A.

Effect of thickness and diameter

Various process conditions have been applied to synthesize VACNT samples with different morphologies on Si wafer. Besides, aluminium foil being a substrate of choice for roll-to-roll synthesis of VACNTs, and the nature of the substrate being a strong influencing parameter in catalytic CVD processes, we have also included one sample of VACNTs grown on Al foil in this comparative study.

From SEM observation, the main differences between these samples are the thickness of the film (i.e. the length of the VACNTs) and the apparent diameter of the VACNTs. It is worth nothing that apparent diameters measured from SEM images may not coincide to individual carbon nanotube diameters as carbon nanotubes may exhibit a carbonaceous surface layer and/or they can assemble together to form bundles. The morphological and optical reflexion properties of these samples are summarized in Table 1. Some typical SEM cross-sections are shown in Fig. 2 and the reflectance spectra are plotted in Fig. 3.

Table 1.

Morphological and optical reflexion properties of VACNTs grown by atmospheric pressure, aerosol assisted CCVD process on Si substrate under various process conditions.

Fig. 2.

SEM X-sections of samples D(left), A (center) and B (right).

Fig. 3.

Optical reflection spectra. The missing data have been intentionally removed due to low S/N ratio.

These results suggest that, at least for this set of samples, the main morphological parameter driving the optical reflectance in the visible and near infrared spectral range is the apparent diameter of the nanotubes. Indeed, the samples exhibiting a distribution of apparent diameters ranging between 30 and 40nm are less reflective than other samples containing larger apparent diameter carbon nanotubes. The thickness of the VACNT film doesn’t play a crucial role in the reflectance level. One sample however does not follow this trend: sample H with VACNTs grown on Al substrate. This points towards another influencing parameter for the reflectance of VACNTs, possibly the density of carbon nanotubes. Work is in progress to clarify this point.

In agreement with literature data on the reflectance of VACNTs [3, 4, 5], the total hemispherical reflectance is below 2% and the specular reflectance is below 0,02% for most samples at wavelength above 400nm. This confirms the diffusive behaviour of VACNTs as a very low reflectance coating.

B.

Top vs Bottom: transfer onto destination substrate

The thermal CVD synthesis of VACNTs can only be applied to substrates withstanding temperature above 450°C. This can be a serious restriction for coating optical or structural parts without deterioration. One possibility to solve this issue is to transfer the VACNT film from its original substrate onto a destination substrate of interest (the part to be coated). When the original substrate exhibits some flexibility (as metal foils do), this technique can be applied to destination substrates with curvature or complex shape.

To assess the optical reflexion properties of VACNTs after transfer onto a destination substrate we applied a thin layer of epoxy glue on the top surface of a VACNT film, then we placed a plastic plate on top and, after drying the glue in ambient air for 24 hours, the plate was removed. The VACNT film was stuck on the plastic plate with its original bottom face now facing up.

Optical and SEM characterizations have been performed on this sample and compared to the same VACNT film before transfer. The results are presented in Table 2, Fig. 4 and Fig. 5. They show a decrease in the reflexion properties (both specular and total) associated with a slight increase in the carbon nanotube apparent diameter and a better alignment of the carbon nanotubes. This suggest that alignment plays a significant role in the reflectance of VACNT films thus providing a path for improvement. It also demonstrates that transferred VACNTs are still excellent black coatings, even better than as-grown VACNTs.

Table 2.

Morphological and optical reflexion properties of VACNTs before (A) and after (A Bottom) transfer on a plastic substrate.

Fig. 4.

SEM X-sections (left) and top views (right) of VACNTs before (A) and after (A Bottom) transfer on a plastic substrate.

Fig. 5.

Optical reflection spectra of VACNTs before (A) and after (A Bottom) transfer on a plastic substrate.

C.

Effect of plasma etching

In order to reduce further the reflectance of VACNTs, we tried to modify the top surface of this black coating by etching with a radio frequency oxygen plasma. We applied this reactive ion etching process to sample G and found conditions leading to a 23% decrease of the specular reflectance in the visible spectral range (from 0,013% to 0,010% at 600nm). The specular reflectance spectra and the SEM cross section view are shown in Fig. 6 and Fig. 7 respectively.

Fig. 6.

Specular reflection spectra of VACNTs (sample G) before and after O2 plasma etch.

Fig. 7.

SEM cross section of the top surface of VACNTs (sample G) before and after O2 plasma etch.

The SEM observations indicate that plasma etching did not significantly change the thickness, apparent diameter nor the density of the VACNTs. However, an improvement in the alignment of the carbon nanotubes at the top surface of the film is perceptible. This again points towards the vertical alignment being a strong influencing parameter for the reflectance of VACNTs and demonstrates that plasma etching provides an efficient post-process to improve this parameter.