REVIEW 4 major objections 5 minor 46 references
Spectroscopic Signatures of Nonlocal Interfacial Coupling in Superconducting FeSe/SrTiO3 Heterostructures
T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read A SrTiO3 substrate nonlocally alters the phonons and the effective electron mass of a superconducting FeSe film within about one nanometer of the interface.
desk verdict The gated Raman data are the real new result; the VEELS effective-mass claim is an acknowledged ambiguity that the abstract overstates. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is ambipolar phonon softening: the frequency of a Raman-active phonon decreases for both electron and hole doping by a gate voltage, which the paper observes in FeSe modes and in STO modes and attributes to interlayer coupling rather than to doping alone. The second leg is the spatial map of the valence plasmon, whose frequency depends on the ratio of electron density to effective mass; the observed red-shift near the interface is the observable that would reveal an effective-mass enhancement if the density is not depleted. Together the two spectroscopies tie a phonon-mediated interfacial effect to the superconducting layers.
What would settle it
Measure the local carrier density across the FeSe/STO interface independently of the plasmon frequency, for example with spatially resolved Hall or quantum-oscillation measurements on patterned one-unit-cell devices, and compare the density profile with the VEELS red-shift. If the red-shift persists where the density is flat, the effective-mass enhancement is supported; if the red-shift tracks a density dip, the red-shift is electron depletion and the central electron-side claim fails.
Extended reading notes
Core claim
The central claim is that the STO substrate exerts a nonlocal interfacial coupling on FeSe: voltage-gated Raman shows the FeSe Eg, A1g, and B1g phonons soften ambipolarly with gate voltage just as the STO modes do, while FeTe modes in the same stack remain almost unaffected; and VEELS shows both STO and FeSe plasmon peaks red-shift as the interface is approached, opposite to the slight blue-shift expected from dielectric theory. The authors propose that the FeSe red-shift is at least partly a local enhancement of the FeSe electron effective mass mediated by STO phonons, confined to a roughly one-nanometer region comparable to the first two FeSe layers. They present this as evidence that interfacial electron-phonon coupling, not just charge transfer, contributes to the enhanced superconductivity of 1UC FeSe/STO.
Load-bearing premise
The load-bearing premise is that the red-shift of the FeSe valence-electron-loss peak near the interface reflects a heavier electron effective mass, not a drop in local electron density; the paper itself says additional experiments are needed to tell these apart, and if the shift is depletion the electron-side evidence for interfacial phonon coupling collapses.
Editorial extensions
If this is right
- The STO substrate actively participates in FeSe lattice dynamics, so models that treat it as an inert template omit a relevant interaction.
- The interfacial response is confined to roughly one nanometer, matching the FeSe thickness range that thickness-dependent studies associate with superconductivity; the coupling is therefore localized in the layers that matter for Tc.
- A positive backgate slightly strengthens the FeSe B1g softening, indicating that the interfacial coupling can be tuned by drawing electrons to the FeSe/STO interface.
- The observed red-shift, if interpreted as effective-mass enhancement, provides a microscopic route by which STO phonons could raise the superconducting transition temperature of 1UC FeSe.
- Low-temperature gated Raman and VEELS can be applied to other heterostructures to reveal similar nonlocal substrate couplings.
Reading between the lines
- Inference: applying the same two-probe protocol to FeSe on other polar oxide substrates would test whether the nonlocal coupling is a general property of such interfaces or specific to STO.
- Inference: a density-resolved measurement, for example local Hall or quantum oscillations, could separate the two explanations of the red-shift; the paper's own interpretation predicts the red-shift survives with density held constant.
- Inference: because the STO modes themselves soften under gate voltage via piezoelectric strain, a control experiment on a non-piezoelectric substrate would reveal how much of the FeSe softening is strain-mediated rather than electronically mediated.
