Months 0 - 6
The project started on two parallel research lines: experiments aiming at making CNT-inks for coating, and the prototyping of coating heads and of a small homemade coater.
A screening of scientific literature allowed selecting the most suitable surfactants for the inks, based on their ability to stick to CNTs and a low toxicity: Sodium dodecyl sulfate (SDS) and Sodium dodecyl benzene sulfonate (SDBS). The ink’s fabrication protocol relied on the exfoliation of CNTs/surfactant solutions in a bath sonicator, followed by a centrifugation step to remove largest CNT clusters.
All coating heads were designed via CAD software and then printed on a chromium mask. The first devices were prototyped by soft-lithography, namely by moulding a polymer (PDMS) on patterned photoresist. The homemade coater was designed on CAD to get a 3D model, then its parts were manufactured in acrylic by laser-cutting and assembled into a frame to fix the PDMS heads for initial evaluation of the coating process.
Months 7 – 12
At concentrations above 1wt.%, CNTs form a nematic liquid crystal, which would be ideal for coating aligned CNTs as it provides high-density and preferential direction. By combining Thermogravimetric and UV-Vis spectroscopy, the applicant measured the concentration of the inks and found a CNT content of 0.05 wt.% after centrifugation, namely too low to expect any CNT alignment. To boost this value, he thus looked at ways to improve the exfoliation yield; first, by processing finer CNTs solutions obtained by ball-milling the feedstock. Yet this route became unpractical due to contamination of the CNTs-solutions by the wear of the system. Secondly, the applicant optimized the ink’s formula and the parameters of the bath-sonicator. The results evidenced a strong increase in yield for inks that have a surfactant’s concentration higher than the Critical Micelle Concentration. A concentration of ~0.1 wt.% could be achieved this time.
Attempts to use the homemade coater showed that the tool was not reliable enough to yield coatings better than those obtained by drop-casting. Also, when the inks were concentrated in a rotary evaporator, they became so viscous that their extrusion causes the catastrophic failure of the PDMS head due to excessive pressure. Still, the efforts provided enough understanding to reconsider some of the project priorities. The host-group hired a post-doc researcher to build an improved version of the coater using a metal frame on a motorized stage. The group also purchased an industrial coater from the company Coatema (Germany).
The viscosity of the ink is due to the larger amount of surfactant compared to that of CNTs in it. The applicant thus implemented a new technique to filter out this excess of surfactant, specifically by centrifuging the ink at a temperature below the surfactant’s Krafft’s point, where micelle’s formation is inhibited. Such solution is original and versatile as it can be extended to other surfactants. At present, these findings are considered for publication.
Months 13 - 18
After year-1, it became clear that making CNT-inks is a complex process often giving inconsistent results, and that the protocols in literature are insufficient. To overcome this challenge, still achieving high-throughput, the applicant assembled a modular setup based on a powerful tip-sonicator (courtesy of Prof. Ian Hutchins at the host-institution) coupled to a flow-thought cell (see figure). A pump recirculates the CNT solution from a reservoir to the cell, in closed loop. The cell’s temperature is stabilized by a chiller. A homogeniser was added in series, so to break CNTs clusters similarly to ball-milling. The setup complies also to tight safety margins, as the applicant built a sound-box around the sonicator and fit it into a fume-hood to avoid the risks of exposure to nanoparticles and of noise induced hearing loss.
Overall, this continuous-flow system greatly increases the throughput in CNT-ink production compared to the state-of-the-art. With the setup in place, the applicant developed a protocol to get excellent reproducibility of results on samples, which allowed to identify a model that captures the evolution of the CNT concentration in a simple equation with only two input parameters. Next, the process was scaled up using the continuous-flow setup. Ultimately, we identified a set of scaling rules to link the model to the results on the scaled-up system. This was a major breakthrough as it links a robust protocol for processing small volumes of ink (i.e. samples) to the results of a high-throughput system for larger volumes (i.e. 200ml or more). All these results are now being submitted to a major peer-reviewed journal.
Months 19 – 24
To understand whether our tip-sonicator was powerful enough to exfoliate CNTs as individuals, the applicant screened inks made from CNTs of different producers: MW-CNT from Nanocyl and SW-CNT from Timesnano, Ocsial and Sigma-Aldrich. Only the latter exhibited the fingerprint of individual CNTs, namely isolated absorption peaks in the UV-Vis range, which lead to the conclusion that our system can indeed provide individual CNTs.
In the last part of the project we capitalized the results to guarantee continuity to this research. The CNT-inks triggered interest in a collaboration with Prof. Dal Pont (Grenoble University – France) aiming at embedding CNT in concrete to enhance its mechanical properties. Besides, two publications are under submission to scientific journals: the first on the continuous-flow setup, the second on the purification of inks from the surfactant in excess. Both topics are industrially relevant and perfectly aligned to the objective of the EU to promote new manufacturing processes for nanomaterials.
