Pith. sign in

REVIEW 3 major objections 5 minor 30 references

Hydrogen Passivation Effects on Spatially Resolved Charge Trap Densities in Si(100)-SiO$_2$

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read This paper reports that hydrogen-terminated silicon of the kind used for hydrogen resist lithography hosts far fewer donor-like charge traps at its Si-SiO2 interface than conventionally prepared silicon, with forming-gas annealing adding…

desk verdict Useful new fm-AFM application with a consistent hydrogen trend, but the key pristine-vs-H-terminated comparison is confounded by processing differences and single scans per condition, so the causal claim and the 80% number need more evidence. read the letter →

arxiv 2505.23574 v2 pith:27CXMSHT submitted 2025-05-29 cond-mat.mtrl-sci

classification cond-mat.mtrl-sci
keywords Si-SiO2interfacechargetrapshydrogenpassivationfrequency-modulatedatomicforcemicroscopyresistlithographytwo-levelforminggasannealingrandomtelegraphnoise
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

What the paper tries to establish: that the hydrogen-terminated silicon surface used in hydrogen resist lithography has fewer donor-like charge traps at its Si-SiO2 interface than conventionally prepared 'pristine' silicon, and that annealing in forming gas reduces them further. Frequency-modulated AFM dissipation scans reveal individual traps as rings, allowing counts on six samples: pristine and H-terminated, each with no anneal, N2 anneal, or N2+H2 anneal. The ordering is consistent across three counting methods: more hydrogen in processing means fewer donor-like traps, and two-level donor-like traps are about 80% rarer in H-terminated samples. Acceptor-like trap density is unchanged by hydrogen. If this is right, hydrogen-resist-lithography substrates are not simply a patterning convenience but a cleaner electronic interface for nanoscale and quantum devices.

What carries the argument

The mechanism that carries the argument is the dissipation ring in frequency-modulated atomic force microscopy (fm-AFM): with a metal-coated tip at fixed negative bias, the ionization of a single donor-like trap changes the tip's dissipation, producing a ring centred on the trap in a dissipation scan. Ring size should be comparable across samples because the bulk doping is identical and the bias and frequency-shift setpoint are held constant, so ring area or ring count can stand in for trap density. Three quantification routes are combined: Otsu's automatic thresholding to measure ring area fraction, a YOLOv8 object-detection model trained on synthetic dissipation images to count rings, and manual counting to cross-check both; two-level traps are isolated by taking the second derivative along the slow scan axis, where striped rings appear as features with large curvature.

What would settle it

A decisive test would be to compare the AFM-derived trap densities with an independent, ring-size-insensitive measurement on identically processed material, for example random-telegraph-noise or charge-pumping counts from small transistors fabricated on pristine and H-terminated samples, or to deliberately vary the AFM bias and setpoint and check that the inferred density ordering is stable; if the 80% two-level reduction does not survive either check, the central claim would be undermined.

Watch

Extended reading notes

Core claim

At the paper's center is a measurement made possible by fm-AFM dissipation imaging: a donor-like trap under the tip ionizes and produces a finite dissipation ring, so each ring marks one electrically active defect. Scanning at negative tip bias, the authors compare ring densities on pristine silicon and on silicon that went through the UHV cleaning, hydrogen termination, and epitaxial silicon capping used for hydrogen resist lithography, with each sample receiving either no anneal, a nitrogen anneal, or a forming-gas (N2+H2) anneal. They find that hydrogen content in processing orders the trap density: H-terminated samples have lower donor-like trap density than pristine samples, forming-gas annealing lowers it further, and N2 annealing alone makes no statistical difference. Two-level donor-like traps, seen as striped rings and quantified via the second derivative along the slow scan axis, are strongly reduced in hydrogen-rich samples, with an 80% reduction in H-terminated samples relative to pristine silicon. The authors conclude that the hydrogen layer survives epitaxial encapsulation and passivates donor-like traps, while acceptor-like traps appear unaffected.

Load-bearing premise

The quantitative comparison assumes that a trap produces the same-sized dissipation ring in every sample regardless of processing; if hydrogen passivation changes trap depth or energy levels and thus ring size, the threshold-based and even the deep-learning counts could misstate the true trap-density reduction.