- Inference: if the coupling is indeed confined to the first layer, thickness-dependent Tc studies should show a crossover dominated by the interface layer rather than a uniform enhancement across the film.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports low-temperature, voltage-gated Raman spectroscopy and low-temperature valence electron energy loss spectroscopy (VEELS) measurements on FeTe-capped one-unit-cell (1UC) and eight-unit-cell (8UC) FeSe films grown on SrTiO3 (STO) substrates. The Raman measurements show that several FeSe vibrational modes soften symmetrically for both positive and negative backgate voltages, and the authors note that the STO substrate modes exhibit similar ambipolar softening, which they attribute to an inverse piezoelectric effect in STO that is transmitted to the FeSe layer through interfacial coupling. The VEELS line scans across the FeSe/STO interface reveal a red-shift of both the STO and FeSe plasmon peaks relative to their bulk positions, and the authors propose that this red-shift may reflect an enhancement of the FeSe electron effective mass mediated by interfacial STO phonons. The paper claims that these observations support a nonlocal interfacial coupling between STO and FeSe lattice and electronic degrees of freedom, potentially relevant to the enhanced superconductivity of monolayer FeSe on STO.
Significance. If the observations and their interpretation hold, the paper would provide new mode-resolved and spatially resolved spectroscopic evidence about interfacial coupling in the FeSe/STO system, a central model system for interface-enhanced superconductivity. The raw measurements appear to be new and technically challenging: gated Raman spectroscopy at 4 K on FeTe-capped monolayer FeSe, and low-temperature VEELS with spatial resolution across the interface. The analysis introduces no fitted free parameters and compares against standard dielectric-function formulas, which is a strength. The Raman softening of FeSe modes mirroring STO modes is an interesting empirical finding. However, the central electron-side conclusion, effective mass enhancement inferred from the VEELS red-shift, is explicitly acknowledged in the text to be ambiguous and not uniquely determined by the data, and the abstract overstates this result. The significance therefore rests on the Raman pillar and on the possibility that the VEELS red-shift is a genuine interfacial electronic effect; further experiments are required to disambiguate the interpretation.
major comments (4)
- [Abstract and VEELS results (Figure 3 and following paragraph)] The abstract and the concluding paragraph state that the FeSe electron effective mass enhancement is 'determined from the red-shift in the FeSe VEELS spectrum.' However, the main text explicitly acknowledges that the observed red-shifts 'could be attributed either to local depletion in electron concentration or to local enhancement of the electron effective mass' and that 'additional experiments would be necessary to further clarify this point.' Since the effective-mass interpretation is the electron-side evidence for interfacial electron-phonon coupling, this ambiguity is load-bearing. The paper should either soften the abstract to match the acknowledged ambiguity, or provide an independent measurement of the local carrier density (for example, core-level EELS as in Ref. [33]) to justify the effective-mass assignment.
- [Methods, Eqs. (1)-(3) and the VEELS analysis] Eq. (3) is presented as determining the variation of plasma frequency 'caused by the change in electron density,' and it assumes a fixed effective mass because ΔN is proportional to ω_pe^2 - ω_pc^2 with the bare electron mass m. Yet the paper's central VEELS interpretation is a local change in effective mass, not density. The analysis therefore conflates density and effective-mass effects: the red-shift is used qualitatively to infer mass enhancement, while the formula that would quantify it assumes the mass is unchanged. This internal inconsistency needs to be addressed, for example by writing the full expression with both N(x) and m*(x) and stating what assumptions are needed to extract either quantity.
- [Gated Raman spectroscopy (Figure 2 and Figure S3)] The reported FeSe peak positions are extracted from raw spectra by locating zeros and quasi-discontinuities of the first derivative (Figure 2c), and the peak-position versus gate-voltage curves in Figure 2d are presented without error bars or uncertainties from the Lorentzian fits shown in Figure S3. The ambipolar softening spans only a few cm^-1, so without uncertainty quantification it is not possible to assess whether the effect is statistically significant or to compare quantitatively the positive- and negative-bias asymmetries. The same applies to the STO mode positions in Figure S2c. Adding error bars, or at least a discussion of the measurement precision, is essential for the central Raman claim.
