limitations since wider electronics can cover a larger sensing area. All meshes have
sharps tip of 45ºto be easily loaded into syringes (Fig.5.10b). Two meshes with
different unit cell geometries have been used here to investigate the injection. In
design #1, the transverse ribbons are tilted 45° counterclockwise in transverse
direction on the mesh plane forminga= 45° to longitudinal ribbons. In design #2,
the transverse ribbons are perpendicular to the longitudinal ribbons to forma=0ºto
longitudinal ribbons. Figure5.10c–e show optical micrographs, 3D-reconstructed
and cross-section images of assembled structures for each mesh in glass channels.
Firstly, 5-mm-wide mesh witha= 45° design structure can be smoothly injected
through a channel with ca. 500-μm-inner-diameter (Fig.5.10c, I). The
3D-reconstructed image shows that the mesh rolls into a tubular structure inside the
channel keeping longitudinal ribbons straight and transverse ribbons bended
(Fig.5.10d, I). The cross-section image of 3D reconstruction further confirms this
tubular structure, illustrating that all of the ribbons are closely and uniformly
packed to the inner surface of glass channel. The other half of mesh in the bottom
part of the needle is blocked from imaging by the dense ribbons on the top part of
channel. Secondly, we found that reduction of channel’s ID did not affect the
assembled structure of mesh in the needle. The same mesh can be injected smoothly
in through 200-μm ID channel (Fig.5.10c, II). 3D-reconstructed and cross-section
images further demonstrate the tubular structure of mesh in the needle and closed
packed ribbons to the inner surface of channel (Fig.5.10d, II, e, II). Thirdly,
increasing the width of mesh can also allow the mesh to be smoothly injected
through channels. As a representative example, 15-mm-wide mesh can be injected
through the channel with an ID of ca. 500μm (Fig.5.10c, III). The width-to-ID
ratio is ca. 30. The 3D-reconstructed and cross-section images (Fig.5.10d, III, e,
III) also show a tubular structure of mesh in the channel and closed packed ribbons
to the inner wall of channel. We found that the longitudinal ribbons can remain
straight during injection even further increase the numbers and densities of ribbons.
Fourthly, as a control sample, we found that the mesh witha= 0° design could not
pass through the channel with 500-μm ID and easily forms a jammed structure
(Fig.5.10c, IV). 3D-reconstructed and cross-section imaging further shows the
ribbons entangled together and block the channel (Fig.5.10d, IV, e, IV). Together,
these results highlight the point that longitudinal ribbons need to keep straight
during injection to avoid a high-strain deformation that could damage devices and
buckles that could dramatically decrease the stiffness of the structure in the lon-
gitudinal direction [ 29 ], and therefore, collapse longitudinal ribbons rather than
bend transverse ribbons causing large strain to damage device and block the needle
for the further injection.
To quantitatively explain these experimental results, we define the bending
stiffness for mesh bent in longitudinal direction and transverse direction of injection
asDLandDTrespectively. Firstly, thea= 0° design leads to a non-uniform dis-
tribution of effective bending stiffnessDL. Considering the effective bending stiff-
ness DL of different cross-sections, when the cross-section goes through the
transverse ribbons, the bending stiffness is high (0.0602 nN m), while when the
cross-section does not go through the transverse ribbons, the bending stiffness is
5.3 Results and Discussion 83