Abstract
We use firstprinciples methods to study doped strong ferroelectrics (taking BaTiO_{3} as a prototype). Here, we find a strong coupling between itinerant electrons and soft polar phonons in doped BaTiO_{3}, contrary to Anderson/Blount’s weakly coupled electron mechanism for "ferroelectriclike metals”. As a consequence, across a polartocentrosymmetric phase transition in doped BaTiO_{3}, the total electronphonon coupling is increased to about 0.6 around the critical concentration, which is sufficient to induce phononmediated superconductivity of about 2 K. Lowering the crystal symmetry of doped BaTiO_{3} by imposing epitaxial strain can further increase the superconducting temperature via a sizable coupling between itinerant electrons and acoustic phonons. Our work demonstrates a viable approach to modulating electronphonon coupling and inducing phononmediated superconductivity in doped strong ferroelectrics and potentially in polar metals. Our results also show that the weakly coupled electron mechanism for "ferroelectriclike metals” is not necessarily present in doped strong ferroelectrics.
Introduction
Electronphonon coupling plays an important role in a variety of physical phenomena in solids^{1}. In metals and doped semiconductors, lowenergy electronic excitations are strongly modified by the coupling of itinerant electrons to lattice vibrations, which influences their transport and thermodynamic properties^{2}. Furthermore, electronphonon coupling provides an attractive electronelectron interaction, which leads to conventional (i.e., phononmediated) superconductivity in many metals^{3}. Recent studies on hydrogenrich materials show that when their electronphonon coupling is strong enough, the transition temperature of conventional superconductors can reach as high as 260 K at 180–200 GPa^{4,5,6}. One general way to increase the electronphonon coupling of solids is to find a particular phonon to which itinerant electrons are strongly coupled and whose softening (i.e., the phonon frequency approaches zero) across a structural phase transition may consequently increase the total electronphonon coupling^{7}. However, identifying a strong coupling between a soft phonon and itinerant electrons in real materials is no easy task, which relies on material details. On the other hand, the superconductivity in doped SrTiO_{3} has drawn great interests from both theorists^{8,9,10,11,12,13,14,15} and experimentalists^{16,17,18,19,20,21,22,23,24}. One beautiful experiment is Sr_{1−x}Ca_{x}TiO_{3−δ} in which Ca doping leads to a weak ferroelectric distortion in SrTiO_{3} and oxygen vacancies provide itinerant electrons^{20,25,26}. Increasing the carrier concentration in Sr_{1−x}Ca_{x}TiO_{3−δ} induces a polartocentrosymmetric phase transition and a superconducting “dome” emerges around the critical concentration. The nature of the superconductivity in doped SrTiO_{3} is highly debatable^{8,9,10,11,12,13,14,15,16,17,18,19,20,21,27}, because the superconductivity in doped SrTiO_{3} can persist to very low carrier density^{11,12,28}, which seriously challenges the standard phonon pairing mechanism^{29}. It is not clear why superconductivity in doped SrTiO_{3} vanishes above a critical concentration in spite of an increasing density of states at the Fermi level^{30}. Attention has been paid to recent proposals on soft polar phonons, but the coupling details and strength are controversial^{10,11,15,27,31}. Furthermore, according to Anderson and Blount’s original proposal that inversion symmetry breaking by collective polar displacements in metals relies on the weak coupling between itinerant electrons and soft phonons responsible for inversion symmetry breaking^{32,33,34}, it is not obvious that across the polartocentrosymmetric phase transition the soft polar phonons can be coupled to itinerant electrons in Sr_{1−x}Ca_{x}TiO_{3−δ}, or more generally in doped ferroelectrics and polar metals^{11,15,31,35}.
Motivated by the above experiments and theories, we use firstprinciple methods with no adjustable parameters to demonstrate a large modulation of electronphonon coupling in doped strong ferroelectrics by utilizing soft polar phonons. We study BaTiO_{3} as a prototype, because (1) previous studies found that in ndoped BaTiO_{3}, increasing the carrier density gradually reduces its polar distortions and induces a continuous polartocentrosymmetric phase transition^{36,37}; and (2) the critical concentration for the phase transition is about 10^{21}/cm^{3}, which is high enough so that the electronphonon coupling can be directly calculated within the Migdal’s approximation (in contrast, in doped SrTiO_{3}, superconductivity emerges at a much lower carrier concentration 10^{17}–10^{20}/cm^{3} so that its Debye frequency is comparable to or even higher than the Fermi energy ℏω_{D}/ϵ_{F} ~ 1 − 10^{2}^{38}, which invalidates the Migdal’s approximation and Eliashberg equation)^{29}. The key result from our calculation is that, contrary to Anderson/Blount’s argument for "ferroelectriclike metals”^{32,33,34}, we find that the phonon bands associated with the soft polar optical phonons are strongly coupled to itinerant electrons across the polartocentrosymmetric phase transition in doped BaTiO_{3}. As a consequence, the total electronphonon coupling of doped BaTiO_{3} can be substantially modulated via carrier density and in particular is increased to about 0.6 around the critical concentration. Eliashberg equation calculations find that such an electronphonon coupling is sufficiently large to induce phononmediated superconductivity of about 2K. In addition, we find that close to the critical concentration, lowering the crystal symmetry of doped BaTiO_{3} by imposing epitaxial strain further increases the superconducting temperature via a sizable coupling between itinerant electrons and acoustic phonon bands.
