(579f) Effect of Body Geometry On Motility of Bacteria-Powered Microrrobots (BACTERIABOTS)

 

Abstract? Motility of micro-scale
swimming robots falls in the realm of low Reynolds number, where viscous forces
exerted on the robots are dominant over inertia. In this work, the importance
of body geometry on the dynamics of bacteria propelled swimming microrobots
(BacteriaBots) is investigated. We have experimentally demonstrated that oblong
geometries enhance directionality of the micro-particles compared with
spherical geometry.

 
I.    
INTRODUCTION

Mobile micro-robots have unique
advantages such as the ability to access small spaces and the potential to be
employed in large numbers as inexpensive agents of distributed systems for
swarm robotic applications. Such micro-robots are envisioned to impact a
diverse range of applications, including minimally invasive diagnosis and
localized treatment of diseases, environmental monitoring, and homeland
security [1]. Amongst the most significant obstacles to
realization of mobile robots at micro-scale are the miniaturization of on-board
actuators, power sources, and communication and control modules. Bio-hybrid
approaches can be employed to address these challenges by integrating
prokaryotic and eukaryotic cells within the robotic system [2-4]. BacteriaBot, a bio-hybrid
micro-robot is constructed here by interfacing a microfabricated robot
body with an ensemble of live engineered bacteria with the purpose of using the
bacteria for actuation, sensing, communication and control. Motile behavior of
bacteria is well characterized and falls under two characteristic modes of run
and tumble. This random run and
tumble behavior leads to the characteristic three dimensional random walk of
bacteria and consequently to the stochastic motion of micro-objects propelled
by bacteria.

Motility of spherical
microstructures actuated by an ensemble of attached bacteria has been thoroughly
characterized in previous literature [5-8]
but a systematic study of the effect of microstructure geometry on propulsive
behavior is currently missing. Mobile microrobots with optimal body geometries are
envisioned to impact minimally invasive diagnosis, localized treatment of
diseases and environmental monitoring. Limited particle diffusion and
directional coefficient of drag are some of the attributes that are enhanced
through such bio-hybrid systems. In this work, we have utilized a low-cost and high
throughput technique to obtain non-spherical mico-particles and investigate the
effect of particle shape on the motile behavior of the BacteriaBots.

 
II.    
Materials and methods
A. BacteriaBot Body Fabrication

A
high throughput spherical particle casting and mechanical stretching under heat
treatment, as previously described in [9], is used to produce non-spherical
polystyrene (PS) particles which act as the engineered synthetic body of the
robots. Briefly, 6 µm PS spheres (Polysciences) are casted in a 35 µm thick
polyvinyl alcohol (PVA) film. The film is uniformly stretched in one dimension to
generate voids around the micro-spheres. The casted PS micro-spheres are
liquefied using a bath of hot mineral oil. For making football shaped
particles, 0.5% glycerol is added to the PVA film as a plasticizer. After
removal from bath and cooling, particles are released from the PVA film by
soaking in 30% isopropyl alcohol (IPA): DI water solution at 80˚C, and are
washed by centrifugation in a 30% IPA solution.

 
B. Bacteria culture
Escherichia coli (E. coli) strain
MG1655 was cultured in Luria Broth (1% tryptone, 0.5% yeast extract, and 0.5%
sodium chloride). The culture was grown to an optical density OD₆₀₀=0.5
at 37 ˚C.

 
C. BacteriaBot Construction

The mixture of
Poly-_L-lysine and micro-particles is incubated on a vortex mixer for
one hour. Bacteria are centrifuged at 3000g for 5 min and then resuspended in
motility medium (0.01M potassium phosphate, 0.067M sodium chloride, 10^-4M
EDTA, 0.21M glucose, and 0.002% Tween-20). The poly-_L-lysine coated
micro-particles are added to the bacteria allowing the bacteria to self-assemble
on the particles.

D. Two-dimensional Single Particle Tracking

The
motion of the microbeads was captured at 20 frames per second using a Zeiss
AxioObserver Z1 inverted microscope equipped with an AxioCam HSm camera. The
images were analyzed using a two-dimensional (2D) particle tracking routine
developed in MATLAB (MathWorks, Natick, MA). Briefly, using cell segmentation
and image restoration, the artifacts existing in most of the captured images
were removed. This was followed by noise removal and cell boundary recognition
using a border following algorithm. Finally, the nearest-neighbor method was
used to link segmented cells in successive frames and to generate the
trajectories.

