pith. sign in

arxiv: 2605.26837 · v1 · pith:EYZID5U3new · submitted 2026-05-26 · ❄️ cond-mat.mes-hall · cond-mat.mtrl-sci

Terahertz spin-current transparency through rough interfaces

Ji\v{r}\'i Jechumt\'al , Jakub Z\'azvorka , Ond\v{r}ej Nov\'ak , Martin Rejhon , Peter Kuba\v{s}\v{c}\'ik , Luk\'a\v{s} Nowak , Petr N\v{e}mec , Eva Schmoranzerov\'a
show 3 more authors
Martin Veis Luk\'a\v{s} N\'advorn\'ik Zden\v{e}k Ka\v{s}par
This is my paper

Pith reviewed 2026-06-29 16:17 UTC · model grok-4.3

classification ❄️ cond-mat.mes-hall cond-mat.mtrl-sci
keywords spin transportterahertz emissioninterface roughnessCoPt heterostructuresspin current transparencyultrafast spin dynamicsspintronics
0
0 comments X

The pith

Spin-current transparency at Co|Pt interfaces falls only 30 percent even when roughness and grain size triple.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that spin transport across interfaces stays largely intact on ultrafast timescales despite substantial morphological disorder. Researchers tune interface roughness in Co|Pt stacks by varying an Au buffer layer while holding other growth conditions fixed. They measure the emitted terahertz electric field and extract the spin-current transparency ts after correcting for impedance and absorption changes in the full stack. The result shows only a modest drop in ts and no spectral shift, indicating that rough interfaces do not strongly hinder spin flow at these speeds. This matters for spintronic devices because real interfaces are rarely atomically flat yet must still carry spin information quickly.

Core claim

In Co|Pt heterostructures, the interface spin-current transparency ts decreases by approximately 30 percent as root-mean-square roughness and lateral grain size each increase by a factor of three, with no measurable change in the terahertz spectrum, after corrections for sample impedance and optical absorption.

What carries the argument

the extracted interface spin-current transparency ts, obtained from terahertz emission spectroscopy after impedance and absorption corrections

If this is right

  • Interfacial spin transport remains relatively robust against morphological variations on ultrafast timescales.
  • Terahertz emission spectroscopy can serve as a reliable probe of spin dynamics even across imperfect interfaces.
  • Device performance in spintronic applications may tolerate moderate increases in interface roughness without major loss of ultrafast spin transmission.
  • The terahertz spectrum itself carries no signature of the roughness-induced changes in transparency.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Fabrication processes that allow some roughness could still yield functional ultrafast spintronic elements if the 30 percent drop is acceptable.
  • The lack of spectral change suggests that roughness does not introduce strong frequency-dependent scattering in the terahertz range.
  • Similar measurements on other material pairs could test whether the observed robustness is specific to Co|Pt or more general.

Load-bearing premise

The corrections for changes in sample impedance and optical absorption isolate the interface spin-current transparency without leftover effects from the Au buffer layer or other layers.

What would settle it

An independent measurement of spin current magnitude, obtained without using the impedance or absorption corrections, that shows a drop larger than 30 percent when roughness triples.

Figures

Figures reproduced from arXiv: 2605.26837 by Eva Schmoranzerov\'a, Jakub Z\'azvorka, Ji\v{r}\'i Jechumt\'al, Luk\'a\v{s} N\'advorn\'ik, Luk\'a\v{s} Nowak, Martin Rejhon, Martin Veis, Ond\v{r}ej Nov\'ak, Peter Kuba\v{s}\v{c}\'ik, Petr N\v{e}mec, Zden\v{e}k Ka\v{s}par.

