/at.}

10.62542  0.00000  0.00000
0.00000  14.91779  0.00000
0.00000  0.00000  34.35054
96 atoms
0.25000  0.51535  0.23141  14  1   1
0.75000  0.48465  0.23141  14  2   2
0.25000  0.11281  0.95396  14  3   3
0.75000  0.88719  0.95396  14  4   4
0.25000  0.12141  0.87777  14  5   5
0.75000  0.87859  0.87777  14  6   6
0.13684  0.63142  0.14602  14  7   7
0.86316  0.36858  0.14602  14  8   8
0.36316  0.63142  0.14602  14  9   9
0.63684  0.36858  0.14602  14  10  10
0.13684  0.86858  0.99724  14  11  11
0.86316  0.13142  0.99724  14  12  12
0.36316  0.86858  0.99724  14  13  13
0.63684  0.13142  0.99724  14  14  14
0.05522  0.56261  0.91432  14  15  15
0.94478  0.43739  0.91432  14  16  16
0.44478  0.56261  0.91432  14  17  17
0.55522  0.43739  0.91432  14  18  18
0.06192  0.56696  0.08724  14  19  19
0.93808  0.43304  0.08724  14  20  20
0.43808  0.56696  0.08724  14  21  21
0.56192  0.43304  0.08724  14  22  22
0.06162  0.56646  0.20451  14  23  23
0.93838  0.43354  0.20451  14  24  24
0.43838  0.56646  0.20451  14  25  25
0.56162  0.43354  0.20451  14  26  26
0.06764  0.71671  0.99726  14  27  27
0.93236  0.28329  0.99726  14  28  28
0.43236  0.71671  0.99726  14  29  29
0.56764  0.28329  0.99726  14  30  30
0.06763  0.78329  0.14600  14  31  31
0.93237  0.21671  0.14600  14  32  32
0.43237  0.78329  0.14600  14  33  33
0.56763  0.21671  0.14600  14  34  34
0.06161  0.93353  0.93876  14  35  35
0.93839  0.06647  0.93876  14  36  36
0.43839  0.93353  0.93876  14  37  37
0.56161  0.06647  0.93876  14  38  38
0.06191  0.93304  0.05602  14  39  39
0.93809  0.06696  0.05602  14  40  40
0.438090.933040.05602144141
0.561910.066960.05602144242
0.055230.937390.22893144343
0.944770.062610.22893144444
0.444770.937390.22893144545
0.555230.062610.22893144646
0.056330.318490.94166144747
0.943670.681510.94166144848
0.443670.318490.94166144949
0.556330.681510.94166145050
0.053620.318070.05335145151
0.946380.681930.05335145252
0.446380.318070.05335145353
0.553620.681930.05335145454
0.083040.325640.22620145555
0.916960.674360.22620145656
0.416960.325640.22620145757
0.583040.674360.22620145858
0.083050.174370.91707145959
0.916950.825630.91707146060
0.416950.174370.91707146161
0.583050.825630.91707146262
0.053620.181930.08991146363
0.946380.818070.08991146464
0.446380.181930.08991146565
0.553620.818070.08991146666
0.056330.181510.20160146767
0.943670.818490.20160146868
0.443670.181510.20160146969
0.556330.818490.20160147070
0.250000.381140.96221147171
0.750000.618860.96221147272
0.250000.379500.03194147373
0.750000.620500.03194147474
0.250000.387200.18930147575
0.750000.612800.18930147676
0.250000.378590.26550147777
0.750000.621410.26550147878
0.250000.532730.94144147979
0.750000.467270.94144148080
0.250000.532480.05212148181
0.750000.467520.05212148282
0.250000.626050.99685148383
0.750000.373950.99685148484
0.250000.873950.14640148585
0.750000.126050.14640148686
0.250000.120500.11132148787
0.750000.879500.11132148888
0.250000.118870.18105148989

Table 62.

0.75000  0.88113  0.18105  14  90  90
0.25000  0.98466  0.91187  14  91  91
0.75000  0.01534  0.91187  14  92  92
0.25000  0.96752  0.09113  14  93  93
0.75000  0.03248  0.09113  14  94  94
0.25000  0.96728  0.20182  14  95  95
0.75000  0.03272  0.20182  14  96  96

Si.iso-II-5

ρ=0.8272 A˚−2=5.347 ML\rho= 0.8272\ \text{\AA}^{-2} = 5.347\ \mathrm{ML}, E−Ebulk=186 meV/at.E - E_{\mathrm{bulk}} = 186\ \mathrm{meV/at.}

10.57224  0.00000  0.00000
0.00000  14.86573  0.00000
0.00000  0.00000  43.59367
130 atoms
0.25000  0.62159  0.21259  14  1   1
0.25000  0.62159  0.78741  14  2   2
0.75000  0.37841  0.78741  14  3   3
0.75000  0.37841  0.21259  14  4   4
0.25000  0.48433  0.18583  14  5   5
0.25000  0.48433  0.81417  14  6   6
0.75000  0.51567  0.81417  14  7   7
0.75000  0.51567  0.18583  14  8   8
0.25000  0.61260  0.15253  14  9   9
0.25000  0.61260  0.84747  14  10  10
0.75000  0.38740  0.84747  14  11  11
0.75000  0.38740  0.15253  14  12  12
0.25000  0.03275  0.83797  14  13  13
0.25000  0.03275  0.16203  14  14  14
0.75000  0.96725  0.16203  14  15  15
0.75000  0.96725  0.83797  14  16  16
0.63655  0.63102  0.11807  14  17  17
0.86345  0.63102  0.88193  14  18  18
0.13655  0.36898  0.88193  14  19  19
0.36345  0.36898  0.11807  14  20  20
0.36345  0.36898  0.88193  14  21  21
0.13655  0.36898  0.11807  14  22  22
0.86345  0.63102  0.11807  14  23  23
0.63655  0.63102  0.88193  14  24  24
0.63592  0.86935  0.00000  14  25  25
0.86408  0.86935  0.00000  14  26  26
0.13592  0.13065  0.00000  14  27  27
0.36408  0.13065  0.00000  14  28  28
0.58265  0.32544  0.18156  14  29  29
0.91735  0.32544  0.81844  14  30  30
0.08265  0.67456  0.81844  14  31  31
0.41735  0.67456  0.18156  14  32  32

