MIT’s mini batteries to power microscopic robots with high energy

Researchers at the Massachusetts Institute of Technology have created microscopic zinc (Zn)-based batteries capable of delivering high energy in volumes of only two picoliters each.

What’s more impressive is that one 2-inch silicon wafer could be used to create 10,000 of these micro-batteries, which would have enough energy to power tiny sensors and robotic parts.

The device uses oxygen from the air or solution to trigger a zinc oxidation reaction, achieving an energy density of 760 to 1070 watt-hours per liter in batteries smaller than 100 micrometers wide and 2 micrometers thick.

MIT researchers believe that this approach may contribute to the downscaling of energy storage systems for incorporation into devices the size of cells.

Advancing micro batteries

Technological advances have consistently reduced the size of robotic devices, but shrinking their energy sources remains a challenge.

Traditional battery materials and fabrication methods often don’t align with microelectronics, limiting micro battery sizes to the millimeter range—much larger than the devices they power.

Additionally, alternative power sources like solar cells restrict functionality in low-light conditions, such as underground or inside pipelines.

To overcome these limitations, innovative battery solutions that are compatible with the size and fabrication techniques of microelectronics are needed to ensure reliable operation across diverse environments and conditions.

As a solution, researchers looked at designing a Zn-air picoliter battery that can be photolithographically etched onto a silicon wafer and then released into a solution.

The team developed a tiny zinc/platinum/SU-8 battery using a process called photolithography, which allowed them to create a very high-energy-density micro battery at a picoliter (10⁻¹² liter) scale.

These batteries use oxygen from their surroundings or dissolved in solutions to power the zinc oxidation reaction. The batteries have an energy density of 760 to 1070 watt-hours per liter and measure less than 100 micrometers wide and 2 micrometers thick.

The photolithography technique is efficient, enabling researchers to produce 10,000 of these batteries on a single wafer. When released, these batteries can be suspended in liquid as colloidal particles, each storing energy on board.

According to the team, each battery is only 2 picoliters in volume and can produce voltages of around 1.05 volts, with total energies between 5.5 to 7.7 microjoules, and maximum power close to 2.7 nanowatts.

Micropower for tiny robots

Tests showed that these batteries could power very small devices, such as a micrometer-sized memristor circuit, which is useful for nonvolatile memory.

Researchers also used the batteries to power microscale actuators that bend back and forth at a rate of 0.05 hertz, which is useful for robotic movement. Additionally, they powered two different types of nanosensors and a clock circuit.

The colloidal batteries are designed for use in large electrolyte reservoirs, but some applications require them to work in dry environments or without available ionic species. To address this, the researchers conducted two proof-of-concept experiments.

First, they used a micro ink-jet printer to deposit 500 picoliters of ionic liquid electrolyte on the batteries, which resulted in lower voltage due to higher viscosity and lower conductivity. These batteries delivered 0.82 microjoules of energy at 0.1 mA cm−2.

In the second experiment, 20 picoliters of salt were dried onto the battery. When immersed in deionized water, these batteries released salt, producing 0.6 to 1.2 microjoules of energy. According to researchers, these experiments show that Zn-air batteries can function in different environments, even without a large electrolyte supply.

Thanks to their high energy density, small size, and straightforward design, these picoliter zinc-air batteries are promising for mass production and use in tiny robots that function independently.

The details of the team’s study were published in the journal Science Robotics.