Li-ion batteries have high energy and power density, although their design, wherein sandwiched battery components are “jelly-rolled” into metal canisters (see Figure), puts constraints on the ultimate form factors of the devices they power. Flexible batteries, stretchable textile energy storage, and paper batteries are some of the unconventional battery designs that were devised for devices without form-factor constraints, although their seamless integration remains a challenge. Self-powered electronics with integrated energy storage also require batteries that can be directly incorporated into the device design.

Moving toward these goals, P.M. Ajayan and a team of researchers from Rice University have fabricated multilayer Li-ion batteries by sequentially coating their components onto commonly encountered materials by spray painting, as reported in the June 28 issue of Scientific Reports (DOI: 10.1038/srep00481). Component materials were formulated into liquid dispersions (paints), and the researchers chose lithium cobalt oxide (LCO, LiCoO2) as the cathode and lithium titanium oxide (LTO, Li4Ti5O12) as the anode, for which the effective cell voltage is about 2.5 V. While a commercially available conductive Cu paint was used for the negative current collector (analogous to Cu foil in Li-ion batteries), Al paint could not be employed for the positive current collector because Al micro-powders form explosive aerosols. In its place, the researchers made a viscous, highly consistent ink containing high concentrations (0.5–1% w/v) of single-walled carbon nanotubes (SWNTs) that was suitable for spray painting. Using commercial polymers, the researchers were also able to make a spray-paintable separator with a morphology that allowed optimal electrolyte uptake and formation of a microporous gel electrolyte with sufficient adhesion to adjacent layers to ensure mechanical stability.

figure 1

(a) A jellyroll assembly of the layered components of a conventional Li-ion battery; and (b) direct fabrication of a multilayer battery achieved by sequential spraying of component paints, using stencil masks tailored for a desired geometry and surface.

After preheating nonconducting substrates (glass, ceramics, and polymer sheets) to 120°C, the researchers spray- painted the component paints layer by layer, starting with the SWNT (cathode charge collector) paint, then the LCO (cathode) paint. After drying and preheating to 105°C, the polymer separator was sprayed on in several light coats up to a final thickness (~200 µm) that prevented internal shorting from solvent penetration from the LTO (anode) layer, which was deposited after preheating to 95°C. The Cu (anode current collector) paint was then deposited. After drying the cell in vacuum, it was transferred to an Argon-filled glove box, soaked in electrolyte, and then packaged with laminated poly(ethylene)-aluminumpoly(ethylene terephthalate) sheets.

For a typical cell, the SWNT layer is ~25 µm, the LCO layer ~120 µm, the polymer separator ~180 µm, the LTO layer ~90 µm, and Cu layer ~5 µm. Pla- teau potentials of about 2.4 V for charge and about 2.3 V for discharge were observed in the galvanostatic charge-discharge curves, with a discharge capacity of about 120 mAh per g of LTO. After 60 cycles the battery retained 90% of its capacity with about 98% Coulombic efficiency, which suggested that all integrated components were working efficiently.

The researchers also demonstrated the versatility of their battery design by connecting nine of them in parallel for a total energy of about 0.65 Wh. The research team glued an inexpensive polycrystalline silicon solar-cell array to the top of one cell and connected it with a current-limiter circuit. When fully charged—with white light illumination for the single cell and with a galvanostat for the other eight—the device powered 40 red light-emitting diodes for more than 6 h (at 40 mA) and could be easily reconfigured to supply different voltages and current capacities.