Characterizing Single Carbon Nanofibers

Quantitative Analysis by Transmission Electron Microscopy

  • Fig.1: Principles of the quantitative diffraction pattern analysis: (a) bright field and HRTEM image together with SAED pattern including diffraction vector k and azimuthal angle φ, (b) scheme for determination of apex angle 2θ and interlayer spacing dhkl, (c) undulation and azimuthal width Δθ, (d) domain size L and correlating peak width Δk, (e) radial intensity profile with identified reflections and corresponding dhkl, (f) azimuthal profiles of the identified reflections. The domain size L is calculated from the full width at half maximum of the reflections in (e) applying the Scherrer equation. Apex angle 2θ and azimuthal width Δθ are determined by Gaussian fits of the azimuthal profile in (f).Fig.1: Principles of the quantitative diffraction pattern analysis: (a) bright field and HRTEM image together with SAED pattern including diffraction vector k and azimuthal angle φ, (b) scheme for determination of apex angle 2θ and interlayer spacing dhkl, (c) undulation and azimuthal width Δθ, (d) domain size L and correlating peak width Δk, (e) radial intensity profile with identified reflections and corresponding dhkl, (f) azimuthal profiles of the identified reflections. The domain size L is calculated from the full width at half maximum of the reflections in (e) applying the Scherrer equation. Apex angle 2θ and azimuthal width Δθ are determined by Gaussian fits of the azimuthal profile in (f).
  • Fig.1: Principles of the quantitative diffraction pattern analysis: (a) bright field and HRTEM image together with SAED pattern including diffraction vector k and azimuthal angle φ, (b) scheme for determination of apex angle 2θ and interlayer spacing dhkl, (c) undulation and azimuthal width Δθ, (d) domain size L and correlating peak width Δk, (e) radial intensity profile with identified reflections and corresponding dhkl, (f) azimuthal profiles of the identified reflections. The domain size L is calculated from the full width at half maximum of the reflections in (e) applying the Scherrer equation. Apex angle 2θ and azimuthal width Δθ are determined by Gaussian fits of the azimuthal profile in (f).
  • Fig. 2: Variation of structural parameters along the axis of a single CNF: (a) bright field image of the analyzed CNF, (b,c) NBED pattern and HRTEM image at position 3, (d) interplanar spacing d002 and d100, domain size L, apex angle 2θ and azimuthal width Δθ. The areas in (a) probed by NBED and HRTEM are numbered corresponding to the position axis in (d).
  • Fig.3: Impact of processing on CNF structure: (a-d) temperature dependencies of interlayer spacing d002, fiber diameter D, apex angle 2θ, and undulation Δθ, (e) TEM images of CNFs obtained at different synthesis temperatures (red) grown on ZrO2 using C2H4. Each combination of synthesis gas and substrate material is represented by a specific color code in the plots (a-d) according to the legend in (a).

Electron Diffraction is a powerful technique for an extended structural evaluation of individual carbon nanofibers in the transmission electron microscope. Quantitative analyses of reflection shapes and positions in diffraction patterns facilitate a precise quantification of local structural parameters of single fibers, providing a more detailed understanding that can e.g. be exploited for an optimization of fiber growth conditions.

Introduction

Transmission electron microscopy (TEM) is the standard method to characterize structural details of carbon nanofibers (CNF) at the individual level, in particular, to visualize the stacked graphene layers by high-resolution transmission electron microscopy (HRTEM). Electron diffraction techniques like selected area electron diffraction (SAED) and nano-beam electron diffraction (NBED) allow for a thorough crystallographic characterization of the structure of CNFs [1-3]. The diffraction patterns represent the structural information of the single fiber in the reciprocal space. Shape and position of reflections provide quantitative information on cone apex angle, interlayer spacing, size of coherent domains and undulation of graphene layers [3]. Crystallographic interpretation of higher order reflections beyond the basal (002) reflections that correspond to the orientation of the graphene layers reveals specific structural features of CNFs, particularly the stacking sequence of successive graphene layers [3].