While ferroelectricity and superconductivity have little in common, our work demonstrates an experimentally viable approach to modulating electronphonon coupling and inducing phononmediated superconductivity in doped strong ferroelectrics and potentially in polar metals^{32,39}. Our results show that the weakly coupled electron mechanism in “ferroelectriclike metals” is not necessarily present in doped strong ferroelectrics and as a consequence, the soft polar phonons can be utilized to induce phononmediated superconductivity across a structural phase transition.
Results
Structural phase transition induced by electron doping
In this study, electron doping in BaTiO_{3} is achieved by adding extra electrons to the system with the same amount of uniform positive charges in the background. For benchmarking, our calculation of the undoped tetragonal BaTiO_{3} gives the lattice constant a = 3.930 Å and c/a = 1.012, polarization P = 0.26 C/m^{2}, and TiO and BaO relative displacements of 0.105 Å and 0.083 Å, respectively, consistent with the previous calculations^{40,41,42}. We note that upon electron doping, BaTiO_{3} becomes metallic and its polarization is illdefined^{43}. Therefore, we focus on analyzing ionic polar displacements and c/a ratio to identify the critical concentration^{36}.
We test four different crystal structures of BaTiO_{3} with electron doping: the rhombohedral structure (space group R3m with Ti displaced along 〈111〉 direction), the orthorhombic structure (space group Amm2 with Ti displaced along 〈011〉 direction), the tetragonal structure (space group P4mm with Ti displaced along 〈001〉 direction) and the cubic structure (space group \(Pm\bar{3}m\) with Ti at the center of oxygen octahedron). Figure 1a shows that as electron doping concentration n increases from 0 to 0.15e/f.u., BaTiO_{3} transitions from the rhombohedral structure to the tetragonal structure, and finally to the cubic structure. The critical concentration is such that the crystal structure of doped BaTiO_{3} continuously changes from tetragonal to cubic (see Supplementary Note 6). While the structural transition from tetragonal to cubic is continuous, the transition from rhombohedral to tetragonal is firstorder and thus does not show phonon softening (see Supplementary Note 7). Furthermore the low electron concentration in the rhombohedral structure invalidates Migdal’s theorem and electronphonon coupling cannot be calculated within Migdal’s approximation (see Supplementary Note 5).
Figure 1b shows c/a ratio and TiO cation displacements δ as a function of the concentration n in the range of 0.06–0.14e/f.u. It is evident that the critical concentration n_{c} of doped BaTiO_{3} is 0.10e/f.u. (about 1.6 × 10^{21} cm^{−3}), at which the polar displacement δ is just completely suppressed and the c/a ratio is reduced to unity. This result is consistent with the previous theoretical studies^{36,37}. Experimentally, in metallic oxygendeficient BaTiO_{3−δ}, the lowsymmetry polar structure can be retained up to an electron concentration of 1.9 × 10^{21} cm^{−3} (close to the theoretical result)^{44,45}. However, weak localization and/or phase separation may exist in oxygendeficient BaTiO_{3}, depending on sample quality^{44,46}.
Electronic structure and phonon properties
Figure 2a shows the electronic structure of doped BaTiO_{3} in the tetragonal structure at a representative concentration (n = 0.09e/f.u., close to the critical value). Undoped BaTiO_{3} is a wide gap insulator. Electron doping moves the Fermi level slightly above the conduction band edge of the three Ti t_{2g} orbitals and thus a Fermi surface is formed. We use three Wannier functions to reproduce the Ti t_{2g} bands, upon which electronphonon coupling is calculated. Figure 2b shows the phonon spectrum of doped BaTiO_{3} in the tetragonal structure at 0.09e/f.u. concentration. We are particularly interested in the zonecenter (Γpoint) polar optical phonons, which are highlighted by the green dots in Fig. 2b. The vibrational modes of those polar phonons are explicitly shown in Fig. 2c. In the tetragonal structure of BaTiO_{3}, the two polar phonons with the ion displacements along x and y directions (ω_{x} and ω_{y}) are degenerate, while the third polar phonon with the ion displacements along z direction (ω_{z}) has higher frequency. Figure 2d shows that electron doping softens the zonecenter polar phonons of BaTiO_{3} in the tetragonal structure until it reaches the critical concentration where the three polar phonon frequencies become zero. With further electron doping, the polar phonon frequencies of BaTiO_{3} increase in the cubic structure (see Supplementary Note 13 for a discussion about doping’s effect on polar phonon behavior).
Electronphonon coupling and phononmediated superconductivity
The continuous polartocentrosymmetric phase transition in doped BaTiO_{3} is similar to the one in “ferroelectriclike metals” proposed by Anderson and Blount^{32}. They first argued, later recast by Puggioni and Rondinelli^{33,34}, that inversion symmetry breaking by collective polar displacements in a metal relies on a weak coupling between itinerant electrons and soft phonons responsible for removing inversion symmetry. According to this argument, one would expect that across the polartocentrosymmetric phase transition, the soft polar phonons are not strongly coupled to itinerant electrons in doped BaTiO_{3}. In order to quantify the strength of electronphonon coupling and make quantitative comparison, we introduce the moderesolved electronphonon coupling λ_{qν} and aroundzonecenter branchresolved electronphonon coupling λ_{ν}:
where \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) is the imaginary part of electronphonon selfenergy, ω_{qν} is the phonon frequency, N_{F} is the density of states at the Fermi level and q_{c} is a small phonon momentum. The reason we define λ_{ν} within ∣q∣ < q_{c} is because: (1) exactly at the zonecenter Γ point, the acoustic phonon frequency is zero and thus the contribution from the acoustic mode is illdefined at Γ point; (2) when q_{c} is sufficiently small, there are no phonon band crossings within ∣q∣ < q_{c} and hence each branch ν can be assigned to a welldefined phonon mode (for a general q point, it is not trivial to distinguish which phonon band corresponds to polar modes and which to other optical modes). We choose \({q}_{c}=0.05\frac{\pi }{a}\) where a is the lattice constant (the qualitative conclusions do not depend on the choice of q_{c}, as long as no phonon band crossings occur within ∣q∣ < q_{c}).
