Since their very first discovery, carbon nanostructures attracted the interest of the scientific community due to their unique and tunable properties from the atomic to the meso scale. Among all the carbon allotropes, one of the most fascinating is the so-called Carbon-Nano Onion (CNO), consisting in carbon nanoparticles with concentric curved graphene sheets, forming several layers of an onion like structure. CNO has shown remarkable characteristics in many field spanning from electrochemistry to tribological applications[Palkar, A.J. et al. Chemistry of Nanoparticles, 478 (2010)]. Despite this, CNOs are still produced by annealing at temperature higher than 1500°C nanodiamonds in high vacuum condition. This process is considered not only time consuming, but also the production of nanodiamonds still represents a limiting process [Kuznetsov, V.L. et al. Carbon 32, 873-882 (1994)]. Another proposed method for a high-quality CNO is by water arc discharge between two graphitic electrodes [Chhowalla, M. et al. Nature 414, 506–507 (2001)]. Nevertheless, a high throughput process has yet to be developed.

In this work we propose a new production method of CNOs, called Nanoparticle Jet Deposition, NanoJeD, which promises high throughput while assuring fine tunability of mesostructure, defectiveness and graphitic content. Moreover with this system it is possible to deposit hierarchical assemblies of CNOs tens of microns thick over any substrate.

The NanoJeD source comprises two chambers separated by a high aspect ratio slit. The system follows a vertical development in which a precursor gas in injected through a porous plug from the top and a vacuum system is connected from the bottom. The slit allow the establishment of a high pressure ratio between the two chambers and hence the formation of a supersonic jet. The precursor gas is a mixture of Aargon (98.4%) and Aacetylene (1.6%). The gas stream, coaxial with the slit, flows between two electrodes where a RF signal (13.56 MHz) is fed. The radio frequency ignites a non-thermal plasma in which the precursor molecules are ionized and dissociated into radicals. Depending on the residence time and other discharge parameters, these radical polymerize forming clusters and nanoparticles (NP) of different size. Size distributions follow the typical lognormal distribution of a plasma chemical reaction [Milani P. et al. Surface Science 402, 441-444 (1998)]. Once formed, NPs are transported in the slit nozzle where they are quenched by the supersonic expansion. Hence, depending on the ration between the particle inertia and gas drag on the same, defined as the Stokes number, St, nanoparticles can follow different paths: dragged down to the nozzle by the gas stream where the gas is compressed. Particles with a St >= 1 will have enough kinetic energy to impact on the substrate, forming a film, while particle with a St<1 will follow the gas streamline to the vacuum pump. In this process, the terminal velocity of particles, and hence their kinetic energy, can be tuned in many ways thus controlling the final density and mesostructure of the growing film. Highly graphitic nanoparticles can be readily deposited at high plasma powers and low temperatures.

Alternatively, CNOs can be formed by annealing nanoparticles of highly hydrogenated carbon in vacuum at 1000°C. Graphitization starts from the outer-sphere towards the core of the particle with release of light hydrocarbons in two steps: first at about 500 °C and then above 800 °C as probed by TGA. The geometrical constrain of the NP (average diameter of 10nm), allow only the formation of curved graphitic planes, in a highly defective CNO structure. For 800 °C<T<1200 °C this process grants an high level defectiveness on the graphitic planes, while preserving high electronic conductivity and specific surface areas in excess of 700 square meter per gram. For T<800 °C the hydrogen contant in the nanoparticles is still too high, whilst for T>1200°C an excessive graphitization is observed with the loss of activity and surface area. The above-described process has been characterized by RAMAN, TGA, FTIR, in situ TEM, BET, XPS.

The CNOs electroactive film was characterized for several electrochemical applications like: redox flow batteries, H2O2 production, non-Pt catalyst for HER, HRR, ORR and supercapacitors. In particular, the mesostructured CNO electrode showed remarkable performances in vanadium redox flow battery applications. In fact the forementioned film, deposited over a carbon paper 39 AA produce by Sigracet, has been tested first in a symmetric cell, to evaluate electrode overpotential, then in a full cell. From charge/discharge curves of full cell test, the system efficiencies have been calculated and a nominal current density as high as 600mA/cm2 with 70.8% energy efficiency was obtained.