Flexible radiofrequency carbon nanotube transistors operating at frequencies above 100 GHz

Chao Ma; Eric Pop; Fan Xia; Guanhua Long; Haotian Su; Lanyue Gan; Lian-Mao Peng; Qianlan Hu; Ruyi Huang; Shan X. Wang

arxiv: 2502.02485 · v4 · pith:UQCZQWDHnew · submitted 2025-02-04 · ⚛️ physics.app-ph · cond-mat.mes-hall· cond-mat.mtrl-sci

Flexible radiofrequency carbon nanotube transistors operating at frequencies above 100 GHz

Fan Xia , Tian Xia , Haotian Su , Lanyue Gan , Qianlan Hu , Wanyi Wang , Ruyi Huang , Tianshun Bai

show 7 more authors

Yufan Chen Chao Ma Guanhua Long Shan X. Wang Eric Pop Lian-Mao Peng Youfan Hu

This is my paper

Pith reviewed 2026-05-23 03:38 UTC · model grok-4.3

classification ⚛️ physics.app-ph cond-mat.mes-hallcond-mat.mtrl-sci

keywords carbon nanotubeflexible electronicsradiofrequency transistorcutoff frequencyelectro-thermal co-designflexible substrateRF amplifierK-band

0 comments

The pith

Aligned carbon nanotube arrays on flexible substrates yield radiofrequency transistors with f_T above 150 GHz via electro-thermal co-design.

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

The paper sets out to demonstrate that carbon nanotube transistors on flexible substrates can reach current and power gain cutoff frequencies above 100 GHz. Poor heat conduction in those substrates normally limits performance, so the work introduces electro-thermal co-design to manage dissipation while preserving high-frequency operation. Reported metrics include an on-state current of 0.947 mA per micrometer, transconductance of 0.728 mS per micrometer, peak extrinsic f_T of 152 GHz, and peak extrinsic f_max of 102 GHz at under 200 mW per mm. The same devices form flexible amplifiers delivering 64 mW per mm output power with 11 dB gain in the K-band. A sympathetic reader would care because these numbers target the frequency range required for sixth-generation wireless terminals that must also bend and consume little power.

Core claim

Radiofrequency transistors based on aligned carbon nanotube arrays on flexible substrates exhibit current gain cutoff frequencies f_T and power gain cutoff frequencies f_max above 100 GHz when electro-thermal co-design is used to improve heat dissipation and radiofrequency performance. The transistors show an on-state current of 0.947 mA μm^{-1}, transconductance of 0.728 mS μm^{-1}, peak extrinsic f_T of 152 GHz, peak extrinsic f_max of 102 GHz, and power consumption under 200 mW mm^{-1}. The devices can form flexible radiofrequency amplifiers with 64 mW mm^{-1} output power and 11 dB power gain in the K-band.

What carries the argument

electro-thermal co-design on aligned carbon nanotube arrays, which simultaneously manages heat flow and maintains radiofrequency characteristics on low-conductivity flexible substrates.

If this is right

The transistors deliver an on-state current of 0.947 mA per micrometer and transconductance of 0.728 mS per micrometer.
Extrinsic current-gain cutoff reaches a peak of 152 GHz and power-gain cutoff reaches 102 GHz.
Power draw remains below 200 mW per millimeter at these frequencies.
Amplifiers fabricated from the transistors produce 64 mW per mm output power with 11 dB gain inside the K-band.

Where Pith is reading between the lines

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

The approach may allow flexible terminals for 6G systems that must conform to curved surfaces while operating at millimeter-wave frequencies.
Repeating the co-design on other low-thermal-conductivity materials would test whether the performance gain is tied specifically to carbon nanotube alignment.
The reported power levels suggest the transistors could be integrated into battery-limited wearable or portable high-frequency circuits.

Load-bearing premise

The claimed cutoff frequencies above 100 GHz arise from the electro-thermal co-design's heat-dissipation benefit rather than from unaccounted parasitics or measurement effects on the flexible substrate.