Figure 1
Figure 1. Figure 1: FIG. 1: THz emission from FM [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: (a) Raw THz signals measured by electro-optical sampling from a set of samples Au(5) [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Sample properties as a function of the Au [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Spin transparency [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
read the original abstract

Spin transport across interfaces is critical for spintronic devices, yet remains difficult to probe on ultrafast timescales. We use terahertz emission spectroscopy on Co|Pt heterostructures whose interface roughness is tuned through the thickness of an underlying Au buffer layer, while leaving other growth parameters unchanged. From the measured THz electric field, we extract the interface spin-current transparency ts after correcting for the changes in sample impedance and optical absorption of the stack. Surprisingly, we find that ts decreases by only approximately 30% as the interface root-mean-square roughness and the lateral grain size both increase by a factor of three, with no measurable change in the THz spectrum. These results demonstrate that interfacial spin transport is relatively robust against morphological variations on ultrafast timescales, establishing terahertz emission spectroscopy as a reliable probe of spin dynamics across imperfect interfaces.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript reports an experimental study of spin-current transparency ts across Co|Pt interfaces whose roughness is varied by changing the thickness of an underlying Au buffer layer. Using terahertz emission spectroscopy, the authors measure the emitted THz electric field, apply corrections for sample impedance and optical absorption of the stack, and extract ts. They find that ts decreases by only ~30% when both RMS roughness and lateral grain size increase by a factor of three, with no measurable change in the THz spectrum, concluding that interfacial spin transport is relatively robust to morphological variations on ultrafast timescales.

Significance. If the impedance and absorption corrections are shown to isolate the interface contribution without significant residuals from the Au buffer, the result would establish THz emission spectroscopy as a practical probe of ultrafast spin dynamics at imperfect interfaces and indicate that spin transport remains efficient despite substantial roughness changes. This would be relevant for spintronic device design where interface morphology is difficult to control perfectly.

major comments (2)
  1. [Abstract and ts extraction section] Abstract and the section describing ts extraction: the claim that corrections for impedance and optical absorption isolate an interface-only ts rests on the unverified assumption that Au-buffer contributions to THz generation and propagation are fully subtracted. No quantitative validation (control samples with fixed roughness but varied Au thickness, or error propagation through the corrections) is described, which directly affects the reliability of the reported 30% trend.
  2. [Results section] Results section (the paragraph reporting the 30% decrease): the central claim that ts changes by only ~30% for a 3× roughness/grain-size increase lacks reported uncertainties, statistical significance tests, or sensitivity analysis showing that this change exceeds possible systematic residuals from the Au layer. Without these, it is unclear whether the observed trend supports the robustness conclusion.
minor comments (2)
  1. Notation for the transparency parameter should be standardized (e.g., consistently use t_s with subscript) across text, figures, and equations.
  2. Figure captions should explicitly state the number of independent samples or measurements averaged for each roughness condition to allow assessment of reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful and constructive review. The two major comments both concern the validation and statistical presentation of the ts extraction procedure. We address each below and will incorporate the requested quantitative checks in a revised manuscript.

read point-by-point responses

  1. Referee: [Abstract and ts extraction section] Abstract and the section describing ts extraction: the claim that corrections for impedance and optical absorption isolate an interface-only ts rests on the unverified assumption that Au-buffer contributions to THz generation and propagation are fully subtracted. No quantitative validation (control samples with fixed roughness but varied Au thickness, or error propagation through the corrections) is described, which directly affects the reliability of the reported 30% trend.

    Authors: We agree that the current manuscript does not contain an explicit error-propagation analysis or control-sample series that isolates the Au-buffer contribution at fixed roughness. The corrections themselves are performed using measured sheet resistance and optical transmission/reflection data for each complete stack; however, we did not propagate the uncertainties in those measured quantities through to ts nor test residual Au-driven generation. In the revision we will add both an error-propagation calculation and a sensitivity analysis that assumes plausible residual THz generation or propagation effects from the Au layer. These additions will directly quantify the uncertainty in the extracted ts values. revision: yes

  2. Referee: [Results section] Results section (the paragraph reporting the 30% decrease): the central claim that ts changes by only ~30% for a 3× roughness/grain-size increase lacks reported uncertainties, statistical significance tests, or sensitivity analysis showing that this change exceeds possible systematic residuals from the Au layer. Without these, it is unclear whether the observed trend supports the robustness conclusion.

