Electro-Thermal Transport in Metallic Single-Wall Carbon Nanotubes for Interconnect Applications

David Mann; Eric Pop; Hongjie Dai; John Reifenberg; Kenneth Goodson

REVIEW

Electro-Thermal Transport in Metallic Single-Wall Carbon Nanotubes for Interconnect Applications

Eric Pop, David Mann, John Reifenberg, Kenneth Goodson, Hongjie Dai

Mesoscale and Nanoscale Physics · Submitted 5 Mar 2006 · 18 references

Pith's verdict: A useful electro-thermal model for micron-scale metallic SWNT interconnects, with a plausible but not fully nailed-down optical-phonon absorption mechanism.

T0 · AI review

How this review was made: An AI referee (deepseek-v4-flash) read the paper on August 28, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

T0 AI review An AI read the paper and answered a fixed list of questions.
T1 Audit Automatic checks that the references and files the paper relies on exist and match.
T2 Compute The paper's code or calculations re-run with the same results.
T3 Formal statement The claims written out in a formal math language a computer can read.
T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 3 major objections · 4 minor · Ask about this paper · The compiled record · Read the full review →

Challenge this review Comment (0) Read on arXiv ↗

Cite ▾

arXiv IDcond-mat/0603110 (version 1)

Pith IDEMRJJBF6, Pith's permanent record of this paper and its review

CitationDavid Mann, Eric Pop, Hongjie Dai, John Reifenberg, Kenneth Goodson. "Electro-Thermal Transport in Metallic Single-Wall Carbon Nanotubes for Interconnect Applications." arXiv:cond-mat/0603110 (2006). https://pith.science/paper/EMRJJBF6

Cite this reviewBibTeX for Pith's review of this paper

Claim paper

Why Pith says this

Highlight any text to comment on it.

Editor's note

read the letter

Colleague,

Quick take: this is a solid, useful paper. It couples a self-consistent electro-thermal model to measurements of metallic SWNTs and makes a specific claim—optical-phonon (OP) absorption affects low-bias resistance above 250 K—that is physically plausible but rests on a single-device comparison with no error bars or sensitivity analysis. Worth reading, worth citing, and worth sending to referees, but I would not treat the OP absorption mechanism as proven until it is reproduced across several tubes and lengths.

What is genuinely new: the first electro-thermal treatment of metallic SWNTs as interconnect-like elements, the OP absorption mechanism for low-bias R(T), and length-dependent breakdown scaling. The model is not a blind curve fit; it solves the heat equation with self-consistent Joule heating and scattering, and it reproduces I-V curves at three ambient temperatures plus breakdown voltages for lengths up to about 5 µm. The design guideline (keep power density below 5 µW/µm) is a useful rule of thumb. The assumption that substrate contacts suppress severe OP non-equilibrium is reasonable and clearly stated.

The soft spots: the OP absorption claim in Fig. 3 is the load-bearing result, yet it comes from one tube, with R_C ≈ 24 kΩ estimated and λ_AC,300 = 1600 nm taken from prior work. The dashed 'excluding OP' curve depends on both of these. The paper gives no error bars and no sensitivity sweep over λ_AC,300 or R_C(T). The stress-test note worries that a temperature-dependent R_C could absorb the excess; I think that specific direction is not quite right—if R_C decreases with T, the mismatch with the acoustic-only model would get worse, not better. But the broader point stands: the attribution hinges on R_C being temperature-independent, and that is not experimentally verified. The length-scaling prediction in Fig. 5 is a nice testable consequence, but no length-dependent R(T) data are shown. The breakdown data in Fig. 9 mostly constrains the substrate thermal conductance, not the OP absorption. Self-citation to the authors' own PRL is fine; the scattering parameters come from that work and are used consistently.

Overall, the central argument holds up as a plausible mechanism, not a proven one. The equations are transparent enough for a referee to reproduce, but the raw data and parameter files are not included.

For peer review: yes, send it. A good referee can push for more devices, error bars, and a sensitivity analysis. I would accept after minor to moderate revision.