Editorial extensions

If this is right

  • Hydrogen resist lithography substrates should give nanoscale transistors and quantum dots a quieter interface, with lower charge noise and more stable threshold voltages, because the donor-like trap density is lower.
  • The roughly 80% drop in two-level donor traps implies less random telegraph noise and, for nearby qubits, potentially longer coherence times, since two-level systems are a known decoherence source.
  • Adding a forming-gas (N2+H2) anneal after processing is an independent passivation step that further reduces donor-like traps, while a pure nitrogen anneal is not sufficient.
  • fm-AFM can act as a non-destructive, spatially resolved inspection tool for interfacial traps, locating individual defects relative to device features rather than reporting a sample-wide average.
  • Because acceptor-like traps do not respond to hydrogen, cleaning the Si-SiO2 interface will require a separate or complementary passivation route for acceptor defects.

Reading between the lines

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

  • The paper does not measure device-level noise; if trap density translates directly into noise, the 80% two-level reduction should show up as lower random telegraph noise in small transistors built on H-terminated substrates, which would be a direct test.
  • Because traps switching faster or slower than the pixel dwell time are invisible at the fixed scan rate, the reported reduction applies to a limited switching-rate window; variable scan-rate fm-AFM could reveal whether hydrogen suppresses the full spectrum of two-level fluctuators.
  • The same imaging protocol could be used to screen other proposed passivating species, such as chlorine resists or deuterium annealing, and determine whether they reproduce hydrogen's donor-trap suppression without its lithographic side effects.
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

3 major / 5 minor

Summary. This manuscript uses frequency-modulated AFM dissipation imaging to map donor-like charge traps at Si(100)-SiO2 interfaces. It compares two sample classes: pristine (unprocessed) silicon and hydrogen-terminated silicon prepared for hydrogen resist lithography (UHV clean, 1200°C flash anneal, H termination, 3 nm epitaxial Si cap, then native oxide), each subjected to no anneal, N2 anneal, or N2+H2 forming-gas anneal. Trap densities are quantified by Otsu thresholding, a YOLOv8 detector trained on synthetic images, and manual counting. The authors report that increased hydrogen exposure lowers donor-like trap density and that H-terminated samples show an 80% reduction in two-level traps relative to pristine silicon, concluding that hydrogen passivates these traps.

Significance. If the causal attribution holds, the result is significant because it suggests HRL-prepared substrates have cleaner Si-SiO2 interfaces, with implications for nanoscale and quantum devices. The paper has several genuine strengths: three independent counting approaches (Otsu, YOLOv8, manual) show the same qualitative ordering; the N2+H2 versus N2 anneal within each sample type provides an internal control for forming-gas passivation; and the ML code is publicly released. The qualitative trend is plausible, but the causal attribution to hydrogen in the cross-sample comparison is not supported by the current experimental design, and the quantitative claims rest on single scans per condition. The manuscript needs additional controls and statistical support before the central claim can be accepted.

major comments (3)
  1. [Sec. II and Sec. V] The comparison between H-terminated and pristine samples is confounded by processing steps other than hydrogen termination. The H-terminated samples receive 1200°C flash anneals in UHV, a slow cool to form the 2x1 reconstruction, and a 3 nm epitaxial Si cap before native-oxide formation, while the pristine samples are unprocessed wafer pieces. The donor traps are imaged at the top Si-SiO2 interface, i.e., above the cap, while the H layer is buried 3 nm below. The lower trap density observed in H-terminated samples could therefore be caused by the UHV clean, the fresh epitaxial interface, or the capping process itself. The within-sample N2+H2 versus N2 anneal demonstrates passivation by forming gas, but it does not isolate the buried H layer's contribution. A no-H capped control (identical UHV clean and epi cap, without hydrogen termination) is needed. Without it, the conclusion in Sec. V that 'this hydrogen contributes to the passivation' is not established.
  2. [Sec. III, Figs. 3-4; Sec. IV, Fig. 7] Each reported condition is represented by a single dissipation scan; Fig. 2 shows six images total, one per sample/anneal combination. The error bars in Figs. 3, 4, and 7 are within-image uncertainties (Monte Carlo quarter-window sampling for Otsu; dropout-based Bayesian approximation for YOLOv8) and do not include scan-to-scan or sample-to-sample variability. Consequently, the word 'significantly' in the abstract and Sec. III is not backed by a statistical test across repeated scans. At minimum, the authors should repeat scans on multiple regions and report inter-scan statistics, or explicitly reframe the claims as descriptive rather than inferential.
  3. [Sec. III, Fig. 5] The conversion from Otsu ring-area fraction to trap counts assumes a consistent ring-size distribution across samples, as the paper acknowledges. The DL method was intended to test this assumption, but it is trained only on synthetic data and is acknowledged to undercount rings in H-terminated samples (false negatives), so it cannot validate the conversion in the sample class where the claimed reduction is largest. The magnitude of the reported reduction, including the 80% two-level-trap reduction in Sec. V, therefore depends on an assumption that remains unverified for the H-terminated samples. Please provide evidence on ring-size distributions per sample, or otherwise bound the resulting bias.
minor comments (5)
  1. [Sec. II] The phrase 'all performed 250 ◦C for 5 minutes' should read 'all performed at 250 °C for 5 minutes.'
  2. [Fig. 2] The dissipation color scales differ strongly between panels (0-239, 0-128, 0-145, 0-148, 0-50, 0-100 meV/cycle). Please clarify how Otsu thresholding is affected by these scaling differences or state that the scaling was normalized before analysis.
  3. [Supplementary Material] The table numbering appears inconsistent: the main supplementary text refers to 'Table S2' for the model metrics, but the caption says 'TABLE S1', and the manual counts are then also called 'Table S2'. Please renumber and cross-check all references.
  4. [Sec. IV, Fig. 7] The phrase 'a very small number of two-level traps' should be replaced by the actual counts or densities; the bars in Fig. 7 are not annotated with values.
  5. [Fig. S2] The figure label 'H-Capped Silicon' is inconsistent with the main text's 'H-terminated silicon'; please unify the terminology.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: trap counts are directly measured and externally calibrated; acknowledged confounds are correctness risks, not circular reductions.