 
III.    
Results and Discussions

Two
different particle geometries were fabricated using a casting and mechanical
stretching method that was described earlier. Parameters used for the manufacturing
and the scanning electron microscopy (SEM) images of the resulting geometries
are given in Table 1 and Fig. 1, respectively. Representative images of the
bacteriabots with different geometries are depicted in Fig. 2. The 2D image tracking
routine was utilized to characterize the motion of the BacteriaBots. Of
particular interest is how the body geometry will affect the directionality of
the motion of BacteriaBot without the need for active steering. Directionality
is defined as the ratio of the magnitude of the displacement vector to the
total distance traveled. To prove that the bacteria attached to the mobile microbeads are
the source of propulsion, a control experiment was performed. Minimal displacement
of the control bead confirms that the bacteria attached to the mobile
microbeads are the main source of propulsion. Therefore, any directed movement
observed for this control bead would be neither due to diffusion nor due to the
flow field generated by the free- swimming bacteria present in the background.
The experimental results for directionality of the bacteria- propelled bodies with
spherical, barrel, and football-shaped geometries are shown in Fig. 3. It
can be seen that the barrel and football shaped bodies are propelled with a
higher degree of directionality compared to the spherical robots. Two
representative trajectories of the BacteriaBots are shown in Fig. 4. Number of attached bacteria for all geometries varies
between 1-6; however, our experimental results do not seem to be significantly
dependent upon the number of bacteria attached. This is consistent with our
previous observations if the attached bacteria are uniformly distributed over
the body, the overall force is expected to remain largely unchanged regardless
of the number of attached bacteria [7].

 

It
should be noted that more than the number of bacteria attached to the
microbead, the location of attachment is expected to affect the overall
behavior. If the areal attachment density becomes significantly nonuniform, we
expect to see a change in the average net resultant force and consequently
observe a change in overall speed and directionality [3].
IV.    
Conclusion

Variation in
motile behavior of BacteriaBots due to their body geometry can be very complex
and can only be determined experimentally. These complexities are due to: (1)
non-spherical geometries have varying coefficients of drag depending on their
aspect ratio and the direction of motion, and (2) the varying radius of curvature
on the surface of non-spherical geometries leads to preferential bacterial
adhesion in certain locations. By utilizing a high throughput PS micro-particle
manufacturing method, we characterized the motile behavior of BacteriaBots with
spherical, barrel and football shaped bodies. It was shown that for
BacteriaBots with uniform areal attachment density, body shape strongly affects
the directionality of the motion.

We are currently examining the
behavior of other geometrical shapes such as bullets. We are also investigating
if the particles size will have an impact on the observed trend.

Acknowledgment

The
authors would like to thank Professor Birgit Scharf in Biological Sciences
Department at Virginia Tech for gifting the bacteria. This work was in part
supported by the National Science Foundation (IIS-117519).

References

[1] B. Behkam, M. Sitti, "Bacteria Integrated
Swimming Micro-robots", in M. Lungarella, et al., (eds.), 50 years of AI,
Festschrift, Lecture Notes in Artificial Intellignece 4850, pp. 154-163, 2007.

[2] S. Martel, et al.,
"Controlled manipulation and actuation of micro-objects with magnetotactic
bacteria," Applied Physics Letters, vol. 89, pp. 233904-233904-3,
2006.

[3] A. A. Julius, M. S. Sakar, E. Steager,
U. K. Cheang, M. Kim, V. Kumar,and G. J. Pappas, ?Harnessing bacterial power
for micro scale manipulation and locomotion,? IEEE International Conference on
Robotics and Automation (IEEE, Kobe, Japan, 2009).

[4] D. B. Weibel, et al.,
"Microoxen: Microorganisms to move microscale loads," Proceedings
of the National Academy of Sciences of the United States of America, vol.
102, p. 11963, 2005.

[5] B. Behkam and M. Sitti, "Bacterial
flagella-based propulsion and on/off motion control of microscale
objects," Applied Physics Letters, vol. 90, p. 023902, 2007.

[6] B. Behkam and M. Sitti, "Effect of
quantity and configuration of attached bacteria on bacterial propulsion of
microbeads," Applied Physics Letters, vol. 93, p. 223901, 2008.

[7] M. A. Traoré, et al.,
"Computational and experimental study of chemotaxis of an ensemble of
bacteria attached to a microbead," Physical Review E, vol. 84, p.
061908, 2011.

[8] B. Behkam and M. Sitti,
"Characterization of bacterial actuation of micro-objects," Proceedings
of IEEE Conference on Robotics and Automation (ICRA) 2009, pp. 1022-1027.

[9] J. A. Champion, et al.,
"Making polymeric micro-and nanoparticles of complex shapes," Proceedings
of the National Academy of Sciences, vol. 104, p. 11901, 2007.

---