betaFeedback

WorksPhysics and Astronomy

Simultaneous Superconducting and Topological Properties in Mg-Li Electrides at High Pressures

Abstract

Electrides as a unique class of emerging materials exhibit fascinating properties and hold important significance for understanding the matter under extreme conditions, which is characterized by valence electrons localized into the interstitial space as quasi-atoms (ISQs). In this work, using crystal structure prediction and first-principles calculations, we identified seven stable phases of Mg-Li that are electride with novel electronic properties under high pressure. Among them, MgLi10 is a semiconductor with a band gap of 0.22 eV; and Pm-3m MgLi is superconductor with a superconducting transition temperature of 22.8 K. The important role played by the localization degree of ISQ in the superconducting transition temperature of these electrides is revealed by systematic comparison of Mg-Li with other Li-rich electride superconductors. Furthermore, we proved that Pm-3m MgLi and Pnma MgLi also have distinct topological behavior with metallic surface states and the non-zero Z2Z_2 invariant. The simultaneous coexistence of superconductivity, electronic band topology and electride property in the same structure of Pm-3m MgLi and Pnma MgLi demonstrates the feasibility of realizing multi-quantum phases in a single material, which will stimulate further research in these interdisciplinary fields.

simultaneous coexistence of superconductivity, electronic band topology and electride property in the same structure of Pm-3mPm\text{-}3m MgLi and PnmaPnma MgLi demonstrates the feasibility of realizing multi-quantum phases in a single material, which will stimulate further research in these interdisciplinary fields.

Introduction

The exploration of superconductors is of great interest in condensed matter physics and materials science because of their unique potential value in civil and industrial application[1, 2, 3, 4].

Particularly, alkali metals as electrides[5, 6, 7] exhibit superconductivity which is quite different from hydrogen-rich superconductors[8, 9]. In electrides, the highly-localized electrons in the interstices of the lattice serve as anions, which are known as interstitial quasi-atom (ISQ)[10] and lead to strong core-ISQ polarization, resulting in exotic phenomena such as LA-TA splitting[11], universal surface interface metallic states[12], and colossal charge state of impurity atoms[7]. The presence of ISQs also can induce electronic phase transitions from metal to semiconductor, even to insulator, and with possible superconducting phase in them[13, 14, 15, 16, 17].

Recent high-pressure studies indicated that some Li-rich electrides might exhibit superconductivity generally, including Li5C\mathrm{Li}_5\mathrm{C}[18], Li5N\mathrm{Li}_5\mathrm{N}[19], Li6P\mathrm{Li}_6\mathrm{P}[20], Li9Te\mathrm{Li}_9\mathrm{Te}[21], and Li8As\mathrm{Li}_8\mathrm{As}[22], in which it seems ISQs enhance superconductivity. However, in electrides Li6C\mathrm{Li}_6\mathrm{C}[23] and Li10Se\mathrm{Li}_{10}\mathrm{Se}[24], the high superconducting critical temperature is attributed to the strong electron-phonon coupling derived from the synergy of interatomic coupling effect, while ISQs impede superconductivity.

On the other hand, ISQs also contribute to the fusion and stability of intermetallic electride superconductors such as Be2Li\mathrm{Be}_2\mathrm{Li}[25], LiCa\mathrm{LiCa}[26], Li6Al\mathrm{Li}_6\mathrm{Al}[27], Li8Cs\mathrm{Li}_8\mathrm{Cs}[28]. Recently, in the design of electrides Li8Au29\mathrm{Li}_{8}\mathrm{Au}_{29} and Li8H4\mathrm{Li}_{8}\mathrm{H}_{4}, the ISQs are adopted as a feature to achieve diverse topology to improve superconducting temperatures. Overall, the strong charge polarization in electrides provides a versatile tool to tune the electronic structure and superconductivity[22]. However, the role played by ISQs in superconductivity remains contentious. To establish a clear relation between ISQs and superconductivity is still imperative for to realize high-temperature electride superconductors.

Due to the localization of electron to the interstitial sites to achieve partial independence (i.e., only weakly associated with the nucleus), all insulating and semiconducting electrides are predicted having universal and robust metallic surface or interface states[12], even though some of them might be topologically trivial in terms of conventional topological band structure theory[31], which indicates electride might have unknown de facto real-space topology that beyond the conventional theories. Nevertheless, some electrides are indeed topological nontrivial in the electronic band structure[32, 33] according to conventional topological band theory. For instance, Ca3Pb\mathrm{Ca}_{3}\mathrm{Pb} possesses a topologically nontrivial band structure[34]. In Cs3O\mathrm{Cs}_{3}\mathrm{O} and Ba5N\mathrm{Ba}_{5}\mathrm{N}, the anionic electrons induce an ss-pp band inversion and result in a nontrivial band topology[35]. Two dimensional electrides Y2C\mathrm{Y}_{2}\mathrm{C}, Sr2Bi\mathrm{Sr}_{2}\mathrm{Bi}, and HfBr\mathrm{HfBr} have also been reported being topological semimetals[36, 37]. Furthermore, the electride phase of simple metals also display nontrivial semimetallic electronic structure under pressure[38]. The high-pressure polymorphs of lithium are found to be topological nodal loop semimetals[39]; the alkali earth metals (Be, Mg, Ca, and Sr) exhibit Dirac nodal line and topological surface states[40].

In principle, the superconducting and conventional band topological states could be envisioned to emerge in the same electride. Their interplay might lead to interesting properties,enders_mid=1 such as the interaction between superconducting surface/edge current and colossal charge states[7] or the stationary longitudinal acoustic phonon mode[11, 12]. The search for novel topological superconducting electride thus is one of the major focuses in material science. As the lightest metallic element, lithium has a rich phase diagram under pressure, and exhibits abundant physicochemical properties[38, 41, 42]. In addition, its melting behaves like hydrogen (for example, the “U” shape in the melting curve after the melting temperature reaches a maximum[43]) under compression; but it maintains electride states that favor superconductivity before entering an semiconducting phase.

In this work, we report seven stable compounds with a wide stoichiometries spaning from 7:1 to 1:10 in Mg-Li system up to 500 GPa. Particularly, Pm-3m MgLi and Pnma MgLi become superconducting electrides with a noteworthy TcT_c of 22.80 K and 8.91 K at high pressure, respectively. By composing and analyzing the variation of superconductivity in typical Li-contained electride, we obtain a practical relationship between the localization degree of ISQs and TcT_c, which could be employed as a guide to tune the superconducting temperatures by adjusting the localization of ISQs. Furthermore, it is demonstrated that Pm-3m MgLi and Pnma MgLi are also topological materials with a nontrivial Z2Z_2 invariant and gapless surface states, thus as the first example of hosting superconductivity and topological nontrivial states in the same electride at the same time, which provides a platform for exploring the interplay between superconducting, band topological states, and non-nuclear centering of electrons.

Computational details

The thermodynamically stable structures of Mg-Li compounds under pressure are explored by using the ab initio evolutionary algorithm USPEX[44, 45, 46]. Structural optimization and electronic property calculations are performed by using the density functional theory (DFT)[47], [48] with the projector augmented wave (PAW)[49] method and the Perdew-Burke-Ernzerhof (PBE) parameterization of the generalized gradient approximation (GGA)[50] functional as implemented in the Vienna Ab-initio Simulation Package (VASP)[51], [52]. The energy cutoff for the plane wave basis set is chosen as 700 eV in all calculations. The largest spacing between k-points is set as 2π×0.0152\pi\times0.015 Å−1^{-1} to generate automatic k-point meshes for the Brillouin zone sampling. All calculations are guaranteed to converge to 10−510^{-5} eV/atom for energy in the self-consistent field iterations and 0.01 eV/Å for the Hellmann-Feynman forces on atoms, respectively. The 1s22s11s^{2}2s^{1} and 2s22p63s22s^{2}2p^{6}3s^{2} electrons are treated in the valence space for Li and Mg, respectively. The Bader charge analysis[53] is employed to characterize the charge state of atoms and the charge transfer between atoms and ISQs. Lattice dynamics are calculated by using the small displacement and supercell approach as implemented in our homemade MyPHON code[54], as well as the open source PHONOPY[55] code. The formation enthalpy (ΔH\Delta H) of MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n} with respect to the elemental Mg and Li at each pressure is defined as:

ΔH(MgmLin)=H(MgmLin)−mH(Mg)−nH(Li)m+n(1)\Delta H(\mathrm{Mg}_{m}\mathrm{Li}_{n}) = \frac{H(\mathrm{Mg}_{m}\mathrm{Li}_{n}) - mH(\mathrm{Mg}) - nH(\mathrm{Li})}{m+n} \tag*{(1)}

where HH is the enthalpy of the studied compound structure or the most stable elemental reference structure at the given pressure, respectively.