- [VEELS comparison with calculated spectra (Methods, Ref. [43])] The comparison between experimental and 'calculated' VEELS spectra relies on a dielectric model (Moreau et al., Ref. [43]) that assumes the FeSe/STO interface is an ideal geometrical plane with a step-function dielectric response. The authors themselves note that the actual intermediate layer cannot be sharp due to charge transfer. A graded interface can produce interface-loss features or apparent plasmon shifts that are not captured by the step-function reference calculation. Therefore, the observed red-shift could be an artifact of the graded interface rather than a change in the bulk plasma frequency of FeSe. To support the claim that the red-shift is an intrinsic FeSe property, the analysis should compare against a graded-interface model or provide an independent local probe of the carrier density and dielectric profile.
minor comments (5)
- [Methods, Eq. (1)] The symbol δε0 is used in Eq. (1) but appears as δ0 in the OCR text; please ensure the notation is consistent and define all symbols in the text surrounding the equation.
- [Introduction, first paragraph] The phrase 'the absence of hole pockets in intercalated or 1UC FeSe' is stated without qualification. Some ARPES studies have reported small hole-like bands in related systems; a more nuanced formulation with citations would avoid overgeneralization.
- [Figure 1 and Figure S1] The temperature-dependent Raman peak positions for FeSe and FeTe modes are plotted only in the lower panels of Figure S1, but the main text refers to them in a way that suggests a quantitative discussion; please state explicitly what change with temperature is observed and how it compares with bulk values.
- [Transport section (Figure S5)] The transport data are reproduced from Ref. [S1] without a description of the measurement uncertainty or the criteria used to define Tc. A brief statement about how Tc was determined (onset, midpoint, zero resistance) would improve the reproducibility of the superconductivity claim.
- [General notation] The paper uses '1UC' and '8UC' without defining 'UC' in the main text; please spell out 'unit cell' at first use for readers outside the immediate field.
Circularity Check
No significant circularity: the Raman and VEELS results are direct measurements analyzed with standard formulas, and the effective-mass interpretation is explicitly presented as one of two alternatives rather than a fitted prediction.
full rationale
The paper's central claims are new experimental observations, not outputs of a fitted model. The gated Raman measurements are raw spectra whose peak positions are read from first-derivative features, and the ambipolar softening is observed directly rather than being derived from an input assumption. The VEELS analysis compares measured plasmon peaks against a dielectric-function model with an assumed step interface; the model has no free parameters adjusted to reproduce the observed red-shift, so the discrepancy is a genuine model-data comparison rather than a quantity equivalent to its input. Equation (3) only converts a measured plasma-frequency shift into a density variation, and the paper explicitly states that the FeSe red-shift could be due to local depletion or to effective-mass enhancement and that additional experiments are needed; thus the effective-mass interpretation is an admitted ambiguity, not an input-derived conclusion. The only self-citation, [33]/[S1] by overlapping authors (W. Zhao, C.-Z. Chang), is used to confirm superconductivity and to support a secondary interpretation of the backgate asymmetry; it is not the source of the Raman or VEELS data and does not force the central claim. The abstract's phrasing that effective-mass enhancement is 'determined' overstates the strength of the VEELS inference, but this is an epistemic overreach, not a circular derivation. Accordingly, the derivation chain is self-contained and no circular step is present.
Assumptions & free parameters
assumptions (4)
- standard math Plasma frequency formula (Eq. 1) and density-variation formula (Eq. 3) apply to the measured VEELS spectra, and the effective electron number N_eff captures all relevant valence electrons.
- domain assumption The dielectric model of Moreau et al. [43] with a step-function interface and bulk dielectric constants produces a reliable calculated VEELS spectrum.
- domain assumption The observed FeSe Raman peak shifts are due to lattice or electronic changes in FeSe rather than spectral artifacts, and the gate-induced shifts are mediated by the STO substrate.
- domain assumption Superconductivity persists at all gate voltages and temperatures used in the measurements.