Figure 3a, b show the imaginary part of electronphonon selfenergy \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) for each phonon mode qν of doped BaTiO_{3} along a highsymmetry path (panel a corresponds to 0.09e/f.u. doping in a tetragonal structure and panel b corresponds to 0.11e/f.u. doping in a cubic structure). Since within the double delta approximation \({\rm{Im}}{{{\Sigma }}}_{{\bf{q}}\nu }\) is positive definite (see Supplementary Note 2), the point size in panels a and b is chosen to be proportional to the value of \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\). Our calculations find that, contrary to Anderson/Blount’s weak coupled electron mechanism^{32}, the phonon bands associated with the zonecenter polar phonons have the strongest coupling to itinerant electrons, while the couplings of other phonon bands are weaker. Specifically, in the case of 0.09e/f.u. doping:
and in the case of 0.11e/f.u. doping:
In both cases, λ_{polar} is larger than λ_{acoustic} and λ_{others}. An intuitive picture for the strong coupling is that in doped BaTiO_{3}, the soft polar phonons involve the cation displacements of Ti and O atoms, and in the meantime itinerant electrons derive from Tid states which hybridize with Op states (see Supplementary Note 14 for an alternative demonstration of this strong coupling, and Supplementary Note 12 for a discussion about doping’s effect on this pd hybridization). This is in contrast to the textbook example of polar metals LiOsO_{3} where the soft polar phonons involve Li displacements while the metallicity derives from Os and O orbitals^{33}. More quantitatively, we find λ_{polar} = 0.50 for LiOsO_{3}, which is substantially smaller than λ_{polar} of about 5–10 for doped BaTiO_{3}. In short, because the itinerant electrons and polar phonons are associated with the same atoms in doped BaTiO_{3}, the coupling is strong, while in LiOsO_{3} the itinerant electrons and polar phonons involve different atoms and thus the coupling is weak. As a consequence of the strong interaction between the polar phonons and itinerant electrons, we expect that the total electronphonon coupling of doped BaTiO_{3} can be increased by softening the polar phonons across the structural phase transition.
Figure 3c shows the total electronphonon spectral function α^{2}F(ω) and accumulative electronphonon coupling λ(ω) of doped BaTiO_{3} at 0.09e/f.u. and 0.11e/f.u. concentrations. α^{2}F(ω) is defined as:
where Ω_{BZ} is the volume of phonon Brillouin zone. With α^{2}F(ω), it is easy to calculate the accumulative electronphonon coupling λ(ω):
The total electronphonon coupling λ is obtained by taking the upper bound ω to ∞ in Eq. (5). The green shades are α^{2}F(ω) and the dashed lines are the corresponding cumulative electronphonon coupling. The total electronphonon coupling λ of doped BaTiO_{3} in the tetragonal structure at 0.09e/f.u. concentration is 0.61, while that in the cubic structure at 0.11e/f.u. concentration is 0.50. Both λ are sufficiently large to induce phononmediated superconductivity with measurable transition temperature. Figure 3d shows the total electronphonon coupling λ of doped BaTiO_{3} for a range of electron concentrations (exactly at the critical concentration, we find some numerical instabilities and divergence in the electronphonon calculations, rendering the result unreliable). An increase of λ around the critical concentration is evident, consistent with the strong coupling between the soft polar phonons and itinerant electrons in doped BaTiO_{3}.