The electron-phonon coupling (EPC) and superconductivity are computed using density functional perturbation theory (DFPT), as implemented in the QUANTUM ESPRESSO (QE) code[56]. The ultrasoft pseudopotential for Li and Mg with a kinetic energy cutoff of 80 Ry are employed. Self-consistent electron density and EPC are calculated by employing carefully checked different kk-meshes and qq-meshes: 24×24×2424 \times24 \times24 kk-meshes and 12×12×1212 \times12 \times12 qq-meshes for Pm-3mPm\text{-}3m MgLi, 20×20×2020 \times20 \times20 kk-meshes and 5×5×55 \times5 \times5 qq-meshes for PnmaPnma MgLi, 16×16×616 \times16 \times6 kk-meshes and 8×8×38 \times8 \times3 qq-meshes for I4/mmmI4/mmm Mg2_2Li, respectively. The superconducting critical temperature TcT_c is estimated using the Allen–Dynes-modified McMillan formula[57, 58] with correction given as follows:

Tc=ωlog⁡1.2exp⁡[−1.04(1+λ)λ−μ∗(1+0.62λ)]λ≤1.5(2)T_c = \frac{\omega_{\log}}{1.2} \exp\left[-\frac{1.04(1+\lambda)}{\lambda-\mu^*(1+0.62\lambda)}\right] \qquad\lambda\le1.5 \tag*{(2)}
Tc=f1f2ωlog⁡1.2exp⁡[−1.04(1+λ)λ−μ∗(1+0.62λ)]λ>1.5(3)T_c = \frac{f_1 f_2 \omega_{\log}}{1.2} \exp\left[-\frac{1.04(1+\lambda)}{\lambda-\mu^*(1+0.62\lambda)}\right] \qquad\lambda> 1.5 \tag*{(3)}

where the effective Coulomb repulsion is taken as μ∗=0.1\mu^* = 0.1. The EPC parameter λ\lambda, the logarithmic average frequency ωlog⁡\omega_{\log}, the strong coupling and shape correction factors f1f_1 and f2f_2 are:

λ=2∫α2F(ω)ω dω(4)\lambda= 2 \int\frac{\alpha^2 F(\omega)}{\omega} \, d\omega \tag*{(4)}
ωlog⁡=2λ∫α2F(ω)ln⁡ωω dω(5)\omega_{\log} = \frac{2}{\lambda} \int\frac{\alpha^2 F(\omega) \ln\omega}{\omega} \, d\omega \tag*{(5)}
f1=1+[λ2.46(1+3.8μ∗)]3/23(6)f_1 = \sqrt[3]{1+\left[\frac{\lambda}{2.46(1+3.8\mu^*)}\right]^{3/2}} \tag*{(6)}
f2=1+(ωˉ2ωlog⁡−1)λ2λ2+[1.82(1+6.3μ∗)(ωˉ2ωlog⁡)]2(7)f_2 = 1 + \frac{\left(\frac{\bar{\omega}_2}{\omega_{\log}}-1\right)\lambda^2}{\lambda^2+\left[1.82(1+6.3\mu^*)\left(\frac{\bar{\omega}_2}{\omega_{\log}}\right)\right]^2} \tag*{(7)}
ωˉ2=2λ∫ωα2F(ω) dω(8)\bar{\omega}_2 = \sqrt{\frac{2}{\lambda}\int\omega\alpha^2 F(\omega) \, d\omega} \tag*{(8)}

Furthermore, using Wannier representations, we also calculate the surface states of the studied compounds using the iterative Green’s function as provided in the WANNIERTOOLS package[59, 60].

Results and discussion

The phase stability of Mgm_mLin_n system is investigated by calculating their enthalpy of formation up to 500 GPa. The compounds located on the convex hull are thermodynamically stable against any type of decomposition. As shown in Figure 1(a), the predicted convex hulls suggest that Mg-rich compounds are easier to form in a wide range of pressures, while the only Li-rich compound MgLi10_{10} with space group P1‾P\overline{1} is stable up to 118 GPa. Figure 1(b) shows the phase diagram and crystal structure of the stable phases of Mgm_mLin_n. The ground-state phase of MgLi is found to be in the CsCl-type[61] structure (space group Pm3‾mPm\overline{3}m, Pearson symbol cP2) within a pressure range of 0-213 GPa. At higher pressures, MgLi decomposes into Mg2_2Li and Li element, as shown in Figure 1(a), Figure S1, and Figure S2. Interestingly, MgLi becomes the ground state again and is stabilized into the orthorhombic MnP-type[62] structure (oP8, space group PnmaPnma) when between 350 GPa and 500 GPa.

Stability of Mg_mLi_n compounds under pressure

Figure 1. (Color online) Stability of Mgm_mLin_n compounds under pressure. (a) Enthalpy of formation of Mgm_mLin_n at selected pressures up to 500 GPa, in which the thermodynamically stable compounds are shown in solid symbols, while metastable ones are in open symbols; (b) Predicted phase diagram for stable Mgm_mLin_n compounds, where the black slashed regions indicate that the corresponding phases are metastable within the given pressure range. Though that the compounds of MgLi5_5, MgLi3_3, MgLi2_2 at 0 GPa, MgLi2_2 at 300 GPa, Mg5_5Li at 400 GPa, and MgLi10_{10}, Mg3_3Li, Mg4_4Li, Mg5_5Li at 500 GPa are very close to the convex hull (within 3∼9 meV/atom), they are denoted as metastable phases here.

Enthalpy differences and energy-cutoff convergence for MgLi phases

Figure S1. (Color online) (a) Enthalpy difference of MgLi in Fmm2\mathit{Fmm2} and P21/c\mathit{P2_{1}/c} phases with respect to the Pm3ˉm\mathit{Pm\bar{3}m} phase for Ecut: 400 eV; The dashed-lines indicate the pressure of the phase transition; (b) Enthalpy difference of MgLi in Fmm2\mathit{Fmm2}, P21/c\mathit{P2_{1}/c}, Cmcm\mathit{Cmcm}, and Pnma\mathit{Pnma} phases with respect to the Pm3ˉm\mathit{Pm\bar{3}m} phase for Ecut: 700 eV; (c) Convergence test for the energy cutoff (Ecut) of the plane wave basis.

Image not extracted

Figure S2. (Color online) Enthalpy of formation of MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n} under 0 GPa, 50 GPa, 150 GPa, 250 GPa, 350 GPa, and 450 GPa. The thermodynamically stable compounds are shown by solid symbols.

For Mg-rich compositions, Mg3_3Li2_2 of Os2_2Al3_3-type[63] and Mg2_2Li of MoSi2_2-type[64] are stable up to 174 GPa and 335 GPa, respectively. Both of them are ordered in a body-centered tetragonal (bct) lattice with space group I4/mmmI4/mmm. Mg3_3Li crystallizes into the Na3_3Cl-type[65] structure with space group P4/mmmP4/mmm and maintains its stability between 4 GPa and 414 GPa. At elevated pressures, the bct-Mg4_4Li becomes stable from 50 to 476 GPa. The structure of Mg4_4Li also derives from Os2_2Al3_3[63] and crystallizes in the same space group (I4/mmmI4/mmm). Their phase diagram at 0 K is summarized in Figure 1(b).

Furthermore, we also identified several metastable phases based on the fact that their formation enthalpies are not on the convex hull but is very close to it and exhibits dynamical stability (i.e., without any imaginary frequency in the phonon spectrum, see Figure S3). The corresponding phases are: P3‾m1P\overline{3}m1 MgLi2_2 (170−500 GPa), Pm3‾mPm\overline{3}m MgLi (213−250 GPa), PnmaPnma MgLi (315−350 GPa), I4/mmmI4/mmm Mg3_3Li2_2 (174−300 GPa), I4/mmmI4/mmm Mg2_2Li (335−410 GPa), P4/mmmP4/mmm Mg3_3Li (0–4 GPa, 414–500 GPa), I4/mmmI4/mmm Mg4_4Li (0–50 GPa, 476–500 GPa), and P4/mmmP4/mmm Mg5_5Li (0–500 GPa).

Calculated phonon spectra of Mg-Li compounds

Figure S3. (Color online) Calculated phonon spectra of P1‾P\overline{1} MgLi10\mathrm{MgLi}_{10} at (a) 0 GPa and (b) 100 GPa; Calculated phonon spectra of P3‾m1P\overline{3}m1 MgLi2\mathrm{MgLi}_{2} at (c) 170 GPa and (d) 500 GPa; Calculated phonon spectra of Pm3‾mPm\overline{3}m MgLi at (e) 0 GPa, (f) 200 GPa, and (g) 250 GPa; Calculated phonon spectra of PnmaPnma MgLi at (h) 350 GPa, (i) 500 GPa; Calculated phonon spectra of I4/mmmI4/mmm Mg3Li2\mathrm{Mg}_{3}\mathrm{Li}_{2} at (j) 0 GPa, (k) 100 GPa, and (l) 300 GPa; Calculated

As shown in Figure 2 (b), each Li atom in Pm3‾mPm\overline{3}m MgLi (B2 type structure) is surrounded by eight nearest Mg atoms and six second-nearest Li atoms. The corresponding Li–Li and Li–Mg distances (2.40 Å and 2.07 Å) are shorter than those in Pm3ˉmPm\bar{3}m LiAl (2.54 Å and 2.20 Å)[66] at 200 GPa due to the stronger bonding in the former. In PnmaPnma-MgLi, the Li atoms are bonded to six Mg atoms, with the Li-Mg bond distance of 1.64-1.73 Å at 500 GPa as shown in Figure S4.

Structures of stable MgₘLiₙ compounds under high pressure

Figure 2. (Color online) Structures of stable MgₘLiₙ compounds under high pressure: (a) P1ˉP\bar{1} MgLi₁₀ at 100 GPa; (b) Pm3ˉmPm\bar{3}m MgLi at 200 GPa; (c) PnmaPnma MgLi at 500 GPa; (d) I4/mmmI4/mmm Mg₃Li₂ at 100 GPa; (e) I4/mmmI4/mmm Mg₂Li at 300 GPa; (f) P4/mmmP4/mmm Mg₃Li at 400 GPa; (g) I4/mmmI4/mmm Mg₄Li at 400 GPa. Orange and green spheres

Histogram of interatomic separations for nine Mg–Li structures

Figure S4. (Color online) Histogram of interatomic separations for (a) P-1P\text{-}1 MgLi10\mathrm{MgLi}_{10} at 100 GPa; (b) Pm-3mPm\text{-}3m MgLi\mathrm{MgLi} at 200 GPa; (c) PnmaPnma MgLi\mathrm{MgLi} at 500 GPa; (d) I4/mmmI4/mmm Mg3Li2\mathrm{Mg}_3\mathrm{Li}_2 at 100 GPa; (e) I4/mmmI4/mmm Mg2Li\mathrm{Mg}_2\mathrm{Li} at 300 GPa; (f) P4/mmmP4/mmm Mg3Li\mathrm{Mg}_3\mathrm{Li} at 400 GPa; (g) I4/mmmI4/mmm Mg4Li\mathrm{Mg}_4\mathrm{Li} at 400 GPa; (h) P-3m1P\text{-}3m1 MgLi2\mathrm{MgLi}_2 at 500 GPa; (i) P4/mmmP4/mmm Mg5Li\mathrm{Mg}_5\mathrm{Li} at 500 GPa, respectively. Note that those structures derived from BCC phase (b, d, e-g, i) have similar histogram.

