Growth of glassy carbon microneedles

Previously, glassy carbon microneedles have been made by the pyrolysis of commercially available polymers. The polymer material was pre-patterned as an array, which was then converted to glassy carbon microneedle patches via a conventional carbon-microelectromechanical system (C-MEMS) process [14]. In this article we report the fabrication of freestanding glassy carbon microneedles in a single step achieved by the pyrolysis of methane on a curved alumina surface. The surface provides the catalyst as well as the “strain” required to direct nucleation and growth.

Figure 1a is a scanning electron microscopy (SEM) overview image showing a number of glassy carbon microneedles, which grow in the direction of the gas flow. Figure 1b is a SEM detail image showing glassy carbon microneedles, nucleating microneedles, and “blisters”, which correspond to the early stages of the microneedle growth. Figure 1 shows that the microneedles grown under the given pyrolysis conditions are uniformly ca. 25 µm in diameter.

Figure 1: (a) SEM overview image showing a number of glassy carbon microneedles, (b) SEM detail image of glassy carbon microneedles, nucleating microneedles, and nucleation “blisters”.

The fullerene-like tips of the microneedles can clearly be seen in the SEM micrographs in Figure 2a and Figure 2c. Figure 2b shows a model of a fullerene, which highlights the requirement of pentagons as well as hexagons in order to close the microneedle tips.

Figure 2: (a) SEM overview image of two glassy carbon tubules; the one on the right has its fullerene dome intact and the one on the left has a fractured fullerene dome giving insights into the internal concentric structure; (b) inset of a model of a fullerene; (c) SEM overview image of two glassy carbon tubules; the one on the left has its dome intact and the one on the right is fractured showing some of the inner intact fullerene domes.

SEM-EDX measurements (EDX INCA 400, Oxford Instruments, at 20 kV) showed that the materials are only carbon with no other detectable elements. The rounded caps of the glassy carbon tubules suggest that they have characteristics of fullerenes regarding the need for pentagons in addition to hexagons to close the cap [15] (see the model of fullerene C60 in Figure 2b). Therefore, we consider the most likely growth mechanism to be that proposed by S. Amelinckx and co-workers [16]. Their model explains the formation of multishell closure domes in which nucleation is attributed to the initial formation of fullerene domes. These originate from the “blisters” observed in Figure 1b near the glassy carbon microneedles.

The SEM images in Figure 3 show two typical glassy carbon microneedles fractured at the tip revealing their internal structure. This is clear experimental evidence that the glassy carbon microneedles are hollow in nature.

Figure 3: SEM images of two typical glassy carbon microneedles fractured at the tip showing that the microneedles have a hollow core and are tubular in structure.

Characterization of the glassy carbon microneedles

Figure 4 shows a typical Raman spectrum of the glassy carbon microneedles. The D-band is at 1352 cm−1, and the G-band is at 1589 cm−1. The D-band, the so-called defect band, originates from a hybridized vibrational mode associated with local defects and disorder. In this case, it results from the curvature of the glassy carbon tubules. The G-band, the so-called graphitic or tangential band, is characteristic of graphite and originates from the in-plane tangential stretching of the C–C bonds. The intensity of the D-band is much higher than that of the G-band, showing the local crystalline structure [17]. This is in good agreement with Raman spectrum data for glassy carbon [17,18] and confirms that the carbon microneedles fabricated here are glassy in nature.

Figure 4: Raman spectrum of glassy carbon tubules. Both D-band and G-band are sharp and well defined, which indicates graphitic rather than amorphous carbon [19].

Figure 5 shows the XRD measurement of the glassy carbon tubules. The single sharp peak is indicative of graphitic carbon with long-range crystalline order. The interlayer spacing is calculated to have a d-spacing of 4.89 Å. Table 1 shows the interlayer spacing of graphite and other selected carbon materials and is further evidence that the tubules are glassy carbon.

Figure 5: XRD of glassy carbon tubules, including the calculation of the interlayer spacing.

In Table 1, the interlayer spacing data characteristic for selected carbon materials is shown. Graphitic carbon is known to have an interlayer spacing of 3.354 Å [20,21]. Glassy carbon has a much larger interlayer spacing than turbostratic carbon [22] and carbon fibres [23]. The interlayer spacing of the glassy carbon microneedles in this work is calculated from the XRD data to have a d-spacing of 4.89 Å, which is further experimental evidence that the microneedles are of glassy carbon.

Table 1: Interlayer spacing data characteristic for selected carbon materials.

Carbon material d-spacing (Å) Reference graphite 3.354 [20,21] turbostratic carbon 3.44–3.67 [22] carbon fibre 3.40–3.53 [23] glassy carbon (polymer-derived carbon) 3.45–3.60
3.46–3.70 [24,25] glassy carbon (H+ irradiated graphite whiskers) 4.7–6.9 [26] glassy carbon (CH4 derived carbon) 4.89 this work

Most glassy carbons prepared to date have been made by pyrolysis of polymeric materials. These generally do not have long-range crystalline order and undergo glass-like fracture. However, more recently, it has been shown that glassy carbons can graphitize under high-stress conditions [27,28]. Furthermore, it has been shown that intercalated species can increase the interlayer spacing of graphitic carbons [29]. It has also been shown that this increase in interlayer spacing in glassy carbon corresponds to the nature of the intercalating species [26]. Here the “stress” arises from the constraints of the growing glassy carbon layers from the curved alumina gas-flow tube and the intercalating species, which are gaseous species, generated when methane undergoes pyrolysis. These are transitory species and do not remain in the glassy carbon tubules. The early stages of this process can be seen, in Figure 1b, as the nucleation “blisters” on the surface of the glassy carbon formed on the alumina. These then develop into glassy carbon microneedles.