SciPost Submission Page
Interface-driven superconductivity in FeSe/SrTiO$_3$ from first-principles
by Riccardo Reho, Nils Wittemeier, Arnold Hermann Kole, Andrés Rafael Botello Méndez, Zeila Zanolli
Submission summary
| Authors (as registered SciPost users): | Riccardo Reho |
| Submission information | |
|---|---|
| Preprint Link: | scipost_202504_00009v1 (pdf) |
| Date submitted: | April 4, 2025, 3:53 p.m. |
| Submitted by: | Reho, Riccardo |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Theoretical, Computational |
Abstract
We investigate the superconducting properties of monolayer FeSe, both freestanding (ML FeSe) and on SrTiO$_3$ (STO), by simultaneously solving the Kohn-Sham Density Functional Theory and Bogoliubov--de Gennes equations. Our results demonstrate that the substrate profoundly alters both the normal-state and superconducting properties of FeSe. We identify proximity-induced superconductivity in the interfacial TiO$_2$ layer of STO, due to hybridization between Fe $d$ and O $p$ orbitals. This hybridization results in a fivefold increase in the superconducting gap width and confines superconducting states to the $M$ point in the Brillouin Zone. This is in contrast to ML FeSe, where superconductivity emerges at both the $\Gamma$ and $M$ points. Furthermore, the substrate modifies the orbital character of the states responsible for superconductivity, which change from Fe $d_{z^2}$ in ML FeSe to Fe $d_{xz}/d_{yz}$ in FeSe/STO. In both systems, we demonstrate an anisotropic superconducting gap with multiple coherence peaks, originating at different k-points in the Brillouin Zone. Additionally, in FeSe/STO, we identify emerging states unique to the superconducting phase arising from electron-hole hybridization at $M$, in agreement with experiments. Our findings highlight the decisive impact of substrate (hybridization, strain, charge transfer, magnetic order) on the superconducting properties of FeSe. We suggest potential pathways for engineering novel high-temperature FeSe-based superconductors by leveraging interfacial interactions in substrates with high electron affinity.
Author indications on fulfilling journal expectations
- Provide a novel and synergetic link between different research areas.
- Open a new pathway in an existing or a new research direction, with clear potential for multi-pronged follow-up work
- Detail a groundbreaking theoretical/experimental/computational discovery
- Present a breakthrough on a previously-identified and long-standing research stumbling block
Current status:
Reports on this Submission
Strengths
Weaknesses
Report
The manuscript titled "Interface-driven superconductivity in FeSe/SrTiO₃ from first-principles" reports a first-principles investigation of monolayer FeSe and FeSe/STO using a combined Kohn-Sham DFT and Bogoliubov-de Gennes (BdG) framework. The authors aim to elucidate how substrate effects influence the superconducting properties of FeSe through changes in electronic structure and orbital hybridization. While the methodology is technically ambitious and the problem is of high interest, the results and conclusions are built upon structural and electronic models that significantly deviate from experimental findings. As such, the reliability of the conclusions is seriously undermined. My major concerns include: 1. The model assumes a simple monolayer FeSe on a single TiO₂-terminated STO slab. However, experimental studies consistently report a double TiOₓ layer at the FeSe/STO interface. This structural motif plays a critical role in determining both the electronic doping and the interfacial phonon environment. The absence of this feature in the simulation compromises the relevance of both the normal and superconducting state calculations. 2. There are serious discrepancies between the calculated band structures and experimental ARPES observations: (1) In freestanding FeSe, the computed hole pocket along Γ–X is not observed experimentally in bulk FeSe or thick films, where instead a small elliptical hole pocket centered at Γ dominates. (2) In FeSe/STO, the calculation shows a hole pocket at M, which contradicts experimental data where only electron pockets at M are observed. These deviations indicate that the DFT base used to construct the BdG framework does not accurately represent the low-energy electronic structure, undermining the applicability of the resulting superconducting state analysis. 