On the other hand, the Mg-rich phase can be viewed as a host-guest structure, similar to Li-Al alloys[27]. The I4/mmmI4/mmm Mg3_3Li2_2 phase can be considered as the cubic lattice of Li-Mg inserted into the two-layer Mg cubic lattice as a guest. The Li atom is bonded in an eight-coordinated geometry of two inequivalent Mg sites. The first site is 4e4e (0.5, 0.5, 0.68865), and each Mg atom is bonded to eight equivalent Li atoms in a body-centered cubic geometry. The second site is 2a2a (0.0, 0.0, 0.0), and each Mg atom is bonded to four Li atoms. The corresponding Li-Mg distance at 100 GPa is 2.23 Å and 2.26 Å, respectively.

represent Mg and Li atoms, respectively.

The structural features of Mg2Li\mathrm{Mg}_{2}\mathrm{Li}-I4/mmmI4/mmm, Mg3Li\mathrm{Mg}_{3}\mathrm{Li}-P4/mmmP4/mmm, and Mg4Li\mathrm{Mg}_{4}\mathrm{Li}-I4/mmmI4/mmm are similar, where each Li atom is coordinated with four Li atoms and eight Mg atoms. In Mg2Li\mathrm{Mg}_{2}\mathrm{Li}-I4/mmmI4/mmm phase, the shortest Li–Li and Li–Mg distance at 300 GPa is 2.31 Å and 1.96 Å, respectively. In Mg3Li\mathrm{Mg}_{3}\mathrm{Li}-P4/mmmP4/mmm and Mg4Li\mathrm{Mg}_{4}\mathrm{Li}-I4/mmmI4/mmm, there are also two inequivalent Mg sites, with the Li–Li, Li–Mg1 and Li–Mg2 bond lengths of 2.22 Å, 1.89 Å, and 2.48 Å at 400 GPa, respectively. The detailed structural information of metastable MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n} compounds can be found in Figure S5 and Table S1 of Supporting Information (SI).

Structures of metastable MgLi2 and Mg5Li

Figure S5. (Color online) (a) Structure of metastable P3ˉm1P\bar{3}m1 MgLi2\mathrm{MgLi}_{2} at 500 GPa; (b) Structure of metastable P4/mmmP4/mmm Mg5Li\mathrm{Mg}_{5}\mathrm{Li} at 500 GPa; Orange and green spheres represent Mg and Li atoms, respectively.

PhaseLattice parameters (Å)AtomSiteAtomic coordinates
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.34099 0.88704 0.72062)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.05131 0.67931 0.88126)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.76970 0.00255 0.59576)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.11877 0.23071 0.46101)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.79363 0.60177 0.44728)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.39365 0.28971 0.17849)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.42905 0.37526 0.68948)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.39605 0.64581 0.32254)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.49193 0.90920 0.03875)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circLi1a1a(0.93797 0.92672 0.23680)
PP-1 MgLi10_{10} (100 GPa)a=4.2731a=4.2731<br>b=4.3904b=4.3904<br>c=4.3942c=4.3942<br>α=87.97∘\alpha=87.97^\circ<br>β=88.97∘\beta=88.97^\circ<br>γ=61.24∘\gamma=61.24^\circMg1a1a(0.85923 0.33964 0.95506)
PP-3mm1 MgLi2_2 (500 GPa)a=b=2.1069a=b=2.1069<br>c=3.2546c=3.2546<br>α=β=90∘\alpha=\beta=90^\circ<br>γ=120∘\gamma=120^\circLi2d2d(0.66667 0.33333 0.91395)
PP-3mm1 MgLi2_2 (500 GPa)a=b=2.1069a=b=2.1069<br>c=3.2546c=3.2546<br>α=β=90∘\alpha=\beta=90^\circ<br>γ=120∘\gamma=120^\circMg1b1b(0.00000 0.00000 0.50000)
PmPm-3mm MgLi (200 GPa)a=b=c=2.395a=b=c=2.395<br>α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi1b1b(0.50000 0.50000 0.50000)
PmPm-3mm MgLi (200 GPa)a=b=c=2.395a=b=c=2.395<br>α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg1a1a(0.00000 0.00000 0.00000)
PhaseLattice parametersAtomWyckoff positionCoordinates
Pnma MgLi (500 GPa)a=3.3692a=3.3692; b=3.3006b=3.3006; c=3.3146c=3.3146; α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi4c4c(0.30321 0.25000 0.51493)
Pnma MgLi (500 GPa)a=3.3692a=3.3692; b=3.3006b=3.3006; c=3.3146c=3.3146; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg4c4c(-0.21068 0.25000 0.48872)
I4/mmmI4/mmm Mg3_3Li2_2 (100 GPa)a=b=2.6098a=b=2.6098; c=13.5705c=13.5705; α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi4e4e(0.50000 0.50000 0.09576)
I4/mmmI4/mmm Mg3_3Li2_2 (100 GPa)a=b=2.6098a=b=2.6098; c=13.5705c=13.5705; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg2a2a(0.00000 0.00000 0.00000)
I4/mmmI4/mmm Mg3_3Li2_2 (100 GPa)a=b=2.6098a=b=2.6098; c=13.5705c=13.5705; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg4e4e(0.50000 0.50000 0.68865)
I4/mmmI4/mmm Mg2_2Li (300 GPa)a=b=2.3127a=b=2.3127; c=7.3017c=7.3017; α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi2a2a(0.00000 0.00000 0.00000)
I4/mmmI4/mmm Mg2_2Li (300 GPa)a=b=2.3127a=b=2.3127; c=7.3017c=7.3017; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg4e4e(0.00000 0.00000 0.35126)
P4/mmmP4/mmm Mg3_3Li (400 GPa)a=b=2.2241a=b=2.2241; c=4.9553c=4.9553; α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi1b1b(0.00000 0.00000 0.50000)
P4/mmmP4/mmm Mg3_3Li (400 GPa)a=b=2.2241a=b=2.2241; c=4.9553c=4.9553; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg1a1a(0.00000 0.00000 0.00000)
P4/mmmP4/mmm Mg3_3Li (400 GPa)a=b=2.2241a=b=2.2241; c=4.9553c=4.9553; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg2h2h(0.50000 0.50000 0.71065)
I4/mmmI4/mmm Mg4_4Li (400 GPa)a=b=2.2196a=b=2.2196; c=12.7959c=12.7959; α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi2a2a(0.50000 0.50000 0.50000)
I4/mmmI4/mmm Mg4_4Li (400 GPa)a=b=2.2196a=b=2.2196; c=12.7959c=12.7959; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg4e4e(0.50000 0.50000 0.91775)
I4/mmmI4/mmm Mg4_4Li (400 GPa)a=b=2.2196a=b=2.2196; c=12.7959c=12.7959; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg4e4e(0.50000 0.50000 0.30575)
P4/mmmP4/mmm Mg5_5Li (500 GPa)a=b=2.1437a=b=2.1437; c=7.6985c=7.6985; α=β=γ=90∘\alpha=\beta=\gamma=90^\circLi1d1d(0.50000 0.50000 0.50000)
P4/mmmP4/mmm Mg5_5Li (500 GPa)a=b=2.1437a=b=2.1437; c=7.6985c=7.6985; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg1a1a(0.00000 0.00000 0.00000)
P4/mmmP4/mmm Mg5_5Li (500 GPa)a=b=2.1437a=b=2.1437; c=7.6985c=7.6985; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg2h2h(0.50000 0.50000 0.81727)
P4/mmmP4/mmm Mg5_5Li (500 GPa)a=b=2.1437a=b=2.1437; c=7.6985c=7.6985; α=β=γ=90∘\alpha=\beta=\gamma=90^\circMg2g2g(0.00000 0.00000 0.63353)

Table S1. Lattice parameters, atomic coordinates and Wyckoff site occupation of PP-1 MgLi10_{10}, PP-3mm1 MgLi2_2, PmPm-3mm MgLi, PnmaPnma MgLi, I4/mmmI4/mmm Mg3_3Li2_2, I4/mmmI4/mmm Mg2_2Li, P4/mmmP4/mmm Mg3_3Li, I4/mmmI4/mmm Mg4_4Li, and P4/mmmP4/mmm Mg5_5Li at the given pressure.

at 400 GPa, (isosurface=0.06 e/Bohr3^{3}). Orange, green and pink spheres represent Mg atoms, Li atoms and ISQ, respectively. Note that the structures in Figure 2 (b) and (d-g) are all derived from BCC lattice, their ELF as shown in Figure 3(bFigure 1 and (d1-g1) are similar to the BCC phase of lithium for the subunit that is of B2 type, and are similar to the uniaxially tensile BCC phase of lithium along z direction for those subunits with mixed occupation of atoms other than the B2 type.

