Ion Flow under an Applied Electric Field in Semiconducting and Metallic Carbon Nanotubes

J. B. Sokoloff

REVIEW 4 major objections 4 minor 14 references

The paper explains the higher ionic conductivity of long semiconducting carbon nanotubes by their slow electronic equipotentialization in an applied field.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review

2026-08-04 00:59 UTC pith:NES5NZRZ

load-bearing objection Timescale-based explanation for why ions flow faster through long semiconducting carbon nanotubes — plausible and clearly argued, but the numbers rest on parameters the paper does not measure. the 4 major comments →

arxiv 2608.00211 v1 pith:NES5NZRZ submitted 2026-07-31 cond-mat.mes-hall

Ion Flow under an Applied Electric Field in Semiconducting and Metallic Carbon Nanotubes

J. B. Sokoloff This is my paper

classification cond-mat.mes-hall

keywords carbon nanotubesionic conductivitysemiconducting nanotubesmetallic nanotubeselectric field screeningequipotentializationnanofluidicsion transport

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper proposes a single mechanism to reconcile two sets of experiments: metallic carbon nanotubes cancel an applied electric field almost instantly by redistributing conduction electrons, while semiconducting nanotubes, with far fewer conduction electrons, take much longer to become equipotential. In 11 nm tubes, both types reach equipotential within the measurement time, so ionic flow is nearly identical. In 100 μm tubes, the semiconducting tube retains an internal electric field during the measurement, adding a field-driven contribution to the ion current that the metallic tube lacks. The paper derives an equilibration time from a Drude current model and the polarization charge of a conducting wire, estimating about 4×10^-5 s for the short semiconductor but about 10^3 s for the long one. If correct, the measured conductivity gap is a transient effect that should vanish if the measurement were extended long enough.

Core claim

On its own terms, the discovery is that the reported enhancement of KCl ionic conductivity in 100 μm semiconducting nanotubes over metallic ones can be explained by the time-dependent screening of the applied electric field by mobile electrons on the nanotube wall. Metallic tubes have a very high conduction-electron density and become equipotential quickly, so ions inside feel no applied field; semiconducting tubes have much lower density and an equilibration time long compared with the 10 s measurement window, so an electric field persists inside the tube and adds a field-driven current. The paper gives this explicitly as one possible explanation, and it implies that waiting long enough wou

What carries the argument

The central object is the screening time t0, defined by the exponential decay E(t)=E(0)e^{-t/t0}. It is estimated from the Drude conductivity relation and the derivative of the nanotube's total polarization charge with respect to the applied field, dQ/dE, taken from the Landau-Lifshitz solution for the charge density on a conducting wire in a uniform field. Metallic tubes have a huge conduction-electron density, making t0 tiny, while semiconducting tubes have a small density and a large t0.

Load-bearing premise

The timescale estimates assume the nanotube behaves as an isolated conducting wire in vacuum, so it must supply the full polarization charge calculated by Landau-Lifshitz; in reality the KCl electrolyte may screen the applied field or provide another path for charge redistribution, which could change t0 and erase the predicted difference.

What would settle it

Measure the time-resolved ionic current through a 100 μm semiconducting nanotube after switching on a fixed voltage. If the current does not decay toward the metallic value on a timescale of about 10^3 s, or if increasing the electrolyte concentration to strengthen screening does not suppress the semiconducting enhancement, the central claim is falsified.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

In 11 nm tubes, both metallic and semiconducting tubes become equipotential within the measurement time, so ion flow rates match—this is the paper's explanation for the short-tube experiment.
In 100 μm tubes, an internal electric field persists in the semiconducting tube and adds n0 e^2 V/(2hλ) to the ionic current, explaining why semiconducting conductivity is higher.
If the 100 μm semiconducting tubes were measured for times much longer than about 10^3 s, the semiconducting and metallic conductivities would tend to converge.
For metallic tubes, the fast equipotentialization means ion current is controlled by entry and pressure-driven momentum transfer rather than by a direct electric force on ions inside the tube.