Referee report

Major objections (3)

[Electrical Transport, Eqs. (3)–(4) and Fig. 3] The attribution of the low-bias resistance rise above 250 K to optical-phonon absorption rests on the residual between the data and an acoustic-phonon-only model computed with λ_AC,300 = 1600 nm taken from prior work and a temperature-independent RC ≈ 24 kΩ. No sensitivity analysis is given for λ_AC,300, and RC(T) is not measured; a smaller λ_AC,300 or a contact resistance that decreases with increasing temperature could absorb part of the apparent OP-absorption contribution. The authors should provide a sensitivity analysis and, if possible, a length-scaling test using at least one additional device.
[Thermal Transport, Eq. (8) and following paragraph] The model assumes that the optical-phonon population in the tube remains in thermal equilibrium with the lattice because the substrate facilitates fast OP decay. This assumption is asserted rather than tested; if a non-equilibrium OP population exists also for substrate-supported tubes, the thermal N_OP(T) used in Eq. (4) would overestimate the OP-absorption contribution above 250 K. The authors should state the experimental or prior evidence that supports this equilibrium assumption for their geometry, or quantify the error if it fails.
[Comparison with Data and Predictions; Fig. 9] The thermal conductance to the substrate is used as an adjustable input (g ≈ 0.15 W K−1 m−1 for resistance data and g ≈ 0.2 W K−1 m−1 for breakdown), and the breakdown temperature is assumed to be T_BD ≈ 800 K rather than measured for these devices. Because the self-heating portion of the model and the breakdown-scaling claim are validated with these adjusted inputs, the paper should explicitly quantify the sensitivity of the conclusions to these choices and justify the difference between the two g values used.

Minor objections (4)

[Figure 3 caption] The caption reads 'Data (sym bols)' and should read 'Data (symbols)'.
[Electrical Transport, second paragraph] The text contains a grammatical error: 'although previous neglected' should be 'although previously neglected'.
[Fig. 2 and text] The ambient temperatures are listed both as 'T = 100, 200, 293 K' and 'T0 = 100, 200, 293 K'; please use consistent notation throughout.
[Electrical Transport, Eq. (1)] A table listing all symbols, their definitions, and units would help readers, particularly for the distinction between RC, the electrical contact resistance, and the quantum contact resistance h/4q² mentioned later.

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

The paper establishes that in micron-length metallic carbon nanotubes, optical-phonon absorption raises low-bias resistance above 250 K, and Joule self-heating becomes significant beyond roughly 1 V, yielding a 5 µW/µm power-density…

The reading

This paper asks whether Joule heating and temperature-dependent electron scattering matter for metallic single-wall carbon nanotubes long enough to be plausible on-chip interconnects, roughly one to ten microns. The answer is yes on both counts. A thermally self-consistent transport model, matched to measured current-voltage curves at 100, 200, and 293 K, shows that the low-bias resistance rises above about 250 K because electrons absorb optical phonons, a mechanism earlier work set aside because optical phonons have high energy. The same model shows that high-bias self-heating is significant for tubes shorter than about 10 µm, that breakdown in air follows from the tube midpoint reaching roughly 800 K, and that keeping power density below about 5 µW/µm avoids noticeable self-heating.

What carries the argument

The central object is a thermally self-consistent resistance model. The total resistance is written as $R = R_C + (h/4q^2)(1 + L/\lambda_{\mathrm{eff}})$, with the effective electron mean free path $\lambda_{\mathrm{eff}}$ combining acoustic-phonon scattering, optical-phonon emission (both field-driven and after an absorption event), and the newly included optical-phonon absorption: $\lambda_{\mathrm{eff}}^{-1} = \lambda_{\mathrm{AC}}^{-1} + \lambda_{\mathrm{OP,ems}}^{-1} + \lambda_{\mathrm{OP,abs}}^{-1}$. Each mean free path is expressed through phonon occupation numbers scaled to 300 K values, and the temperature dependence enters through $N_{\mathrm{OP}}(T) = 1/[\exp(\hbar\omega_{\mathrm{OP}}/k_BT)-1]$ with $\hbar\omega_{\mathrm{OP}}\approx 0.18$ eV. This resistance feeds a one-dimensional heat equation, $A\nabla\cdot(k_{th}\nabla T) - g(T-T_0) + p' = 0$, where $p' = I^2(R-R_C)/L$ is Joule heating per unit length and $g \approx 0.15$ W K$^{-1}$ m$^{-1}$ is the thermal conductance to the substrate; the two equations are iterated until the temperature converges to 0.1 K at every bias point. The load-bearing identity is that the OP absorption length is about 10 $\mu$m at room temperature, which is why the mechanism matters only for micron-length tubes.