Electron localization function and charge density of Mg_mLi_n compounds

Figure 3. (Color online) Electron localization function (ELF) of MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n}: (a) PP-1 MgLi10 at 100 GPa, (isosurface=0.75); (b1) PmPm-3mm MgLi at 200 GPa, (isosurface=0.60); (c) PnmaPnma MgLi at 500 GPa, (isosurface=0.80); (d1) I4/mmmI4/mmm Mg3Li2 at 100 GPa, (isosurface=0.60); (e1) I4/mmmI4/mmm Mg2Li at 300 GPa, (isosurface=0.65); (f1) P4/mmmP4/mmm Mg3Li at 400 GPa, (isosurface=0.65); (g1) I4/mmmI4/mmm Mg4Li at 400 GPa, (isosurface=0.65). Charge density of MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n}: (b2) PmPm-3mm MgLi at 200 GPa, (isosurface=0.0425 e/Bohr3^{3}); (d2) I4/mmmI4/mmm Mg3Li2 at 100 GPa, (isosurface=0.0325 e/Bohr3^{3}); (e2) I4/mmmI4/mmm Mg2Li at 300 GPa, (isosurface=0.052 e/Bohr3^{3}); (f2) P4/mmmP4/mmm Mg3Li at 400 GPa, (isosurface=0.06 e/Bohr3^{3}); (g2) I4/mmmI4/mmm Mg4Li

The electron localization function (ELF) and charge density are calculated and shown in Figure 3. It is evident that all Mgm_{m}Lin_{n} compounds show unambiguous localized interstitial electron, indicating they are indeed electrides according to the “ELF-charge” criterion advocated in Ref. [7]. Moreover, the calculated Bader charge (listed in Table S2 of SI) reveals that the valence states of both Li and Mg atoms are positive, confirming the charge transfer from atomic nuclei to ISQs. In the Li-rich phase (MgLi10_{10}), the integrated number of electrons on Li is approximately -0.97 electrons, in other words, with a nominal oxidation state of +1. The localized electrons in interstices mainly come from Li atoms rather than from Mg, since Li has a smaller electronegativity than Mg at 100 GPa67^{67}. In Pm-3m MgLi, the anionic electrons are localized at two non-equivalent octahedral interstitial sites, forming ISQ1 and ISQ2 with a Bader charge of 0.44 e and 0.14 e, respectively. By contrast, the ISQs (with a charge state of 0.92 e) are arranged in layers in Pnma MgLi. The I4/mmm Mg3_3Li2_2 is derived from BCC lattice, with its ELF similar to the pristine or uniaxially tensile BCC phase of lithium, depending on whether the corresponding subunit is of the B2 type or not. The averaged Bader charge on the ISQs in this phase is: 0.55 e (ISQ1), 0.64 e (ISQ2), 0.68 e (ISQ3), 0.50 e (ISQ4), respectively. There are three inequivalent interstitial sites in I4/mmm Mg2_2Li, leading to ISQ1 with 0.64 e, mathrmISQ2\\mathrm{ISQ2} with 0.84 e, and mathrmISQ3\\mathrm{ISQ3} with 0.64 e, respectively. Excess electrons in both of them are localized in the vertex-sharing octahedral interstitials. The Bader charge analysis of mathrmMg3Li\\mathrm{Mg_3Li}-P4/mmmP4/mmm gives 1.11 e (mathrmISQ1\\mathrm{ISQ1}), 0.89 e (mathrmISQ2\\mathrm{ISQ2}), 0.61 e (mathrmISQ3\\mathrm{ISQ3}); and 1.15 e (mathrmISQ1\\mathrm{ISQ1}), 0.85 e (mathrmISQ2\\mathrm{ISQ2}), 0.61 e (mathrmISQ3\\mathrm{ISQ3}) for mathrmMg4Li\\mathrm{Mg_4Li}-I4/mmmI4/mmm, respectively.

PhaseMg (per atom)Li (per atom)ISQ (e/site)ISQ (e/cell)
PP-1 MgLi10_{10} (100 GPa)+0.65+0.97/7.51
PmPm-3m3m MgLi (200 GPa)+1.16+0.6ISQ1: 0.44; ISQ2: 0.141.76
PnmaPnma MgLi (500 GPa)+1.27+0.560.927.32
I4/mmmI4/mmm Mg3_3Li2_2 (100 GPa)+1.18+0.65ISQ1: 0.55; ISQ2: 0.64; ISQ3: 0.68; ISQ4: 0.509.71
I4/mmmI4/mmm Mg2_2Li (300 GPa)+1.20+0.56ISQ1: 0.64; ISQ2: 0.84; ISQ3: 0.645.92
P4/mmmP4/mmm Mg3_3Li (400 GPa)+1.20+0.53ISQ1: 1.11; ISQ2: 0.89; ISQ3: 0.614.11

Table S2. The valence state of Mg and Li atoms and the Bader charge of ISQs in Mgm_mLin_n compounds.

Electrides can exhibit interesting electronic properties, such as the semiconducting phases in compressed Li (C2C2[68] and Aba2Aba2[69]), Na (hP4hP4[70] and oP8oP8[71]), Li-Na compounds$[72], mathrmCa2N\\mathrm{Ca_2N}-II, mathrmSr2N\\mathrm{Sr_2N}-II, and mathrmBa2N\\mathrm{Ba_2N}-IV$[73]. We thus explored the band structure and projected density of states (PDOS) of the discovered Mg-Li electrides, as shown in Figure 4. The electride mathrmMgLi10\\mathrm{MgLi_{10}} is discovered as a semiconductor with an energy gap of 0.08 eV at 60 GPa, which increases to 0.22 eV at 100 GPa, as calculated by the revised Heyd-Scuseria-Ernzerhof screened hybrid functional (HSE06). As shown in Figure 4(b)-(Figure 4h), the dominant contribution for the DOS at the Fermi level (EFE_{\mathrm{F}}) comes from the interstitial electron. Moreover, there is not only a strong hybridization between ISQ and the Li-2p, Mg-2p, and Mg-3d electrons, but also remarkable van Hove singularities (vHs) dominated by the anionic electrons near the Fermi level. The increased DOS at the EFE_{\mathrm{F}} of MgLiMgLi and Mg2LiMg_{2}Li suggest the possibility of the emergence of superconductivity in such compounds[74].

Band structure and projected density of states of Mg-Li electrides

Figure 4. (Color online) (a) Band structure of PP-11 mathrmMgLi10\\mathrm{MgLi_{10}} at 100 GPa; PDOS for (b) PP-11 mathrmMgLi10\\mathrm{MgLi_{10}} at 100 GPa, (c) PmPm-3m3m MgLi at 200 GPa, (d) PnmaPnma MgLi at 500 GPa, (e) I4/mmmI4/mmm mathrmMg3Li2\\mathrm{Mg_3Li_2} at 100 GPa, (f) I4/mmmI4/mmm mathrmMg2Li\\mathrm{Mg_2Li} at 300 GPa, (g) P4/mmmP4/mmm mathrmMg3Li\\mathrm{Mg_3Li} at 400 GPa, (h) I4/mmmI4/mmm mathrmMg4Li\\mathrm{Mg_4Li} at 400 GPa, respectively.

To investigate the phonon-mediated superconductivity in Mg-Li alloys, we calculated their phonon dispersion spectra and electron-phonon coupling (EPC) constants λ\lambda, as depicted in Figure 5, Figure S6, Figure S7, and Table 1. Using a Coulomb pseudopotential of μ∗=0.1\mu^{*}=0.1, the calculated EPC constant λ\lambda of the Pm-3mPm\text{-}3m MgLi is around 1.35 at 200 GPa (Figure 5(c)), and the superconductivity critical temperature TcT_{\mathrm{c}} is estimated to be around 22.80 K. It is worth noting that the main contribution to the superconductivity comes from the low-frequency phonon modes along the R-M-Γ\Gamma directions (see Figure 5(a), especially at the M point). From the PHDOS and Eliashberg spectral function, one can determine that the low-frequency region below 13 THz contribute 78.6% to the EPC constant due to the hybridization of the vibrations of Li with Mg atom. The modes at higher frequency contribute 21.4% to the total λ\lambda, where the vibration of Li plays a critical role. As for Pnma MgLi and I4/mmm Mg2_2Li, the EPC is weaker, with λ=0.79\lambda=0.79 and λ=0.45\lambda=0.45, resulting in a Tc=8.91T_c=8.91 K and Tc=3.24T_c=3.24 K, respectively. The low-frequency modes of the phonon spectrum also dominate in their total EPC.

Phonon dispersion spectrum, projected phonon density of states, and Eliashberg spectral function

Figure 5. (Color online) (a) Phonon dispersion spectrum, (b) projected phonon density of states (PHDOS), (c) Eliashberg spectral function α2F(ω)\alpha^{2}F(\omega) and electron-phonon coupling constant λ\lambda of Pm-3mPm\text{-}3m MgLi at 200 GPa, respectively. The size of the solid red circles in (a) is proportional to the electron phonon coupling strength.

PhasePressure (GPa)λ\lambdaωlog⁡\omega_{\log}TcT_c (K)
Pm-3m MgLi200 GPa1.35174.20122.80
Pnma MgLi500 GPa0.79176.4088.91
I4/mmm Mg2_2Li300 GPa0.45254.7683.24

Table 1. The calculated total EPC parameter λ\lambda, logarithmic average phonon frequency ωlog⁡\omega_{\log}, and superconducting critical temperature TcT_c of Pm-3m MgLi at 200 GPa, Pnma MgLi at 500 GPa, and I4/mmm Mg2_2Li at 300 GPa, respectively.

Phonon dispersion relations, projected phonon density of states, and Eliashberg spectral function for Pnma MgLi at 500 GPa

Figure S6. (Color online) (a) Phonon dispersion relations, (b) projected phonon density of states (PHDOS), (c) Eliashberg spectral function α2F(ω)\alpha^{2}F(\omega) and electron-phonon coupling constant λ\lambda of Pnma MgLi at 500 GPa. The sizes of the solid red circles are proportional to the electron phonon coupling strength.