0.41735 0.67456 0.81844 14 33 33 0.08265 0.67456 0.18156 14 34 34 0.91735 0.32544 0.18156 14 35 35 0.58265 0.32544 0.81844 14 36 36 0.56182 0.56654 0.16429 14 37 37 0.93818 0.56654 0.83571 14 38 38 0.06182 0.43346 0.83571 14 39 39 0.43818 0.43346 0.16429 14 40 40 0.43818 0.43346 0.83571 14 41 41 0.06182 0.43346 0.16429 14 42 42 0.93818 0.56654 0.16429 14 43 43 0.56182 0.56654 0.83571 14 44 44 0.55537 0.93719 0.18366 14 45 45 0.94463 0.93719 0.81634 14 46 46 0.05537 0.06281 0.81634 14 47 47 0.44462 0.06281 0.18366 14 48 48 0.44462 0.06281 0.81634 14 49 49 0.05537 0.06281 0.18366 14 50 50 0.94463 0.93719 0.18366 14 51 51 0.55537 0.93719 0.81634 14 52 52 0.56699 0.71656 0.00000 14 53 53 0.93301 0.71656 0.00000 14 54 54 0.06699 0.28344 0.00000 14 55 55 0.43301 0.28344 0.00000 14 56 56 0.56726 0.78317 0.11800 14 57 57 0.93274 0.78317 0.88200 14 58 58 0.06726 0.21683 0.88200 14 59 59 0.43274 0.21683 0.11800 14 60 60 0.43274 0.21683 0.88200 14 61 61 0.06726 0.21683 0.11800 14 62 62 0.93274 0.78317 0.11800 14 63 63 0.56726 0.78317 0.88200 14 64 64 0.56200 0.93286 0.04673 14 65 65 0.93800 0.93286 0.95327 14 66 66 0.06200 0.06714 0.95327 14 67 67 0.43800 0.06714 0.04673 14 68 68 0.43800 0.06714 0.95327 14 69 69 0.06200 0.06714 0.04673 14 70 70 0.93800 0.93286 0.04673 14 71 71 0.56200 0.93286 0.95327 14 72 72 0.56201 0.56710 0.07162 14 73 73 0.93799 0.56710 0.92838 14 74 74 0.06201 0.43290 0.92838 14 75 75 0.43799 0.43290 0.07162 14 76 76 0.43799 0.43290 0.92838 14 77 77 0.06201 0.43290 0.07162 14 78 78 0.93799 0.56710 0.07162 14 79 79 0.56201 0.56710 0.92838 14 80 80 0.55344 0.18209 0.07366 14 81 81 0.94656 0.18209 0.92634 14 82 82

0.053440.817910.92634148383
0.446560.817910.07366148484
0.446560.817910.92634148585
0.053440.817910.07366148686
0.946560.182090.07366148787
0.553440.182090.92634148888
0.553220.317900.04458148989
0.946780.317900.95542149090
0.053220.682100.95542149191
0.446780.682100.04458149292
0.446780.682100.95542149393
0.053220.682100.04458149494
0.946780.317900.04458149595
0.553220.317900.95542149696
0.556210.181550.16191149797
0.943790.181550.83809149898
0.056210.818450.83809149999
0.443790.818450.1619114100100
0.443790.818450.8380914101101
0.056210.818450.1619114102102
0.943790.181550.1619114103103
0.556210.181550.8380914104104
0.250000.881080.1455614105105
0.250000.881080.8544414106106
0.750000.118920.8544414107107
0.750000.118920.1455614108108
0.250000.125850.1182914109109
0.250000.125850.8817014110110
0.750000.874150.8817014111111
0.750000.874150.1182914112112
0.250000.879510.0906714113113
0.250000.879510.9093314114114
0.750000.120490.9093314115115
0.750000.120490.0906714116116
0.250000.032680.0746314117117
0.250000.032680.9253714118118
0.750000.967320.9253714119119
0.750000.967320.0746314120120
0.250000.467580.0437614121121
0.250000.467580.9562414122122
0.750000.532430.9562414123123
0.750000.532430.0437614124124
0.250000.620670.0275314125125
0.250000.620670.9724714126126
0.750000.379330.9724714127127
0.750000.379330.0275314128128
0.250000.374680.0000014129129
0.750000.625320.0000014130130

Table 65.

Ge-Clathrate/InN

Thinner Ge Clathrate Film on InN (initial configuration)

See Figure S6.2.

10.89643   0.00000   -0.00000
-5.44821   9.43658    0.00000
 0.00000   0.00000   26.29850
95 atoms
0.26262   0.13131   0.17022   32   1   1
0.86869   0.13131   0.17022   32   2   2
0.86869   0.73738   0.17022   32   3   3
0.66667   0.33333   0.21704   32   4   4
0.33333   0.66667   0.21710   32   5   5
0.41997   0.20999   0.24593   32   6   6
0.79002   0.20999   0.24593   32   7   7
0.79002   0.58003   0.24593   32   8   8
0.20999   0.41997   0.24594   32   9   9
0.20999   0.79002   0.24594   32   10   10
0.58003   0.79002   0.24594   32   11   11
0.39163   0.99998   0.28812   32   12   12
0.39163   0.39165   0.28812   32   13   13
0.60835   0.99998   0.28812   32   14   14
0.60835   0.60837   0.28812   32   15   15
0.00002   0.39165   0.28812   32   16   16
0.00002   0.60837   0.28812   32   17   17
0.00052   0.38429   0.04618   32   18   18
0.00052   0.61623   0.04618   32   19   19
0.38377   0.38429   0.04618   32   20   20
0.38377   0.99948   0.04618   32   21   21
0.61571   0.61623   0.04618   32   22   22
0.61571   0.99948   0.04618   32   23   23
0.41845   0.20923   0.09415   32   24   24
0.79077   0.20923   0.09415   32   25   25
0.79077   0.58155   0.09415   32   26   26
0.20914   0.41829   0.09469   32   27   27
0.20914   0.79086   0.09469   32   28   28
0.58171   0.79086   0.09469   32   29   29
0.66667   0.33333   0.12381   32   30   30
0.33333   0.66667   0.12389   32   31   31
0.13090   0.26180   0.17048   32   32   32
0.13090   0.86910   0.17048   32   33   33
0.73820   0.86910   0.17048   32   34   34
-0.77000   0.00000   0.35200   32   35   35
-0.76077   0.23947   0.35200   32   36   36
-0.23989   0.76044   0.35200   32   37   37
-0.23989   0.00000   0.35200   32   38   38
-0.030030.760520.35200323939
-0.030030.239560.35200324040
0.000000.000000.30200324141
0.666670.333330.73784494242
0.333330.666670.73561494343
0.000000.000000.73969494444
0.003850.332740.72373494545
0.003850.671110.72373494646
0.328890.332740.72373494747
0.328890.996150.72373494848
0.667260.671110.72373494949
0.667260.996150.72373495050
0.438570.219280.83677495151
0.780720.219280.83677495252
0.780720.561430.83677495353
0.110750.221510.84237495454
0.110750.889250.84237495555
0.778500.889250.84237495656
0.111700.555850.83822495757
0.444150.555850.83822495858
0.444150.888300.83822495959
0.560970.121930.94440496060
0.560970.439030.94440496161
0.878070.439030.94440496262
0.227100.454210.95022496363
0.227100.772900.95022496464
0.545800.772900.95022496565
0.215620.107810.95497496666
0.892190.107810.95497496767
0.892190.784380.95497496868
0.111980.223960.7564376969
0.111980.888020.7564377070
0.776040.888020.7564377171
0.439900.219950.7502877272
0.780050.219950.7502877373
0.780050.560100.7502877474
0.109270.554630.7519277575
0.445370.554630.7519277676
0.445370.890730.7519277777
0.555920.111840.8604077878
0.555920.444080.8604077979
0.888160.444080.8604078080
0.224270.112130.8693078181
0.887870.112130.8693078282
0.887870.775730.8693078383
0.222540.445080.8655478484
0.222540.777460.8655478585
0.554920.777460.8655478686
0.338840.000240.9752278787

Table 9.

0.33884  0.33860  0.97522  7  88  88
0.66140  0.00024  0.97522  7  89  89
0.66140  0.66116  0.97522  7  90  90
0.99976  0.33860  0.97522  7  91  91
0.99976  0.66116  0.97522  7  92  92
0.00000  0.00000  0.98093  7  93  93
0.66667  0.33333  0.97516  7  94  94
0.33333  0.66667  0.98001  7  95  95

Thicker Ge Clathrate Film on InN (initial configuration)

See Figure S6.6.