Based on the electronphonon spectrum α^{2}F(ω), we use a threeorbital Eliashberg equation (see Supplementary Note 3) to calculate the superconducting gap Δ(T) and estimate the superconducting transition temperature T_{c} as a function of electron concentration. Because the three Ti t_{2g} orbitals become identical at the critical concentration, when solving the threeorbital Eliashberg equation, we set MorelAnderson pseudopotential \({\mu }_{ij}^{* }\) to be 0.1 for each orbital pair (i.e., i, j = 1, 2, 3)^{47}. Figure 3e shows the superconducting gap Δ(T) of doped BaTiO_{3} as a function of temperature T at two representative concentrations (0.09e/f.u. in the tetragonal structure and 0.11e/f.u. in the cubic structure). Since both concentrations are close to the critical value, the three Ti t_{2g} orbitals are almost degenerate in doped BaTiO_{3}. For clarification, we show the superconducting gap of one orbital for each concentration. From the Eliashberg equation, we find that at 0.09e/f.u. concentration, Δ(T = 0) = 0.27 meV and T_{c} = 1.75 K; and at 0.11e/f.u. concentration, Δ(T = 0) = 0.11 meV and T_{c} = 0.76 K. Thus Δ(T = 0)/(k_{B}T_{c}) = 1.79 at 0.09e/f.u. concentration and 1.68 at 0.11e/f.u. concentration, both close to the BCS prediction of 1.77. Figure 3f shows the estimated superconducting transition temperature T_{c} of doped BaTiO_{3} for a range of electron concentrations. T_{c} notably exhibits a domelike feature as a function of electron concentration. The origin of the superconducting “dome” is that the electronphonon coupling of doped BaTiO_{3} is increased by the softened polar phonons around the critical concentration. When the electron concentration is away from the critical value, the polar phonons are “hardened” (i.e., phonon frequency increases) and the electronphonon coupling of doped BaTiO_{3} decreases. We note that the estimated T_{c} strongly depends on \({\mu }_{ij}^{* }\). Therefore in the inset of Fig. 3f, we study doped BaTiO_{3} at a representative concentration (0.09e/f.u.) and calculate its superconducting transition temperature T_{c} as a function of \({\mu }_{ij}^{* }\). As \({\mu }_{ij}^{* }\) changes from 0 to 0.3, the estimated T_{c} decreases from 9.3 K to 0.4 K, the lowest of which (0.4 K) is still measurable in experiment^{28}. We make two comments here: (1) The superconducting transition temperature is only an estimation due to the uncertainty of MorelAnderson pseudopotential \({\mu }_{ij}^{* }\) and other technical details. But the picture of an increased electronphonon coupling around the structural phase transition in doped BaTiO_{3} is robust. (2) Experimentally in Sr_{1−x}Ca_{x}TiO_{3}, the optimal doping for superconductivity is larger than the “ferroelectric” critical concentration^{20}, while in our calculations of doped BaTiO_{3}, the two critical concentrations (one for optimal superconducting T_{c} and the other for suppressing polar displacements) just coincide due to polar phonon softening and an increased electronphonon coupling. Comparison of these two materials implies that the microscopic mechanism for superconductivity in doped SrTiO_{3} is probably not purely phononmediated.
Crystal symmetry and acoustic phonons
We note that in Figure 3a, b, in addition to the large \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) in the polar optical phonon bands, there is also sizable \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) in the acoustic phonon bands (from Γ to X) in the tetragonal structure at 0.09e/f.u. concentration. Since the moderesolved electronphonon coupling \({\lambda }_{{\bf{q}}\nu }\propto {\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }/{\omega }_{{\bf{q}}\nu }^{2}\), the small frequency of acoustic phonons can lead to a substantial λ_{qν}, given a sizable \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\). However, in the cubic structure at 0.11e/f.u. concentration, \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) in the acoustic phonon bands almost vanishes from Γ to X. To exclude that the concentration difference may have an effect, we perform a numerical experiment: we start from the cubic structure doped at 0.11e/f.u. concentration (space group \(Pm\bar{3}m\)), and then we impose a slight (001) compressive biaxial 0.8% strain by fixing the two inplane lattice constants (a and b) to a smaller value. This compressive strain makes the crystal structure of doped BaTiO_{3} tetragonal and polar (space group P4mm). Figure 4a shows the optimized crystal structures of the two doped BaTiO_{3}. For doped BaTiO_{3} at 0.11e/f.u. concentration, without strain, the ground state structure is cubic and the optimized lattice constant a is 3.972 Å; under a 0.8% biaxial (001) compressive strain, the ground state structure becomes tetragonal with the inplane lattice constants a and b being fixed at 3.940 Å and the optimized long lattice constant c being 4.019 Å. We find that the total electronphonon coupling λ increases from 0.50 in the \(Pm\bar{3}m\) structure to 0.57 in the P4mm structure. Figure 4b, c compare the imaginary part of the electronphonon selfenergy \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) and the moderesolved electronphonon coupling λ_{qν} along the Γ → X path for the two doped BaTiO_{3}. Similar to Fig. 3a and b, we find that there is a notable difference in \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) from the acoustic phonon bands. The difference in \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) is further “amplified” by the low phonon frequencies ω_{qν}, which results in
At the same time, we find that for polar modes,
This shows that under 0.8% (001) compressive strain, λ_{polar} remains substantial (albeit reduced by about 40%), but λ_{acoustic} is increased by one order of magnitude, which altogether leads to an enhancement of the total electronphonon coupling λ. Note that in the numerical experiment, the two doped BaTiO_{3} have exactly the same electron concentration, indicating that the additional increase in \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) of the acoustic phonons arises solely from the crystal structure difference. A possible explanation, which is based on our calculations, is that in the cubic structure, some electronphonon vertices \({g}_{ij}^{\nu }({\bf{k}},{\bf{q}})\) are exactly equal to zero because some atoms are frozen in the acoustic phonons, while in the lowsymmetry structure, those \({g}_{ij}^{\nu }({\bf{k}},{\bf{q}})\) become nonzero. Because \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\propto  {g}_{ij}^{\nu }({\bf{k}},{\bf{q}}){ }^{2}\)^{48,49}, this leads to an increase in \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\). In addition, the frequencies of acoustic phonon modes ω_{qν} are very small and \({\lambda }_{{\bf{q}}\nu }\propto {\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }/{\omega }_{{\bf{q}}\nu }^{2}\), therefore even a slight increase in \({\rm{Im}}{{{\Pi }}}_{{\bf{q}}\nu }\) results in a substantial enhancement in λ_{qν} (see Supplementary Note 16 for the demonstration of a specific acoustic phonon). Our numerical experiment also implies that in doped BaTiO_{3}, when the electron concentration is close to the critical value, a small (001) compressive strain that lowers the crystal symmetry may also enhance its superconducting transition temperature due to the increased electronphonon coupling, similar to doped SrTiO_{3}^{18,22}.
