Layer-dependent superconductivity in iron-based superconductors

Ke Meng, Xu Zhang, Boqin Song, Bai-Zhuo Li, Xiangming Kong, Sicheng Huang, Xiaofan Yang, Xiaobo Jin, Yiyuan Wu, Jiaying Nie, Guang‐Han Cao, Shiyan LiarXiv 2023 · DOI · arXiv

TL;DR

Researchers have uncovered a "rule of thumb" for how superconductivity fades as materials are thinned down to a single atomic layer. By studying iron-based superconductors, they established an empirical law that predicts the drop in transition temperature ($T_c$) based on how much the material's electron pairs "leak" between layers.

What Problem Does This Paper Address?

In the world of condensed matter physics, the Hohenberg-Mermin-Wagner theorem is a bit of a killjoy: it suggests that truly two-dimensional (2D) systems shouldn't be able to maintain long-range order (like superconductivity) at temperatures above absolute zero.

However, experimental reality is inconsistent. When you thin the superconductor $NbSe_2$ down to a monolayer, its $T_c$ plummets by 60%. Yet, the cuprate superconductor $Bi-2212$ shows almost no change in $T_c$ when reduced to a single layer. This discrepancy leaves a massive question: What actually dictates how dimensionality affects superconductivity? Is it the material's chemical makeup, its internal "connectedness" (anisotropy), or something else entirely?

Methods

The team focused on two iron-based superconductors (IBS) with very different levels of anisotropy: $CsCa_2Fe_4As_4F_2$ (Cs-12442), which is highly layered and anisotropic, and $CaKFe_4As_4$ (1144), which has much stronger coupling between its layers.

  • $Al_2O_3$-assisted Exfoliation: Standard "Scotch tape" methods often fail for these brittle materials. The authors used a more advanced technique, evaporating Aluminum in an Oxygen environment to create a sticky $Al_2O_3$ film that helps peel off large, intact monolayers.
  • Ionic Liquid Gating: Thinning materials often causes surface atoms (like Cesium) to escape, which changes the "doping" (electron/hole concentration) and can artificially lower $T_c$. To fix this, the researchers used an ionic liquid "gate" to inject protons ($H^+$) into the monolayer, tuning it back to its optimal electronic state. This ensured the observed $T_c$ changes were due to physics, not just chemical degradation.
  • Transport Measurements: They used four-probe electrical resistance measurements to track the superconducting transition and confirmed 2D behavior by identifying the Berezinskii-Kosterlitz-Thouless (BKT) transition, a hallmark of 2D physics.

Datasets

This is an experimental physics paper rather than a data-science study. The "dataset" consists of:

  • Original transport data (Resistance vs. Temperature) for Cs-12442 (monolayer to bulk) and CaKFe4As4 (3-layer to bulk).
  • Comparative literature data for $Bi-2212$, $NbSe_2$, and $FeSe$ to validate their proposed empirical law.

Key Insights and Ideas

  • The "Connectedness" Factor: The defining feature isn't just being "thin"; it's the c-axis coherence length ($\xi_c$)—essentially how far a superconducting Cooper pair "reaches" vertically.
  • Effective Dimensionality: In $Bi-2212$, the coherence length is so tiny ($<0.6$ Å) compared to the layer thickness that even the "bulk" crystal behaves like a stack of independent 2D sheets. That’s why thinning it further doesn't change the $T_c$.
  • The Logarithmic Law: The authors discovered a beautiful empirical relationship: the relative drop in $T_c$ is proportional to the logarithm of the ratio between the vertical coherence length and the monolayer thickness ($-\Delta T_c/T_c^{bulk} \sim \log(\xi_c/d_{monolayer})$).

Key Results

  • Cs-12442 Monolayer: Successfully fabricated and tuned to an optimal $T_c$ of 24 K, which is about 20% lower than its bulk counterpart (30.5 K).
  • CaKFe4As4 (1144): Because this material has stronger interlayer coupling (larger $\xi_c$), the dimensional effect was much harsher. A 3-layer flake saw a 46% drop in $T_c$.
  • Universal Trend: The proposed empirical law held true for a variety of quasi-2D superconductors, provided their anisotropy (ratio of vertical to horizontal resistance) was roughly greater than 100.

Limitations and Future Work

  • Low Anisotropy Limit: The empirical law breaks down for materials with very low anisotropy (like $FeSe$), where the interlayer coupling is so strong that the "layered" model no longer applies simply.
  • Substrate Interference: The study notes that external factors like the substrate (e.g., growing $FeSe$ on $SrTiO_3$) can introduce new effects like interfacial phonon coupling that override the intrinsic dimensional effects discussed here.
  • Predictive Power: The authors suggest this law can now be used to extrapolate the $T_c$ of new monolayer superconductors before they are even fabricated, which is a major boon for the field of van der Waals heterostructures.

Summary updated

Abstract

The Hohenberg-Mermin-Wagner theorem states that a two-dimensional system cannot spontaneously break a continuous symmetry at finite temperature. This is supported by the observation of layer-dependent superconductivity in the quasi-two-dimensional superconductor NbSe2, in which the superconducting transition temperature (Tc) is reduced by about 60% in the monolayer limit. However, for the extremely anisotropic copper-based high-Tc superconductor Bi2Sr2CaCu2O8+δ (Bi-2212), the Tc of the monolayer is almost identical to that of its bulk counterpart. To clarify the effect of dimensionality on superconductivity, here we successfully fabricate ultrathin flakes of CsCa2Fe4As4F2, a highly anisotropic iron-based high-Tc superconductor, down to monolayer. The monolayer flake exhibits the highest Tc of 24 K (after tuning to the optimal doping by ionic liquid gating), which is about 20% lower than that of the bulk crystal. We also fabricate ultrathin flakes of CaKFe4As4, another iron-based superconductor with much smaller anisotropy. The Tc of the 3-layer flake decreases by 46%, showing a more pronounced dimensional effect than that of CsCa2Fe4As4F2. By carefully examining their anisotropy and the c-axis coherence length, we reveal the general trend and empirical law of the layer-dependent superconductivity in these quasi-two-dimensional superconductors. From this, the Tc of a new monolayer superconductor can be extrapolated.
BibTeX
@misc{meng2023layer,
  author = {Ke Meng and Xu Zhang and Boqin Song and Bai-Zhuo Li and Xiangming Kong and Sicheng Huang and Xiaofan Yang and Xiaobo Jin and Yiyuan Wu and Jiaying Nie and Guang‐Han Cao and Shiyan Li},
  title = {{Layer-dependent superconductivity in iron-based superconductors}},
  year = {2023},
  eprint = {2311.14559},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2311.14559},
}
Figure