10.89643  0.00000  -0.00000
-5.44821  9.43658  0.00000
0.00000  0.00000  50.00000
129 atoms
1.00700  1.00700  0.36991  32  1  1
0.34644  0.06187  0.16999  32  2  2
-0.04787  0.29157  0.16999  32  3  3
-0.27757  -0.33244  0.16999  32  4  4
-0.27757  0.06187  0.16999  32  5  5
0.34644  0.29157  0.16999  32  6  6
-0.04787  -0.33244  0.16999  32  7  7
-0.04176  0.28546  0.03126  32  8  8
0.33423  0.05576  0.03126  32  9  9
-0.27146  -0.32023  0.03126  32  10  10
0.33423  0.28546  0.03126  32  11  11
-0.27146  0.05576  0.03126  32  12  12
-0.04176  -0.32023  0.03126  32  13  13
-0.37510  -0.38120  0.21851  32  14  14
0.00089  0.38910  0.21851  32  15  15
0.39520  0.01311  0.21851  32  16  16
0.00089  -0.38120  0.21851  32  17  17
0.39520  0.38910  0.21851  32  18  18
-0.37510  0.01311  0.21851  32  19  19
-0.45167  0.46567  0.05657  32  20  20
-0.45167  0.08966  0.05657  32  21  21
-0.07566  0.46567  0.05657  32  22  22
0.21500  -0.20100  0.24382  32  23  23
0.21500  0.42299  0.24382  32  24  24
-0.40899  -0.20100  0.24382  32  25  25
-0.24367  -0.11834  0.14468  32  26  26
0.13234  0.25767  0.14468  32  27  27
0.13234  -0.11834  0.14468  32  28  28
0.00700  0.00700  0.07722  32  29  29
-0.32633  0.34033  0.26447  32  30  30
-0.326330.340330.12403323131
-0.449070.463070.14127323232
-0.449070.094860.14127323333
-0.080860.463070.14127323434
-0.23847-0.115740.05998323535
0.129740.252470.05998323636
0.12974-0.115740.05998323737
0.428190.217600.24723323838
-0.20360-0.414190.24723323939
-0.203600.217600.24723324040
0.007000.007000.12893324141
-0.326330.340330.07232324242
0.34033-0.326330.25957324343
0.47177-0.457770.09958324444
0.47177-0.063460.09958324545
0.07746-0.457770.09958324646
0.13844-0.124440.28683324747
0.138440.269870.28683324848
-0.25587-0.124440.28683324949
-0.39679-0.194900.10167325050
0.208900.410790.10167325151
0.20890-0.194900.10167325252
0.013110.395200.35723325353
-0.38120-0.375100.35723325454
0.389100.000890.35723325555
0.389100.395200.35723325656
0.01311-0.375100.35723325757
-0.381200.000890.35723325858
0.422990.215000.33193325959
-0.20100-0.408990.33193326060
-0.201000.215000.33193326161
0.34033-0.326330.31128326262
0.21760-0.203600.32852326363
0.217600.428190.32852326464
-0.41419-0.203600.32852326565
-0.326330.340330.31618326666
0.269870.138440.28892326767
-0.12444-0.255870.28892326868
-0.124440.138440.28892326969
0.767441.767430.39949327070
0.767441.007000.39949327171
1.007001.767430.39949327272
1.007001.246560.39949327373
0.246561.007000.39949327474
0.246561.246560.39949327575
0.666670.33333-0.13789497676
0.333330.66667-0.13906497777
0.000000.00000-0.13692497878
0.003850.33274-0.14531497979
0.003850.67111-0.14531498080
0.328890.33274-0.14531498181
0.328890.99615-0.14531498282
0.667260.67111-0.14531498383
0.667260.99615-0.14531498484
0.438570.21928-0.08586498585
0.780720.21928-0.08586498686
0.780720.56143-0.08586498787
0.110750.22151-0.08291498888
0.110750.88925-0.08291498989
0.778500.88925-0.08291499090
0.111700.55585-0.08509499191
0.444150.55585-0.08509499292
0.444150.88830-0.08509499393
0.560970.12193-0.02925499494
0.560970.43903-0.02925499595
0.878070.43903-0.02925499696
0.227100.45421-0.02618499797
0.227100.77290-0.02618499898
0.545800.77290-0.02618499999
0.215620.10781-0.0236849100100
0.892190.10781-0.0236849101101
0.892190.78438-0.0236849102102
0.111980.22396-0.128117103103
0.111980.88802-0.128117104104
0.776040.88802-0.128117105105
0.439900.21995-0.131347106106
0.780050.21995-0.131347107107
0.780050.56010-0.131347108108
0.109270.55463-0.130497109109
0.445370.55463-0.130497110110
0.445370.89073-0.130497111111
0.555920.11184-0.073437112112
0.555920.44408-0.073437113113
0.888160.44408-0.073437114114
0.224270.11213-0.068757115115
0.887870.11213-0.068757116116
0.887870.77573-0.068757117117
0.222540.44508-0.070727118118
0.222540.77746-0.070727119119
0.554920.77746-0.070727120120
0.338840.00024-0.013037121121
0.338840.33860-0.013037122122
0.661400.00024-0.013037123123
0.661400.66116-0.013037124124
0.999760.33860-0.013037125125
0.999760.66116-0.013037126126
0.000000.00000-0.010037127127
0.666670.33333-0.013077128128
0.333330.66667-0.010517129129

Table 10.

Tabulated E(ρ)E(\rho) Dependencies

E(ρ)E(\rho) of Sn

ρ\rho [at./Å2^2]ρ\rho [ML]EKSE_{\mathrm{KS}} [eV/at.]E−EbulkE-E_{\mathrm{bulk}} [meV/at.]EgE_g [eV]sample designation
0.14221.356-3.60432650.038db-monolayer (oP10)
0.17411.660-3.61592530.000db-monolayer (hP11)
0.14251.358-3.62682420.065Sn9-web (40 atoms)
0.17491.667-3.65302160.000Sn9-web (100 atoms)
0.22242.120-3.67071980.000triangles (bilayer)
0.44494.241-3.70921600.000triangles (2×2 \times bilayer)
0.24482.334-3.71031590.000squares (bilayer)
0.48504.623-3.73841300.000squares (2×2 \times bilayer)
0.71156.783-3.74991190.000squares (3×3 \times bilayer)
0.34283.268-3.72681420.000Sn.equi-II-3
0.55015.244-3.7731960.250Sn.equi-II-5
0.79817.608-3.7934750.263Sn.equi-II-7.5
1.04709.981-3.8041650.266Sn.equi-II-10
0.26292.506-3.65682120.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (bilayer)
0.25782.458-3.62192470.000(2×2)(2 \times2)-dia (bilayer)
0.35783.411-3.69061780.000(2×2)(2 \times2)-dia (trilayer)
0.46184.402-3.72571430.000(2×2)(2 \times2)-dia (fourlayer)
0.56635.398-3.75121180.000(2×2)(2 \times2)-dia (fivelayer)
0.67056.392-3.76741010.000(2×2)(2 \times2)-dia (sixlayer)
0.77547.392-3.7810880.000(2×2)(2 \times2)-dia (sevenlayer)
0.88038.392-3.7909780.000(2×2)(2 \times2)-dia (eightlayer)
0.98549.394-3.7986700.000(2×2)(2 \times2)-dia (ninelayer)

Table S10.1. Table S10.1: Tabulated content of Figure 6 (main text) for Sn thin films. “KS” denotes total Kohn-Sham energy as a direct output of VASP (without dispersion vdW correction). For unit conversion: thickness of 1ML of (111)-diamond cF8 Sn is ρ0=0.10490\rho_0 = 0.10490at./Å2^2 and the respective bulk Kohn-Sham per-atom energy Ebulk=−3.8688E_{\mathrm{bulk}} = -3.8688 eV/at. EgE_g stands for the width of electronic band gap.