Discussion
Finally we discuss possible experimental verification. Chemical doping^{44,45,50,51,52,53,54} and epitaxial strain^{55,56} have been applied to ferroelectric materials such as BaTiO_{3}. Ladoped BaTiO_{3} has been experimentally synthesized. Hightemperature transport measurements show that Ba_{1−x}La_{x}TiO_{3} exhibits polar metallic behavior but ultralowtemperature transport measurements are yet to be performed^{50,51,52,53,54}. We note that La doping in BaTiO_{3} may result in some chemical disorder. While the randomness of La distribution in La_{x}Ba_{1−x}TiO_{3} may affect the transport properties in the normal state, Anderson’s theorem asserts that superconductivity in a conventional superconductor is robust with respect to nonmagnetic disorder in the host material^{57}. As a consequence, the superconducting transition temperature T_{c} of a conventional superconductor barely depends on the randomness of defects. In our case, the superconductivity in doped BaTiO_{3} is phononmediated (i.e., conventional) and La is a nonmagnetic dopant. Therefore Anderson’s theorem applies and we expect that even if chemical disorder may arise in actual experiments, it does not affect the superconducting properties of doped BaTiO_{3}. In addition, we perform supercell calculations which include real La dopants. We find that even in the presence of real La atoms, the conduction electrons on Ti atoms are almost uniformly distributed in La_{x}Ba_{1−x}TiO_{3} (see Supplementary Note 8 and Supplementary Note 9 for details). Since our simulation does not consider dopants explicitly, a more desirable doping method is to use electrostatic carrier doping^{58,59,60}, which does not involve chemical dopants and has been successfully used to induce superconductivity in KTaO_{3}^{61}. We clarify two points concerning the electrostatic doping method. (1) The electrostatic gating by ionic liquid can achieve a twodimensional carrier density as high as 8 × 10^{14} cm^{−2}^{62}. The induced electrons are usually confined in a narrow region that is a few nanometers from the surface/interface, which leads to an effective threedimensional carrier density of about 1 × 10^{21} ~ 5 × 10^{21} cm^{−3}^{61,63}. In our current study, the critical concentration of doped BaTiO_{3} is about 1.6 × 10^{21} cm^{−3}, which is feasible by this approach. (2) While the electrostatic doping method induces the carriers in the surface/interface area, we show that our results on bulk doped BaTiO_{3} can still be used as a guidance to search for superconductivity in the surface area of BaTiO_{3}. We perform calculations of Pt/BaTiO_{3} interface (see Supplementary Note 10) and find that just in the second unit cell of BaTiO_{3} from the interface, the TiO displacement saturates and a bulklike region emerges with almost uniform cation displacements. In addition, we calculate the electronphonon properties of bulk KTaO_{3} at 0.14e/f.u. doping (based on the experiment^{61}) (see Supplementary Note 15). We find that the total electronphonon coupling of KTaO_{3} at 0.14e/f.u. doping is 0.36. Using McMillian equation (take μ^{*} = 0.1) as a rough estimation of superconducting transition temperature T_{c}, we obtain a T_{c} of about 68 mK, which is in reasonable agreement with the experimental value of 50 mK. While there is definitely room for improvement, our results demonstrate that for a given target material, its desirable bulk electronphonon property can point to the right direction in which superconductivity is found in surface/interface regions.
In summary, we use firstprinciples calculations to demonstrate a large modulation of electronphonon coupling and an emergent superconducting “dome” in ndoped BaTiO_{3}. Contrary to Anderson/Blount’s weak electron coupling mechanism for “ferroelectriclike metals”^{32,33,34}, our calculations find that the soft polar phonons are strongly coupled to itinerant electrons across the polartocentrosymmetric phase transition in doped BaTiO_{3} and as a consequence, the total electronphonon coupling increases around the critical concentration. In addition, we find that lowering the crystal symmetry of doped BaTiO_{3} by imposing epitaxial strain can also increase the electronphonon coupling via a sizable coupling between acoustic phonons and itinerant electrons. Our work provides an experimentally viable method to modulating electronphonon coupling and inducing phononmediated superconductivity in doped strong ferroelectrics. Our results indicate that the weak electron coupling mechanism for “ferroelectriclike metals”^{32,33,34} is not necessarily present in doped strong ferroelectrics. We hope that our predictions will stimulate experiments on doped ferroelectrics and search for the phononmediated superconductivity that is predicted in our calculations.
