The sheath thickness for a typical plasma density (n p ≈ 1017 to 1018 m−3) may be assumed to be of the order of a few Debye lengths [34] (2) where ϵ 0 is a dielectric constant, λ D is the electron Debye length and k p is the constant, typically in the range between 1 and 5. The estimates using Equation 2 give the sheath thickness of the order of 10 μm to 0.1 mm, that is, much larger than the average diameter of the alumina membrane channels. This means that the ions extracted from the plasma edge will not be significantly deflected by the electric field distorted by nanosized features on the membrane surface. Hence, the ions move along straight trajectories and could penetrate deeply into the channels. As a result, one can expect

that the surface of the channels will be treated

by the ion flux penetrating relatively deeply under the upper surface of the membrane. The Raman spectra of the nanotubes grown using C2H4 Small molecule library and C2H4 precursors (Figure 6c,d) show D and G bands that are typical for multi-walled carbon nanotubes and a relatively low number of defects. The spectra of other samples are also very similar to those shown in Figure 6c,d, thus exhibiting relatively low defect level irrespective of the specific process conditions (see Additional file 1: Figure S5 for the Raman spectrum of nanotubes grown without S1813 photoresist). PR-171 ic50 Further TEM analysis of the carbon nanotubes grown on top of alumina membrane with S1813 photoresist has demonstrated a rather good quality of the grown nanostructures with relatively thin walls consisting of approximately 10 atomic carbon layers (Figure 6a,b). More TEM images can be found in Additional file 1: Figures S4 and S6. Figure 6 TEM and Raman characterization. (a, b) High-resolution TEM images of the carbon nanotubes grown on top of the alumina membrane with S1813 photoresist. A relatively thin wall consisting of 10 atomic carbon layers can be seen in (b). (c, d) The Raman spectra of the nanotube grown using C2H4 and C2H2 precursors show D and G bands and a relatively low presence of defects. Conclusions

To conclude, we have demonstrated that effective PtdIns(3,4)P2 control of nucleation and growth of carbon nanotubes in nanopores of alumina membranes is possible by using plasma posttreatment of the membrane and application of S1813 photoresist as an additional carbon precursor. A few options to control the growth of nanotubes inside the membrane channels or on the upper membrane surface were considered and successfully demonstrated. In particular, we have demonstrated the fabrication of multi-walled carbon nanotubes on plasma-treated membranes. The nanotubes conformally filled the membrane channels and did not form mats on the membrane top. Thus, the growth mode can be controlled, and complex structures on the basis of nanotubes can be produced for various applications. A plausible nucleation and growth mechanism was also proposed on the basis of analysis of the plasma parameters.