Cite this review
Pith. "Pith review of Spectroscopic Signatures of Nonlocal Interfacial Coupling in Superconducting FeSe/SrTiO3 Heterostructures." pith.science (2026). https://pith.science/paper/NR6YQYRW
@misc{pith2026190805648,
author = {Pith},
title = {Pith review of: Spectroscopic Signatures of Nonlocal Interfacial Coupling in Superconducting FeSe/SrTiO3 Heterostructures},
year = {2026},
howpublished = {\url{https://pith.science/paper/NR6YQYRW}},
note = {Machine review of arXiv:1908.05648}
}
read the original abstract
The mechanism of enhanced superconductivity in the one unit-cell (1UC) FeSe film on a SrTiO3 (STO) substrate has stimulated significant research interest but remains elusive. Using low-temperature, voltage-gated Raman spectroscopy and low-temperature valence electron energy loss spectroscopy (VEELS), we characterize the phonon behavior and interfacial charge transfer in single- and few-layer FeSe films on STO. Raman measurements reveal ambipolar softening of the FeSe vibrational modes, mimicking that of the underlying STO substrate. We attribute this behavior to an interfacial coupling effect of STO on FeSe lattice dynamics. This interfacial coupling effect is further supported by local electron effective mass enhancement, which is determined from the red-shift in the FeSe VEELS spectrum near the FeSe/STO interface. Our work sheds light on the possible interfacial mechanisms contributing to the enhanced superconductivity across the FeSe/STO interface and further unveils the potential of low-temperature gated Raman spectroscopy and VEELS in clarifying a broad category of quantum materials.
Figures
Reference graph
Works this paper leans on
-
[33]
Zhao et al., Science Advances 4 (2018), 10.1126/sciadv.aao2682
W. Zhao et al., Science Advances 4 (2018), 10.1126/sciadv.aao2682
- [43]
-
[1]
Kamihara, T
Y. Kamihara, T. Watanabe, M. Hirano, and H. Hosono, Journal of the American Chemical Society 130, 3296 (2008)
2008
-
[2]
X. H. Chen et al., Nature 453, 761 (2008)
work page 2008
-
[3]
G. F. Chen et al., Phys. Rev. Lett. 100, 247002 (2008)
work page 2008
-
[4]
Z. A. Ren et al., Chinese Physics Letters 25, 2215 (2008)
work page 2008
-
[5]
Wang et al., EPL (Europhysics Letters) 83, 67006 (2008)
C. Wang et al., EPL (Europhysics Letters) 83, 67006 (2008)
work page 2008
-
[6]
Cheng et al., EPL (Europhysics Letters) 85, 67003 (2009)
P. Cheng et al., EPL (Europhysics Letters) 85, 67003 (2009)
work page 2009
Show all 46 references
-
[7]
de la Cruz et al., Nature 453, 899 (2008)
C. de la Cruz et al., Nature 453, 899 (2008)
2008
-
[8]
Dong et al., EPL (Europhysics Letters) 83, 27006 (2008)
J. Dong et al., EPL (Europhysics Letters) 83, 27006 (2008)
2008
-
[9]
Paglione and R
J. Paglione and R. L. Greene, Nature Physics 6, 645 (2010)
2010
-
[10]
Ding et al., EPL (Europhysics Letters) 83, 47001 (2008)
H. Ding et al., EPL (Europhysics Letters) 83, 47001 (2008)
2008
-
[11]
I. I. Mazin, D. J. Singh, M. D. Johannes, and M. H. Du, Phys. Rev. Lett. 101, 057003 (2008)
2008
-
[12]
Kuroki et al., Phys
K. Kuroki et al., Phys. Rev. Lett. 101, 087004 (2008)
2008
-
[13]
Qian et al., Phys
T. Qian et al., Phys. Rev. Lett. 106, 187001 (2011)
2011
-
[14]
Zhang et al., Nature Materials 10, 273 (2011)
Y. Zhang et al., Nature Materials 10, 273 (2011)
2011
-
[15]
Mou et al., Phys
D. Mou et al., Phys. Rev. Lett. 106, 107001 (2011)