Extended reading notes

Core claim

The paper's core claim is that metallic SWNTs in the interconnect-relevant length range are not isothermal resistors. At low bias, the resistance increases with temperature above roughly 250 K, and the increase is attributed to optical-phonon absorption: at 300 K the OP absorption mean free path approaches 10 µm, so the mechanism is negligible for sub-micron tubes but becomes important exactly when L > 1 µm. At high bias, Joule heating raises the tube temperature above the ambient, and the measured I-V curves are reproduced only when the heat equation is solved along the tube with heat loss to the substrate and then iterated together with the resistance. The same calculation predicts that breakdown in air occurs when the highest-temperature point, the midpoint, reaches about 800 K, and that the breakdown voltage scales roughly linearly with length. These results are collected into a model the paper presents as ready for circuit simulators.

Load-bearing premise

The result assumes the optical-phonon population in a substrate-supported tube stays in thermal equilibrium with the lattice because the substrate decays optical phonons quickly; if hot optical phonons accumulate instead, the predicted resistance rise above 250 K would be overestimated.

Editorial extensions

If this is right

At operating temperatures above about 250 K, circuit models of micron-length metallic SWNTs that include only acoustic-phonon and optical-phonon emission will underpredict the low-bias resistance; optical-phonon absorption must be added.
A power-density guideline emerges: keep Joule power below about 5 µW/µm in 1–10 µm tubes to avoid noticeable self-heating.
Longer tubes (L > 10 µm) run cooler than short tubes at the same total power, because the power density is lower and heat escapes into the substrate all along the length.
The breakdown voltage of SWNTs in air should scale approximately linearly with length, with failure at roughly 800 K at the tube midpoint; this can be used as a non-electrical check of the thermal model.

Reading between the lines

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

A natural test of the model's substrate assumption would be to measure low-bias resistance versus temperature for the same tube both supported on SiO₂ and suspended; the model implies a weaker or delayed resistance rise above 250 K when the optical-phonon population can leave equilibrium.
If the model is right, bundle-level interconnect power budgets should scale roughly as the number of parallel tubes times the single-tube 5 µW/µm limit, provided inter-tube thermal coupling is weak.
Because the onset temperature depends on the optical-phonon energy through the Bose factor, a systematic diameter or chirality sweep that shifts $\hbar\omega_{\mathrm{OP}}$ should move the temperature at which the resistance rise appears; measuring that shift would pin down the effective OP energy more tightly than the 0.18 eV estimate used here.

Original abstract

This work represents the first electro-thermal study of metallic single-wall carbon nanotubes (SWNTs) for interconnect applications. Experimental data and careful modeling reveal that self-heating is of significance in short (1 < L < 10 um) nanotubes under high-bias. The low-bias resistance of micron scale SWNTs is also found to be affected by optical phonon absorption (a scattering mechanism previously neglected) above 250 K. We also explore length-dependent electrical breakdown of SWNTs in ambient air. Significant self-heating in SWNT interconnects can be avoided if power densities per unit length are limited to less than 5 uW/um.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13-20260917 via OpenRouter, question bank abstract-p1-20260923, 2026-09-23: 1 request, 4,245 input tokens, $0.0002 reported.