Where Pith is reading between the lines

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

A direct, unexplored test: apply a voltage step to a 100 μm semiconducting nanotube and record the ionic current over time—if the slow-equipotentialization picture is right, the current should decay with a roughly 10^3 s time constant.
The model ignores ionic screening by the surrounding KCl solution; including Debye screening could shorten or lengthen t0 and might change the predicted magnitude of the semiconducting enhancement.
The mechanism suggests the semiconducting enhancement should grow with tube length, since the required polarization charge and hence t0 increase with length; a non-monotonic length dependence would challenge the explanation.
If correct, this transient electronic screening could serve as a controllable knob in nanofluidic devices, tuning ion current through the tube's electronic character and the measurement window.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

Timescale-based explanation for why ions flow faster through long semiconducting carbon nanotubes — plausible and clearly argued, but the numbers rest on parameters the paper does not measure.

read the letter

Short version: this is a hypothesis paper, not a calculation paper. It proposes that the longer a carbon nanotube takes to become equipotential, the longer an internal axial field persists to drive ions. For the 100 μm tubes in Cui et al., that timescale is ~10^3 s for a semiconducting tube and ~10^-3 s for a metallic one, so during a 10 s measurement the semiconductor still has a driving field while the metal does not. That would explain the observed higher ionic conductivity in semiconducting tubes, and it naturally connects to the earlier Li et al. experiment with 11 nm tubes, where both tube types equipotentialize quickly and flows are equal. The idea is genuinely new, at least to me: the electron equilibration timescale of the tube wall is not something the nanofluidics literature has been thinking about. The paper also states a crisp falsifiable prediction — wait longer than t0 and the semiconducting/metal difference should disappear.

What it does well: the author is explicit that this is one possible explanation, not the explanation. The short-tube case is grounded in the prior Lau–Sokoloff model, and the long-tube case has a quantitative skeleton: Landau–Lifshitz charge distribution for a wire, Drude conductivity, exponential relaxation. The estimates are at least self-consistent order-of-magnitude arguments.

Soft spots, in order of seriousness. First, the wire-in-vacuum electrostatics ignores the KCl electrolyte entirely. The surrounding ions will partially screen the applied field and offer a double-layer charging path, so the required charge and the timescale could shift by orders of magnitude. The paper gives no argument for why this is negligible. Second, the scattering time τ is not measured; the text explicitly says it should be determined from the empty-tube conductivity. The stress-test concern is valid: high-quality semiconducting CNTs can have mean free paths of hundreds of nm to μm, not 10 nm. Since t0 ∝ 1/τ, that alone drops t0 from 10^3 s into the 10–100 s range, dangerously close to or below the measurement window. Doping by electrolyte would shorten it further. So the central quantitative claim is fragile. Third, the derivation from Eq. (4) to Eq. (11) is compressed; dQ/dE is asserted rather than derived, and the circumference enters without much setup. None of this is fatal to the mechanism — it could survive with revised parameters — but as written the numbers should be read as illustrative, not predictive. There are also presentational issues: the abstract says "100 long" (missing μm), equation numbering is duplicated, and some references are inconsistent.

Who should read it: anyone working on ion transport through nanotubes or membranes. It deserves a serious referee. My recommendation: send it out, but expect the author to either measure or bracket τ, address electrolyte screening, and show the timescale estimate is robust over a plausible parameter range. If the mechanism survives that, it is a worthwhile contribution.

Referee Report

4 major / 4 minor

Summary. The paper proposes an electronic mechanism to explain why, in 100-μm-long carbon nanotubes, the ionic conductivity of KCl is higher in semiconducting than in metallic nanotubes, whereas in the 11-nm-long tubes of Li et al. the flow rates are almost the same. The author calculates the electronic equilibration time t0 needed for a nanotube in an applied field to become an equipotential, using a Drude relaxation model and the Landau-Lifshitz charge distribution on a conducting wire. For short semiconducting tubes t0 ≈ 4.56×10^-5 s, so both metallic and semiconducting tubes equilibrate quickly and ionic flow is similar. For 100-μm semiconducting tubes the paper estimates t0 ≈ 10^3 s, longer than the ~10 s measurement window, so an internal electric field persists inside the semiconducting tube and drives ions, explaining the larger ionic conductivity. For metallic long tubes t0 is much shorter, so the internal field is cancelled.