Phonon dispersion relations, projected phonon density of states, and Eliashberg spectral function for I4/mmm Mg2Li at 300 GPa

Figure S7. (Color online) (a) Phonon dispersion relations, (b) projected phonon density of states (PHDOS), (c) Eliashberg spectral function α2F(ω)\alpha^{2}F(\omega) and electron-phonon coupling constant λ\lambda of I4/mmmI4/mmm Mg2Li\mathrm{Mg}_{2}\mathrm{Li} at 300 GPa. The sizes of the solid red circles are proportional to the electron-phonon coupling strength.

at the Fermi level, (c) the DOS of ISQs at the Fermi level, (d) the DOS of Li-ss orbital at the Fermi level, (e) the DOS of Li-pp orbital at the Fermi level, (f) the DOS of Au-pp orbital at the Fermi level in Li8Au\mathrm{Li}_8\mathrm{Au}, respectively; (g) The evolution of TcT_c with localization intensity q/Qq/Q in typical Li-based electrides. The bule squares represent weakly coupled electrides (0.1<λ≤0.80.1<\lambda\leq0.8), red spheres represent moderately coupled electrides (λ>0.8\lambda>0.8).

As mentioned above, Li-rich compounds exhibit general superconductivity. They in fact can be viewed as variants derived from the elemental Li that is also a superconducting electride. They should have a universal rule that governs the electron localization and superconductivity. To explore this, we took Li8Au\mathrm{Li}_8\mathrm{Au} as an example to illustrate the relationship between ISQs and superconductivity, which has the highest TcT_c reported to date for all electrides. There are two stages as unveiled in Figure 6(a)-(Figure 6f): in the first stage (TcT_c: 55 K→65K), the ISQs enhance the density of states (DOS) at the Fermi level, and the TcT_c values are positively correlated with the ISQs; In the second stage (TcT_c: 65 K→40K), the over-localized ISQs start to weaken the electron-phonon coupling, which lowers down the superconducting temperature. The two competing mechanisms result in a maximum critical point which bridges the free-electron gas model of metals to highly localized electrides.

Evolution of superconducting critical temperature

Figure 6. (Color online) The evolution of TcT_c with (a) the charge of ISQs, (b) the total density of states (DOS)

In order to quantify this, let us define the ratio of Bader charge of ISQs (qq) and the valence electron (QQ) as the localization intensity, and divide the typical electride superconductors into two groups based on EPC constants λ\lambda: weakly coupled materials (0.1<λ≤0.80.1<\lambda\leq0.8) and moderately coupled materials (λ>0.8\lambda>0.8). The superconducting critical temperatures of the weakly coupled electrids are below 15K, for which the localization intensity of ISQs hardly affects the superconductivity [see the blue area in Figure 6(g)]. On the other hand, for the moderately coupled electrides, the superconducting temperatures increase and then decrease notably and finally become flattened with increasing q/Qq/Q. As shown in Figure 6(g) and Table S3, the localization intensity of ISQs can be used to judge whether the superconducting temperature reaches the optimal value or not. For moderately coupled electrides, the superconductivity can be tuned by adjusting the localization intensity of ISQs. For the Mg-Li system we studied, the PmPm-3mm MgLi is moderately coupled electride, in which ISQs are over-localized (q/Q=0.59q/Q = 0.59, see the Table S3 in the Supporting Information) thus has a relatively low TcT_{\mathrm{c}}. In this regard, weakening the localization may increase the superconducting critical temperatures. On the other hand, PnmaPnma MgLi and I4/mmmI4/mmm Mg2Li\mathrm{Mg}_{2}\mathrm{Li} are weakly coupled electrides, the localization intensity of ISQs does not affect their TcT_{\mathrm{c}}.

Phaseqq (e/cell)QQ (e/cell)q/Qq/Qλ\lambdaTcT_c (K)
Li8_8Au-Fm-3m-250 GPa3.0496760.04268.5[3]
Li5_5N-P6/mmm-150 GPa1.2475100.124751.3948.97[4]
MgLi-Pm-3m-200 GPa1.7630.586671.3522.8
Li10_{10}Se-C2/m-50 GPa9.8228320.306961.2816[5]
Li8_8Au-Fm-3m-200 GPa3.2044760.042161.2651.5[3]
Li5_5C-P6/mmm-210 GPa0.583890.064871.2648.3[6]
LiCa-Fd-3m-80 GPa3.8592240.16081.2518.9[7]
Li6_6As-C2/c-270 GPa0.92440.020911.245.2[8]
Li10_{10}Sb-C2/m-100 GPa3.52300.117330.89.2[8]
MgLi-Pnma-500 GPa7.32120.610.798.91
Li11_{11}Sb2_2-C2/m-300 GPa0.8772420.020890.599.4[9]
Li8_8As-P6/mmm-50 GPa3.27130.251540.566.3[8]
Li-Im-3m-30 GPa1.440620.72030.5114[10]
Li9_9Te-C2/m-50 GPa7.6912300.256370.54.01[11]
Li-Im-3m-20.3 GPa1.487820.74390.495.47[10]
Li10_{10}Te-C2/m-100 GPa3.51320.109690.484[8]
Li-Im-3m-10 GPa1.539220.76960.472.3[10]
Mg2_2Li-I4/mmm-300 GPa5.92100.5920.453.24

Table S3. Table S3. The charge state of ISQs and valence electron, q/Qq/Q, EPC parameter λ\lambda, and superconducting critical temperatures TcT_c of typical electrides that contain Li.

the SOC effect. Inset of (c) shows a zoom in of the DP with a small gap opened by SOC. The Fermi level is set to zero. The pressure is 200 GPa.

In the band structure of bulk Pm−3mPm\mathord{-}3m MgLi, we found that the conduction and valence bands cross each other at the RR point, leading to the formation of a band touching point (i.e., a Dirac point). Interestingly, the gapless Dirac point (DP) appears near the Fermi level if without spin-orbit coupling (SOC); whereas the DP becomes gapped when the SOC is turned on. Here, we have investigated its topological properties by studying the surface states (Figure 7(b)) and the topologically invariant Z2Z_2 indices (Table 2). The latter can be determined from the parity of the wave function at the time reversal invariant momentum (TRIM) points in the presence of time-reversal symmetry and inversion symmetry. The product of the parity for the occupied states at the TRIM points are calculated by δ(ki)=∏N=14ξ2N(ki)\delta(\mathbf{k}_i) = \prod_{N=1}^{4} \xi_{2N}(\mathbf{k}_i), where ξ2N(ki)\xi_{2N}(\mathbf{k}_i) represents the parity of the occupied valence bands at ki\mathbf{k}_i. The strong topological insulators are characterized by a Z2Z_2 invariant (ν0;ν1ν2ν3)=(1;111)(\nu_0; \nu_1\nu_2\nu_3) = (1; 111). The gapless surface states can be identified near the Fermi level according to the bulk-boundary correspondence. The Pm−3mPm\mathord{-}3m phase of MgLi electride thus could be described as a strong topological material, as evidenced by the calculated Z2=(1;111)Z_2 = (1; 111) indices and the nontrivial surface states (Figure 7(d)). For the PnmaPnma phase, the band-crossing point is located at the SS point (see Figure S8) and the gap is also opened up by SOC. The calculated Z2Z_2 invariant (see Table S4 of SI) and metallic surface-states also confirm the nontrivial topological nature of MgLi-PnmaPnma. These results indicate that the MgLi electrides (Pm−3mPm\mathord{-}3m and PnmaPnma) all exhibit topologically nontrivial characteristics. Since superconductivity, band topology and interstitial localization are all derived from the quantum behavior of electrons. Pm-3m MgLi and Pnma MgLi demonstrate the multi-quantum state realized in this system.

Bulk band structures and surface states of Pm-3m MgLi

Figure 7. (Color online) (a) Bulk band structure and (b) (001) surface states of PmPm-3mm MgLi calculated without the SOC effec. (c) Bulk band structure and (d) (001) surface states of PmPm-3mm MgLi calculated with

TRIMParity productsν\nuZ2Z_2
X-ν1=0\nu_1=0, ν1′=1\nu_1'=1
3R+ν2=0\nu_2=0, ν2′=1\nu_2'=1
3M-ν3=0\nu_3=0, ν3′=1\nu_3'=1(1;111)
Γ\Gamma-ν0=1\nu_0=1

Table 2. Parity products at TRIMs and the Z2Z_2 indices of Pm-3m MgLi at 200 GPa.

Band structures and surface states of MgLi-Pnma without and with SOC effect

Figure S8. (Color online) (a) Band structure calculated without the SOC effect for MgLi-PnmaPnma; (b) Surface states on the (001) surface of MgLi-PnmaPnma without the SOC effect; (c) Band structure calculated with the SOC effect for MgLi-PnmaPnma, the inset shows a zoomed-in view of the plot of the bands at DP with a gap; (d) Surface states on the (001) surface of MgLi-PnmaPnma with the SOC effect; the Fermi level is set to zero. The pressure is 500 GPa.

TRIMParity Productsν\nuZ2Z_2
Y-ν1=0\nu_1=0, ν1′=1\nu_1'=1
3S+ν2=0\nu_2=0, ν2′=1\nu_2'=1
3X-ν3=0\nu_3=0, ν3′=1\nu_3'=1(1;111)(1;111)
U-ν0=1\nu_0=1

Table S4. Parity products at TRIMs and the Z2Z_2 Index of MgLi-PnmaPnma at 500 GPa.

Finally, it is worth mentioning that the long-range interaction between atoms and ISQs could lead to an anomalous LA-TA splitting in electride, analogous to the LO-TO splitting in ionic compounds[11]. For the isotropic Pm-3m MgLi, the magnitude of such splitting is (Δω=ωLA−ωTA\Delta\omega= \omega_{\mathrm{LA}}-\omega_{\mathrm{TA}}) 34.5 cm−1^{-1} at 200 GPa (see Figure S10 in the Supporting Information), which might have an impact on the electrical or optical properties of the surface state of this class of materials[12].