What would settle it

Independent thermal imaging of the channel temperature at the reported bias points combined with complete S-parameter de-embedding that reproduces the stated extrinsic f_T and f_max values.

read the original abstract

The development of the sixth generation of wireless communications technology (6G) requires terminals that can operate at frequencies above 100 GHz. For human-centric applications, these terminals should also be flexible and have low power. However, current flexible radiofrequency transistors typically have lower maximum frequencies, in part due to the poor thermal conductivity of flexible substrates. Here, we report radiofrequency transistors that are based on aligned carbon nanotube arrays on flexible substrates and have current gain cutoff frequencies ($f_{\text{T}}$) and power gain cutoff frequencies ($f_{\text{max}}$) above 100 GHz. This is achieved by using electro-thermal co-design to improve the heat dissipation and radiofrequency performance of the devices. The transistors exhibit an on-state current of 0.947 mA $\mu$m$^{-1}$, a transconductance of 0.728 mS $\mu$m$^{-1}$, a peak extrinsic $f_{\text{T}}$ of 152 GHz, a peak extrinsic $f_{\text{max}}$ of 102 GHz, and a power consumption under 200 mW mm$^{-1}$. We also show that the devices can be used to create flexible radiofrequency amplifiers with an output power of 64 mW mm$^{-1}$ and a 11 dB power gain in the K-band.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper shows CNT transistors on flexible substrates hitting extrinsic f_T=152 GHz and f_max=102 GHz, but the abstract leaves the thermal measurements and de-embedding details out, so the electro-thermal co-design claim needs the full data to land.

read the letter

The headline result here is extrinsic cutoff frequencies above 100 GHz on a flexible substrate using aligned CNT arrays. That would matter for 6G flexible electronics if the numbers survive scrutiny, since prior flexible RF devices have stayed well below that range mainly due to substrate heating. The work does a clean job of stating the device metrics (on-current 0.947 mA/μm, gm 0.728 mS/μm, power under 200 mW/mm) and showing a simple amplifier demo in the K-band. Those numbers are presented as new for this platform, and the electro-thermal co-design idea is a reasonable way to tackle the thermal bottleneck on low-conductivity films. Credit for getting the devices to work at all on plastic and reporting extrinsic rather than intrinsic figures. The soft spot is exactly what the stress-test flagged: the abstract attributes the high frequencies directly to improved heat dissipation from the co-design, yet supplies no temperature maps, thermal resistance data, or the open/short structures and de-embedding steps used for the S-parameters. Without those, it is impossible to rule out pad parasitics or measurement choices that commonly push reported f_T/f_max upward. The paper is experimental rather than theoretical, so the circularity burden is low, but the central performance claim rests on unshown supporting evidence. This is the kind of result that belongs in a reading group for the RF flexible electronics crowd; people working on 6G terminals or CNT RF would want to see the full methods and raw data. It is worth sending to peer review because the target metrics are important and the approach is plausible, even if the current write-up leaves the key verification steps for the referees to check.

Referee Report

2 major / 2 minor

Summary. The manuscript reports radiofrequency transistors fabricated from aligned carbon nanotube arrays on flexible substrates. Through an electro-thermal co-design approach intended to enhance heat dissipation, the devices are stated to reach an on-state current of 0.947 mA μm^{-1}, transconductance of 0.728 mS μm^{-1}, peak extrinsic f_T of 152 GHz, and peak extrinsic f_max of 102 GHz while consuming under 200 mW mm^{-1}. The work further demonstrates their use in flexible K-band amplifiers delivering 64 mW mm^{-1} output power with 11 dB gain, targeting 6G flexible terminals.

Significance. If the reported extrinsic cutoff frequencies are robustly supported by the underlying measurements, the result would constitute a meaningful step forward for flexible RF electronics by showing >100 GHz operation on low-thermal-conductivity substrates. The combination of aligned CNT arrays with explicit co-design for thermal management offers a concrete route to high-frequency flexible devices with modest power draw.

major comments (2)

[Abstract] Abstract: The central performance claims (peak extrinsic f_T = 152 GHz, f_max = 102 GHz) are presented without error bars, without description of the de-embedding procedure (open/short structures, calibration method), and without raw S-parameter data or pad/substrate parasitic extraction. These omissions are load-bearing because the headline >100 GHz result on a flexible substrate is directly dependent on the extrinsic figures being free of measurement artifacts.
[Abstract] Abstract: The manuscript attributes the achieved f_T and f_max directly to electro-thermal co-design that improves heat dissipation on the flexible substrate, yet supplies no experimental thermal data (channel temperature under bias, thermal resistance measurements, or comparison devices) to confirm that the co-design produces a measurable temperature reduction rather than merely a modeled one. This attribution is load-bearing for the claimed mechanism.

minor comments (2)