Significance. If quantitatively reliable, the proposed mechanism would be a valuable and falsifiable explanation of chirality-dependent ionic transport in carbon nanotubes. It explicitly connects two experiments and makes a specific numerical prediction for the equilibration time, and it identifies the measurement (conductivity of the empty nanotube) needed to fix the scattering time. However, the central conclusion is not currently supported: the key timescale depends on unmeasured parameters and on an electrostatic model that omits the electrolyte environment. The paper is also very rough in presentation. The idea is worth pursuing, but the quantitative case requires substantial additional work.

major comments (4)

[Section II, Eq. (11)] The key timescale t0 for the 100-μm semiconducting nanotube is not robust. Eq. (11) gives t0 ∝ 1/(n τ). The text assumes a mean free path of 10 nm (τ = 1.25×10^-14 s) but explicitly states that 'The correct value of τ can be determined by measuring the conductivity of the empty nanotube' — this measurement is not provided. High-quality semiconducting CNTs can have mean free paths of hundreds of nm to a micron, which changes τ (and hence t0) by one to two orders of magnitude. Doping by the KCl electrolyte could increase n and further shorten t0. With such variations, t0 could fall below the ~10 s measurement window of Cui et al., eliminating the internal field and the proposed explanation. The manuscript needs in-situ bounds or measurements of n and τ, or an argument that the assumed values are conservative lower bounds on t0.
[Section II, Eqs. (6)–(8)] The charge needed to cancel the external field is taken from the Landau-Lifshitz solution for an isolated conducting wire in vacuum. The nanotube, however, is immersed in a KCl electrolyte. Ionic screening and double-layer charging at the nanotube/electrolyte interface provide an additional path for charge redistribution and modify the effective capacitance per unit length. No estimate of the Debye length, double-layer capacitance, or comparison with the vacuum-wire capacitance is given. If the screening is significant, the calculated t0 could be changed by orders of magnitude, and the central explanation would fail. This point is load-bearing and should be addressed quantitatively.
[Section II, Eqs. (10a)–(11)] The derivation from the Drude current to the exponential relaxation is compressed and dimensionally unclear. Eq. (4) defines a three-dimensional current density j = n e^2 τ E / m*, but Eq. (10a) multiplies by the circumference C_h without specifying the conducting cross-section or the dimensionality of n (per unit area or per unit volume). The relation between dQ/dE, C_h, and the current is asserted rather than derived. Since the central quantitative estimate t0 comes from this equation, the derivation should be redone with a clear accounting of geometry and units.
[Section II, Eq. (8) and t0 for the long tube] The stated value t0 ≈ 1.00×10^3 s for a 100-μm semiconducting tube does not obviously follow from the displayed equations. Using Eq. (8) with Q ∝ E L/[6 ln(L/a) - 14], the ratio of t0 for L = 50 μm to t0 for the short tube (L = 5.5 nm) is roughly 300, giving t0 ≈ 0.01 s, not 10^3 s. If a different expression for dQ/dE is intended (e.g., one proportional to L^2), it should be stated explicitly and evaluated for both geometries. This is not a cosmetic issue: the entire argument for the long-tube experiment requires t0 to exceed the ~10 s measurement time.

minor comments (4)

[Throughout] There are many OCR/presentation errors: duplicate equation number (Eq. (4) is used for both the effective mass and the Drude current), '100 long' missing 'μm' in the abstract, references to 'Ref. 3' where the intended reference appears to be 'Ref. 11', and a reference to 'Fig. 9' when the plot appears to be Fig. 1.
[Eqs. (1)–(3)] The equations for the carrier density are garbled in the submitted version (symbols such as β, ε, η are not typeset correctly). This makes it difficult to verify the numerical values n0 = 3.44×10^7 nm^-2 and n0 = 6.53×10^14 nm^-2. The author should provide a cleanly typeset derivation and check the dimensions.
[References] Ref. 14 is a private communication with Ming Ma. If experimental details (e.g., the 10 s measurement time) rely on this private communication, they should be either published or stated in the text so that the reader can assess the basis of the numbers.
[Section III] The conclusion refers to 'Ref. 1' and 'Ref. 3' where the context indicates Refs. 9 and 11. This should be corrected.