Phonon dispersions of MgLi

Figure S10. (Color online) Phonon dispersions of Pm3ˉmPm\bar{3}m MgLi at 200 GPa, which are calculated using Bader charges (dot-dash-ed lines), and compared to that without taking nucleus-ISQ polarization into account (dotted lines in the right panel).

Conclusion

In summary, we have extensively explored the ground-state structures and phase diagram of Mg-Li system up to 500 GPa by combining crystal structure prediction and first-principles calculations. Seven compounds have been found to be stable, including MgLi10_{10}, Pm-3m MgLi, Pnma MgLi, Mg3_3Li2_2, Mg2_2Li, Mg3_3Li, and Mg4_4Li, respectively. The Mg-Li system seems to favor Mg-rich compounds compared to the Li-rich side in a wide range of pressures. Remarkably, all high-pressure compounds are electrides with various morphologies of ISQs. Calculations of the electronic properties reveal that MgLi10\mathrm{MgLi}_{10} is a semiconductor with almost all of the 2s electrons of lithium having been transferred into the interstitial sites. Interestingly, our calculations show that Pm-3m\mathit{Pm}\text{-}3m MgLi\mathrm{MgLi} is a superconductor with a critical temperature TcT_{\mathrm{c}} of 22.8 K at 200 GPa; and Pnma\mathit{Pnma} MgLi\mathrm{MgLi} and Mg2Li\mathrm{Mg}_{2}\mathrm{Li} also manifest superconductivity. We found in the moderately coupled electrides (such as Pm-3m\mathit{Pm}\text{-}3m MgLi\mathrm{MgLi}), the superconducting temperatures increase firstly and then decrease notably and finally become flattened with increasing localization intensity of ISQs. By contrast, in the weakly coupled electrides (such as Pnma\mathit{Pnma} MgLi\mathrm{MgLi} and Mg2Li\mathrm{Mg}_{2}\mathrm{Li}), the ISQs hardly affect the superconductivity.

In particular, the valance and conduction bands of MgLi\mathrm{MgLi} compounds (Pm-3m\mathit{Pm}\text{-}3m and Pnma\mathit{Pnma}) cross each other at the DPs if without SOC; but the gap at the DPs opens up near the Fermi level when with SOC. Both the gapless surface states and the calculated Z2Z_{2} invariant clearly confirm that these two phases of MgLi\mathrm{MgLi} are band topologically nontrivial. The Pm-3m\mathit{Pm}\text{-}3m MgLi\mathrm{MgLi} electride exhibits both superconductivity and topological behaviors, which also has the highest superconducting TcT_{\mathrm{c}} among all reported band topological electrides to date. The simultaneous emergence of the superconductivity and the band topology in the same electride may lead to potential applications in electride-based quantum devices.

Author contributions

Dan Wang: Investigation, Methodology, Writing - original draft, Writing - review & editing. Hong X. Song: Methodology, Writing - review & editing. Qi D. Hao: Methodology, Writing - review & editing. Guang F. Yang: Methodology, Writing - review & editing. Hao Wang: Methodology. Lei L. Zhang: Writing - review & editing. Xiang R. Chen: Supervision, Project administration. Hua Y. Geng: Idea conceiving, Project design, Writing, Reviewing, and Editing, Supervision, Project administration, Software.

Conflicts of interest

The authors declare no competing interests.

Data availability

All data are included in the manuscript and supporting information.

Code availability

The VASP code used in this paper is commercial software provided by VASP Software GmbH. All others are open-source codes that can be obtained via internet.

Supporting Information

The Enthalpy difference for different EcutE_{\mathrm{cut}}; The enthalpy of formation, calculated phonon spectra, and histogram of interatomic separations of MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n}; The structure of P-3m1P\text{-}3m1 MgLi2_2 and P4/mmmP4/mmm Mg5_5Li; The structure information of MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n}; The valence state of Mg and Li atoms and the Bader charge of ISQs in MgmLin\mathrm{Mg}_{m}\mathrm{Li}_{n} compounds. The phonon dispersion relations, PHDOS, α2F(ω)\alpha^{2}F(\omega) and EPC parameter λ\lambda of PnmaPnma MgLi and I4/mmmI4/mmm Mg2_2Li; The charge state of ISQs(qq), valence electrons(QQ), q/Qq/Q, EPC parameter λ\lambda, and superconducting critical temperature TcT_{\mathrm{c}} of typical electrides that contain Li; The topological behaviors of MgLi-PnmaPnma; Band structures of I4/mmmI4/mmm Mg3_3Li2_2, I4/mmmI4/mmm Mg2_2Li, P4/mmmP4/mmm Mg3_3Li, and I4/mmmI4/mmm Mg4_4Li; LA-TA splitting of Pm-3mPm\text{-}3m MgLi.

Acknowledgments

This work was supported by the National Key R&D Program of China under Grant No. 2021YFB3802300, the Foundation of National key Laboratory of shock wave and detonation physics (No. 2023JCJQLB05401), the National Natural Science Foundation of China under Grant No. 12074274, and the NSAF under Grant No. U1730248. Part of the computation was performed using the supercomputer at the Center for Computational Materials Science (CCMS) of the Institute for Materials Research (IMR) at Tohoku University, Japan.

TOC Graphic

Figure 8 (PDF p. 26). TOC Graphic

Supporting Information for Superconducting and Topological Properties in Mg−Li Electrides at High Pressures

Dan Wang; Hong X. Song; Qi D. Hao; Guang F. Yang; Hao Wang; Lei L. Zhang; Ying Chen; Xiang R. Chen; Hua Y. Geng

As shown in Figure S1, the Pm3ˉm\mathit{Pm\bar{3}m} MgLi transforms to the P21/c\mathit{P2_{1}/c} phase at 225 GPa and then transforms to the Fmm2\mathit{Fmm2} phase at 405 GPa if with a cut-off energy of 400 eV, which are consistent with the results of the previous theoretical study[1]. However, when the cut-off energy for the plane-wave basis is increased to 700 eV, the picture is changed: Pm3ˉm\mathit{Pm\bar{3}m} MgLi decomposes into Mg2_2Li and Li at 213 GPa and then becomes stable again into the Pnma\mathit{Pnma} phase at 315 GPa. Based on the convergence test for the energy cutoff of plane wave basis, the results of Ecut: 700 eV are more reliable.

Formation enthalpies of Mg-Li compounds at different pressures

Figure 9 (PDF p. 28).

phonon spectra of I4/mmmI4/mmm Mg2Li\mathrm{Mg}_2\mathrm{Li} at (m) 0 GPa, (n) 300 GPa, and (o) 410 GPa; Calculated phonon spectra of P4/mmmP4/mmm Mg3Li\mathrm{Mg}_3\mathrm{Li} at (p) 0 GPa, (q) 400 GPa, and (r) 500 GPa; Calculated phonon spectra of I4/mmmI4/mmm Mg4Li\mathrm{Mg}_4\mathrm{Li} at (s) 0 GPa, (t) 400 GPa, and (u) 500 GPa; Calculated phonon spectra of P4/mmmP4/mmm Mg5Li\mathrm{Mg}_5\mathrm{Li} at (v) 0 GPa and (w) 500 GPa, respectively.

The characteristics of newly predicted structures can be described by their interatomic separation histograms, as shown in Figure S4. In MgLi10\mathrm{MgLi}_{10}, the coordination numbers of Li atoms vary in the ranges 6–10 with distances from 1.5 Å to 2.5 Å. Li atoms in MgLi\mathrm{MgLi}-Pm3ˉmPm\bar{3}m have coordination number 14, which is similar to bcc-Li. Li atoms in MgLi\mathrm{MgLi}-PnmaPnma are connected with six Mg atoms with a separation distance less than 1.75 Å as the nearest neighboring shell; and just above 2 Å, Li atoms coordinate with twelve other Li atoms and four Mg atoms. In Mg3Li2\mathrm{Mg}_{3}\mathrm{Li}_{2}, Li atoms have coordination number 13, they are connected with eight Mg atoms in the nearest shell and five Li atoms in the second shell. The averaged local atomic environment of Li atoms in Mg2Li\mathrm{Mg}_{2}\mathrm{Li}-I4/mmmI4/mmm is similar to that of Mg3Li\mathrm{Mg}_{3}\mathrm{Li}-P4/mmmP4/mmm and Mg4Li\mathrm{Mg}_{4}\mathrm{Li}-I4/mmmI4/mmm, it contains a body-centered motif with coordination numbers of Li atoms equal to 8 (cubic coordination).

The new structure MgLi2\mathrm{MgLi}_{2} and Mg5Li\mathrm{Mg}_{5}\mathrm{Li} are dynamically stable at 170–500 GPa and 0–500 GPa, respectively. As shown in Figure S5, MgLi2\mathrm{MgLi}_{2} is found to be CdI2_{2}-type{sup}2 structure (hP3 in the Pearson’s symbol) and crystallizes in the trigonal $P\bar{3}m1$ space group. As shown in {xref}Fig. S4(h)`, Li atoms are bonded to three Li atoms and three Mg atoms, the bond lengths for Li–Li and Li–Mg are 1.34 Å and 1.82 Å, respectively. The Li atoms are located at body center and bonded with eight equivalent Mg atoms in P4/mmmP4/mmm Mg5_5Li, and the Li–Mg cubic geometry inserted into the two-layer Mg cubic lattice as a guest. The coordination environment is similar to Mg2_2Li-I4/mmmI4/mmm, and the Li–Li, Li–Mg1, and Li–Mg2 bonds are 2.14 Å, 1.83 Å, and 2.44 Å, respectively.

I4/mmmI4/mmm Mg4Li\mathrm{Mg}_{4}\mathrm{Li} (400 GPa)+1.20+0.53ISQ1: 1.15; ISQ2: 0.85<br>ISQ3: 0.6110.64

Table 5.