[Abstract] The power-consumption figure (<200 mW mm^{-1}) is stated without specifying bias conditions, duty cycle, or whether it refers to DC or RF operation.
[Abstract] The amplifier demonstration (64 mW mm^{-1} output power, 11 dB gain) would benefit from a direct comparison to a reference device without the co-design to isolate its contribution.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful review and constructive feedback. The comments highlight important aspects of measurement transparency and mechanism validation. We address each point below and indicate planned revisions.

read point-by-point responses

Referee: [Abstract] Abstract: The central performance claims (peak extrinsic f_T = 152 GHz, f_max = 102 GHz) are presented without error bars, without description of the de-embedding procedure (open/short structures, calibration method), and without raw S-parameter data or pad/substrate parasitic extraction. These omissions are load-bearing because the headline >100 GHz result on a flexible substrate is directly dependent on the extrinsic figures being free of measurement artifacts.

Authors: The abstract is length-limited, but the full manuscript details the de-embedding procedure (open/short structures and calibration) in the Methods section, with raw S-parameter data, pad parasitic extraction, and error bars provided in the main figures and supplementary information. We will revise the abstract to include a concise reference to these protocols and the extrinsic nature of the reported values, directing readers to the detailed measurement sections. This strengthens transparency without changing the reported numbers. revision: yes
Referee: [Abstract] Abstract: The manuscript attributes the achieved f_T and f_max directly to electro-thermal co-design that improves heat dissipation on the flexible substrate, yet supplies no experimental thermal data (channel temperature under bias, thermal resistance measurements, or comparison devices) to confirm that the co-design produces a measurable temperature reduction rather than merely a modeled one. This attribution is load-bearing for the claimed mechanism.

Authors: The attribution rests on electro-thermal simulations that are validated against measured electrical performance and power-handling limits. Direct experimental thermal data (e.g., nanoscale channel temperature) are not included because such measurements on flexible substrates require specialized techniques not available in this study. In revision we will expand the discussion to clarify the simulation basis, add indirect supporting evidence from device stability under bias, and note the distinction between modeled and measured thermal effects. We partially revise to improve clarity on this point. revision: partial

Circularity Check

0 steps flagged

No circularity: experimental device results with no derivation chain

full rationale

This is an experimental paper reporting measured device metrics (on-state current, transconductance, extrinsic f_T=152 GHz, f_max=102 GHz) from fabricated CNT transistors on flexible substrates. The abstract and full text contain no equations, models, or first-principles derivations that reduce the reported frequencies or performance numbers to fitted parameters, self-citations, or ansatzes defined within the paper. The electro-thermal co-design is presented as the fabrication/measurement approach used to achieve the results, but the headline numbers are direct empirical outputs, not predictions equivalent to inputs by construction. No load-bearing self-citation chains or uniqueness theorems appear. The work is self-contained as a measurement demonstration.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

This is an experimental materials and device paper. The central claim rests on standard semiconductor transport and RF measurement assumptions rather than new mathematical derivations or postulated entities.

axioms (2)

standard math Standard definitions of extrinsic current-gain and power-gain cutoff frequencies apply to these CNT devices after standard de-embedding
Invoked when the abstract reports peak extrinsic f_T and f_max values
domain assumption Electro-thermal co-design improves heat dissipation on low-thermal-conductivity flexible substrates without introducing dominant new loss mechanisms
Central attribution of performance gain in the abstract

pith-pipeline@v0.9.0 · 5817 in / 1460 out tokens · 39613 ms · 2026-05-23T03:38:44.858717+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

6 extracted references · 6 canonical work pages

[1]

1 Rappaport, T. S. et al. Wireless communications and applications above 100 GHz: Opportunities and challenges for 6G and beyond. IEEE Access 7 , 78729 - 78757 (2019). 2 Akyildiz, I. F., Kak, A. & Nie, S. 6G and beyond: The future of wireless communications systems. IEEE Access 8 , 133995 - 134030 (2020). 3 Saad, W., Bennis, M. & Chen, M. A vision of 6G w...

work page 2019
[2]

5 | Flexibility test of flexible RF transistors and performance of flexible power amplifier s

f T.ext (GHz) | V ds | (V) f max,ext (GHz) a b c d 20 GHz 50 GHz 100 GHz 75 nm 120 nm 160 nm 80 GHz 16 Fig . 5 | Flexibility test of flexible RF transistors and performance of flexible power amplifier s . a , Current gain of a typical flexible RF transistor with L g = 75 nm under different bending radii of curvature (left) and the relationship between f T...

work page 2020
[3]