3. The superconducting state is introduced using a semi-empirical fixed-Δ method, where the pairing potential is manually initialized as spherical wells centered on Fe atoms. This approach is not derived self-consistently and lacks physical justification. Moreover, the large value of Δ (e.g., 73 meV for FeSe/STO) significantly exceeds typical experimental estimates and appears arbitrary. Without a rigorous derivation of the pairing potential from microscopic mechanisms (e.g., spin or phonon-mediated pairing), the results risk being model-dependent and speculative. 4. The claim that Fe dz2 orbitals dominate the superconducting pairing contradicts ARPES experiments and theoretical studies showing negligible dz2 weight near the Fermi surface. In FeSe/STO, the superconducting states are predominantly associated with dxz/dyz orbitals at the M point. The prominent role assigned to dz2 in the manuscript is thus not supported by existing evidence and casts doubt on the interpretation of the calculated SC gaps and coherence peaks. Although the authors employ an advanced theoretical framework and present detailed computational results, the fundamental assumptions regarding structure and electronic structure are inconsistent with experiments. As the BdG results are built directly upon these flawed foundations, the derived conclusions regarding superconducting gap structures, orbital character, and proximity effects are unlikely to be reliable. I cannot recommend publication of the manuscript in its current form. A substantial revision involving: (1) accurate structural modeling of the FeSe/STO interface, (2) consistent reproduction of the experimentally observed band structures, and (3) a more justified treatment of superconducting pairing, would be required before the scientific claims can be meaningfully evaluated.
Recommendation
Ask for major revision
Strengths
Weaknesses
Report
They drew several conclusions about the nature of the superconducting gap analyzing in terms of Fe d orbital character. They also emphasize the important role of Fe hybridizing with O in modifying the electronic structure, and thus the superconductivity.
All that is well and good, and the importance of Fe hybridizing with O is eminently reasonable. My problem with this paper, and it is a serious one, is that the model (DFT + assumed form of what they call the superconducting strength parameter entering into the BdG equations.
It has long been known that DFT predicts a very poor band structure of FeSe. Much, though by no means all of the failure is connected with the magnetism. It is also known that DFT isn't able to predict the structure well, especially the all-important height of Se above the Fe plane. See for example, https://doi.org/10.1103/PhysRevB.94.195146 . Much depends on that height, especially the coherence of the dxy orbital. In a DFT context, Zunger and coworkers were able to "save" it by making supercells and accounting (in the spectral function) for the all-important spin fluctuations statically, through spatial variations. See https://link.aps.org/doi/10.1103/PhysRevB.102.235121
This is all very important because the nature of the gap depends the position of the d orbitals relative to Ef, and it also depends strongly on the incoherence of the dxy orbital, which is completely absent in the theory. This is explained in some detail in the works by Acharya they cite.
Turning to the Fe/STO interface, it is also well known that DFT does a poor job of predicting the Fe(d)-O(p) alignment. Therefore any DFT study of the interface must be taken with a grain of salt, especially as a shift of 10 or so meV in a d level causes a significant change in the superconducting properties.
Finally, regarding the two particle sector, the title Ref 11 by Acharya reads "Vertex dominated superconductivity in intercalated fese." For them, the vertex is the all-important property, and the paper emphasizes its matrix structure and frequency dependence.
Since the level of theory there is vastly better than this work and covers the same system (Ref 9 address the Fe/STO interface as well), to me this work must be considered retrograde. What is new here is their computation of the supercondcting gap state from from the BdG equations. But to be credible, they should compare their results to the three papers by Acharya. It seems unlikely they will be similar, since DFT theory doesn't include the ingredients what seem to be essential when a many-body effects are taken into account.
In light of this, I do not think this paper deserves publication.
Requested changes
To make this work credible, the authors would have to explain a great deal about why the theory is relevant to this system.
Recommendation
Reject