E(ρ)E(\rho) of Ge

ρ\rho [at./Å2^2]ρ\rho [ML]EKSE_{\mathrm{KS}} [eV/at.]E−EbulkE-E_{\mathrm{bulk}} [meV/at.]EgE_{\mathrm{g}} [eV]sample designation
0.18841.359-4.20833080.100db-monolayer (oP10)
0.23671.707-4.16363520.000GS-bilayer
0.35342.549-4.26482510.256Ge.iso-II-2.5
0.40372.911-4.29412220.400Ge.equi-II-3
0.47813.448-4.33281830.428Ge.equi-II-3.5
0.54613.938-4.34041760.217Ge.iso-II-4
0.80425.800-4.39571200.517Ge.equi-II-6
0.34552.492-4.25352620.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (bilayer)
0.33312.402-4.22522910.133(2×2)(2 \times2)-dia (bilayer)
0.46833.377-4.29502210.000(2×2)(2 \times2)-dia (trilayer)
0.60594.370-4.34041760.000(2×2)(2 \times2)-dia (fourlayer)
0.74405.366-4.37361420.000(2×2)(2 \times2)-dia (fivelayer)
0.88226.362-4.39481210.000(2×2)(2 \times2)-dia (sixlayer)
1.02087.362-4.41181040.022(2×2)(2 \times2)-dia (sevenlayer)
1.15858.355-4.4245910.057(2×2)(2 \times2)-dia (eightlayer)
1.29829.362-4.4343820.108(2×2)(2 \times2)-dia (ninelayer)

Table S10.2.1. Table S10.2.1: Tabulated content of Figure 6 *(main text) for Ge thin films. “KS” denotes total Kohn-Sham energy as a direct output of VASP (without dispersion vdW correction). For unit conversion: thickness of 1ML of (111)-diamond cF8 Ge is ρ0=0.13866\rho_0 = 0.13866at./Å2^2 and the respective bulk Kohn-Sham per-atom energy Ebulk=−4.5159E_{\mathrm{bulk}} = -4.5159 eV/at. EgE_{\mathrm{g}} stands for the width of electronic band gap.

ρ\rho [at./Å2^2]ρ\rho [ML]EKSE_{\mathrm{KS}} [eV/at.]E−EbulkE-E_{\mathrm{bulk}} [meV/at.]EgE_{\mathrm{g}} [eV]sample designation
0.19241.361-4.38103890.000db-monolayer (oP10)
0.35912.540-4.44613230.000Ge.iso-II-2.5
0.56123.970-4.53432350.255Ge.iso-II-4
0.79085.595-4.58551840.435Ge.iso-II-5.5
0.41072.905-4.48442850.455Ge.equi-II-3
0.49123.475-4.53132380.490Ge.equi-II-3.5
0.81375.756-4.59521740.450Ge.equi-II-6
0.35102.483-4.44443250.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (bilayer)
0.47713.374-4.50692630.000(2×2)(2 \times2)-dia (trilayer)
0.63204.471-4.56092090.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (fourlayer)
0.77125.456-4.59591740.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (fivelayer)
1.04927.422-4.64331260.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (sevenlayer)
0.42763.025-4.46323070.000(3×3)(3 \times3)-DAS (trilayer)
0.56503.997-4.52962400.000(3×3)(3 \times3)-DAS (fourlayer)

Table S10.2.2. Table S10.2.2: Tabulated content of Figure S7.1 *(SM) for Ge thin films. “KS” denotes total Kohn-Sham energy as a direct output of VASP, with dispersion vdW correction (PBE-

D3, IVDW=12). For unit conversion: thickness of 1 ML of (111)-diamond cF8 Ge is ρ0=0.141357\rho_0 = 0.141357 at./Å2^2 and the respective bulk Kohn-Sham per-atom energy Ebulk=−4.7696E_{\mathrm{bulk}} = -4.7696 eV/at. EgE_{\mathrm{g}} stands for the width of electronic band gap.

E(ρ)E(\rho) of Si

ρ\rho [at./Å2^2]ρ\rho [ML]EKSE_{\mathrm{KS}} [eV/at.]E−EbulkE-E_{\mathrm{bulk}} [meV/at.]EgE_{\mathrm{g}} [eV]sample designation
0.21101.364-4.99064300.650db-monolayer (oP10)
0.27091.751-4.95664640.000GS-bilayer
0.38512.489-5.09873220.467Si.iso-II-2.5
0.50323.253-5.13992800.950Si.equi-II-3
0.60563.915-5.18852320.837Si.iso-II-4
0.82725.347-5.23381860.762Si.iso-II-5
0.38662.499-5.06433560.138(3×3)(\sqrt{3} \times\sqrt{3})-dia (bilayer)
0.36902.385-5.00804120.000(2×2)(2 \times2)-dia (bilayer)
0.52633.402-5.12452960.450(2×2)(2 \times2)-dia (trilayer)
0.67884.388-5.18342370.300(2×2)(2 \times2)-dia (fourlayer)
0.83395.391-5.22791920.465(2×2)(2 \times2)-dia (fivelayer)
0.98746.383-5.25611640.436(2×2)(2 \times2)-dia (sixlayer)
1.14237.384-5.27921410.450(2×2)(2 \times2)-dia (sevenlayer)
1.29698.384-5.29611240.473(2×2)(2 \times2)-dia (eightlayer)
1.45159.383-5.30941110.480(2×2)(2 \times2)-dia (ninelayer)

Table 11.

ρ\rho [at./Å2^2]ρ\rho [ML]EKSE_{\mathrm{KS}} [eV/at.]E−EbulkE-E_{\mathrm{bulk}} [meV/at.]EgE_{\mathrm{g}} [eV]sample designation
0.21271.351-5.17515500.508db-monolayer (oP10)
0.39032.479-5.32733980.371Si.iso-II-2.5
0.61313.894-5.42563000.787Si.iso-II-4
0.83765.320-5.47622490.731Si.iso-II-5.5
0.51053.242-5.37733480.950Si.equi-II-3
0.53813.417-5.39573290.847Si.equi-II-3.5
0.39192.488-5.30444210.000(3×3)(\sqrt{3} \times\sqrt{3})-dia (bilayer)
0.53563.402-5.38743380.420(2×2)(2 \times2)-dia (trilayer)
0.84625.374-5.50512200.334c(2×4)c(2 \times4)-dia (fivelayer)
0.47553.020-5.34603790.000(3×3)(3 \times3)-DAS (trilayer)
0.63054.004-5.42742980.000(3×3)(3 \times3)-DAS (fourlayer)
0.44352.817-5.26924560.034a-Si (trilayer)
0.61433.901-5.36523600.070a-Si (fourlayer, cF8 nuclei)

Table S10.3.1. Tabulated content of Figure 6 (main text) for Si thin films. “KS” denotes total Kohn-Sham energy as a direct output of VASP (without dispersion vdW correction). For unit conversion: thickness of 1ML of (111)-diamond cF8 Si is ρ0=0.15469\rho_0 = 0.15469 at./Å2^2 and the respective bulk Kohn-Sham per-atom energy Ebulk=−5.4202E_{\mathrm{bulk}} = -5.4202 eV/at. EgE_{\mathrm{g}} stands for the width of electronic band gap.