Our approach thus provides an upper - bound estimate for the peak device temperature

to model the worst - case scenario. Our approach thus provides an upper - bound estimate for the peak device temperature. Supplementary Notes 2 | De - embedding Procedure of Flexible RF transistors A standard open - short de - embedding method was employed for pad and intrinsic de - embedding. The process is outlined as follows: (1) Measure the S - parame...

work page 2020
[4]

5 | SEM images of a flexible CNT - based RF transistor fabricated without electro - thermal co - design after RF measurements

25 V Supplementary Fig. 5 | SEM images of a flexible CNT - based RF transistor fabricated without electro - thermal co - design after RF measurements. a , SEM image of the damaged two - fingered flexible RF transistor. Scale bar: 1 μm. b , Zoomed - in SEM image of the area highlighted by the red box in ( a ). Scale bar: 400 nm. Supplementary Fig. 6 | S - ...

work page 2000
[5]

Nominal thermal properties used in simulations. Some thermal boundary conductance (TBC) values not available in the literature were approximated by TBCs for pairs of similar and/or better - studied materials, using the lower bound estimates to model the worst - case scenario. The actual thermal properties vary depending on the material and interface quali...

work page 2006
[6]

IEDM Technical Digest. 4 pp. - 256 (IEEE). 87 Miyazaki, K., Kuriyama, K. & Yabuki, T. Printable thermoelectric device. J. Phys. Conf. Ser. 1407 , 012057 (2019). 88 Cappella, A. et al. High temperature thermal conductivity of amorphous A l 2 O 3 thin films grown by low temperature ALD . Adv. Eng. Mater. 15 , 1046 - 1050 (2013). 89 Monachon, C., Weber, L. &...

work page 2019

[1] [1]

1 Rappaport, T. S. et al. Wireless communications and applications above 100 GHz: Opportunities and challenges for 6G and beyond. IEEE Access 7 , 78729 - 78757 (2019). 2 Akyildiz, I. F., Kak, A. & Nie, S. 6G and beyond: The future of wireless communications systems. IEEE Access 8 , 133995 - 134030 (2020). 3 Saad, W., Bennis, M. & Chen, M. A vision of 6G w...

work page 2019

[2] [2]

5 | Flexibility test of flexible RF transistors and performance of flexible power amplifier s

f T.ext (GHz) | V ds | (V) f max,ext (GHz) a b c d 20 GHz 50 GHz 100 GHz 75 nm 120 nm 160 nm 80 GHz 16 Fig . 5 | Flexibility test of flexible RF transistors and performance of flexible power amplifier s . a , Current gain of a typical flexible RF transistor with L g = 75 nm under different bending radii of curvature (left) and the relationship between f T...

work page 2020

[3] [3]

Our approach thus provides an upper - bound estimate for the peak device temperature

to model the worst - case scenario. Our approach thus provides an upper - bound estimate for the peak device temperature. Supplementary Notes 2 | De - embedding Procedure of Flexible RF transistors A standard open - short de - embedding method was employed for pad and intrinsic de - embedding. The process is outlined as follows: (1) Measure the S - parame...

work page 2020

[4] [4]

5 | SEM images of a flexible CNT - based RF transistor fabricated without electro - thermal co - design after RF measurements

25 V Supplementary Fig. 5 | SEM images of a flexible CNT - based RF transistor fabricated without electro - thermal co - design after RF measurements. a , SEM image of the damaged two - fingered flexible RF transistor. Scale bar: 1 μm. b , Zoomed - in SEM image of the area highlighted by the red box in ( a ). Scale bar: 400 nm. Supplementary Fig. 6 | S - ...

work page 2000

[5] [5]

Nominal thermal properties used in simulations. Some thermal boundary conductance (TBC) values not available in the literature were approximated by TBCs for pairs of similar and/or better - studied materials, using the lower bound estimates to model the worst - case scenario. The actual thermal properties vary depending on the material and interface quali...

work page 2006

[6] [6]

IEDM Technical Digest. 4 pp. - 256 (IEEE). 87 Miyazaki, K., Kuriyama, K. & Yabuki, T. Printable thermoelectric device. J. Phys. Conf. Ser. 1407 , 012057 (2019). 88 Cappella, A. et al. High temperature thermal conductivity of amorphous A l 2 O 3 thin films grown by low temperature ALD . Adv. Eng. Mater. 15 , 1046 - 1050 (2013). 89 Monachon, C., Weber, L. &...

work page 2019