Circularity Check

0 steps flagged

No circularity: the long-tube internal-field explanation is derived from electrostatics and Drude transport, not restated from the measured conductivity difference; unmeasured τ is a robustness issue.

full rationale

The central derivation in Section II computes the equipotential time t0 from Landau-Lifshitz's isolated-wire charge density (Ref. 13), Eqs. (6)-(8), and Drude transport, Eqs. (10)-(11), using independently stated parameters: band gap 0.5 eV, v_F = 8e5 m/s, L and a from the experiments, and an assumed mean free path l_c = 10 nm giving τ = 1.25e-14 s. The predicted inequality t0 (about 10^3 s for the 100-micron semiconducting tube) > 10 s measurement window is then compared with the reported conductivity difference; it is not obtained by fitting that difference. The paper explicitly flags that τ is not measured ('The correct value of τ can be determined by measuring the conductivity of the empty nanotube'), which is a parameter-identification/robustness limitation rather than a circular reduction. The reliance on the author's prior Ref. 10 is for the short-tube equal-flow-rate interpretation, but the long-tube claim is separately derived here. Electrolyte screening of the external field (not included in the vacuum-wire Landau-Lifshitz solution) is an assumption, not a restatement of the target result. Hence no step reduces by construction to its input, and the central claim has independent content.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

The central claim rests on a small number of parameters and assumptions. The key free parameter is the scattering time τ, which directly controls the equilibration time, and the assumed band gap g controls the carrier concentration. The most fragile assumption is that the nanotube can be treated as an isolated wire in vacuum for calculating the induced charge in an electrolyte environment. No new entities are introduced.

free parameters (2)

Scattering time τ (mean free path l)
Appears in Eq. (10)-(11); t0 is inversely proportional to τ. The paper says τ can be determined by measuring the empty nanotube conductivity, but no value is fixed, making the equilibration time adjustable.
Band gap g = 0.5 eV
Assumed for the semiconducting nanotube. The intrinsic carrier concentration n0 in Eq. (2) depends exponentially on g, so t0 is highly sensitive to this choice; the paper does not measure or justify g for the specific nanotubes in Ref. 11.

axioms (4)

domain assumption The charge density on a nanotube in an applied field is given by the Landau-Lifshitz solution for an isolated metallic wire in vacuum (Eq. 6-8).
Used to compute the charge Q that must flow along the nanotube to cancel the external field. This ignores the electrolyte environment and ionic screening, which could alter the induced charge and the equilibration time.
domain assumption The Drude model applies to the conduction electrons in the nanotube wall (Eq. 4).
The electron current that redistributes charge along the tube is assumed to follow the Drude formula with a scattering time τ; this is a simple model, and its applicability to subnanometer nanotubes is not discussed.
domain assumption Ion transport through an equipotential nanotube proceeds as a line of ions transferring momentum by collisions (Ref. 10, Appendix B).
Adopted from the author's prior work to argue that the flow rate in short metallic/semiconducting tubes is independent of wall friction. The validity of this model is not established in the present paper.
standard math The intrinsic carrier concentration in the semiconducting tube is given by thermal excitation across a fixed gap g (Eq. 2).
Uses standard semiconductor statistics. Assumes undoped tube and no gate or electrochemical doping, which may not hold in the experiment.

pith-pipeline@v1.3.0-alltime-deepseek · 4999 in / 14317 out tokens · 139842 ms · 2026-08-04T00:59:35.868047+00:00 · methodology

0 comments

read the original abstract

Measurements made by Li, et. al., showed that the flow rates of potassium ions through 11nm long, subnanometer diameter, metallic and semiconducting carbon nanotubes under an applied electric field are almost the same. In contrast, measurements of the electrical conductivity of potassium chloride solution through 100 long metallic and semiconducting carbon nanotubes with diameters between 2.6-5.4nm by Cui, et. al., show that the ionic conductivity in the semiconducting nanotubes is larger than that of the metallic nanotubes. Possible theoretical origins of these differences are explored in this article.

Figures

Figures reproduced from arXiv: 2608.00211 by J. B. Sokoloff.