Figures S6 and S7 show the role of phonon dispersion on superconductivity. The calculated Eliashberg spectral function reflects a 77.7% contribution of phonons below 16 THz to the EPC constant, implying that the superconductivity of the electride MgLi-Pnma is dominated by low-frequency phonons. In MgLi-Pnma, the calculated EPC parameter (λ\lambda) is 0.79, and the logarithmic average frequency ωlog⁡\omega_{\log} is 176.408, which corresponds to a TcT_c of 8.91 K. Similarly, the low-frequency phonons below 15 THz play a dominant role in the superconductivity for Mg2_2Li, due to their large contribution to λ\lambda (78.8%). We conclude that low-frequency phonons jointly contributed by Li and Mg atoms, and the PHDOS of the Li atom is more concentrated at lower frequencies than Mg atoms.

We report a discovery of nontrivial Z2Z_2 topology in the electronic structures of superconducting electride MgLi-PnmaPnma, which is similar to MgLi-Pm3ˉmPm\bar{3}m where the band-crossing point (Dirac point) is located at S points near the Fermi level and gapped with SOC effect. In Table S4 and Figure S8, we find a topological Z2Z_2 invariant of (1; 111) for the bulk band and topologically protected surface states in the (001) surfaces, signifying its nontrivial electronic topology.

Band structures for Mg-Li compounds

Figure S9. (Color online) Band structures for (a) I4/mmmI4/mmm Mg3_3Li2_2 at 100 GPa; (b) I4/mmmI4/mmm Mg2_2Li at 300 GPa; (c) P4/mmmP4/mmm Mg3_3Li at 400 GPa; (d) I4/mmmI4/mmm Mg4_4Li at 400 GPa, respectively.

Supplementary references

  1. Xu, Y.; Wang, S.; Chen, C.; Li, H.; Zhou, D., Superconductivity at 23 K in MgLi compound at ultrahigh pressure. Comp. Mater. Sci. 2019, 164, 158-165.

  1. Bozorth, R. M., The crystal structure of cadmium iodide. J. Am. Chem. Soc. 1922, 44 (10), 2232-2236.

  1. Zhang, X. G.; Yao, Y. S.; Ding, S. C.; Bergara, A.; Li, F.; Liu, Y.; Zhou, X. F.; Yang, G. C., Superconductivity in Li8_8Au electride. Phys. Rev. B 2023, 107 (10), L100501.

  1. Wan, Z.; Zhang, C.; Yang, T.; Xu, W.; Zhang, R., Predicted superconductivity and superionic state in the electride Li5_5N under high pressure. New J. Phys. 2022, 24 (11), 113012.