full rationale

The derivation chain here is experimental rather than formal: the reported quantities (Otsu ring-area fraction, YOLOv8 bounding-box counts, manually counted rings, second-derivative two-level features) are direct statistics on fm-AFM dissipation images, not outputs of a model fitted to the conclusion. The Otsu-to-count conversion is anchored to an external label: 'Otsu's method follows a linear trend when compared to manual counting, with a fitted slope of 590±30 nm2' (Sec. III, Fig. 5), and Table S2 gives manual counts that reproduce the same ordering (pristine 250/230/80 vs H-terminated 120/85/35). The DL count is trained on synthetic data and explicitly lacks direct real-data validation ('direct quantitative validation on real data is not available,' Supplementary Sec. B.2), but the paper does not use the DL count alone to force the central claim; the manual counts and Otsu trends agree. The paper explicitly flags its two genuine limitations. First, the Otsu comparison 'assumes a consistent ring size distribution across all samples' (Sec. III), a measurement assumption that could bias the magnitude of the reduction; because manual counts confirm the trend, this is a quantitative caveat, not a definitional circularity. Second, the H-terminated versus pristine comparison is confounded by UHV cleaning and epitaxial capping: 'There is an open question as to what happens to the hydrogen layer when the silicon capping layer is deposited after hydrogen lithography in the H-terminated samples' (Sec. III), yet 'we conclude that this hydrogen contributes to the passivation of donor-like traps' (Sec. V). This is a causal-attribution gap (a no-H capped control is missing), which belongs in correctness risk, not circularity. The self-citations (refs. 4, 22: Cowie et al.) establish the ring/trap correspondence; they are published prior work with independent content and are reinforced here by manual counting, so the central hydrogen-passivation claim does not reduce to an unverified self-citation. No step in the paper makes a predicted quantity equal to its input by construction.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

No new physical entities are introduced. The analysis relies on calibrations and assumptions rather than new theoretical constructs. The main free parameter is the Otsu-to-count conversion slope fitted to manual counts; the axioms concern the physical interpretation of dissipation rings, tip-height reproducibility, and ring-size consistency.

free parameters (1)
  • Otsu to count conversion slope = 590 +/- 30 nm^2
    Linear fit between ring area fraction and manual ring count (Fig. 5); used to convert area coverage to ring density. This is a calibration constant, not a physical parameter.
assumptions (3)
  • domain assumption fm-AFM dissipation rings correspond to individual charge traps
    The paper builds on Cowie et al. (refs. 22, 4) and does not re-derive the correspondence; all trap counts depend on it.
  • domain assumption Equal frequency-shift setpoint yields equal tip height across samples
    Sec. II: 'scanning the samples with the same frequency shift (Delta f) setpoint is assumed to result in the same tip heights for all measurements'. Used to justify comparable ring sizes across samples.
  • domain assumption Consistent ring size distribution across samples for Otsu method
    Sec. III explicitly states this assumption; required for converting ring area to density. The paper notes it may not hold if passivation alters the trap-type mix.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hydrogen Passivation Effects on Spatially Resolved Charge Trap Densities in Si(100)-SiO$_2$." pith.science (2026). https://pith.science/paper/27CXMSHT

