REVIEW 4 major objections 6 minor 5 references
Poleval: A Python package for HAXPES analysis
T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read Grouped fitting makes noisy HAXPES fits more confident
desk verdict A useful, openly licensed HAXPES analysis package with a clever grouped-fitting feature; the paper underdelivers on validation of that feature, but it deserves referee time. 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 data model is built from Region, Sequence, SequenceGroup, and Config objects, which organise spectra by core level, by loaded file, and by experimental group so that trends in time, temperature, or pressure can be explored in a few lines of code. The load-bearing fitting mechanism is a grouped peak fit that operates on all Regions of the same core level in a SequenceGroup at once, using dependency constraints between parameters to encode prior knowledge; the paper states this is the feature that no published XPS software provides. For overlayer thickness, the key identity is the effective probing depth $1/\lambda' = 1/\lambda_p + 1/\lambda_e$, which enters the signal ratio $S_a/S_b$ (Eq. 5); the package solves this ratio for d with a numerical root finder and supplies a closed-form approximation when the two effective probe depths in the adsorbate are equal.
What would settle it
Simulate a series of HAXPES spectra from a known ground truth with controlled noise and a peak width that drifts linearly across the series, then fit with the grouped constraint that widths are equal and compare with unconstrained fits. If the constrained fits yield confidence intervals that exclude the true drifting width, or if the unconstrained fits recover the drift with no larger uncertainty, the central claim that grouping gives more confident and more consistent fits is falsified.
Extended reading notes
Core claim
On the paper's own terms, the contribution is a working software implementation of grouped HAXPES analysis. The novel step is the fitting routine: a single fit operates on every Region of the same core level within a SequenceGroup, and the user can impose dependencies such as equal peak widths, fixed positions, or fixed ratios between peaks across all spectra in the set. This transfers prior physical knowledge into the fit, restricting the information that must be extracted from noisy data and, the authors contend, reducing the uncertainty of the results. The second quantitative contribution is the overlayer thickness model (Eqs. 5-7), which estimates the adsorbate thickness d by equating the measured signal ratio $S_a/S_b$ to an expression involving effective probing depths and material densities, solving numerically for d and using the simplified closed form (Eq. 6) as an initial guess.
Load-bearing premise
The grouped fitting routine assumes that the parameters it constrains to be equal are physically common across all spectra in the group; if a peak width, position, or ratio actually drifts with temperature, pressure, or time, forcing equality biases every fit, and the paper provides no simulation or statistical test of this assumption.
Editorial extensions
If this is right
- Beamtime users can treat a full measurement campaign as one fit, so a noisy single spectrum does not have to be fitted alone; shared parameters across the series stabilise the result.
- Analysis reports become reproducible because the notebook records every data treatment step, intermediate spectrum, and fit constraint, and can be handed off between researchers mid-campaign.
- Quantitative comparisons between different core levels are possible with cross-section normalization from literature values, without leaving the analysis environment.
- Adsorbate thicknesses on the nanometre scale can be estimated directly from signal ratios, including for a composite substrate measured at two core levels (Eq. 7).
Reading between the lines
- If grouped fitting reduces fit uncertainty as claimed, the same constraint-based strategy should apply to other spectroscopy series where parameters are known to drift smoothly, such as temperature-dependent X-ray absorption or resonant photoemission; the paper does not discuss these settings.
- The benefit of grouping should grow with the number of spectra and the amount of shared structure; a quantitative benchmark against synthetic data would let users know how many spectra justify the constraints, which the paper does not provide.
- The thickness model's neglect of roughness and standing-wave effects, which the authors acknowledge, means its estimates are only as good as the flat, semi-infinite geometry assumption; extending the ratio formalism to account for a rough interface would be a natural next step that the paper leaves undone.
- The claim that no published software provides grouped fitting is a strong comparative statement; it could be tested only by a wider literature survey, and if a similar feature exists elsewhere, the novelty claim would narrow.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. POLEVAL is presented as a Python toolbox for HAXPES/XPS analysis, organized around Jupyter notebooks and custom data classes (Region, Sequence, SequenceGroup) to support persistent, collaborative, and reproducible workflows. The paper describes three quantitative analysis functions: cross-section normalization using haxquantpy, a grouped fitting routine built on lmfit that enforces shared peak parameters across spectra, and an adsorbate-overlayer thickness model based on signal ratios (Eqs. 1-7). The central claimed advantage is that grouped fitting with equality constraints yields "more confident fit results" for noisy data, and the authors further assert that no published software provides this grouping functionality. The paper contains no experimental data, synthetic benchmarks, or statistical validation of these claims; the thickness model equations appear standard and correct, and the flat-homogeneous geometry assumption is explicitly acknowledged in §3.
Significance. If the central claims are validated, POLEVAL would be a useful contribution to the HAXPES community: it is open source under an MIT license, integrates literature photoionization parameters via haxquantpy, and offers a notebook-based workflow that could improve reproducibility in synchrotron-based experiments. The overlayer-thickness formalism in Eqs. (1)-(7) is a standard textbook result and is presented clearly. The paper's main novelty, the grouped fitting routine with dependency constraints, is conceptually appealing, and the free availability of the code is a concrete strength. However, the paper currently does not demonstrate the key advantage with any example: there is no fit output, no comparison against unconstrained fitting, no uncertainty calibration, and no discussion of when the equality constraints are physically valid. The literature claim of novelty is also under-supported. These gaps make the present significance claims speculative rather than established.
major comments (4)
- [Abstract and §2 (Implementation)] The central claim that grouped fitting with equality constraints yields "more confident fit results" is not validated anywhere in the manuscript. No fit results, synthetic-data experiments, or statistical analyses are provided. Please add a benchmark with known ground-truth parameters, compare constrained versus unconstrained fits in terms of both bias and reported uncertainty, and show that the uncertainties are calibrated. Without such evidence, the main novelty of the package is only asserted, not demonstrated.
- [§2 (Implementation)] The premise that constrained parameters such as peak widths, positions, or ratios are common across all spectra in a group is stated without qualification. If any constrained parameter varies with temperature, pressure, time, or sample condition, forcing equality will introduce systematic bias even while reported uncertainties shrink. Please state the physical conditions under which the constraint is valid, and provide a diagnostic or a sensitivity analysis showing the effect of violating the commonality assumption on fitted parameters and their uncertainties.
- [§1 (Statement of Need)] The claim that "no published software so far provides this functionality" is too strong given the limited survey in the paper (only two open-source tools, LG4X and PyARPES, and one commercial package are cited). Many XPS analysis packages and general peak-fitting libraries support parameter linking or shared constraints; a broader literature and software search should be reported, or the claim should be narrowed to "no published HAXPES-specific software that we are aware of." Otherwise, the novelty statement is not substantiated.
- [§4 (Availability)] The reproducibility argument would be much stronger if the manuscript referenced a runnable example notebook, a test suite, or a documented API. The current text only points to a GitLab repository. Please add a link to documentation and include a minimal worked example (e.g., a synthetic or small experimental dataset fitted with the grouped routine) either in the paper or as supplementary material, so that readers and reviewers can verify the claimed behavior.
minor comments (6)
- [§1] The phrase "common open-source source algorithms" contains a typo; it should likely read "open-source algorithms" or "open-source analysis algorithms."
- [Figure 1 caption] The caption reads "showing the key classes and and their most important properties"; the duplicate "and" should be removed.
- [§3, Table 1] The sentence preceding Table 1 uses "absorbate" instead of "adsorbate" in the description of the subscripts.
- [Throughout] The name "POLEV AL" appears with an extra space in several places (e.g., the title, abstract, and running text); this appears to be a formatting artifact and should be corrected to "POLEVAL."
- [§3] The sentence beginning "The rationale for this is that the differences in X-ray intensity..." is unclear: please rephrase to specify what is negligible and why, since the scale argument as written is ambiguous.
- [References] Reference [1] contains an extra space in the author name "V . Yarzhemsky" and in the volume/page formatting; please ensure consistent citation style.
Circularity Check
No significant circularity: the paper is a software/toolbox description whose quantitative model is an independently stated physical formula, not a fitted re-statement of its inputs.
full rationale
The paper's derivation chain is self-contained in the relevant sense. The overlayer thickness model in Section 3 starts from exponential attenuation integrals (Eqs. 1-4), forms the signal ratio (Eq. 5), and solves for the thickness d by root finding or by the closed-form Eq. 6 under a stated approximation. The inputs (densities, cross-sections, escape depths, measured signal ratio) are external or measured; d is not fitted to data in a way that is later renamed as a prediction. The grouped fitting routine enforces equality or fixed-ratio constraints across spectra, and the claim of 'more confident fit results' is a statistical property of constrained fitting rather than a quantity predicted by those fits. No parameter is fitted to a subset and then presented as a prediction of the same subset, no central premise is justified solely by a self-citation, and no uniqueness theorem is imported from the authors' prior work. The acknowledged flat-homogeneous geometry assumption is a modeling limitation, not circularity. Therefore the score is 0.
Assumptions & free parameters
assumptions (5)
- domain assumption The effective probing depth is given by 1/λ' = 1/λp + 1/λe (Eq. 1).
- domain assumption For a semi-infinite substrate, the detected signal is S = I0 ρ σ λ' (Eq. 2).
- domain assumption The sample is a flat, homogeneous adsorbate on a flat, homogeneous semi-infinite substrate, with no surface roughness, reflections, or standing-wave effects (§3).
- domain assumption For the simplified thickness formula, λ'_a,A ≈ λ'_a,B (§3.1).
- ad hoc to paper Constrained peak parameters (widths, positions) are truly common across spectra in a group (§2).
Cite this review
Pith. "Pith review of Poleval: A Python package for HAXPES analysis." pith.science (2026). https://pith.science/paper/D5CIVFQG
@misc{pith2026250514194,
author = {Pith},
title = {Pith review of: Poleval: A Python package for HAXPES analysis},
year = {2026},
howpublished = {\url{https://pith.science/paper/D5CIVFQG}},
note = {Machine review of arXiv:2505.14194}
}
read the original abstract
POLEVAL provides a software toolbox for collaborative, persistent and reproducible analysis of XPS experiments. It allows to treat, analyse and visualise the results of an extended experimental campaign in a single python notebook in a consistent manner. Managing experimental data in adequate objects enables experimentalists to process and analyse measurements in very few lines of code, so as to provide decision aids through online data analysis during e.g. beamtime experiments. The persistent and self-documentary style of the notebook-based analysis allows for easy communication of intermediate results and enables progressive refinements into publishable figures or exporting the results to other programs. The toolbox facilitates various routines for data treatment (normalization, cropping, etc.) and aggregation of spectra into groups to analyse trends. It also enables quantitative analysis with three major functions: First, normalization to the photoionization cross-section and probability of emission into the analyser cone allows for quantitative comparisons between intensities from different core levels. The integrated haxquantpy package allows easy retrieval of literature values for this purpose. Second, an extensive fitting functionality is implemented to treat groups of spectra together, rather than spectrum-by-spectrum. This grouping allows reinforcing the fit algorithm with prior knowledge, such as the equivalence of peak widths or positions between spectra, which enables for more consistent, and importantly, more confident fit results for sets of potentially noisy spectra. Third, a simple formalism to estimate the thickness of adsorbate layers based on the ratio between the substrate's and adsorbate's XPS signal is implemented.
Figures
Reference graph
Works this paper leans on
-
[1]
Dirac–Fock photoionization parameters for HAXPES applications,
M. Trzhaskovskaya and V . Yarzhemsky, “Dirac–Fock photoionization parameters for HAXPES applications,” Atomic Data and Nuclear Data Tables , vol. 119, pp. 99–174, 01 2018. [Online]. Available: https: //linkinghub.elsevier.com/retrieve/pii/S0092640X16300596
work page 2018
-
[2]
H. Nakajima. LG4X. Zenodo. [Online]. Available: https://zenodo.org/doi/10.5281/zenodo.10477914
-
[3]
PyARPES: An analysis framework for multimodal angle-resolved photoemission spectroscopies,
C. Stansbury and A. Lanzara, “PyARPES: An analysis framework for multimodal angle-resolved photoemission spectroscopies,” SoftwareX, vol. 11, p. 100472, 1 2020. [Online]. Available: https: //www.sciencedirect.com/science/article/pii/S2352711019301633
work page 2020
-
[4]
Practical guide for curve fitting in x-ray photoelectron spectroscopy,
G. H. Major, N. Fairley, P. M. A. Sherwood, M. R. Linford, J. Terry, V . Fernandez, and K. Artyushkova, “Practical guide for curve fitting in x-ray photoelectron spectroscopy,” Journal of V acuum Science & Technology A: V acuum, Surfaces, and Films , vol. 38, no. 6, p. 061203, 12 2020. [Online]. Available: https://pubs.aip.org/jva/article/38/6/061203/1023...
work page 2020
- [5]
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.