Table S10.3.2: Tabulated content of Figure S7.1 (SM) for Si thin films. “KS” denotes total Kohn-Sham energy as a direct output of VASP, with dispersion vdW correction (PBE-D3, IVDW=12). For unit conversion: thickness of 1 ML of (111)-diamond cF8 Si is ρ0=0.157457 at./A˚2\rho_0 = 0.157457\ \mathrm{at./\AA^2} and the respective bulk Kohn-Sham per-atom energy Ebulk=−5.7251 eV/atE_{\mathrm{bulk}} = -5.7251\ \mathrm{eV/at}. EgE_g stands for the width of electronic band gap.

References

  1. [1]C. J. Pickard, R. J. Needs: Hypothetical Low-Energy Chiral Framework Structure of Group 14 Elements. Physical Review B, VOLUME 81, ISSUE 1: 014106, 2010. URL https://link.aps.org/doi/10.1103/PhysRevB.81.014106.DOI
  2. [2]E. Pospíšilová, M. Mihalkovič: Prediction of Coherent Interfaces between Diamond and Clathrate Structures. Computational Materials Science, VOLUME 226: 112228, 2023. URL https://doi.org/10.1016/j.commatsci.2023.112228.
  3. [3]Y. Zhang, A. Rubio and G. Le Lay: Emergent Elemental Two-Dimensional Materials beyond Graphene. Journal of Physics D: Applied Physics, VOLUME 50, ISSUE 5: 053004, 2017. URL https://doi.org/10.1088/1361-6463/aa4e8b.
  4. [4]J. Deng, B. Xia, X. Ma, et al.: Epitaxial Growth of Ultraflat Stanene with Topological Band Inversion. Nature Materials, VOLUME 17: 1081, 2018. URL https://doi.org/10.1038/s41563-018-0203-5.
  5. [5]F. Matusalem, M. Marques, L. K. Teles and F. Bechstedt: Stability and Electronic Structure of Two-Dimensional Allotropes of Group-IV Materials. PHYSICAL REVIEW B, VOLUME 92, ISSUE 4: 045436, 2015. URL https://doi.org/10.1103/PhysRevB.92.045436.
  6. [6]P. Borlido, C. Rödl, M. A. L. Marques and S. Botti: The Ground State of Two-Dimensional Silicon. 2D Materials, VOLUME 5, ISSUE 3: 035010, 2018. URL https://doi.org/10.1088/2053-1583/aab9ea.
  7. [7]S. Cahangirov, O. Özçelik, L. Xian, J. Avila: Atomic Structure of the √3 × √3 Phase of Silicene on Ag(111). Physical Review B, VOLUME 90: 035448, 2014. URL https://doi.org/10.1103/PhysRevB.90.035448.
  8. [8]S. Saxena, R. P. Chaudhary & S. Shukla: Stanene: Atomically Thick Free-standing Layer of 2D Hexagonal Tin. Scientific Reports, VOLUME 6: 31073, 2016. URL https://doi.org/10.1038/srep31073.
  9. [9]Peizhe Tang et al.: Stable Two-Dimensional Dumbbell Stanene: A Quantum Spin Hall Insulator. Physical Review B, VOLUME 90, ISSUE 12: 121408, 2014. URL https://link.aps.org/doi/10.1103/PhysRevB.90.121408.DOI
  10. [10]Yong Xu et al.: Large-Gap Quantum Spin Hall Insulators in Tin Films. Physical Review Letters, VOLUME 111: 136804, 2013. URL https://doi.org/10.1103/PhysRevLett.111.136804.
  11. [11]M. Liao, Y. Zang, Z. Guan et al.: Superconductivity in Few-Layer Stanene. Nature Physics, VOLUME 14: 344, 2018. URL https://doi.org/10.1038/s41567-017-0031-6.
  12. [12]M. E. Dávila, L. Xian, S. Cahangirov, A. Rubio and G. L. Lay: Germanene: A Novel Two-Dimensional Germanium Allotrope Akin to Graphene and Silicene. New Journal of Physics, VOLUME 16: 095002, 2014. URL https://doi.org/10.1088/1367-2630/16/9/095002.
  13. [13]P. Vogt et al.: Silicene: Compelling Experimental Evidence for Graphene-Like Two-Dimensional Silicon. Physical Review Letters, VOLUME 108: 155501, 2012. URL https://doi.org/10.1103/PhysRevLett.108.155501.
  14. [14]V. O. Özçelik, E. Durgun, and S. Ciraci: New Phases of Germanene. The Journal of Physical Chemistry Letters, VOLUME 5, ISSUE 15: 2694, 2014. URL https://doi.org/10.1021/jz500977v.
  15. [15]A. Popescu, P. Rodriguez-Lopez, and L. M. Woods: Composition and Stacking Dependent Topology in Bilayers from the Graphene Family. Physical Review Materials, VOLUME 3: 064002, 2019. URL https://doi.org/10.1103/PhysRevMaterials.3.064002.
  16. [16]H. Fu, J. Zhang, Zijing Ding, Hui Li, Sheng Meng: Stacking-Dependent Electronic Structure of Bilayer Silicene. Applied Physics Letters, VOLUME 104, NUMBER 13: 131904, 2014. URL https://doi.org/10.1063/1.4870534.
  17. [17]J. E. Padilha and R. B. Pontes: Free-Standing Bilayers Silicene: The Effect of Stacking Order on the Structural, Electronic and Transport Properties. The Journal of Physi- [18] B. Feng, Z. Ding, S. Meng, Y. Yao, X. He, P. Cheng, L. Chen and K. Wu: Evidence of Silicene in Honeycomb Structures of Silicon on Ag(111). Nano Letters, VOLUME 12, NUMBER 7: 3507, 2012. URL https://doi.org/10.1021/nl301047g.
  18. [19]R. Arafune, C.-L. Lin, K. Kawahara, N. Tsukahara, E. Minamitani, Y. Kim, N. Takagi, M. Kawai: Structural Transition of Silicene on Ag(111). Surface Science, VOLUME 608: 297, 2013. URL https://doi.org/10.1016/j.susc.2012.10.022.
  19. [20]Q. Xiong, W. Cen, X. Wu, C. Chen, L. Lv: Effect of Stacking Mode on Electronic Structure and Optical Properties of Bilayer Germanene. Solid State Communications, VOLUME 360: 115036, 2023. URL https://doi.org/10.1016/j.ssc.2022.115036.
  20. [21]P. Borlido et al.: Structural Prediction of Stabilized Atomically Thin Layers. npj 2D Materials and Applications, VOLUME 3, ISSUE 21, 2019. URL https://doi.org/10.1038/s41699-019-0103-9.
  21. [22]Materials Project, Materials Virtual Lab. http://crystalium.materialsvirtuallab.org/. Accessed: 2025-02-15.
  22. [23]A. A. Stekolnikov and F. Bechstedt: Shape of Free and Constrained Group-IV Crystallites: Influence of Surface Energies. Physical Review B, VOLUME 72, ISSUE 12: 125326, 2005. URL https://link.aps.org/doi/10.1103/PhysRevB.72.125326.DOI
  23. [24]A. A. Stekolnikov, J. Furthmüller, and F. Bechstedt: Absolute Surface Energies of Group-IV Semiconductors: Dependence on Orientation and Reconstruction. Physical Review B, VOLUME 65, ISSUE 11: 115318, 2002. URL https://link.aps.org/doi/10.1103/PhysRevB.65.115318.DOI