@misc{pith2026250523574,
  author       = {Pith},
  title        = {Pith review of: Hydrogen Passivation Effects on Spatially Resolved Charge Trap Densities in Si(100)-SiO$_2$},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/27CXMSHT}},
  note         = {Machine review of arXiv:2505.23574}
}
abstract

As silicon-based devices continue to shrink to the nanoscale, traps at the Si-SiO$_2$ interface pose increasing challenges to device performance. These traps reduce channel carrier mobility and shift threshold voltages in integrated circuits, and introduce charge noise in quantum systems, reducing their coherence times. Knowledge of the precise location of such traps aids in understanding their influence on device performance. In this work, we demonstrate that frequency-modulated atomic force microscopy (fm-AFM) allows the detection of individual traps. We use this to study how sample preparation, specifically the introduction of a buried hydrogen termination layer, and post-processing annealing in forming gas (N$_2$+H$_2$), affects the density of donor-like traps in Si(100)-SiO$_2$ systems. We spatially map and quantify traps in both conventionally prepared ("pristine") silicon samples and those processed under ultra-high vacuum for hydrogen resist lithography (HRL). We confirm previous studies demonstrating hydrogen passivation of traps and find that hydrogen termination further reduces the donor-like trap density. We also observe a significant reduction in two-level donor-like traps in the hydrogen-terminated samples compared to pristine silicon samples. These findings suggest that HRL-prepared silicon may offer advantages for high-performance nanoscale and atomic-scale devices due to reduced trap densities.

Figures

Figures reproduced from arXiv: 2505.23574 by the authors.

Figure 1
Figure 1. FIG. 1. fm-AFM dissipation scans reveal rings due to traps. Shown is [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Negative bias dissipation scans reveal rings (blue regions) [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. FIG. 4. A YOLOv8 medium sized model was trained to predict [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figures from the paper (2 more)
Figure 6
Figure 6. Figure 6: FIG. 6. Taking the second derivative along the slow scan axis (here [PITH_FULL_IMAGE:figures/full_fig_p004_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7. The histogram shows the spatial density of two-level traps [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

30 extracted references · 27 canonical work pages

  1. [1]

    Kato , author T

    author author K. Kato , author T. Yamasaki , \ and\ author T. Uda ,\ title title Origin of P _ b1 center at S i O _ 2 / S i(100) interface: First-principles calculations , \ @noop journal journal Phys. Rev. B \ volume 73 ,\ pages 073302 ( year 2006 ) NoStop

  2. [2]

    author author J. P. \ Campbell , author P. M. \ Lenahan , author C. J. \ Cochrane , author A. T. \ Krishnan , \ and\ author S. Krishnan ,\ title title Atomic-scale defects involved in the negative-bias temperature instability , \ 10.1109/TDMR.2007.911379 journal journal IEEE Transactions on Device and Materials Reliability \ volume 7 ,\ pages 540--557 ( y...

  3. [3]

    Kirton \ and\ author M

    author author M. Kirton \ and\ author M. Uren ,\ title title Noise in solid-state microstructures: A new perspective on individual defects, interface states and low-frequency (1/ƒ) noise , \ @noop journal journal Advances in Physics \ volume 38 ,\ pages 367--468 ( year 1989 ) NoStop

  4. [4]

    Cowie , author P

    author author M. Cowie , author P. C. \ Constantinou , author N. J. \ Curson , author T. J. Z. \ Stock , \ and\ author P. Gr \"u tter ,\ title title Spatially resolved random telegraph fluctuations of a single trap at the S i/ S i O _2 interface , \ 10.1073/pnas.2404456121 journal journal Proceedings of the National Academy of Sciences \ volume 121 ,\ pag...

  5. [5]

    \ Chen , author S.-W

    author author S.-H. \ Chen , author S.-W. \ Lian , author T. R. \ Wu , author T.-R. \ Chang , author J.-M. \ Liou , author D. D. \ Lu , author K.-H. \ Kao , author N.-Y. \ Chen , author W.-J. \ Lee , \ and\ author J.-H. \ Tsai ,\ title title Impact of semiconductor permittivity reduction on electrical characteristics of nanoscale mosfets , \ 10.1109/TED.2...

