Recently, a bio-inspired, synthetic membrane-based hair cell sensor was fabricated and characterized. This sensor
generates current in response to mechanical stimuli, such as airflow or free vibration, which perturb the sensor’s hair.
Vibration transferred from the hair to a lipid membrane (lipid bilayer) causes a voltage-dependent time rate of change in
electrical capacitance of the membrane, which produces measurable current. Studies to date have been performed on
systems containing only two droplets and a single bilayer, even though an array of multiple bilayers can be formed with
more than 2 droplets. Thus, it is yet to be determined how multiple lipid bilayers affect the sensing response of a
membrane-based hair cell sensor. In this work, we assemble serial droplet arrays with more than 1 bilayer to
experimentally study the current generated by each membrane in response to perturbation of a single hair element. Two
serial array configurations are studied: The first consists of a serial array of 3 bilayers formed using 4 droplets with the
hair positioned in an end droplet. The second configuration consists of 3 droplets and 2 bilayers in series with the hair
positioned in the central droplet. In serial arrays of up to four droplets, we observe that mechanotransduction of the
hair’s motion into a capacitive current occurs at every membrane, with bilayers positioned adjacent to the droplet
containing the hair generating the largest sensing current. The measured currents suggest the total current generated by
all bilayers in a 4-droplet, 3-bilaye array is greater than the current produced by a single-membrane sensor and similar in
magnitude to the sum of currents output by 3, single-bilayer sensors operated independently. Moreover, we learned that
bilayers positioned on the same side of the hair produce sensing currents that are in-phase, whereas bilayers positioned
on opposite sides of the droplet containing the hair generate out-of-phase responses.
Recent research has shown that a new class of mechanical sensor, assembled from biomolecules and which features an
artificial cell membrane as the sensing element, can be used to mimic basic hair cell mechanotransduction in vertebrates.
The work presented in this paper is motivated by the need to increase sensor performance and stability by refining the
methods used to fabricate and connect lipid-encapsulated hydrogels. Inspired by superficial neuromasts found on fish,
three hydrogel materials are compared for their ability to be readily shaped into neuromast-inspired geometries and
enable lipid bilayer formation using self-assembly at an oil/water interface. Agarose, polyethylene glycol (PEG,
6kg/mole), and hydroxyethyl methacrylate (HEMA) gel materials are compared. The results of this initial study
determined that UV-curable gel materials such as PEG and HEMA enable more accurate shaping of the gel-needed for
developing a sensor that uses a gel material both for mechanical support and membrane formation-compared to
agarose. However, the lower hydrophobicity of agarose and PEG materials provide a more fluid, water-like environment
for membrane formation-unlike HEMA. In working toward a neuromast-inspired design, a final experiment demonstrates that a bilayer can also be formed directly between two lipid-covered PEG surfaces. These initial results suggest that candidate gel materials with a low hydrophobicity, high fluidity, and a low modulus can be used to provide membrane support.