  24. [25]G.-H. Lu, M. Huang, M. Cuma, F. Liu: Relative Stability of Si Surfaces: A First-Principles Study. Surface Science, VOLUME 588, ISSUE 1-3: 61, 2005. URL https://doi.org/10.1016/j.susc.2005.05.028.
  25. [26]G. V. Gadiyak, A. A. Karpushin, A. S. Kushkova, Yu. N. Morokov, S. M. Repinskiy, A. A. Shklyaev: Wulff’s Diagram for the Surfaces of Silicon and Germanium. Quantum Chemical Calculation. Physics, Chemistry and Mechanics of Surfaces, VOLUME 3, 1988. URL https://www.researchgate.net/profile/Yuri-Morokov/publication/301617910_Wulff’s_diagram_for_the_surfaces_of_silicon_and_germanium_Quantum_chemical_calculation/links/571d8fca08ae408367be5b07/Wulffs-diagram-for-the-surfaces-of-silicon-and-germanium-Quantum-chemical-calculation.pdf.
  26. [27]G. V. Gadiyak, Yu. N. Morokov, A. S. Kushkova, S. M. Repinskii, and A. A. Shklyaev: Calculation of Wulff’s Diagram for Germanium. Physics, Chemistry and Mechanics of Surfaces, VOLUME 3, NUMBER 7: 1978, 1985. URL https://www.researchgate.net/profile/Yuri-Morokov/publication/295087759_Calculation_of_Wulff’s_diagram_for_germanium/links/571a664b08ae408367bc8b24/Calculation-of-Wulffs-diagram-for-germanium.pdf.
  27. [28]J.-Y. Lee et al.: The Surface Energy and Stress of Metals. Surface Science, VOLUME 674: 51, 2018. URL https://doi.org/10.1016/j.susc.2018.03.008.
  28. [29]R. Zhachuk, J. Coutinho, A. Dolbak, V. Cherepanov, and B. Voigtländer: Si(111) Strained Layers on Ge(111): Evidence for c(2×4) Domains. Physical Review B, VOLUME 96, ISSUE 8: 085401, 2017. URL https://link.aps.org/doi/10.1103/PhysRevB.96.085401.DOI
  29. [30]A. A. Baski, S. C. Erwin, L. J. Whitman: The Structure of Silicon Surfaces from (001) to (111). Surface Science, VOLUME 392, ISSUES 1-3: 69, 1997. URL https://doi.org/10.1016/S0039-6028(97)00499-8.
  30. [31]R. A. Zhachuk, and J. Coutinho: Crucial Role of Vibrational Entropy in the Si(111)-7 × 7 Surface Structure Stability. Physical Review B, VOLUME 105, ISSUE 24: 245306, 2022. URL https://link.aps.org/doi/10.1103/PhysRevB.105.245306.DOI
  31. [32]R. Zhachuk, S. Teys, J. Coutinho: Strain-Induced Structure Transformations on Si(111) and Ge(111) Surfaces: A Combined Density-Functional and Scannning Tunnelling Microscopy Report. The Journal of Chemical Physics, VOLUME 138, ISSUE 22: 224702, 2013. URL https://doi.org/10.1063/1.4808356.
  32. [33]V. K. Singh, E. Pospišilová, M. Mihalkovič, S. R. Barman et al.: Decagonal Sn Clathrate on d-Al-Ni-Co. Physical Review B, VOLUME 107, ISSUE 4: 045410, 2023. URL https://link.aps.org/doi/10.1103/PhysRevB.107.045410.DOI
  33. [34]V. K. Singh, M. Mihalkovič, M. Krajčí, S. Sarkar, P. Sadhukhan, M. Mankraj, A. Rai, K. Pussi, D. L. Schlagel, T. A. Lograsso, A. K. Shukla, S. R. Barman: Quasiperiodic Ordering in Thick Sn Layer on i-Al-Pd-Mn: A Possible Quasicrystalline Clathrate. Physical Review Research, VOLUME 2, ISSUE 1: 013023, 2020. URL https://doi.org/10.1103/PhysRevResearch.2.013023.
  34. [35]R. Kumar, Y. Hazama, F. Ohashi, H. S. Jha, T. Kume: A Fabrication Method for Type-II Ge Clathrate Film by Annealing of Ge Film Covered with Na Layer. Thin Solid Films, VOLUME 734: 138859, 2021. URL https://doi.org/10.1016/j.tsf.2021.138859.
  35. [36]T. Kume et al.: A Thin Film of a Type II Ge Clathrate Epitaxially Grown on a Ge Substrate. CrystEngComm, VOLUME 18, ISSUE 30: 5630, 2016. URL https://doi.org/10.1039/C6CE01148A.
  36. [37]E. G. Noya, I. Zubieta, D. J. Pine, and F. Sciortino: Assembly of Clathrates from Tetrahedral Patchy Colloids with Narrow Patches. The Journal of Chemical Physics, VOLUME 151, ISSUE 9: 094502, 2019. URL https://doi.org/10.1063/1.5109382.
  37. [38]M. H. Factorovich, P. M. Naullage, and V. Molinero: Can Clathrates Heterogeneously Nucleate Ice? The Journal of Chemical Physics, VOLUME 151, ISSUE 11: 114707, 2019. URL https://doi.org/10.1063/1.5119823.
  38. [39]V. Molinero and E. B. Moore: Water Modeled As an Intermediate Element between Carbon and Silicon. The Journal of Physical Chemistry B, VOLUME 113, NUMBER 13: 4008, 2008. URL https://doi.org/10.1021/jp805227c.
  39. [40]A. D. Martinez, L. Krishna, L. L. Baranowski, M. T. Lusk, E. S. Toberer, and A. C. Tamboli: Synthesis of Group IV Clathrates for Photovoltaics. IEEE Journal of Photovoltaics, VOLUME 3, NUMBER 4: 1305, 2013. URL https://doi.org/10.1109/JPHOTOV.2013.2276478.
  40. [41]D. Connétable, V. Timoshevskii, E. Artacho, and X. Blase: Tailoring Band Gap and Hardness by Intercalation: An ab initio Study of I8@Si-46 and Related Doped Clathrates. Physical Review Letters, VOLUME 87, NUMBER 20: 206405, 2001. URL https://doi.org/10.1103/PhysRevLett.87.206405.
  41. [42]K. Moriguchi, S. Munetoh, and A. Shintani: First-Principles Study of Si₃₄₋ₓGeₓ Clathrates: Direct Wide-Gap Semiconductors in Si-Ge Alloys. Physical Review B, VOLUME 62, NUMBER 11: 7138, 2000. URL https://doi.org/10.1103/PhysRevB.62.7138.
  42. [43]L. Grigorian, P. C. Eklund, S. Fang: Clathrate Structure for Electronic and Electro-Optic Applications. University of Kentucky, Physics and Astronomy Faculty, NUMBER 6103403, 2000. URL https://uknowledge.uky.edu/physastron_patents/5.
  43. [44]G. Kresse and J. Furthmüller: Efficient Iterative Schemes for ab initio Total-Energy Calculations Using a Plane-Wave Basis Set. Physical Review B, VOLUME 54, ISSUE 16: 11169, 1996. URL https://link.aps.org/doi/10.1103/PhysRevB.54.11169.DOI
