Printing of aqueous carbon nanotube (CNT) inks for flexible electronics


Carbn nanotubes are a wonder material discovered in a research lab in 1991 and improved since then. They're exceptionally strong, electrically and thermally conductive, and can be better than normal (metal) electronics at high frequencies, high temperatures, and small scales.

To advance the development of new-age technologies, we are developing manufacturing methods to make mm-cm scale circuits from carbon nanotubes.

For more information, see our article:

Owens CE, RJ Headrick, SM Williams, AJ Fike, M Pasquali, GH McKinley, AJ Hart (2021) Substrate-Versatile Direct-Write Printing of Carbon Nanotube-Based Flexible Conductors, Circuits, and Sensors.

The full text can be accessed for free here:

Here is a related presentation abstract.

Extrusion printer design

For this project, a standard Makergear M2 was modified to allow fine-scale extrusion by connecting a linear actuator (Thorlabs) with 3D printed syringe mounts for small-volume syringes (blue in this image).

As the nozzle moves, fluid is extruded from the syringe onto our 2D substrate. By controlling the motion path, motion speed, and extrusion speed, we can create designs with tunable conductivity and geometry along a surface.

(a) As one demonstration, we can draw out arbitrary 2D structures like text. (It's cursive because the paths must be connected in order to conduct electricity!) (b) An LED in the center shows that conductivity is maintained when our sticker is folded around different items in lab (e-h).

d) We measured the resistance divided by the resistance when it is flat, and how that varies as it is wrapped around items with larger curvature (smaller radius).

The conductivity of features, measured as sigma*Ac = specific conductivity, could be modified by the linear density (rho*Ac) of carbon nanotubes as printed.

For example, printing with a lower-concentration ink (like the pink data, 0.06 mg of carbon nanotubes per mL of water) generally made lines with lower conductivity and density, while printing with higher-concentration ink (like the blue data, 7.6 mg of carbon nanotubes per mL of water, 100x more concentrated) made lines with higher conductivity and density.

Printing the pink solution very slowly would create lines of equal conductivity to printing the blue solution very quickly, as these properties were dependent on the amount of carbon nanotubes, not on the ink itself.

Power law trends of 2.17 and 1.08 indicate transitions in percolation scaling of this conductive material.

By applying our understanding of process properties we could avoid adding plastic fillers to the carbon nanotube inks. Altogether, we printed carbon nanotube wires with 50x higher conductivity than previously printed carbon nanotube materials (seen in the yellow oval).

Here, the horizontal axis is elastic compliance = 1/young's modulus, an indicator of flexibility.

The vertical axis is conductivity (sigma) divided by density (rho), meaning conductivity per weight.

Metals are more conductive but stiff, affecting flexibility. Conductive polymers like PEDOT:PSS and metal aerogels are more flexible but less conductive.

(a-b) We finally managed to print onto a variety of different plastic and paper substrates, and (c) use wetting interactions (Wenzel wetting) to predict the feature size (width w divided by the nozzle size D0 and the contact angle theta-star) based on how much the ink spread.

See our paper for full details!