Analysis of CNF Diffraction Patterns

Representative bright field image, HRTEM image, and SAED pattern of a single herringbone CNF are displayed in figure 1a. The principles for quantitative diffraction pattern evaluation are illustrated by the schemes in figure 1b-d. Each set of parallel lattice planes is represented by a pair of centrosymmetric reflections with diffraction vectors perpendicular to the planes (red and green in figure 1b, respectively). The reflections closest to the center correspond to the (002) lattice planes of graphite according to the stacking of graphene layers in the fiber structure. In herringbone CNFs consisting of stacked graphite cones, the (002) diffraction vectors span the cone apex angle 2θ.

From the lengths of the diffraction vectors, the average interplanar spacings dhkl are calculated. The arched shape of the reflections corresponds to an orientation range of the graphene layers that possess undulations or distortions instead of a perfectly parallel alignment. The stronger the curvature of the graphene layers, the larger is the azimuthal width of the arc Δθ (fig. 1c). Due to the high density of layer distortions, CNFs consist of nm-sized coherent domains as shown from a dark-field image in figure 1d. The average size of the coherently scattering domains L is calculated from the full width at half maximum of the reflections along the radial direction Δk.

The determination of the structural parameters is based on a transformation of the diffraction pattern to polar coordinates and subsequent extraction of radial (fig.1e) and azimuthal profiles (fig. 1f). Figure 1a displays the SAED pattern containing the azimuthal angle φ and the diffraction vector k as radial coordinate. From the radial intensity profile interplanar spacings dhkl and domain size L are calculated (fig. 1e). For each dhkl, the corresponding azimuthal intensity profile is extracted to quantify apex angle 2θ and undulation Δθ (fig. 1f).This methodology is not only applicable to SAED and NBED patterns, but also to power spectra obtained by Fourier transformations of HRTEM images. Details of the automated procedure can be found in ref. [3].

Variation of Structural Parameters along a Single CNF

To quantify structural variations in a single CNF along its axis and to understand their correlation with the growth conditions during chemical vapor deposition (CVD), a series of point analyses by NBED and HRTEM was carried out (fig. 2b and c). The probed areas are numbered corresponding to their position on the CNF (fig. 2a and x-axis in fig. 2d). A clear trend of increasing interlayer spacing d002 and decreasing domain size L along the fiber axis is found (fig. 2d), which may be related to the decrease of the performance of the Pd catalyst (CH4 conversion) during CNF growth. The CH4 conversion was determined by time-resolved in-situ Fourier transformed infrared spectroscopy [4]. The other structural parameters (2θ and Δθ) fluctuate along the fiber axis, possibly due to local and temporary fluctuations of the CVD parameters like turbulences in the gas flow or a change of the CNF growth direction.

Insight into Structure Processing Relationship of CNFs

The characterization method has already been utilized to evaluate the influence of CVD processing parameters such as carbon source, synthesis temperature, and substrate material on the structural details of CNFs in a systematic study [4,5]. Interlayer spacing dhkl, apex angle 2θ, undulation of carbon layers Δθ, and fiber diameter D were quantified in a statistical manner, allowing a comprehensive assessment of the structure processing relationships. Figure 3 summarizes the results of CNFs grown at different temperatures on ZrO2 and Al2O3 substrates using CH4 and C2H4 as carbon feedstock. The interlayer spacing d002 decreases with increasing synthesis temperature (fig. 3a), irrespective of carbon source and substrate material. The carbon source mainly affects the fiber diameter D, showing opposed temperature dependences for CH4 and C2H4 (fig. 3b). Additionally, by using C2H4 smaller apex angles 2θ were obtained (fig. 3c).

According to Raman spectroscopy, best structural properties of fibers were observed for those generated at 800°C with CH4 and at 750°C with C2H4 as a carbon source [5]. This corresponds to the smallest apex angles 2θ and a low undulation of carbon layers Δθ together with low interlayer spacings d002 (fig. 3c and d).
The structural parameters were determined from more than 10 representative HRTEM images per sample. In figure 3e such HRTEM images of the CNFs grown on ZrO2 using C2H4 at different synthesis temperatures are displayed.
Detailed characterization of the structure processing relationships of CNFs may contribute to synthesize carbon nanomaterials with tailored properties.

Summary

Profound interpretation and quantitative analysis of CNF diffraction patterns or power spectra of HRTEM images provide extensive structural information. Quantitative structural analysis of single CNFs has been carried out in an automated procedure. Utilizing these possibilities yields new insight into structural variations during CNF growth and its relationship with processing parameters. Quantified structural parameters may serve to understand and optimize growth conditions for CNFs and to correlate their resulting properties to their structural characteristics.