2011
-
[16]
Liu et al., Nature Communications 3, 931 (2012)
D. Liu et al., Nature Communications 3, 931 (2012)
2012
-
[17]
He et al., Nature Materials 12, 605 (2013)
S. He et al., Nature Materials 12, 605 (2013)
2013
-
[18]
Tan et al., Nature Materials 12, 634 (2013)
S. Tan et al., Nature Materials 12, 634 (2013)
2013
-
[19]
X. H. Niu et al., Phys. Rev. B 92, 060504 (2015)
2015
-
[20]
Zhao et al., Nature Communications 7, 10608 (2016)
L. Zhao et al., Nature Communications 7, 10608 (2016)
2016
-
[21]
Wang et al., Chinese Physics Letters 29, 037402 (2012)
Q.-Y. Wang et al., Chinese Physics Letters 29, 037402 (2012)
2012
-
[22]
L. Wang, X. Ma, and Q.-K. Xue, Superconductor Science and Technology 29, 123001 (2016)
2016
-
[23]
Zhang et al., Phys
W. Zhang et al., Phys. Rev. B 89, 060506 (2014)
2014
-
[24]
Peng et al., Nature Communications 5, 5044 (2014)
R. Peng et al., Nature Communications 5, 5044 (2014)
2014
-
[25]
Li et al., Journal of Applied Physics 115, 193907 (2014)
B. Li et al., Journal of Applied Physics 115, 193907 (2014)
2014
-
[26]
Xia et al., Phys
T.-L. Xia et al., Phys. Rev. B 79, 140510 (2009)
2009
-
[27]
A. S. Pine and G. Dresselhaus, Phys. Rev. B 4, 356 (1971)
1971
-
[28]
Kumar et al., Solid State Communications 150, 557 (2010)
P. Kumar et al., Solid State Communications 150, 557 (2010)
2010
-
[29]
Gnezdilov et al., Phys
V. Gnezdilov et al., Phys. Rev. B 87, 144508 (2013)
2013
-
[30]
Zhang et al., Phys
S. Zhang et al., Phys. Rev. B 97, 035408 (2018)
2018
-
[31]
W. G. Nilsen and J. G. Skinner, The Journal of Chemical Physics 48, 2240 (1968)
1968
-
[32]
S. Coh, M. L. Cohen, and S. G. Louie, New Journal of Physics 17, 073027 (2015)
2015
-
[34]
F. R. Aussenegg and M. E. Lippitsch, Journal of Raman Spectroscopy 17, 45 (1986)
1986
-
[35]
P. A. Fleury and J. M. Worlock, Phys. Rev. 174, 613 (1968)
1968
-
[36]
Petzelt et al., Phys
J. Petzelt et al., Phys. Rev. B 64, 184111 (2001)
2001
-
[37]
D. E. Grupp and A. M. Goldman, Science 276, 392 (1997)
1997
-
[38]
Zhang et al., Phys
S. Zhang et al., Phys. Rev. Lett. 122, 066802 (2019)
2019
-
[39]
Wang et al., 2D Materials 2, 044012 (2015)
Q. Wang et al., 2D Materials 2, 044012 (2015)
2015
-
[40]
Z. Wang, C. Liu, Y. Liu, and J. Wang, Journal of Physics: Condensed Matter 29, 153001 (2017)
2017
-
[41]
C. Wen et al., Nature Communications 7, 10840 (2016) Methods Sample Growth and Preparation The FeSe films used in this work w ere grown in a n ultrahigh vacuum (UHV) MBE chamber with a base pressure lower than 5×10 -10 mbar. The insulating STO(001) substrates were heat -treate...
2016
-
[42]
Egerton, Electron Energy-Loss Spectroscopy in the Electron Microscope (Springer, 2011)
R. Egerton, Electron Energy-Loss Spectroscopy in the Electron Microscope (Springer, 2011)
2011
-
[44]
Nakagawa, H
N. Nakagawa, H. Y. Hwang, and D. A. Muller, Nature Materials5, 204 (2006)
2006
-
[45]
Pines, Elementary Excitations in Solids (Benjamin,1963)
D. Pines, Elementary Excitations in Solids (Benjamin,1963)
1963
-
[46]
H. R. Philipp and H. Ehrenreich, Phys. Rev.129, 1550(1963). Acknowledgements N.A. acknowledges the support of the National Science Foundation Graduate Research Fellowship Program under Grant No. 1122374. T.N. thanks the support from the MIT SMA -2 Fellowship Program. C.Z.C. ac...
1963
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.