What the abstract itself states, as probabilities: claims first 0.98 · states limitation 0.33 · human participants 0.01 · controlled comparison 0.06 · randomized assignment 0.01 · states quantitative result 0.98 · compares with prior results 0.16 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

This work represents the first electro-thermal study of metallic single-wall carbon nanotubes (SWNTs) for interconnect applications. Experimental data and careful modeling reveal that self-heating is of significance in short (1 < L < 10 um) nanotubes under high-bias. The low-bias resistance of micron scale SWNTs is also found to be affected by optical phonon absorption (a scattering mechanism previously neglected) above 250 K. We also explore length-dependent electrical breakdown of SWNTs in ambient air. Significant self-heating in SWNT interconnects can be avoided if power densities per unit length are limited to less than 5 uW/um.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	This work represents the first electro-thermal study of metallic single-wall carbon nanotubes (SWNTs) for interconnect applications.	contribution 0.97	0.83	0.06	asserted 1.00
s1	Experimental data and careful modeling reveal that self-heating is of significance in short (1 < L < 10 um) nanotubes under high-bias.	result 1.00	0.95	0.27	asserted 1.00
s2	The low-bias resistance of micron scale SWNTs is also found to be affected by optical phonon absorption (a scattering mechanism previously neglected) above 250 K.	result 1.00	0.94	0.09	asserted 1.00
s3	We also explore length-dependent electrical breakdown of SWNTs in ambient air.	contribution 1.00	0.92	0.04	asserted 0.55
s4	Significant self-heating in SWNT interconnects can be avoided if power densities per unit length are limited to less than 5 uW/um.	interpretation 0.53	0.81	0.45	asserted 0.94

What kind of paper the classifier reports this is: empirical study 1.00 · evidence from observation or experiment 1.00.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures

**Figure 2.** Current-voltage (experimental data and model) of a metallic SWNT with L ≈ 3 µm and d ≈ 2.4 nm in ambient T0 = 100, 200 and 293 K from top to bottom. The isothermal model deviates from the data beyond V > 1 V, suggesting SWNT self-heating. T0 = 100, 200, 293 K

**Figure 4.** Estimate of the various MFPs at low-bias as a function of temperature. The total MFP (λeff in Eq. 2) is also plotted (solid line). AC scattering dominates at low bias as expected, but OP scattering (with MFPs ≈ 10 µm at 300 K) cannot be neglected at higher temperatures, especially for longer tubes (see

**Figure 5.** Computed temperature dependence of the low-bias resistance assuming ideal electrical contacts (RC = 0) for various length tubes. The quantum contact resistance (h/4q2 ≈ 6.5 kΩ) is naturally still present (also see Figs. 3 and 4)

**Figure 6.** summarizes the resistance components of a 2 µm single-wall nanotube with 2 µm long contacts. We use the 2 x 10 ~ 15 9 5 x 10 Contacts 6 2 x 10 ~ 15 9 3 x 10 Nanotube 8 Electrical Resis. Electrons (kΩ) Thermal Resis. Electrons (K/W) Thermal Resis. Phonons (K/W) 2 x 10 ~ 15 9 5 x 10 Contacts 6 2 x 10 ~ 15 9 3 x 10 Nanotube 8 Electrical Resis. Electrons (kΩ) Thermal Resis. Electrons (K/W) Thermal Resis. Phonons (K/W)

**Figure 7.** High-bias I-V characteristics computed with (solid lines) and without (dashed lines) taking self-heating into account. The ambient is T0 = 293 K and the resistance parameters listed in

**Figure 8.** Estimated temperature profiles along the SWNTs from

**Figure 9.** Computed and experimentally measured breakdown voltages (VBD) for metallic SWNTs of various lengths exposed to air. The data was obtained in the course of this work and from Refs. [10,18]. Breakdown occurs by oxidation when the peak nanotube temperature (in their middle) reaches ≈ 800 K [17]. Inspection by AFM imaging revealed that tubes were indeed broken at or very near their midpoint, as expected [10]

Showing 7 of the paper's 9 figures. See all figures in the PDF ↗ Figures reproduced from arXiv:cond-mat/0603110 by the authors.