    Authors: The manuscript reports the ~30% reduction without accompanying uncertainties or statistical tests. We will revise the Results section to include (i) error bars derived from the standard deviation across multiple independently grown samples at each Au thickness, (ii) a simple statistical comparison confirming the trend exceeds the combined random and estimated systematic uncertainty, and (iii) the sensitivity analysis described in the response to the first comment. These changes will allow a reader to judge whether the observed variation remains significant once plausible Au residuals are taken into account. revision: yes

Circularity Check

0 steps flagged

No circularity: purely experimental extraction with independent corrections

full rationale

The paper reports direct THz emission measurements on Co|Pt stacks where interface roughness is varied via Au buffer thickness. The quantity ts is obtained by applying impedance and absorption corrections to the measured electric field; no equations, ansatzes, or self-citations are invoked that would make ts equivalent to its own inputs by construction. The result is therefore a measurement against external data rather than a self-referential derivation.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Abstract-only review limits visibility into parameters and assumptions; main domain assumption is that buffer-layer thickness affects only roughness.

free parameters (1)
  • impedance and optical absorption corrections
    Extraction of ts requires these corrections whose exact functional form and any fitted constants are not specified in the abstract.
axioms (1)
  • domain assumption Changes in Au buffer thickness affect only interface roughness and leave other growth parameters and intrinsic spin properties unchanged.
    Explicitly stated in the abstract as the basis for isolating roughness effects.

pith-pipeline@v0.9.1-grok · 5752 in / 1220 out tokens · 47959 ms · 2026-06-29T16:17:37.029298+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

40 extracted references · 2 canonical work pages

  1. [1]

    Wunderlichet al., Science330, 6012 (2010)

    J. Wunderlichet al., Science330, 6012 (2010)

    2010

  2. [2]

    Wanget al., Nature Communications14, 1068 (2022)

    D. Wanget al., Nature Communications14, 1068 (2022)

    2022

  3. [3]

    Kanget al., npj Spintronics3, 8 (2025)

    M.-G. Kanget al., npj Spintronics3, 8 (2025)

    2025

  4. [4]

    V. D. Nguyenet al., npj Spintronics2, 48 (2024)

    2024

  5. [5]

    V. P. Amin and M. D. Stiles, Phys. Rev. B94, 104419 (2016)

    2016

  6. [6]

    Guet al., Physical review B113, 134402 (2026)

    N. Guet al., Physical review B113, 134402 (2026)

    2026

  7. [7]

    Jinet al., Applied Materials & Interfaces11, 35458 (2019)

    L. Jinet al., Applied Materials & Interfaces11, 35458 (2019)

    2019

  8. [8]

    Rojas-S´ anchezet al., Phys

    J.-C. Rojas-S´ anchezet al., Phys. Rev. Lett.112, 106602 (2014)

    2014

  9. [9]

    Guptaet al., Disorder dependence of interface spin memory loss (2020)

    K. Guptaet al., Disorder dependence of interface spin memory loss (2020)

    2020

  10. [10]

    Zhuet al., Physical review B99, 180404 (2019)

    L. Zhuet al., Physical review B99, 180404 (2019)

    2019

  11. [11]

    Mazinet al., Physical review B84, 014425 (2011)

    I. Mazinet al., Physical review B84, 014425 (2011)

    2011

  12. [12]

    Mukhopadhyayet al., npg Asia Materials15, 3835 (2023)

    S. Mukhopadhyayet al., npg Asia Materials15, 3835 (2023)

    2023

  13. [13]

    S. N. Pandaet al., Science Advances5, eaav7200 (2019), https://www.science.org/doi/pdf/10.1126/sciadv.aav7200

    work page doi:10.1126/sciadv.aav7200 2019

  14. [14]

    Jungwirthet al., Nature Nanotechnology11, 231 (2016)