  6. [6]

    \ Wang \ and\ author A

    author author Y.-X. \ Wang \ and\ author A. A. \ Clerk ,\ title title Intrinsic and induced quantum quenches for enhancing qubit-based quantum noise spectroscopy , \ @noop journal journal Nature Communications \ volume 12 ,\ pages 6528 ( year 2021 ) NoStop

  7. [7]

    Ambal , author P

    author author K. Ambal , author P. Rahe , author A. Payne , author J. Slinkman , author C. C. \ Williams , \ and\ author C. Boehme ,\ title title Electrical current through individual pairs of phosphorus donor atoms and silicon dangling bonds , \ @noop journal journal Scientific Reports \ volume 6 ,\ pages 18531 ( year 2016 ) NoStop

  8. [8]

    author author A. Stesmans ,\ title title Passivation of P _ b0 and P _ b1 interface defects in thermal (100)Si/SiO _ 2 with molecular hydrogen , \ @noop journal journal Applied Physics Letters \ volume 68 ,\ pages 2076--2078 ( year 1996 ) NoStop

Show all 30 references
  1. [9]

    Stesmans ,\ title title Dissociation kinetics of hydrogen-passivated P _ b defects at the (111)Si/SiO _ 2 interface , \ @noop journal journal Phys

    author author A. Stesmans ,\ title title Dissociation kinetics of hydrogen-passivated P _ b defects at the (111)Si/SiO _ 2 interface , \ @noop journal journal Phys. Rev. B \ volume 61 ,\ pages 8393--8403 ( year 2000 ) NoStop

  2. [10]

    author author K. L. \ Brower ,\ title title Dissociation kinetics of hydrogen-passivated (111) S i- S i O _ 2 interface defects , \ 10.1103/PhysRevB.42.3444 journal journal Phys. Rev. B \ volume 42 ,\ pages 3444--3453 ( year 1990 ) NoStop

  3. [11]

    Cartier , author J

    author author E. Cartier , author J. H. \ Stathis , \ and\ author D. A. \ Buchanan ,\ title title Passivation and depassivation of silicon dangling bonds at the S i/ S i O _2 interface by atomic hydrogen , \ @noop journal journal Applied Physics Letters \ volume 63 ,\ pages 15...

  4. [12]

    Pantelides , author S

    author author S. Pantelides , author S. Rashkeev , author R. Buczko , author D. Fleetwood , \ and\ author R. Schrimpf ,\ title title Reactions of hydrogen with S i/ S i O _2 interfaces , \ @noop journal journal IEEE Transactions on Nuclear Science \ volume 47 ,\ pages 2262--22...

  5. [13]

    author author J. W. \ Lyding , author K. Hess , \ and\ author I. C. \ Kizilyalli ,\ title title Reduction of hot electron degradation in metal oxide semiconductor transistors by deuterium processing , \ @noop journal journal Applied Physics Letters \ volume 68 ,\ pages 2526--2...

  6. [14]

    author author J. W. \ Lyding , author T.-C. \ Shen , author J. S. \ Hubacek , author J. R. \ Tucker , \ and\ author G. C. \ Abeln ,\ title title Nanoscale patterning and oxidation of H -passivated S i(100)-2 1 surfaces with an ultrahigh vacuum scanning tunneling microscope , \...

  7. [15]

    Fuechsle , author J

    author author M. Fuechsle , author J. A. \ Miwa , author S. Mahapatra , author H. Ryu , author S. Lee , author O. Warschkow , author L. C. L. \ Hollenberg , author G. Klimeck , \ and\ author M. Y. \ Simmons ,\ title title A single-atom transistor , \ 10.1038/nnano.2012.21 jour...

  8. [16]

    Wyrick , author X

    author author J. Wyrick , author X. Wang , author R. V. \ Kashid , author P. Namboodiri , author S. W. \ Schmucker , author J. A. \ Hagmann , author K. Liu , author M. D. \ Stewart Jr. , author C. A. \ Richter , author G. W. \ Bryant , \ and\ author R. M. \ Silver ,\ title tit...

  9. [17]

    Huff , author H

    author author T. Huff , author H. Labidi , author M. Rashidi , author L. Livadaru , author T. Dienel , author R. Achal , author W. Vine , author J. Pitters , \ and\ author R. A. \ Wolkow ,\ title title Binary atomic silicon logic , \ @noop journal journal Nature Electronics \ ...

  10. [18]

    Achal , author M

    author author R. Achal , author M. Rashidi , author J. Croshaw , author D. Churchill , author M. Taucer , author T. Huff , author M. Cloutier , author J. Pitters , \ and\ author R. A. \ Wolkow ,\ title title Lithography for robust and editable atomic-scale silicon devices and ...