References

Works cited (18)

[1]

F. Kreupl et al. Proc. IEDM p. 683 (2004)

View on Pith 2004
[2]

P. McEuen et al., IEEE Trans. Nano. 1, 78 (2002)

View on Pith 2002
[3]

A. Naeemi et al., IEEE EDL 26, 84 (2005)

View on Pith 2005
[4]

A. Raychowdhury and K. Roy, ICCAD p. 237 (2004)

View on Pith 2004
[5]

N. Srivastava and K. Banerjee, VMIC p. 393 (2004)

View on Pith 2004
[6]

C. Yu et al., Nano Letters, 5, 1842 (2005)

View on Pith 2005
[7]

E. Pop et al., Phys. Rev. Lett. 95, 155505 (2005)

View on Pith 2005
[8]

Z. Yao et al., Phys. Rev. Lett. 84, 2941 (2000)

View on Pith 2000
[9]

J. Y. Park et al., Nano Letters 4, 517 (2004)

View on Pith 2004
[10]

A. Javey et al., Phys. Rev. Lett. 92, 106804 (2004)

View on Pith 2004
[11]

J. Small et al., Solid State Comm. 127, 181 (2003)

View on Pith 2003
[12]

S. Huxtable et al., Nature Materials 2, 731 (2003)

View on Pith 2003
[13]

D. Cahill et al., J. Appl. Phys. 93, 793 (2003)

View on Pith 2003
[14]

W. Steinhogl et al., Phys. Rev. B 66, 075414 (2002)

View on Pith 2002
[15]

C. Kittel, Intro Solid-State Phys, Wiley & Sons (1995)

View on Pith 1995
[16]

T. Yamamoto et al., Phys. Rev. Lett. 92, 075502 (2004)

View on Pith 2004
[17]

K. Hata et al., Science 306, 1362 (2004)

View on Pith 2004
[18]

P. Qi et al., J. Am. Chem. Soc. 126, 11774 (2004). 0 1 2 3 4 5 0 5 10 15 20 25 L (µm) VBD (V) Model Data Figure 9. Computed and experimentally measured breakdown voltages ( VBD) for metallic SWNTs of various lengths ex posed to air. The data was obtained in the course of this work and from Refs. [10,18]. Breakdown occurs by oxidation when the peak nanotub...

View on Pith 2004

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.

[1] [1]

F. Kreupl et al. Proc. IEDM p. 683 (2004)

View on Pith 2004

[2] [2]

P. McEuen et al., IEEE Trans. Nano. 1, 78 (2002)

View on Pith 2002

[3] [3]

A. Naeemi et al., IEEE EDL 26, 84 (2005)

View on Pith 2005

[4] [4]

A. Raychowdhury and K. Roy, ICCAD p. 237 (2004)

View on Pith 2004

[5] [5]

N. Srivastava and K. Banerjee, VMIC p. 393 (2004)

View on Pith 2004

[6] [6]

C. Yu et al., Nano Letters, 5, 1842 (2005)

View on Pith 2005

[7] [7]

E. Pop et al., Phys. Rev. Lett. 95, 155505 (2005)

View on Pith 2005

[8] [8]

Z. Yao et al., Phys. Rev. Lett. 84, 2941 (2000)

View on Pith 2000

[9] [9]

J. Y. Park et al., Nano Letters 4, 517 (2004)

View on Pith 2004

[10] [10]

A. Javey et al., Phys. Rev. Lett. 92, 106804 (2004)

View on Pith 2004

[11] [11]

J. Small et al., Solid State Comm. 127, 181 (2003)

View on Pith 2003

[12] [12]

S. Huxtable et al., Nature Materials 2, 731 (2003)

View on Pith 2003

[13] [13]

D. Cahill et al., J. Appl. Phys. 93, 793 (2003)

View on Pith 2003

[14] [14]

W. Steinhogl et al., Phys. Rev. B 66, 075414 (2002)

View on Pith 2002

[15] [15]

C. Kittel, Intro Solid-State Phys, Wiley & Sons (1995)

View on Pith 1995

[16] [16]

T. Yamamoto et al., Phys. Rev. Lett. 92, 075502 (2004)

View on Pith 2004

[17] [17]

K. Hata et al., Science 306, 1362 (2004)

View on Pith 2004

[18] [18]

P. Qi et al., J. Am. Chem. Soc. 126, 11774 (2004). 0 1 2 3 4 5 0 5 10 15 20 25 L (µm) VBD (V) Model Data Figure 9. Computed and experimentally measured breakdown voltages ( VBD) for metallic SWNTs of various lengths ex posed to air. The data was obtained in the course of this work and from Refs. [10,18]. Breakdown occurs by oxidation when the peak nanotub...

View on Pith 2004