Methods
We perform firstprinciples calculations by using density functional theory^{64,65,66,67} as implemented in the Quantum ESPRESSO package^{68}. We use normconserving pseudopotentials^{69} with local density approximation as the exchangecorrelation functional. For electronic structure calculations, we use an energy cutoff of 100 Ry. We optimize both cell parameters and internal coordinates in atomic relaxation, We find that the optimized crystal structures are in good agreement with experiments (see Supplementary Note 1). The detailed structural information is reported in Supplementary Note 7. In the strain calculations, the inplane lattice constants are fixed while the outofplane lattice constant and internal coordinates are fully optimized. The electron Brillouin zone integration is performed with a Gaussian smearing of 0.005 Ry over a Γcentered k mesh of 12 × 12 × 12. The threshold of total energy convergence is 10^{−7} Ry; selfconsistency convergence is 10^{−12} Ry; force convergence is 10^{−6} Ry/Bohr and pressure convergence for variable cell is 0.5 kbar. For phonon calculations, we use density functional perturbation theory^{66} as implemented in the Quantum ESPRESSO package^{68} (see Supplementary Note 11 for the validation of this method on a prototypical oxide SrTiO_{3}). The phonon Brillouin zone integration is performed over a q mesh of 6 × 6 × 6. For the calculations of electronphonon coupling and superconducting gap (see Supplementary Note 2), we use maximally localized Wannier functions and MigdalEliashberg theory, as implemented in the Wannier90^{70} and EPW code^{71}. The Fermi surface of electrondoped BaTiO_{3} is composed of three Ti t_{2g} orbitals. We use three maximally localized Wannier functions to reproduce the Fermi surface. The electronphonon matrix elements \({g}_{ij}^{\nu }({\bf{k}},{\bf{q}})\) are first calculated on a coarse 12 × 12 × 12 kgrid in the electron Brillouin zone and a coarse 6 × 6 × 6 qgrid in the phonon Brillouin zone, and then are interpolated onto fine grids via maximally localized Wannier functions. The fine electron and phonon grids are both 50 × 50 × 50. We check the convergence on the electron kmesh, phonon qmesh and Wannier interpolation and no significant difference is found by using a denser mesh. Details can be found in Supplementary Note 4. We solve a threeorbital Eliashberg equation to estimate the superconducting transition temperature T_{c} (see Supplementary Note 3).
We only use Eliashberg equation when electron doping concentration is high enough so that λT_{D}/T_{F} < 0.1 and Migdal’s theorem is valid^{29} (λ is electronphonon coupling, T_{D} is Debye temperature and T_{F} is Fermi temperature). Validation test of Migdal’s theorem is shown in Supplementary Note 5.
We solve a threeorbital Eliashberg equation to estimate the superconducting transition temperature T_{c}. This method is compared to McMillan Equation. Details of Eliashberg Equation and McMillan Equation can be found in Supplementary Note 3.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Code availability
The electronic structure calculations were performed using the opensource code Quantum Espresso^{68}. Quantum Espresso package is freely distributed on academic use under the Massachusetts Institute of Technology (MIT) License.
References
Grimvall, G. The electronphonon interaction in metals (NorthHolland, Amsterdam, 1981).
Ziman, J. M. Electrons and phonons: the theory of transport phenomena in solids (Clarendon, Oxford, 1960).
Bardeen, J., Cooper, L. N. & Schrieffer, J. R. Theory of superconductivity. Phys. Rev. 108, 1175–1204 (1957).
Somayazulu, M. et al. Evidence for superconductivity above 260 K in lanthanum superhydride at megabar pressures. Phys. Rev. Lett. 122, 027001 (2019).
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. Nature 525, 73–76 (2015).
Drozdov, A. P. et al. Superconductivity at 250 K in lanthanum hydride under high pressures. Nature 569, 528–531 (2019).
Allen, P. B. & Mitrović, B. Theory of superconducting T_{c}. Solid State Phys. 37, 1–92 (1983).
Kozii, V., Bi, Z. & Ruhman, J. Superconductivity near a ferroelectric quantum critical point in ultralowdensity Dirac materials. Phys. Rev. X 9, 031046 (2019).
Dunnett, K., Narayan, A., Spaldin, N. A. & Balatsky, A. V. Strain and ferroelectric softmode induced superconductivity in strontium titanate. Phys. Rev. B 97, 144506 (2018).
Edge, J. M., Kedem, Y., Aschauer, U., Spaldin, N. A. & Balatsky, A. V. Quantum critical origin of the superconducting dome in SrTiO_{3}. Phys. Rev. Lett. 115, 247002 (2015).
Wölfle, P. & Balatsky, A. V. Superconductivity at low density near a ferroelectric quantum critical point: doped SrTiO_{3}. Phys. Rev. B 98, 104505 (2018).
Ruhman, J. & Lee, P. A. Superconductivity at very low density: the case of strontium titanate. Phys. Rev. B 94, 224515 (2016).
Gor’kov, L. P. Phonon mechanism in the most dilute superconductor ntype SrTiO_{3}. Proc. Natl Acad. Sci. 113, 4646–4651 (2016).
Kanasugi, S. & Yanase, Y. Multiorbital ferroelectric superconductivity in doped SrTiO_{3}. Phys. Rev. B 100, 094504 (2019).
van der Marel, D., Barantani, F. & Rischau, C. W. Possible mechanism for superconductivity in doped SrTiO_{3}. Phys. Rev. Res. 1, 013003 (2019).
Lin, X. et al. Critical doping for the onset of a twoband superconducting ground state in SrTiO_{3}. Phys. Rev. Lett. 112, 207002 (2014).
Koonce, C. S., Cohen, M. L., Schooley, J. F., Hosler, W. R. & Pfeiffer, E. R. Superconducting transition temperatures of semiconducting SrTiO_{3}. Phys. Rev. 163, 380–390 (1967).
Russell, R. et al. Ferroelectric enhancement of superconductivity in compressively strained SrTiO_{3} films. Phys. Rev. Mater. 3, 091401 (2019).