    T. Jungwirthet al., Nature Nanotechnology11, 231 (2016)

    2016

  15. [15]

    Baltzet al., Rev

    V. Baltzet al., Rev. Mod. Phys.90, 015005 (2018)

    2018

  16. [16]

    Smejkalet al., Physical Review X12, 040501 (2022)

    L. Smejkalet al., Physical Review X12, 040501 (2022)

    2022

  17. [17]

    Seifertet al., Nature Photon10, 483 (2016)

    T. Seifertet al., Nature Photon10, 483 (2016)

    2016

  18. [18]

    Jechumt´ alet al., Phys

    J. Jechumt´ alet al., Phys. Rev. Lett.132, 226703 (2024)

    2024

  19. [19]

    T. S. Seifertet al., Appl. Phys. Lett.120, 180401 (2022)

    2022

  20. [20]

    D. M. Nennoet al., Sci. Rep.9, 13348 (2019)

    2019

  21. [21]

    Liet al., Phys

    G. Liet al., Phys. Rev. Mater.3, 084415 (2019)

    2019

  22. [22]

    Wanget al., Comm

    Y. Wanget al., Comm. Phys.6, 280 (2023)

    2023

  23. [23]

    Scheueret al., iScience25, 104319 (2022)

    L. Scheueret al., iScience25, 104319 (2022)

    2022

  24. [24]

    Torosyanet al., Sci

    G. Torosyanet al., Sci. Rep.8, 1311 (2018)

    2018

  25. [25]

    Verbenoet al., J

    C. Verbenoet al., J. Magn. Magn. Mater.585, 171124 (2023)

    2023

  26. [26]

    Rouzegaret al., Phys

    R. Rouzegaret al., Phys. Rev. B106, 144427 (2022)

    2022

  27. [27]

    T. S. Seifertet al., Jour. of Phys. D: Applied Physics51, 364003 (2018)

    2018

  28. [28]

    Sinovaet al., Rev

    J. Sinovaet al., Rev. Mod. Phys.87, 1213 (2015)

    2015

  29. [29]

    Taoet al., Science Advances4, eaat1670 (2018)

    X. Taoet al., Science Advances4, eaat1670 (2018)

    2018

  30. [30]

    N´ advorn´ ıket al., Phys

    L. N´ advorn´ ıket al., Phys. Rev. X11, 021030 (2021)

    2021

  31. [31]

    Neˇ cas and P

    D. Neˇ cas and P. Klapetek, centr.eur.j.phys.10, 181–188 (2012)

    2012

  32. [32]

    Moritaet al., J

    M. Moritaet al., J. Appl. Phys.68, 1272–1281 (1990)

    1990

  33. [33]

    Abadiaset al., Appl

    G. Abadiaset al., Appl. Phys. Lett.107, 183105 (2015)

    2015

  34. [34]

    Lianget al., Nanoscale Adv.4, 2533 (2022)

    S. Lianget al., Nanoscale Adv.4, 2533 (2022)

    2022

  35. [35]

    Malinsk´ yet al., Nanoscale Research Letters7, 241 (2012)

    P. Malinsk´ yet al., Nanoscale Research Letters7, 241 (2012)

    2012

  36. [36]

    Gueckstocket al., Advanced Materials33, 2006281 (2021)

    O. Gueckstocket al., Advanced Materials33, 2006281 (2021)

    2021

  37. [37]

    Zhuet al., Phys

    L. Zhuet al., Phys. Rev. Lett.123, 057203 (2019)

    2019

  38. [38]

    T. K. H. Phamet al., Sci. Rep.8, 13907 (2018)

    2018

  39. [39]

    Bass and W

    J. Bass and W. P. Pratt, Jour. of Phys.: Condensed Mat- ter19, 183201 (2007)

    2007

  40. [40]

    Source data available at https://doi.org/10.5281/zenodo.20379022

    work page doi:10.5281/zenodo.20379022