  44. [45]G. Kresse and D. Joubert: From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Physical Review B, VOLUME 59, ISSUE 3: 1758, 1999. URL https://link.aps.org/doi/10.1103/PhysRevB.59.1758.DOI
  45. [46]J. P. Perdew, J. A. Chevary, S. H. Vosko, K. A. Jackson, M. R. Pederson, D. J. Singh, and C. Fiolhais: Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient Approximation for Exchange and Correlation. Physical Review B, VOLUME 46: 6671, 1992. URL https://link.aps.org/doi/10.1103/PhysRevB.46.6671.DOI
  46. [47]B. Haberl, T. A. Strobel, and J. E. Bradby: Pathways to Exotic Metastable Silicon Allotropes. Applied Physics Reviews, VOLUME 3, ISSUE 4: 040808, 2016. URL https://doi.org/10.1063/1.4962984.
  47. [48]D. Selli, I. A. Baburin, R. Martoňák: Novel Metastable Metallic and Semiconducting Germaniums. Scientific Reports, VOLUME 3: 1466, 2013. URL https://doi.org/10.1038/srep01466.
  48. [49]K. Sudoh, M. Naito: Interfacial Reaction During Dewetting of Ultrathin Silicon on Insulator. Comptes Rendus Physique, VOLUME 14, ISSUE 7: 601, 2013. URL https://doi.org/10.1016/j.crhy.2013.06.008.
  49. [50]Ya-Ping Wang et al.: Large-Gap Quantum Spin Hall State in Functionalized Dumbbell Stanene. Applied Physics Letters, VOLUME 108, ISSUE 7: 073104, 2016. URL https://doi.org/10.1063/1.4942380.
  50. [51]S. Hara, S. Izumi, T. Kumagai, S. Sakai: Surface Energy, Stress and Structure of Well-Relaxed Amorphous Silicon: A Combination Approach of ab initio and Classical Molecular Dynamics. Surface Science, VOLUME 585, NUMBER 1: 17, 2005. URL https://doi.org/10.1016/j.susc.2005.03.061.
  51. [52]Haisheng Li et al.: Threadlike Tin Clusters with High Thermal Stability Based on Fundamental Units. The Journal of Physical Chemistry C, VOLUME 116, ISSUE 1: 231, 2012. URL https://doi.org/10.1021/jp208121s.
  52. [53]Haisheng Li et al.: Tin Clusters Formed by Fundamental Units: A Potential Way to Assemble Tin Nanowires. Physical Chemistry Chemical Physics, VOLUME 15, ISSUE 6: 1831, 2013. URL https://doi.org/10.1039/C2CP42948A.
  53. [54]F. C. Frank, J. S. Kasper: Complex Alloy Structures Regarded as Sphere Packings. I. Definitions and Basic Principles. Acta Crystallographica, VOLUME 11: 184, 1958. URL https://doi.org/10.1107/S0365110X58000487.
  54. [55]F. C. Frank, J. S. Kasper: Complex Alloy Structures Regarded as Sphere Packings. II. Analysis and Classification of Representative Structures. Acta Crystallographica, VOLUME 12: 483, 1959. URL https://doi.org/10.1107/S0365110X59001499.
  55. [56]B. Wehinger et al.: Diffuse Scattering in Metallic Tin Polymorphs. Journal of Physics: Condensed Matter, VOLUME 26, ISSUE 11: 115401, 2014. URL https://iopscience.iop.org/article/10.1088/0953-8984/26/11/115401.
  56. [57]R. Kubiak: Evidence for the Existence of the γ Form of Tin. Journal of the Less-Common Metals, VOLUME 116, ISSUE 2: 307, 1986. URL https://doi.org/10.1016/0022-5088(86)90663-6.
  57. [58]G. Bergmann: Upper Critical Field and the Density of States in Amorphous Super Conductors. Physical Review B, VOLUME 7, ISSUE 11: 4850, 1973. URL https://doi.org/10.1007/978-1-4684-2688-5_113.
  58. [59]K. Knorr, N. Barth: Investigation of Disordered Sn-Films by Superconducting Tunneling. Solid State Communications, VOLUME 6, ISSUE 11: 791, 1968. URL https://doi.org/10.1016/0038-1098(68)90121-X.
  59. [60]W. Buckel: Superconductivity and Metastable Phases in Thin Films. Journal of Vacuum Science & Technology, VOLUME 10, ISSUE 5: 599, 1973. URL https://doi.org/10.1116/1.1318397.
  60. [61]M. Vergnat, G. Marchal, and Ph. Mangin: Preparation and Structure of Amorphous Semiconductor Hydrogenated Tin. Applied Physics Letters, VOLUME 57, ISSUE 22: 2300, 1990. URL https://doi.org/10.1063/1.103893.
  61. [62]L. K. Nanver, G. Weyer, and B. I. Deutch: Experimental Study of the Amorphous Phases of Group-IV Semiconductors by the ¹¹⁹mSn Mössbauer Probe. Zeitschrift für Physik B Condensed Matter, VOLUME 47, ISSUE 2: 103, 1982. URL https://doi.org/10.1007/BF01441292.
  62. [63]P. Li, W. Fu, P. Zhuang, Y. Cao, C. Tang, A. B. Watson, P. Dong, J. Shen, M. Ye: Amorphous Sn/Crystalline SnS₂ Nanosheets via In Situ Electrochemical Reduction Methodology for Highly Efficient Ambient N₂ Fixation. Small, VOLUME 15, ISSUE 40: 1902535, 2019. URL https://doi.org/10.1002/smll.201902535.
  63. [64]M. Ryník, S. Leoni, and R. Martoňák: Pressure-Induced Structural Transformation of Clathrate Ge136 via an Ultrafast Recrystallization of an Amorphous Intermediate. Physical Review B, VOLUME 105, ISSUE 13: 134107, 2022. URL https://link.aps.org/doi/10.1103/PhysRevB.105.134107.DOI
  64. [65]V. V. Brazhkin, A. G. Lyapin, S. V. Popova, and R. N. Voloshin: Nonequilibrium Phase Transitions and Amorphization in Si, Si/GaAs, Ge, and Ge/Gash at the Decompression of High-Pressure Phases. Physical Review B, VOLUME 51, ISSUE 12: 7549, 1995. URL https://link.aps.org/doi/10.1103/PhysRevB.51.7549.DOI
  65. [66]Liang Xu et al.: Folded Network and Structural Transition in Molten Tin. Nature Communications, VOLUME 13: 126, 2022. URL https://doi.org/10.1038/s41467-021-27742-2.
  66. [67]P. Biswas, R. Atta-Fynn, and D. A. Drabold: Reverse Monte Carlo Modeling of Amorphous Silicon. Physical Review B, VOLUME 69, ISSUE 19: 195207, 2004. URL https://link.aps.org/doi/10.1103/PhysRevB.69.195207.DOI
  67. [68]J. Fortner and J. S. Lannin: Radial Distribution Functions of Amorphous Silicon. Physical Review B, VOLUME 39, NUMBER 8: 5527, 1989. URL https://link.aps.org/doi/10.1103/PhysRevB.39.5527.DOI