Acknowledgment
Dr. Adrian Simon from Fraunhofer Institute for Ceramic Technologies and Systems (IKTS), Germany is gratefully acknowledged for stimulating discussions and providing all the samples.

Fig.1: Principles of the quantitative diffraction pattern analysis: (a) bright field and HRTEM image together with SAED pattern including diffraction vector k and azimuthal angle φ, (b) scheme for determination of apex angle 2θ and interlayer spacing dhkl, (c) undulation and azimuthal width Δθ, (d) domain size L and correlating peak width Δk, (e) radial intensity profile with identified reflections and corresponding dhkl, (f) azimuthal profiles of the identified reflections. The domain size L is calculated from the full width at half maximum of the reflections in (e) applying the Scherrer equation. Apex angle 2θ and azimuthal width Δθ are determined by Gaussian fits of the azimuthal profile in (f).
Fig. 2: Variation of structural parameters along the axis of a single CNF: (a) bright field image of the analyzed CNF, (b,c) NBED pattern and HRTEM image at position 3, (d) interplanar spacing d002 and d100, domain size L, apex angle 2θ and azimuthal width Δθ. The areas in (a) probed by NBED and HRTEM are numbered corresponding to the position axis in (d).
Fig.3: Impact of processing on CNF structure: (a-d) temperature dependencies of interlayer spacing d002, fiber diameter D, apex angle 2θ, and undulation Δθ, (e) TEM images of CNFs obtained at different synthesis temperatures (red) grown on ZrO2 using C2H4. Each combination of synthesis gas and substrate material is represented by a specific color code in the plots (a-d) according to the legend in (a).

 

Authors
Martin Seyring1 and Markus Rettenmayr1

Affiliation
1 Friedrich Schiller University Jena, Otto Schott Institute of Materials Research (OSIM), Metallic Materials, Jena, Germany

Contact
Dr. Martin Seyring
Friedrich Schiller University Jena
Otto Schott Institute of Materials Research (OSIM)
Metallic Materials
Jena, Germany
martin.seyring@uni-jena.de
www.matwi.uni-jena.de/metalle.html

References
[1] N.A. Kiselev, J.L. Hutchison, A.P. Moravsky, E.V. Rakova, E.V. Dreval,
C.J.D. Hetherington, D.N. Zakharov, J. Sloan, and R.O. Loutfy: Carbon micro- and nanotubes synthesized by PECVD technique: tube structure and catalytic particles crystallography, Carbon 42(1): 149-161 (2004)  doi: 10.1016/j.carbon.2003.10.014
[2] C. van Gulijk, K.M. de Lathouder, and R. Haswell: Characterizing herring bone
structures in carbon nanofibers using selected area electron diffraction and dark field transmission electron microscopy
, Carbon 44(14): 2950-6 (2006) doi: 10.1016/j.carbon.2006.05.036
[3] Martin Seyring, Adrian Simon, Ingolf Voigt, Uwe Ritter, and Markus Rettenmayr: Quantitative crystallographic analysis of individual carbon nanofibers using high resolution transmission electron microscopy and electron diffraction, Carbon, 116: 347-55 (2017) doi: 10.1016/j.carbon.2017.01.107
[4] Adrian Simon, Martin Seyring, Susanne Kämnitz, Hannes Richter, Ingolf Voigt, Markus Rettenmayr, and Uwe Ritter: Carbon nanotubes and carbon nanofibers fabricated on tubular porous Al2O3 substrates, Carbon 90: 25-33 (2015) doi: 10.1016/j.carbon.2015.03.048
[5] Adrian Simon, Martin Seyring, Norman Reger-Wagner, Hannes Richter, Ingolf Voigt, Markus Rettenmayr, and Uwe Ritter: Influence of carbon source and synthesis temperature on structural and morphological properties of Carbon Nanofibers synthesized on tubular porous ZrO2 layers, Diamond & Related Materials (2017) doi: 10.1016/j.diamond.2017.08.006

More on Transmission Electron Microscopy

Read more about crystal analysis

Contact

Friedrich Schiller University Jena

Jena
Deutschland

Register now!

The latest information directly via newsletter.

To prevent automated spam submissions leave this field empty.