  11. [19]

    Weber , author Y

    author author B. Weber , author Y. H. M. \ Tan , author S. Mahapatra , author T. F. \ Watson , author H. Ryu , author R. Rahman , author L. C. L. \ Hollenberg , author G. Klimeck , \ and\ author M. Y. \ Simmons ,\ title title Spin blockade and exchange in coulomb-confined sili...

  12. [20]

    Wang , author E

    author author X. Wang , author E. Khatami , author F. Fei , author J. Wyrick , author P. Namboodiri , author R. Kashid , author A. F. \ Rigosi , author G. Bryant , \ and\ author R. Silver ,\ title title Experimental realization of an extended fermi-hubbard model using a 2d lat...

  13. [21]

    author author T. J. Z. \ Stock , author O. Warschkow , author P. C. \ Constantinou , author J. Li , author S. Fearn , author E. Crane , author E. V. S. \ Hofmann , author A. K \"o lker , author D. R. \ McKenzie , author S. R. \ Schofield , \ and\ author N. J. \ Curson ,\ title...

  14. [22]

    Cowie , author T

    author author M. Cowie , author T. J. Z. \ Stock , author P. C. \ Constantinou , author N. J. \ Curson , \ and\ author P. Gr\"utter ,\ title title Spatially resolved dielectric loss at the S i/ S i O _ 2 interface , \ 10.1103/PhysRevLett.132.256202 journal journal Phys. Rev. L...

  15. [23]

    Cockins , author Y

    author author L. Cockins , author Y. Miyahara , author S. D. \ Bennett , author A. A. \ Clerk , author S. Studenikin , author P. Poole , author A. Sachrajda , \ and\ author P. Grutter ,\ title title Energy levels of few-electron quantum dots imaged and characterized by atomic ...

  16. [24]

    author author N. Otsu ,\ title title A threshold selection method from gray-level histograms , \ 10.1109/TSMC.1979.4310076 journal journal IEEE Transactions on Systems, Man, and Cybernetics \ volume 9 ,\ pages 62--66 ( year 1979 ) NoStop

  17. [25]

    Jocher \ and\ author Ultralytics ,\ @noop title YOLOv8 - ultralytics , \ howpublished https://github.com/ultralytics/ultralytics ( year 2023 ),\ note accessed: 2025-05-07 NoStop

    author author G. Jocher \ and\ author Ultralytics ,\ @noop title YOLOv8 - ultralytics , \ howpublished https://github.com/ultralytics/ultralytics ( year 2023 ),\ note accessed: 2025-05-07 NoStop

  18. [26]

    Gal \ and\ author Z

    author author Y. Gal \ and\ author Z. Ghahramani ,\ title title Dropout as a bayesian approximation: Representing model uncertainty in deep learning , \ in\ @noop booktitle International Conference on Machine Learning \ ( organization PMLR ,\ year 2016 )\ pp.\ pages 1050--1059 NoStop

  19. [27]

    Deng , author P

    author author X. Deng , author P. Namboodiri , author K. Li , author X. Wang , author G. Stan , author A. F. \ Myers , author X. Cheng , author T. Li , \ and\ author R. M. \ Silver ,\ title title Silicon epitaxy on H -terminated S i (100) surfaces at 250 \,^ C , \ @noop journa...

  20. [28]

    Wimmer , author A.-M

    author author Y. Wimmer , author A.-M. \ El-Sayed , author W. G \"o s , author T. Grasser , \ and\ author A. L. \ Shluger ,\ title title Role of hydrogen in volatile behaviour of defects in S i O _2 -based electronic devices , \ @noop journal journal Proceedings of the Royal S...

  21. [29]

    Defrance , author A

    author author F. Defrance , author A. D. \ Beyer , author J. Wheeler , author J. Sayers , \ and\ author S. R. \ Golwala ,\ title title Low two-level-system noise in hydrogenated amorphous silicon , \ @noop journal journal Applied Physics Letters \ volume 126 ,\ pages 122904 ( ...

  22. [30]

    author author T. V. \ Pavlova \ and\ author K. N. \ Eltsov ,\ title title Reactivity of the S i(100)-2 × 1- C l surface with respect to PH 3, PC l3, and BC l3: comparison with PH 3 on S i(100)-2 × 1- H , \ @noop journal journal Journal of Physics: Condensed Matter \ volume 33 ...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.