References

  1. [1]Ashcroft, N. W., Metallic hydrogen: a high-temperature superconductor? Phys. Rev. Lett. 1968, 21 (26), 1748.DOI
  2. [2]Drozdov, A. P.; Eremets, M. I.; Troyan, I. A.; Ksenofontov, V.; Shylin, S. I., Conventional superconductivity at 203 kelvin at high pressures in the sulfur hydride system. Nature 2015, 525 (7567), 73-76.DOI
  3. [3]Somayazulu, M.; Ahart, M.; Mishra, A. K.; Geballe, Z. M.; Baldini, M.; Meng, Y.; Struzhkin, V. V.; Hemley, R. J., Evidence for superconductivity above 260 K in lanthanum superhydride at megabar pressures. Phys. Rev. Lett. 2019, 122 (2), 027001.DOI
  4. [4]Sun, Y.; Lv, J.; Xie, Y.; Liu, H.; Ma, Y., Route to a superconducting phase above room temperature in electron-doped hydride compounds under high pressure. Phys. Rev. Lett. 2019, 123 (9), 097001.DOI
  5. [5]Bergara, A.; Neaton, J. B.; Ashcroft, N. W., Pairing, π-bonding, and the role of nonlocality in a dense lithium monolayer. Phys. Rev. B 2000, 62 (12), 8494.DOI
  6. [6]Rodriguez-Prieto, A.; Bergara, A., Pressure induced complexity in a lithium monolayer: Ab initio calculations. Phys. Rev. B 2005, 72 (12), 125406.
  7. [7]Zhang, L. L.; Wu, Q.; Li, S. R.; Sun, Y.; Yan, X. Z.; Chen, Y.; Geng, H. Y., Interplay of anionic quasi-atoms and interstitial point defects in electrides: abnormal interstice occupation and colossal charge state of point defects in dense fcc-Lithium. ACS Appl. Mater. Interfaces 2021, 13 (5), 6130-6139.DOI
  8. [8]Struzhkin, V. V.; Eremets, M. I.; Gan, W.; Mao, H. K.; Hemley, R. J., Superconductivity in dense lithium. Science 2002, 298 (5596), 1213-1215.
  9. [9]Shimizu, K.; Ishikawa, H.; Takao, D.; Yagi, T.; Amaya, K., Superconductivity in compressed lithium at 20 K. Nature 2002, 419 (6907), 597-599.
  10. [10]Miao, M. S.; Hoffmann, R., High pressure electrodes: a predictive chemical and physical theory. Acc. Chem. Res. 2014, 47 (4), 1311-1317.DOI
  11. [11]Zhang, L. L.; Geng, H. Y.; Wu, Q., Prediction of anomalous LA-TA splitting in electrides. Matter Radiat. Extremes 2021, 6 (3), 038403.DOI
  12. [12]Wang, D.; Song, H.; Zhang, L.; Wang, H.; Sun, Y.; Wu, F.; Chen, Y.; Chen, X.; Geng, H., Universal metallic surface states in electrides. J. Phys. Chem. C 2024, 128 (4), 1845-1854.DOI
  13. [13]Marques, M.; McMahon, M. I.; Gregoryanz, E.; Hanfland, M.; Guillaume, C. L.; Pickard, C. J.; Ackland, G. J.; Nelmes, R. J., Crystal structures of dense lithium: a metal-semiconductor-metal transition. Phys. Rev. Lett. 2011, 106 (9), 095502.pubmed.ncbi.nlm.nih.gov/21405633
  14. [14]Lv, B.; Zhu, X. Y.; Lorenz, B.; Wei, F. Y.; Xue, Y. Y.; Yin, Z. P.; Kotliar, G.; Chu, C. W., Superconductivity in the Mn5Si3-type Zr5Sb3 system. Phys. Rev. B 2013, 88 (13), 134520.DOI
  15. [15]Ge, Y.; Guan, S.; Liu, Y., Two dimensional superconductors in electrides. New J. Phys. 2017, 19 (12), 123020.DOI
  16. [16]Miyakawa, M.; Kim, S. W.; Hirano, M.; Kohama, Y.; Kawaji, H.; Atake, T.; Ikegami, H.; Kono, K.; Hosono, H., Superconductivity in an inorganic electride 12CaO.7Al2O3:e-. J. Am. Chem. Soc. 2007, 129 (23), 7270-7271.
  17. [17]Hosono, H.; Kitano, M., Advances in materials and applications of inorganic electrides. Chem. Rev. 2021, 121 (5), 3121-3185.
  18. [18]Pereira, Z. S.; Faccin, G. M.; Silva, E. Z. D., Predicted superconductivity in the electride Li5C. J. Phys. Chem. C 2021, 125 (16), 8899-8906.
  19. [19]Wan, Z.; Zhang, C.; Yang, T.; Xu, W.; Zhang, R., Predicted superconductivity and superionic state in the electride Li3N under high pressure. New J. Phys. 2022, 24 (11), 113012.
  20. [20]Zhao, Z.; Zhang, S.; Yu, T.; Xu, H.; Bergara, A.; Yang, G., Predicted pressure-induced superconducting transition in electride Li6P. Phys. Rev. Lett. 2019, 122 (9), 097002.DOI
  21. [21]Zhang, X.; Li, F.; Bergara, A.; Yang, G., Pressure-induced superconductivity in Li-Te electrides. Phys. Rev. B 2021, 104 (13), 134505.DOI
  22. [22]Zhao, Y.; Bergara, A.; Zhang, X.; Li, F.; Liu, Y.; Yang, G., Interstitial anionic electrons favoring superconductivity in Li-As electrides. Phys. Rev. B 2023, 108 (10), 104505.DOI
  23. [23]Liu, Z.; Zhuang, Q.; Tian, F.; Duan, D.; Song, H.; Zhang, Z.; Li, F.; Li, H.; Li, D.; Cui, T., Proposed superconducting electrode Li₆C by sp-hybridized cage states at moderate pressures. Phys. Rev. Lett. 2021, 127 (15), 157002.
  24. [24]Zhang, X.; Zhao, Y.; Bergara, A.; Yang, G., Superconducting Li₁₀Se electrode under pressure. J. Chem. Phys. 2022, 156 (19), 194112.DOI
  25. [25]Errea, I.; Martinez-Canales, M.; Bergara, A., Ab initio study of superconducting hexagonal Be₂Li under pressure. Phys. Rev. B 2008, 78 (17), 172501.
  26. [26]Xu, Y.; Chen, C.; Wang, S.; Sun, X., Novel structures and superconductivities of calcium–lithium alloys at high pressures: A first-principles study. J. Alloy. Compd. 2016, 669, 101-107.DOI
  27. [27]Wang, X.; Wang, Y.; Wang, J.; Pan, S.; Lu, Q.; Wang, H. T.; Xing, D.; Sun, J., Pressure stabilized lithium-aluminum compounds with both superconducting and superionic behaviors. Phys. Rev. Lett. 2022, 129 (24), 246403.DOI
  28. [28]Huang, H. M.; Zhu, Q.; Blatov, V. A.; Oganov, A. R.; Wei, X.; Jiang, P.; Li, Y. L., Novel topological motifs and superconductivity in Li-Cs system. Nano Lett. 2023, 23 (11), 5012-5018.DOI
  29. [31]Xu, Y.; Elcoro, L.; Li, G.; Song, Z. D.; Regnault, N.; Yang, Q.; Sun, Y.; Parkin, S.; Felser, C.; Bernevig, B. A., Three-dimensional real space invariants, obstructed atomic insulators and a new principle for active catalytic sites. arXiv 2021, 2111, 02433.arxiv.org/abs/2111.02433
  30. [32]Liu, C.; Nikolaev, S. A.; Ren, W.; Burton, L. A., Electrides: a review. J. Mater. Chem. C 2020, 8 (31), 10551-10567.DOI
  31. [33]Meng, W.; Zhang, X.; Jiang, J.; Li, Z.; Liu, Y.; Dai, X.; Liu, G., Multi-Dimensional Topological Fermions in Electrides. Adv. Phys. Res. 2023, 2 (7), 2200119.DOI
  32. [34]Zhang, X.; Guo, R.; Jin, L.; Dai, X.; Liu, G., Intermetallic Ca₃Pb: a topological zero-dimensional electride material. J. Mater. Chem. C 2018, 6 (3), 575-581.DOI
  33. [35]Park, C.; Kim, S. W.; Yoon, M., First-principles prediction of new elect rides with nontrivial band topology based on one-Dimensional building blocks. Phys. Rev. Lett. 2018, 120 (2), 026401.DOI
  34. [36]Zhang, X.; Xiao, Z.; Lei, H.; Toda, Y.; Matsuishi, S.; Kamiya, T.; Ueda, S.; Hosono, H., Two- dimensional transition-metal electride Y₂C. Chem. Mater. 2014, 26 (22), 6638-6643.
  35. [37]Huang, H.; Jin, K. H.; Zhang, S.; Liu, F., Topological electride Y₂C. Nano Lett. 2018, 18 (3), 1972-1977.DOI
  36. [38]Mack, S. A.; Griffin, S. M.; Neaton, J. B., Emergence of topological electronic phases in elemental lithium under pressure. Proc. Natl. Acad. Sci. U. S. A. 2019, 116 (19), 9197-9201.DOI
  37. [39]Elatresh, S. F.; Zhou, Z.; Ashcroft, N. W.; Bonev, S. A.; Feng, J.; Hoffmann, R., High-pressure lithium as an elemental topological semimetal. Phys. Rev. Mater. 2019, 3 (4), 044203.DOI
  38. [40]Li, R.; Ma, H.; Cheng, X.; Wang, S.; Li, D.; Zhang, Z.; Li, Y.; Chen, X. Q., Dirac node lines in pure alkali earth metals. Phys. Rev. Lett. 2016, 117 (9), 096401.DOI
  39. [41]Deemyad, S.; Schilling, J. S., Superconducting phase diagram of Li metal in nearly hydrostatic pressures up to 67 GPa. Phys. Rev. Lett. 2003, 91 (16), 167001.DOI
  40. [42]Yu, Z.; Geng, H. Y.; Sun, Y.; Chen, Y., Optical properties of dense lithium in electride phases by first-principles calculations. Sci. Rep. 2018, 8 (1), 3868.DOI
  41. [43]Geng, H. Y.; Wu, Q., Predicted reentrant melting of dense hydrogen at ultra-high pressures. Sci. Rep. 2016, 6, 36745.DOI
  42. [44]Oganov, A. R.; Glass, C. W., Crystal structure prediction using ab initio evolutionary techniques: principles and applications. J. Chem. Phys. 2006, 124 (24), 244704.pubmed.ncbi.nlm.nih.gov/16821993
  43. [45]Oganov, A. R.; Lyakhov, A. O.; Valle, M., How evolutionary crystal structure prediction works--and why. Acc. Chem. Res. 2011, 44 (3), 227-237.DOI
  44. [46]Lyakhov, A. O.; Oganov, A. R.; Stokes, H. T.; Zhu, Q., New developments in evolutionary structure prediction algorithm USPEX. Comput. Phys. Commun. 2013, 184 (4), 1172-1182.DOI
  45. [47]Kohn, W.; Sham, L. J., Quantum density oscillations in an inhomogeneous electron gas. Phys. Rev. 1965, 137 (6A), 1697-1705.DOI
  46. [48]Hohenberg, P.; Kohn, W., Inhomogeneous electron gas. Phys. Rev. 1964, 136 (3), B864.
  47. [49]Blochl, P. E., Projector augmented-wave method. Phys. Rev. B Condens. Matter 1994, 50 (24), 17953-17979.pubmed.ncbi.nlm.nih.gov/9976227
  48. [50]Perdew, J. P.; Burke, K.; Ernzerhof, M., Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77 (18), 3865.pubmed.ncbi.nlm.nih.gov/10062328
  49. [51]Kresse, G.; Furthmuller, J., Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 54 (16), 11169.pubmed.ncbi.nlm.nih.gov/9984901
  50. [52]Kresse, G.; Furthmüller, J., Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comp. Mater. Sci. 1996, 6 (1), 15-50.
  51. [53]Yu, M.; Trinkle, D. R., Accurate and efficient algorithm for Bader charge integration. J. Chem. Phys. 2011, 134 (6), 064111.DOI
  52. [54]Geng, H. Y., Full Temperature-Dependent Potential and Anharmonicity in Metallic Hydrogen: Colossal NQE and the Consequences. J. Phys. Chem. C 2022, 126 (45), 19355-19366.DOI
  53. [55]Togo, A.; Tanaka, I., First principles phonon calculations in materials science. Scripta Mater. 2015, 108, 1-5.arxiv.org/abs/1506.08498
  54. [56]Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; et al., QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials. J. Phys.: Condens. Matt. 2009, 21 (39), 395502.DOI
  55. [57]Dynes, R. C., McMillan's equation and the T_c of superconductors. Solid State Commun. 1972, 10 (7), 615-618.DOI
  56. [58]Allen, P. B.; Dynes, R. C., Transition temperature of strong-coupled superconductors reanalyzed. Phys. Rev. B 1975, 12 (3), 905.DOI
  57. [59]Sancho, M. P. L.; Sancho, J. M. L.; Rubio, J., Highly convergent schemes for the calculation of bulk and surface Green functions. J. Phys. F: Met. Phys. 1985, 15 (4), 851-858.
  58. [60]Wu, Q.; Zhang, S.; Song, H. F.; Troyer, M.; Soluyanov, A. A., WannierTools: An open-source software package for novel topological materials. Comput. Phys. Commun. 2018, 224, 405-416.
  59. [61]Ganesan, V.; Girirajan, K. S., Lattice parameter and thermal expansion of CsCl and CsBr by x-ray powder diffraction. Pramana 1986, 27 (3), 475-478.
  60. [62]Rundqvist, S.; Nawapong, P. C.; Widmark, G.; Nielsen, P. H.; Sjöberg, B.; Larsen, E., Crystal structure refinements of some MnP-type phosphides. Acta Chem. Scand. 1965, 19 (19), 1006-1008.DOI
  61. [63]Edshammar, L. E.; Longo, J. M.; Marinder, B. O.; Nielsen, P. H.; Sjöberg, B.; Larsen, E., The crystal structures of Os2Al3 and OsAl2. Acta Chem. Scand. 1965, 19 (19), 871-874.DOI
  62. [64]Harada, Y.; Morinaga, M.; Saso, D.; Takata, M.; Sakata, M., Refinement in crystal structure of MoSi2. Intermetallics 1998, 6 (6), 523-527.DOI
  63. [65]Zhang, W.; Oganov, A. R.; Goncharov, A. F.; Zhu, Q.; Boulfelfel, S. E.; Lyakhov, A. O.; Stavrou, E.; Somayazulu, M.; Prakapenka, V. B.; Konopkova, Z., Unexpected stable stoichiometries of sodium chlorides. Science 2013, 342 (6165), 1502-1505.DOI
  64. [66]Li, C.; Yang, W.; Sheng, H. W., Bridge-bond formation in aluminum and its alloys under high pressure. Phys. Rev. Mater. 2022, 6 (3), 033601
  65. [68]Yao, Y.; Tse, J. S.; Klug, D. D., Structures of insulating phases of dense lithium. Phys. Rev. Lett. 2009, 102 (11), 115503.pubmed.ncbi.nlm.nih.gov/19392214
  66. [69]Miao, M. S.; Hoffmann, R.; Botana, J.; Naumov, I. I; Hemley, R. J., Quasimolecules in compressed Lithium. Angew. Chem. Int. Ed. Engl. 2017, 56 (4), 972-975.
  67. [70]Ma, Y.; Eremets, M.; Oganov, A. R.; Xie, Y.; Trojan, I.; Medvedev, S.; Lyakhov, A. O.; Valle, M.; Prakapenka, V., Transparent dense sodium. Nature 2009, 458 (7235), 182-185.DOI
  68. [71]Li, Y.; Wang, Y.; Pickard, C. J.; Needs, R. J.; Wang, Y.; Ma, Y., Metallic icosahedron phase of sodium at terapascal pressures. Phys. Rev. Lett. 2015, 114 (12), 125501.DOI
  69. [72]Chen, Y. M.; Geng, H. Y.; Yan, X. Z.; Wang, Z. W.; Chen, X. R.; Wu, Q., Predicted novel insulating electr ide compound between alkali metals lithium and sodium under high pressure. Chin. Phys. B 2017, 26 (5), 056102.DOI
  70. [73]Zhang, Y.; Wu, W.; Wang, Y.; Yang, S. A.; Ma, Y., Pressure-stabilized semiconducting elect rides in alkaline-earth-metal subnitrides. J. Am. Chem. Soc. 2017, 139 (39), 13798-13803.
  71. [74]Quan, Y.; Pickett, W. E., Van Hove singularities and spectral smearing in high-temperature superconducting H₃S. Phys. Rev. B 2016, 93 (10), 104526.DOI

Paper details

vlnc:W_ugfhern8
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