Elsevier

Carbon

Volume 201, 5 January 2023, Pages 1025-1029
Carbon

Nanometer-level temperature mapping of Joule-heated carbon nanotubes by plasmon spectroscopy

https://doi.org/10.1016/j.carbon.2022.10.006Get rights and content

Highlights

  • Temperature of suspended Joule-heated CNTs is mapped with 5-nm resolution in a TEM.

  • CNTs can withstand over 2000 K, more than 1000 K higher than previously measured.

  • Hot-spot near smaller electrical contact due to high resistance and low heat transfer.

  • High temperature along the entire CNT length because of high thermal conductivity.

Abstract

We map the temperature distribution of suspended carbon nanotubes down to the 5-nm level inside a transmission electron microscope, by using electron energy-loss plasmon spectroscopy. The nanotubes are Joule heated in-situ, by individually contacting them using a biasing holder. The nanotubes can withstand current densities of over 5 × 107 A/cm2 and temperatures of over 2000 K without breaking. The temperature reaches its maximum around the nanoprobe contact, due to the comparatively large electrical resistance and small thermally-conductive area. The temperature remains high along the entire length of the CNTs, due to the excellent in-plane thermal conductivity (up to 2800 W/m/K) of the graphitic lattice. These results verify the expected robustness of these structures and confirm them as ideal candidates for applications as interconnects under extremely high current densities and temperatures.

Introduction

Carbon nanotubes (CNTs) are a leading replacement candidate of current metallic interconnect materials near the size-scaling end of the semiconductor roadmap and Moore's law [1], due to their unique one-dimensional tubular structure, high thermal conductivity and high current capacity. It is critical to understand the temperature distribution and damage tolerance of the nanotubes under high current density and temperature, especially near the contact area. However, it is a great challenge to measure the temperature of a nanoscale object with high spatial resolution and without contact effects which would affect the temperature distribution. The inelastic interaction of an electron beam with matter provides a remote sensing probe for measuring the temperature. Local temperatures could be obtained from electron energy loss spectroscopy (EELS) by measuring phonon scattering [2,3], excitonic absorption [4], and plasmon excitations [5,6].

The thermometry method in this study involves the measurement of small shifts in the plasmon energy due to lattice expansion. Earlier experiments had the advantage of working with homogeneous materials with very well-known properties [6]. CNTs on the other hand, due to their structural diversity and crystalline anisotropy, pose significant challenges which will be discussed in detail below. High-resolution temperature mapping on in-situ heated CNTs was previously attempted using a thermocouple-cantilever in an AFM [7] or by depositing Indium particles next to the CNT in a transmission electron microscope (TEM) [8]. However, these methods do not provide an absolute-scale measurement of the CNT temperature. In addition, due to the heat-sinking effects of the substrates, the maximum temperatures were limited to hundreds of Kelvin.

Section snippets

Method description

Our sample consists of arc-discharge multi-wall carbon nanotubes (MWCNTs), 20–30 nm in diameter, with narrow channels. Some high-resolution images of typical CNTs used in this work are summarized in Fig. S1. A schematic of an experiment is given in Fig. 1. A sharp Au probe whose tip is visible on the left side of the image has contacted the CNT which is displayed horizontally. The other end of the CNT is supported by a bulk Au wire. Between the contacts, the CNT is suspended in high vacuum (∼10

Conclusions

We have successfully mapped the temperature of Joule-heated suspended carbon nanotubes with nanometer resolution, by measuring their plasmon energy shift using EELS, combined with a rigorous temperature calibration process. By intentionally driving the nanotubes to the limit of their structural stability, we were able to determine the maximum temperatures they could withstand under these conditions; this information is essential for their applications involving high current density and

CRediT authorship contribution statement

Ovidiu Cretu: Conceptualization, Methodology, Formal analysis, Investigation, Writing - original draft, Writing - review & editing. Dai-Ming Tang: Conceptualization, Methodology, Investigation, Resources, Funding acquisition, Writing - original draft, Writing - review & editing. Da-Bao Lu: Formal analysis, Writing - original draft, Writing - review & editing. Bo Da: Formal analysis, Resources, Writing - original draft, Writing - review & editing. Yoshihiro Nemoto: Resources, Writing - review &

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors are grateful to Keiji Kurashima (RNFS, NIMS) for supporting the Jeol GrandARM experiments. This work was supported by JSPS KAKENHI Grant Numbers JP20K05281, JP20H02624, and JP22H05145.

References (29)

  • M. Mecklenburg et al.

    Nanoscale temperature mapping in operating microelectronic devices

    Science

    (2015)
  • L. Shi et al.

    Thermal probing of energy dissipation in current-carrying carbon nanotubes

    J. Appl. Phys.

    (2009)
  • K.H. Baloch et al.

    Remote joule heating by a carbon nanotube

    Nat. Nanotechnol.

    (2012)
  • L. de Knoop

    Development of Quantitative in Situ Transmission Electron Microscopy for Nanoindentation and Cold-Field Emission

    (2014)
  • Cited by (0)

    View full text