**Figure 2.** Figure 2: A sketch of the experimental set-up in Ref. 11. Shown are the two positive and the two negative electrodes, which are poles perpendicular to the planar structure in which the fluid and the ions flow, and the etched guiding structures (illustrated in the sketch) that direct the ions into the nanotube. Let us consider a simplified model of the set-up in Ref. 11, illustrated in [PITH_FULL_IMAGE:figures/full_… view at source ↗

discussion (0)

Reference graph

Works this paper leans on

14 extracted references · 1 linked inside Pith

[1]

Kaira, S

A. Kaira, S. Garde, and G. Hummer, Proc. Nat. Acad. Sci. USA 100, 10175 (2003)

2003
[2]

Fornasiero, et

F. Fornasiero, et. al., Proc. Nat. Acad. Sci. USA 105, 17250 (2008)

2008
[3]

Corry, J

B. Corry, J. Phys. Chem. B 112, 1427 (2008)

2008
[4]

K. D. Kreuer, S. J. Paddison, E. Spohr and M. Schuster, Chem. Rev. 104, 4637 (2004)

2004
[5]

Dellago, M

C. Dellago, M. M. Naor and G. Hummer, Phys. Rev. Lett. 90, 10590 (2003)

2003
[6]

Dellago and G

C. Dellago and G. Hummer, Phys. Rev. Lett.\ 97, 245901 (2006)

2006
[7]

Köfinger, G

J. Köfinger, G. Hummer and C. Dellago, Proc. Natl Acad. Sci. USA 105, 13218 (2008)

2008
[8]

Kavokine, M.-L

N. Kavokine, M.-L. Bocquet, and L. Bocquet, Nature 602, 84 (2022)

2022
[9]

Yuhao Li et al., Nature Materials 23, 1123–1130 (2024)

2024
[10]

A. W. C. Lau and J. B. Sokoloff. Arxiv 2606.00314

Pith/arXiv arXiv
[11]

Guandong Cui, Xu Zhu, Shuchen Zhang, Alessandro Siria and Ming Ma, Science Advances 11, 7410 (2025)

2025
[12]

Charlier, Rev. Mod. Phys. 79, 677 (2007)

2007
[13]

L. D. Landau and E. M. Lifshit͡ s, L. P. Pitaevskiĭ, Electrodynamics of Continuous Media (Oxford ; New York : Pergamon 1984)

1984
[14]

Private communication with Ming Ma, one of the authors of Re. 3. Acknowledgements: I wish to thank Ming Ma for many useful discussions of the results reported in Ref. 3

This paper was first reviewed by deepseek-v4-flash on August 4, 2026.

[1] [1]

Kaira, S

A. Kaira, S. Garde, and G. Hummer, Proc. Nat. Acad. Sci. USA 100, 10175 (2003)

2003

[2] [2]

Fornasiero, et

F. Fornasiero, et. al., Proc. Nat. Acad. Sci. USA 105, 17250 (2008)

2008

[3] [3]

Corry, J

B. Corry, J. Phys. Chem. B 112, 1427 (2008)

2008

[4] [4]

K. D. Kreuer, S. J. Paddison, E. Spohr and M. Schuster, Chem. Rev. 104, 4637 (2004)

2004

[5] [5]

Dellago, M

C. Dellago, M. M. Naor and G. Hummer, Phys. Rev. Lett. 90, 10590 (2003)

2003

[6] [6]

Dellago and G

C. Dellago and G. Hummer, Phys. Rev. Lett.\ 97, 245901 (2006)

2006

[7] [7]

Köfinger, G

J. Köfinger, G. Hummer and C. Dellago, Proc. Natl Acad. Sci. USA 105, 13218 (2008)

2008

[8] [8]

Kavokine, M.-L

N. Kavokine, M.-L. Bocquet, and L. Bocquet, Nature 602, 84 (2022)

2022

[9] [9]

Yuhao Li et al., Nature Materials 23, 1123–1130 (2024)

2024

[10] [10]

A. W. C. Lau and J. B. Sokoloff. Arxiv 2606.00314

Pith/arXiv arXiv

[11] [11]

Guandong Cui, Xu Zhu, Shuchen Zhang, Alessandro Siria and Ming Ma, Science Advances 11, 7410 (2025)

2025

[12] [12]

Charlier, Rev. Mod. Phys. 79, 677 (2007)

2007

[13] [13]

L. D. Landau and E. M. Lifshit͡ s, L. P. Pitaevskiĭ, Electrodynamics of Continuous Media (Oxford ; New York : Pergamon 1984)

1984

[14] [14]

Private communication with Ming Ma, one of the authors of Re. 3. Acknowledgements: I wish to thank Ming Ma for many useful discussions of the results reported in Ref. 3