Baratoff, A. & Binnig, G. Mechanism of superconductivity in SrTiO_{3}. Phys. B+C. 108, 1335–1336 (1981).
Rischau, C. W. et al. A ferroelectric quantum phase transition inside the superconducting dome of Sr_{1−x}Ca_{x}TiO_{3−x}. Nat. Phys. 13, 643–648 (2017).
Swartz, A. G. et al. Polaronic behavior in a weakcoupling superconductor. Proc. Natl Acad. Sci. 115, 1475–1480 (2018).
Ahadi, K. et al. Enhancing superconductivity in SrTiO_{3} films with strain. Sci. Adv. 5, eaaw0120 (2019).
Herrera, C. et al. Strainengineered interaction of quantum polar and superconducting phases. Phys. Rev. Mater. 3, 124801 (2019).
Stucky, A. et al. Isotope effect in superconducting ndoped SrTiO_{3}. Sci. Rep. 6, 37582 (2016).
Wang, J. et al. Charge transport in a polar metal. npj Quantum Mater. 4, 61 (2019).
Bednorz, J. G. & Müller, K. A. Sr_{1−x}Ca_{x}TiO_{3}: an XY quantum ferroelectric with transition to randomness. Phys. Rev. Lett. 52, 2289–2292 (1984).
Kedem, Y., Zhu, J.X. & Balatsky, A. V. Unusual superconducting isotope effect in the presence of a quantum criticality. Phys. Rev. B 93, 184507 (2016).
Lin, X., Zhu, Z., Fauqué, B. & Behnia, K. Fermi surface of the most dilute superconductor. Phys. Rev. X 3, 021002 (2013).
Migdal, A. Interaction between electrons and lattice vibrations in a normal metal. Sov. Phys. J. Exp. Theor. Phys. 7, 996–1001 (1958).
Tomioka, Y., Shirakawa, N., Shibuya, K. & Inoue, I. H. Enhanced superconductivity close to a nonmagnetic quantum critical point in electrondoped strontium titanate. Nat. Commun. 10, 738 (2019).
Ruhman, J. & Lee, P. A. Comment on “superconductivity at low density near a ferroelectric quantum critical point: doped SrTiO_{3}”. Phys. Rev. B 100, 226501 (2019).
Anderson, P. W. & Blount, E. I. Symmetry considerations on martensitic transformations: “ferroelectric” metals? Phys. Rev. Lett. 14, 217–219 (1965).
Laurita, N. J. et al. Evidence for the weakly coupled electron mechanism in an AndersonBlount polar metal. Nat. Commun. 10, 3217 (2019).
Puggioni, D. & Rondinelli, J. M. Designing a robustly metallic noncenstrosymmetric ruthenate oxide with large thermopower anisotropy. Nat. Commun. 5, 3432 (2014).
Wölfle, P. & Balatsky, A. V. Reply to “comment on ‘superconductivity at low density near a ferroelectric quantum critical point: doped SrTiO_{3}’ ”. Phys. Rev. B 100, 226502 (2019).
Wang, Y., Liu, X., Burton, J. D., Jaswal, S. S. & Tsymbal, E. Y. Ferroelectric instability under screened Coulomb interactions. Phys. Rev. Lett. 109, 247601 (2012).
Xia, C., Chen, Y. & Chen, H. Coexistence of polar displacements and conduction in doped ferroelectrics: an ab initio comparative study. Phys. Rev. Mater. 3, 054405 (2019).
Gastiasoro, M. N., Ruhman, J. & Fernandes, R. M. Superconductivity in dilute SrTiO_{3}: a review. Ann. Phys. 417, 168107 (2020).
Shi, Y. et al. A ferroelectriclike structural transition in a metal. Nat. Mater. 12, 1024–1027 (2013).
Yuk, S. et al. Towards an accurate description of perovskite ferroelectrics: exchange and correlation effects. Sci. Rep. 7, 43482 (2017).
Padilla, J. & Vanderbilt, D. Ab initio study of BaTiO_{3} surfaces. Phys. Rev. B 56, 1625–1631 (1997).
Zhang, Y., Sun, J., Perdew, J. P. & Wu, X. Comparative firstprinciples studies of prototypical ferroelectric materials by LDA, GGA, and scan metaGGA. Phys. Rev. B 96, 035143 (2017).
Resta, R. Macroscopic polarization in crystalline dielectrics: the geometric phase approach. Rev. Mod. Phys. 66, 899–915 (1994).
Kolodiazhnyi, T., Tachibana, M., Kawaji, H., Hwang, J. & TakayamaMuromachi, E. Persistence of ferroelectricity in BaTiO_{3} through the insulatormetal transition. Phys. Rev. Lett. 104, 147602 (2010).
Hwang, J., Kolodiazhnyi, T., Yang, J. & Couillard, M. Doping and temperaturedependent optical properties of oxygenreduced BaTiO_{3}. Phys. Rev. B 82, 214109 (2010).
Jeong, I.K. et al. Structural evolution across the insulatormetal transition in oxygendeficient BaTiO_{3} studied using neutron total scattering and rietveld analysis. Phys. Rev. B 84, 064125 (2011).
Morel, P. & Anderson, P. W. Calculation of the superconducting state parameters with retarded electronphonon interaction. Phys. Rev. 125, 1263–1271 (1962).