  68. [69]E. Guerrero, and D. A. Strubbe: Structural Changes and Void Generation in Low-Density Amorphous Silicon: A Computational Study. arXiv preprint arXiv:1907.01327, 2019. URL https://arxiv.org/abs/1907.01327.
  69. [70]A. Pedersen, L. Pizzagalli and H. Jónsson: Optimal Atomic Structure of Amorphous Silicon Obtained from Density Functional Theory Calculations. New Journal of Physics, VOLUME 19, NUMBER 6: 063018, 2017. URL https://iopscience.iop.org/article/10.1088/1367-2630/aa732e/pdf.DOI
  70. [71]R. V. Meidanshahi, S. Bowden and S. M. Goodnick: Electronic Structure and Localized States in Amorphous Si and Hydrogenated Amorphous Si. Physical Chemistry Chemical Physics, VOLUME 21, ISSUE 24: 13248, 2019. URL https://doi.org/10.1039/C9CP01121H.
  71. [72]I. Stich, R. Car, M. Parrinello: Amorphous Silicon Studied by ab initio Molecular Dynamics: Preparation, Structure, and Properties. Physical Review B, VOLUME 44, NUMBER 20: 11092, 1991. URL https://link.aps.org/doi/10.1103/PhysRevB.44.11092.DOI
  72. [73]D. Choudhary and P. Clancy: Characterizing the Nature of Virtual Amorphous Silicon. The Journal of Chemical Physics, VOLUME 122, ISSUE 17: 174509, 2005. URL https://link.aps.org/doi/10.1063/1.1888566.DOI
  73. [74]R. Li, E. Lee, T. Luo: A Unified Deep Neural Network Potential Capable of Predicting Thermal Conductivity of Silicon in Different Phases. Materials Today Physics, VOLUME 12: 100181, 2020. URL https://doi.org/10.1016/j.mtphys.2020.100181.
  74. [75]H. M. Shodja, M. Tabatabaei and K. Esfarjani: First Principles Molecular Dynamics Studies of Elastic Constants, Ideal Tensile Strength, Chemistry of Crack Initiation, and Surface and Cohesive Energies in Amorphous Silicon. Philosophical Magazine, VOLUME 94, NUMBER 25: 2913, 2014. URL https://doi.org/10.1080/14786435.2014.939735.
  75. [76]I. Santos, M. Aboy, L. A. Marqués, P. López, L. Pelaz: Generation of Amorphous Si Structurally Compatible with Experimental Samples through the Quenching Process: A Systematic Molecular Dynamics Simulation Study. Journal of Non-Crystalline Solids, VOLUMES 503-504: 20, 2019. URL https://doi.org/10.1016/j.jnoncrysol.2018.09.024.
  76. [77]A. Pedersen, L. Pizzagalli, and H. Jónsson: Atomic and Electronic Structures of a Vacancy in Amorphous Silicon. Physical Review B, VOLUME 101, ISSUE 5: 054204, 2020. URL https://link.aps.org/doi/10.1103/PhysRevB.101.054204.DOI
  77. [78]Y. Furukawa and Y. Masushima: First-Principles Prediction of Densities of Amorphous Materials: The Case of Amorphous Silicon. Journal of the Physical Society of Japan, VOLUME 87, NUMBER 2: 024701, 2017. URL https://doi.org/10.48550/arXiv.1704.06107.
  78. [79]J. S. Custer et al.: Density of Amorphous Si. Applied Physics Letters, VOLUME 64, ISSUE 4: 437, 1994. URL https://doi.org/10.1063/1.111121.
  79. [80]C. Messmer, J. C. Bilello: The Surface Energy of Si, GaAs, and GaP. Journal of Applied Physics, VOLUME 52, ISSUE 7: 4623, 1981. URL https://doi.org/10.1063/1.329342.
  80. [81]R. J. Jacobsson: Surface Energy of Germanium and Silicon. Journal of The Electrochemical Society, VOLUME 110, NUMBER 6: 524, 1963. URL https://iopscience.iop.org/article/10.1149/1.2425806.
  81. [82]J. J. Gilman: Direct Measurements of the Surface Energies of Crystals. Journal of Applied Physics, VOLUME 31, ISSUE 12: 2208, 1960. URL https://doi.org/10.1063/1.1735524.
  82. [83]A. Mujica, C. J. Pickard, R. J. Needs: Low-Energy Tetrahedral Polymorphs of Carbon, Silicon, and Germanium. Physical Review B, VOLUME 91, ISSUE 21: 214104, 2015. URL https://link.aps.org/doi/10.1103/PhysRevB.91.214104.DOI
  83. [84]J. Conesa: Computer Modeling of allo-Si and allo-Ge Polymorphs. The Journal of Physical Chemistry B, VOLUME 106, ISSUE 13: 3402, 2002. URL https://doi.org/10.1021/jp014115g.
  84. [85]Z. Zhao et al.: Tetragonal Allotrope of Group 14 Elements. Journal of the American Chemical Society, VOLUME 134, NUMBER 30: 12362, 2012. URL https://doi.org/10.1021/ja304380p.
  85. [86]E. Pospíšilová: Search for Stable Clathrate Films within Elements from Column XIV of the Periodic Table (Dissertation Thesis, Comenius University in Bratislava, 2024). https://opac.crzp.sk/?fn=detailBiblioForm&sid=C007B9F4EAC E4D3BE4CB45B8D8B7.
  86. [87]A. J. Karttunen, D. Usvyat, M. Schütz and L. Maschio: Dispersion Interactions in Silicon Allotropes. Physical Chemistry Chemical Physics, VOLUME 19, ISSUE 11: 7699, 2017. URL https://doi.org/10.1039/C6CP08873B.
  87. [88]F. Tran, J. Stelzl, P. Blaha: Rungs 1 to 4 of DFT Jacob’s Ladder: Extensive Test on the Lattice Constant, Bulk Modulus, and Cohesive Energy of Solids. The Journal of Chemical Physics, VOLUME 144, ISSUE 20: 204120, 2016. URL S. Grimme, J. Antony, S. Ehrlich, H. Krieg: A Consistent and Accurate ab initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. The Journal of Chemical Physics, VOLUME 132, ISSUE 15: 154104, 2010. URL https://doi.org/10.1063/1.3382344.
  88. [90]S. Grimme, S. Ehrlich, L. Goerigk: Effect of the Damping Function in Dispersion Corrected Density Functional Theory. Journal of Computational Chemistry, VOLUME 32: 1456, 2011. URL https://doi.org/10.1002/jcc.21759.
  89. [91]S. Grimme: Density Functional Theory with London Dispersion Corrections. WIREs Computational Molecular Science, VOLUME 1: 211, 2011. URL https://doi.org/10.1002/wcms.30.
  90. [92]P. Blaha and F. Tran: DFT Approximations: Which One to Use? https://dokumen.tips/documents/dft-approximations-which-one-to-use-p-blaha-and-f-tran.html. Institute of Materials Chemistry, TU Wien, lecture, pg. 8.

Paper details

vlnc:W_dajx4ewx
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents