Pith. sign in

REVIEW 4 major objections 6 minor 19 references

Emittance Minimization for Aberration Correction II: Physics-informed Bayesian Optimization of an Electron Microscope

T0 review · 4 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that aligning an aberration-corrected scanning transmission electron microscope can be reduced to optimizing a single scalar—the beam emittance growth predicted from a single Ronchigram by a neural network—using Bayesian…

desk verdict Solid simulation-backed automation paper whose headline experimental claim rests on a circular metric; worth refereeing for the simulation and the method, with a demand for independent experimental validation. read the letter →

arxiv 2412.20356 v2 pith:MDRN2SG5 submitted 2024-12-29 physics.ins-det cond-mat.mtrl-sciphysics.acc-ph

Desheng Ma , Steven E. Zeltmann , Chenyu Zhang , Zhaslan Baraissov , Yu-Tsun Shao , Cameron Duncan , Jared Maxson , Auralee Edelen
show 1 more author
David A. Muller
This is my paper · ORCID
classification physics.ins-detcond-mat.mtrl-sciphysics.acc-ph
keywords BayesianoptimizationbeamemittanceaberrationcorrectionscanningtransmissionelectronmicroscopyRonchigramdeepkernellearningGaussianprocesssurrogateautonomousmicroscopetuning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that tuning an aberration-corrected scanning transmission electron microscope—normally a 2-to-3-hour expert procedure—can be recast as a single-number optimization. The authors' key move is to replace the standard practice of measuring many individual aberration coefficients with a quantity called beam emittance growth, which they show is mathematically equivalent to the quality of the aberration correction and is convex in the aberration coefficients (one bowl-shaped landscape to descend). A neural network trained on simulated Ronchigrams—the diffraction shadow patterns formed from an amorphous sample—estimates this emittance from one experimental Ronchigram, and a Bayesian optimizer (with a deep neural-network kernel) then searches the corrector controls to minimize it. In simulations and on three real microscopes, the loop converges in about 50 iterations—roughly 4 minutes—to a lower normalized emittance than the conventional Zemlin-tableau method, suggesting that automated, machine-driven alignment is practical.

What carries the argument

The load-bearing object is the beam emittance growth $\varepsilon_{rms}^2$, a single scalar obtained from the second moments of the Wigner distribution of the 2D electron wave function at the aperture. For an aberration function $\chi(\vec{\alpha})$ expanded in Krivanek notation, it can be computed from $\langle|\nabla\chi|^2\rangle$, $\langle|\vec{\alpha}|^2\rangle$, and $\langle \vec{\alpha}\cdot\nabla\chi\rangle$, which the authors show reduces to an integral over the squared aperture brightness and the gradient of $\chi$. Its useful properties are that it is independent of defocus and convex in the aberration coefficients, so the Hessian with respect to aberration coefficients is positive semidefinite. That convexity makes it a well-behaved black-box objective, and a CNN trained on simulated Ronchigrams supplies the mapping from experimental images to emittance. The optimizer is Bayesian: a Gaussian-process surrogate with a kernel (RBF, Matern, or a deep-kernel network $k(g(\mathbf{x}_1,\mathbf{w}), g(\mathbf{x}_2,\mathbf{w}))$) is updated after each Ronchigram, and the next corrector setting is chosen by an acquisition function (upper confidence bound). The deep-kernel variant is what lets the surrogate learn couplings between separate control channels without measuring the aberration coefficients explicitly.

What would settle it

Take the states the optimizer declares optimal on a real microscope and measure the probe directly—by imaging a known atomic lattice, by 4D-STEM ptychographic reconstruction, or by reconstructing the phase-space emittance from the Wigner distribution—and check whether the CNN's emittance ranking matches the physical probe quality; if the CNN says a state is good while the probe is demonstrably aberrated, the central equivalence between the learned metric and aberration correction fails.

Watch

Extended reading notes

Core claim

The central claim is that minimizing beam emittance growth is equivalent to performing aberration correction, so the whole tuning problem collapses to optimizing a single scalar objective rather than estimating individual Zernike coefficients. Building on the Part I derivation, the authors take the root-mean-square emittance growth $\varepsilon_{rms}^2$ defined from the Wigner distribution of the electron wave function, note that it depends only on the gradient of the aberration function $\chi(\vec{\alpha})$ and is convex in the aberration coefficients, and use a deep convolutional network trained on simulated Ronchigrams to predict it from a single experimental Ronchigram. Around this predictor they build a Bayesian optimization loop, testing generic RBF and Matern kernels and a deep-kernel surrogate that learns correlations between control channels. The reported results are that Bayesian optimization outperforms the Nelder-Mead simplex baseline in simulation, that the deep kernel converges to lower emittance than isotropic kernels given enough iterations, and that on two ThermoFisher microscopes the automated loop reaches a CNN-predicted normalized emittance of $0.0266 \pm 0.0037$ and $0.0264 \pm 0.0045$ in about 50 iterations and 4 minutes, versus $0.0977 \pm 0.0041$ after Zemlin-tableau tuning on the Spectra 300.

Load-bearing premise

The whole demonstration rests on the assumption that the neural network trained on simulated Ronchigrams predicts true emittance on real, imperfect microscopes closely enough that minimizing its output also minimizes actual aberrations.

Editorial extensions

If this is right

  • Aberration correction no longer requires measuring individual Zernike aberration coefficients; a single Ronchigram plus one scalar objective suffices, which removes the multi-image tilt-series bottleneck of the Zemlin tableau.
  • A commercial microscope can be aligned from a random starting state in about 50 iterations and 4 minutes, compared with hours for a human expert or several multi-minute Zemlin measurements, making frequent re-tuning practical.
  • Because the CNN can score every candidate state from a single Ronchigram, the loop can run on an amorphous sample area, allowing re-alignment in the middle of an experiment rather than only at the start.
  • Deep-kernel surrogates that learn input couplings reduce the chance of getting trapped near local optima and reach lower emittance than isotropic RBF/Matern kernels in the long run, according to the simulation benchmarks.
  • The same emittance-minimization-plus-Bayesian-optimization framework is claimed to generalize to other high-dimensional, expensive scientific-instrument tuning tasks.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension, not spelled out in the paper, is to use the same Wigner-based emittance metric for other charged-particle optical systems—accelerator beamlines or electron sources—wherever a CNN predictor can be trained.
  • The paper does not compare the CNN's emittance scores with a direct experimental phase-space measurement; doing so (for example, with 4D-STEM or a Wigner reconstruction of the probe) would independently confirm that the optimized state is genuinely better.
  • The periodic 'deGauss' reset hints that magnetic hysteresis is a confound for the surrogate; an extension would be to model that history explicitly in the kernel or state, potentially removing the need for resets and speeding convergence.
  • Because the CNN was trained on simulation and applied without per-instrument recalibration, a practical extension is to fine-tune it with the first few online runs on each microscope, making the objective robust to instrument-specific optics and detectors.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper presents a Bayesian optimization (BO) framework for automated aberration correction in scanning transmission electron microscopy (STEM). The objective is a scalar beam-emittance-growth metric predicted by a convolutional neural network (CNN) from a single Ronchigram, developed in Part I. The authors propose a deep-kernel Bayesian optimization (DKBO) variant to capture inter-channel correlations. They validate the method in General Particle Tracer (GPT) simulations of a simplified hexapole-corrected microscope, showing that BO outperforms the Nelder-Mead simplex baseline in 2D and 6D tuning tasks, and report online experiments on a ThermoFisher Titan Cryo-S/TEM and a ThermoFisher Spectra 300 (plus a Nion UltraSTEM in Appendix A), where the method converges in about 50 iterations and 4 minutes. They compare the final state to the standard Zemlin tableau method on the Spectra 300 and claim their approach achieves a better optical state (normalized emittance 0.0266 vs. 0.0977) with a higher convergence rate.

Significance. If the claims hold, the method is a substantial practical advance: it replaces a 2–3 hour manual tuning procedure with a fully automated, roughly 4-minute routine that uses a single scalar metric and does not require explicit measurement of aberration coefficients. The simulation results are credible and provide independent support: GPT ray tracing shows a large FWHM reduction (from 0.385 μm to 0.0064 μm in one example), a visibly enlarged Ronchigram flat area, and a reduced phase-space area. The convergence benchmarks over 20 repetitions are a solid feature. However, the central experimental claim that BO 'outperforms conventional approaches' is currently supported only by evaluations using the same CNN that serves as the optimization objective, which is a circular comparison. The paper would be significantly strengthened by an independent optical-state measurement (e.g., residual aberration coefficients, probe size, or image resolution).

major comments (4)
  1. [III.B] The quantitative experimental comparison is circular. In Section III.B, the CNN emittance predictor is used as (i) the objective for Bayesian optimization, (ii) the metric for the final BO state (0.0266 ± 0.0037), and (iii) the metric for the Zemlin-tuned state (0.0977 ± 0.0041). No independent measurement of the optical state—such as residual aberration coefficients, direct probe size, or a resolution test—is reported for the BO-final states. Therefore the headline claim that the proposed method 'outperforms conventional approaches' in experiments is not established by the reported data. The authors should either provide an independent evaluation of the BO-final states or explicitly limit the experimental claim to a qualitative Ronchigram improvement.
  2. [III.B] The comparison with the Zemlin tableau is not a controlled experiment. The text states that the Spectra 300 was tuned 'several times' using the Zemlin method, but gives no details on the number of tuning iterations, the stopping criterion, the operator variance, or whether the same experimental conditions (sample, aperture, camera settings) were used as in the BO runs. The BO results are averaged over 10 runs, whereas no distribution is reported for the Zemlin results. This makes the quantitative margin (0.0266 vs. 0.0977) hard to interpret even if the circularity concern were resolved.
  3. [III.A / Figure 4] The dramatic FWHM reduction (0.3850 μm to 0.0064 μm) is reported for a single illustrative run, not averaged over the 20 repetitions used for the convergence benchmarks in Figure 6. The text claims that 'the optimization of aberration correctors according to the minimization of beam emittance can effectively eliminate lower order aberrations,' but this claim is supported by only one example. Please report the distribution of final FWHM or emittance over the 20 runs and assess statistical significance.
  4. [II.D / Figure 7] The claim that the deep kernel 'eliminates the chance of being trapped at local optima' is not quantitatively supported. Figure 7 shows qualitative scatter plots of queried points, but no metric for exploration/exploitation balance or local-optimum avoidance is provided, and the standard error bars in Figure 6 overlap in several regions. Please quantify exploration (e.g., coverage of the parameter space, GP predictive variance, or final emittance variance across repetitions) or temper the claim.
minor comments (6)
  1. [II.C] There is a duplicated heading: 'C. Physics-informed kernel' appears twice, immediately after 'C. Bayesian optimization'. Also, equation numbers are reused (Equation (5) appears twice, and Equation (2) is used for both the emittance definition and the Matern kernel). Please renumber consistently.
  2. [III.A] Typographical errors: 'Baysian' should be 'Bayesian', 'shruk' should be 'shrunk', and 'posess' should be 'possess'.
  3. [II.C] The paper's title advertises 'physics-informed' optimization, but the implemented DKBO method uses a deep neural network kernel and the Hessian-derived correlation kernel described in Section II.C is not used in the actual experiments or benchmarks. Please clarify how 'physics-informed' applies to the implemented method, or adjust the title/abstract to avoid overclaiming.
  4. [III.A / Figure 6] The text states the budget of iterations for the GPT-6D runs is 300, while the right panel of Figure 6 is labeled 'within 100 iterations.' Please clarify the final budget and whether the plotted results are truncated or the budget was changed.
  5. [III.B] The CNN emittance predictor from Part I is used as the objective for all experiments, but the paper gives only a reference to Part I and does not summarize its architecture, training data, or validation on experimental Ronchigrams. Since the experimental results depend entirely on this predictor, a brief description or at least a validation summary (especially regarding defocus invariance and transfer to different microscopes) should be included.
  6. [General] No data availability statement is provided. Please include a statement on whether the GPT simulation scripts, BO code, and experimental datasets will be made available.

Circularity Check

1 steps flagged · score 7.0 of 10

The experimental comparison is circular: the CNN from Part I is simultaneously the Bayesian optimizer's objective and the only quantitative evaluator, so the claimed superiority over Zemlin tuning is forced by construction.

  1. self definitional [Section III.B (online optimization of real microscopes), with the CNN-objective link established in Section II.C.]
    "we first acquire 10 Ronchigrams with random inputs and predict their emittance growth values with the CNN, upon which we initialize the GP for Bayesian optimization. ... The final normalized emittance values measured by the CNN after Bayesian optimization were 0.0266 ± 0.0037 (Cryo S/TEM) and 0.0264 ± 0.0045 (Spectra 300) averaged across 10 runs. ... Finally, we examine the Ronchigrams acquired from the corrected states and predict final emittance with the same trained deep neural network 0.0977 ± 0.0041 (Spectra 300), averaged across 10 runs."

    The CNN is not an independent evaluator. Section II.C makes it the black-box objective ('a deep learning model can build a mapping from experimentally accessible electron Ronchigrams, which becomes the objective function we optimize here'), and Section III.B uses 'the same trained deep neural network' to score both the BO-final and Zemlin-final states. BO therefore minimizes exactly the function used to declare it superior, while the Zemlin state was never optimized against that CNN metric. No independent optical-state measurement (residual aberration coefficients, probe size, or image resolution) is reported for the experimental comparison, so the quantitative claim that BO outperforms conventional approaches reduces, in the experiment, to the identity objective = evaluation metric.

full rationale

The circularity is confined to the experimental comparison. The simulation benchmarks in Section III.A are grounded externally: they evaluate directly ray-traced beam emittance and probe FWHM in GPT, and the Part I CNN is not the scoring function there. Citations to Part I are companion-paper self-citations carrying mathematical derivations and a separately trained network, but they are not used to forbid alternative approaches, so they are not load-bearing circularity by themselves. However, in Section III.B the same fitted CNN is both the Bayesian optimizer's objective and the only quantitative metric for comparing BO against Zemlin tuning; the reported 0.0266 vs 0.0977 emittance values are therefore a by-construction outcome of minimizing that metric, not an independent demonstration of a better optical state. The qualitative Ronchigram flat-area improvement is supporting but does not establish the quantitative superiority claim. Transfer of the CNN to real microscopes without per-instrument calibration is an accuracy risk, not a circularity argument.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim depends on a fitted CNN objective, an assumed equivalence from Part I, and a transferability assumption to real microscopes. No independent calibration of the CNN on the experimental instruments is reported. The deep kernel and GP hyperparameters are additional fitted components.

free parameters (4)
  • CNN emittance predictor weights = not disclosed
    Trained in Part I on simulated Ronchigrams; used as the expensive black-box objective for all BO runs and for the final evaluation of real microscopes. Its calibration to non-simulated instruments is not established.
  • Deep kernel DNN weights and GP lengthscales = not disclosed
    Learned online during BO; architecture and training details not given in this paper.
  • UCB acquisition beta = 0.2
    Hand-chosen for main benchmarks; sensitivity shown in Appendix C but no principled selection is given.
  • Random phase plate size and Gaussian blur = optimized but unspecified
    Tuned by hand to mimic amorphous scattering in simulations; exact values not provided.
assumptions (4)
  • domain assumption Emittance growth defined in Eq. 2 is convex in aberration coefficients and equivalent to aberration correction
    Taken from the authors' Part I; used to justify why a single scalar metric suffices. Not re-derived in this paper.
  • domain assumption CNN trained on simulated Ronchigrams transfers to real microscopes with no per-instrument calibration
    Explicitly assumed in the experimental workflow: 10 Ronchigrams are used only to seed the GP, not to retrain the CNN.
  • domain assumption CEOS aberration buttons map stationarily to physical lens currents
    Stated in Section II.C footnote: factory calibrated mapping assumed stationary; deGauss only partially handles hysteresis.
  • standard math Standard GP/BO machinery and Wigner-Weyl transform results
    Background from Frazier tutorial and the authors' Part I; not in question.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Emittance Minimization for Aberration Correction II: Physics-informed Bayesian Optimization of an Electron Microscope." pith.science (2026). https://pith.science/paper/MDRN2SG5

@misc{pith2026241220356,
  author       = {Pith},
  title        = {Pith review of: Emittance Minimization for Aberration Correction II: Physics-informed Bayesian Optimization of an Electron Microscope},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MDRN2SG5}},
  note         = {Machine review of arXiv:2412.20356}
}
read the original abstract

Aberration-corrected Scanning Transmission Electron Microscopy (STEM) has become an essential tool in understanding materials at the atomic scale. However, tuning the aberration corrector to produce a sub-{\AA}ngstr\"om probe is a complex and time-costly procedure, largely due to the difficulty of precisely measuring the optical state of the system. When measurements are both costly and noisy, Bayesian methods provide rapid and efficient optimization. To this end, we develop a Bayesian approach to fully automate the process by minimizing a new quality metric, beam emittance, which is shown to be equivalent to performing aberration correction. In part I, we derived several important properties of the beam emittance metric and trained a deep neural network to predict beam emittance growth from a single Ronchigram. Here we use this as the black box function for Bayesian Optimization and demonstrate automated tuning of simulated and real electron microscopes. We explore different surrogate functions for the Bayesian optimizer and implement a deep neural network kernel to effectively learn the interactions between different control channels without the need to explicitly measure a full set of aberration coefficients. Both simulation and experimental results show the proposed method outperforms conventional approaches by achieving a better optical state with a higher convergence rate.

Figures

Figures reproduced from arXiv: 2412.20356 by the authors.

Figure 1
Figure 1. FIG. 1: Workflow of online optimization of an electron microscope. [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Left: 3D contour of predi [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 9
Figure 9. FIG. 9: In comparison, we show the aberration measurements from the Zemlin [PITH_FULL_IMAGE:figures/full_fig_p017_9.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 15 canonical work pages

  1. [1]

    Barthel, A

    J. Barthel, A. Thust, On the optical stability of high-resolution transmission electron microscopes, Ultramicroscopy 134 (2013) 6–17

    work page 2013

  2. [2]

    Schramm, S

    S. Schramm, S. Van Der Molen, R. Tromp, Intrinsic instability of aberration-corrected electron microscopes, Phys. Rev. Lett. 109 (2012) 163901

    work page 2012

  3. [4]

    D. Ma, C. Zhang, Y.-T. Shao, Z. Baraissov, C. Duncan, A. Hanuka, A. Edelen, J. Maxon, D. Muller, Physics-informed Bayesian Optimization of an Electron Microscope, Microsc. Microanal. 29 (2023) 1875–1877. https://doi.org/10.1093/micmic/ozad067.968

    work page doi:10.1093/micmic/ozad067.968 2023

  4. [5]

    Frazier, A Tutorial on Bayesian Optimization, (2018)

    P.I. Frazier, A Tutorial on Bayesian Optimization, (2018). https://doi.org/10.48550/arXiv.1807.02811

    arXiv doi:10.48550/arxiv.1807.02811 2018

  5. [6]

    M.C. Cao, Z. Chen, Y. Jiang, Y. Han, Automatic parameter selection for electron ptychography via Bayesian optimization, Sci. Rep. 12 (2022) 12284. https://doi.org/10.1038/s41598-022-16041-5

    doi:10.1038/s41598-022-16041-5 2022

  6. [7]

    BEACON -- Automated Aberration Correction for Scanning Transmission Electron Microscopy using Bayesian Optimization

    A.J. Pattison, S.M. Ribet, M.M. Noack, G. Varnavides, K. Park, E. Kirkland, J. Park, C. Ophus, P. Ercius, BEACON -- Automated Aberration Correction for Scanning Transmission Electron Microscopy using Bayesian Optimization, (2024). http://arxiv.org/abs/2410.14873 (accessed October 23, 2024)

    work page Pith review arXiv 2024

  7. [8]

    The go-to choice of kernel is the radial basis function (RBF) kernel

    Appendix B: Other Kernels The kernel plays a significant role in the modeling of Gaussian processes, which encodes the covariance between each pair of the inputs. The go-to choice of kernel is the radial basis function (RBF) kernel. However, we further discuss how we can leverage our knowledge of the physics in the microscope system to choose more suitabl...

    work page 2013

  8. [9]

    Duris, D

    J. Duris, D. Kennedy, A. Hanuka, J. Shtalenkova, A. Edelen, P. Baxevanis, A. Egger, T. Cope, M. McIntire, S. Ermon, D. Ratner, Bayesian Optimization of a Free-Electron Laser, Phys. Rev. Lett. 124 (2020) 124801. https://doi.org/10.1103/PhysRevLett.124.124801

    doi:10.1103/physrevlett.124.124801 2020

Show all 19 references
  1. [10]

    Shalloo, S.J.D

    R.J. Shalloo, S.J.D. Dann, J.-N. Gruse, C.I.D. Underwood, A.F. Antoine, C. Arran, M. Backhouse, C.D. Baird, M.D. Balcazar, N. Bourgeois, J.A. Cardarelli, P. Hatfield, J. Kang, K. Krushelnick, S.P.D. Mangles, C.D. Murphy, N. Lu, J. Osterhoff, K. Põder, P.P. Rajeev, C.P. Ridgers...

    2020

  2. [11]

    Jalas, M

    S. Jalas, M. Kirchen, P. Messner, P. Winkler, L. Hübner, J. Dirkwinkel, M. Schnepp, R. Lehe, A.R. Maier, Bayesian Optimization of a Laser-Plasma Accelerator, Phys. Rev. Lett. 126 (2021) 104801. https://doi.org/10.1103/PhysRevLett.126.104801

    2021 doi

  3. [12]

    Zhang, Z

    C. Zhang, Z. Baraissov, C. Duncan, A. Hanuka, A. Edelen, J. Maxson, D. Muller, Aberration Corrector Tuning with Machine-Learning-Based Emittance Measurements and Bayesian Optimization, Microsc. Microanal. 27 (2021) 810–812. https://doi.org/10.1017/S1431927621003214

    2021 doi

  4. [13]

    Zhang, Y.-T

    C. Zhang, Y.-T. Shao, Z. Baraissov, C.J. Duncan, A. Hanuka, A.L. Edelen, J.M. Maxson, D.A. Muller, Bayesian Optimization for Multi-dimensional Alignment: Tuning Aberration Correctors and Ptychographic Reconstructions, Microsc. Microanal. 28 (2022) 3146–3148. https://doi.org/10...

    2022 doi

  5. [14]

    Krivanek, N

    O.L. Krivanek, N. Dellby, A.R. Lupini, Towards sub-Å electron beams, Ultramicroscopy 78 (1999) 1–11. https://doi.org/10.1016/S0304-3991(99)00013-3

    1999 doi

  6. [15]

    De Loos, S.B

    M.J. De Loos, S.B. Van der Geer, General Particle Tracer: A New 3D Code for Accelerator and Beamline Design, (1996). https://cds.cern.ch/record/860825 (accessed April 9, 2024)

    1996

  7. [16]

    Wilson, Z

    A.G. Wilson, Z. Hu, R. Salakhutdinov, E.P. Xing, Deep Kernel Learning, in: Proc. 19th Int. Conf. Artif. Intell. Stat., PMLR, 2016: pp. 370–378. https://proceedings.mlr.press/v51/wilson16.html (accessed April 9, 2024)

    2016

  8. [17]

    Wistuba, J

    M. Wistuba, J. Grabocka, Few-Shot Bayesian Optimization with Deep Kernel Surrogates, (2021). https://doi.org/10.48550/arXiv.2101.07667

    2021 doi arXiv

  9. [18]

    Hartel, V

    P. Hartel, V. Gerheim, M. Linck, H. Müller, S. Uhlemann, Z. Zach, M. Haider, On the residual six-fold astigmatism in DCOR/ASCOR, Ultramicroscopy 206 (2019) 112821. https://doi.org/10.1016/j.ultramic.2019.112821

    2019

  10. [2021]

    https://www.osti.gov/biblio/1811769 (accessed April 9, 2024)

    2024

  11. [6355]

    https://doi.org/10.1038/s41467-020-20245-6

    doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.