Margine, E. R. & Giustino, F. Anisotropic MigdalEliashberg theory using Wannier functions. Phys. Rev. B 87, 024505 (2013).
Giustino, F. Electronphonon interactions from first principles. Rev. Mod. Phys. 89, 015003 (2017).
Ali, A. I. & Kaytbay, S. H. Electrical transport properties of LaBaTiO_{3}. Mater. Sci. Appl. 2, 716–720 (2011).
Ianculescu, A. et al. Dielectric and tunability properties of Ladoped BaTiO_{3} ceramics. J. Alloy. Compd. 509, 10040–10049 (2011).
Morrison, F. D., Sinclair, D. C. & West, A. R. Doping mechanisms and electrical properties of Ladoped BaTiO_{3} ceramics. Int. J. Inorg. Mater. 3, 1205–1210 (2001).
Takahashi, K. S. et al. Polar metal phase stabilized in strained Ladoped BaTiO_{3} films. Sci. Rep. 7, 4631 (2017).
Zhang, A. et al. Conductivity, charge transport, and ferroelectricity of Ladoped BaTiO_{3} epitaxial thin films. J. Phys. D: Appl. Phys. 53, 025301 (2019).
Schlom, D. G. et al. Elastic strain engineering of ferroic oxides. MRS Bull. 39, 118–130 (2014).
Martin, L. W. & Schlom, D. G. Advanced synthesis techniques and routes to new singlephase multiferroics. Curr. Opin. Solid State Mater. Sci. 16, 199–215 (2012).
Anderson, P. Theory of dirty superconductors. J. Phys. Chem. Solids 11, 26–30 (1959).
Moetakef, P. et al. Electrostatic carrier doping of GdTiO_{3}/SrTiO_{3} interfaces. Appl. Phys. Lett. 99, 232116 (2011).
Eyvazov, A. B., Inoue, I. H., Stoliar, P., Rozenberg, M. J. & Panagopoulos, C. Enhanced and continuous electrostatic carrier doping on the SrTiO_{3} surface. Sci. Rep. 3, 1721 (2013).
Ye, J. T. et al. Superconducting dome in a gatetuned band insulator. Science 338, 1193–1196 (2012).
Ueno, K. et al. Discovery of superconductivity in KTaO_{3} by electrostatic carrier doping. Nat. Nanotechnol. 6, 408–412 (2011).
Yuan, H. et al. Highdensity carrier accumulation in ZnO fieldeffect transistors gated by electric double layers of ionic liquids. Adv. Funct. Mater. 19, 1046–1053 (2009).
Copie, O. et al. Towards twodimensional metallic behavior at LaAlO_{3}/SrTiO_{3} interfaces. Phys. Rev. Lett. 102, 216804 (2009).
Hohenberg, P. & Kohn, W. Inhomogeneous electron gas. Phys. Rev. 136, B864–B871 (1964).
Kohn, W. & Sham, L. J. Selfconsistent equations including exchange and correlation effects. Phys. Rev. 140, A1133–A1138 (1965).
Baroni, S., de Gironcoli, S., Dal Corso, A. & Giannozzi, P. Phonons and related crystal properties from densityfunctional perturbation theory. Rev. Mod. Phys. 73, 515–562 (2001).
Marzari, N., Mostofi, A. A., Yates, J. R., Souza, I. & Vanderbilt, D. Maximally localized Wannier functions: theory and applications. Rev. Mod. Phys. 84, 1419–1475 (2012).
Giannozzi, P. et al. QUANTUM ESPRESSO: a modular and opensource software project for quantum simulations of materials. J. Phys.: Condens. Matter 21, 395502 (2009).
van Setten, M. et al. The pseudodojo: training and grading a 85 element optimized normconserving pseudopotential table. Comput. Phys. Commun. 226, 39–54 (2018).
Pizzi, G. et al. Wannier90 as a community code: new features and applications. J. Phys.: Condens. Matter 32, 165902 (2020).
Ponce, S., Margine, E., Verdi, C. & Giustino, F. EPW: electronphonon coupling, transport and superconducting properties using maximally localized Wannier functions. Comput. Phys. Commun. 209, 116–133 (2016).
Acknowledgements
We acknowledge useful discussion with Kevin Garrity, Jia Chen and Jin Zhao. H.C. is supported by the National Natural Science Foundation of China under Project No. 11774236 and NYU University Research Challenge Fund. J.M. is supported by the Student Research Program in Physics of NYU Shanghai. NYU high performance computing at Shanghai, New York and Abu Dhabi campuses provide the computational resources.
Author information
Authors and Affiliations
Contributions
H.C. conceived and supervised the project. J.M. and H.C. performed the calculations. R.Y. contributed to the data analysis. H.C. and J.M. wrote the paper and all the authors commented on the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Peer review information Nature Communications thanks Olivier Delaire and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Ma, J., Yang, R. & Chen, H. A large modulation of electronphonon coupling and an emergent superconducting dome in doped strong ferroelectrics. Nat Commun 12, 2314 (2021). https://doi.org/10.1038/s41467021225411
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467021225411
This article is cited by

Superconductivity enhancement in polar metal regions of Sr0.95Ba0.05TiO3 and Sr0.985Ca0.015TiO3 revealed by systematic Nb doping
npj Quantum Materials (2022)

Superconductivity from energy fluctuations in dilute quantum critical polar metals
Nature Communications (2022)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.