In this study we
show a chitosan:polyaniline (CPA)-based ink, responding
to eco-biofriendly criteria, specifically developed for the manufacturing
of the first organic memristive device (OMD) with an aerosol jet printed
conductive channel. Our contribution is in the context of bioelectronics,
where there is an increasing interest in emulating neuromorphic functions.
In this framework, memristive devices and systems have been shown
to be well suited. In particular organic-based devices are envisaged
as very promising in some applications, such as brain–machine
interfacing, owing to specific properties of organics (e.g., biocompatibility,
mixed ionic–electronic conduction). On the other hand, the
research activities on flexible organic (bio)electronic devices and
direct writing (DW) noncontact techniques increasingly overlap in
the effort of achieving reliable applications benefiting from the
rapid prototyping to accomplish a fast device optimization. In this
context, ink-based techniques, such as aerosol jet printing (AJP),
although particularly well suited to implement 3D-printed electronics
due to advantages it offers in terms of a wide set of allowed printable
materials, still require research efforts aimed at conferring printability
to the desired precursors. The developed CPA composite was characterized
by FTIR, DLS, and MALDI-TOF techniques, while the related aerosol
jet printed films were studied by SEM and profilometry. Taking advantage
of the intrinsic and stable electrical conductivity of CPA films,
which do not necessarily require any acidic treatment to promote a
sustained charge carrier conduction, 10 μm short-channel OMDs
were hence manufactured by interfacing the printed CPA layers with
a solid polyelectrolyte (SPE). We accordingly demonstrated prototypes
of stable and best performing OMD devices with downscaled features,
showing well-defined counterclockwise hysteresis/rectification and
an enhanced durability. These properties pave the way to further improving
performance, as well as to realizing a direct integration of the devices
into hardware neural networks by in-line fabrication routes.