Muscle-driven in vivo nanogenerator.

2010 WILEY-VCH Verlag Gmb Harvesting energy from the environment is crucial for the independent, wireless, and sustainable operation of nanodevices. This is a key requirement for building self-powered nanosystems. The living environment of nanodevices is diverse ranging from natural to in vitro and in vivo. Mechanical energy is one of the most abundant and popular energies in the environment, which can range from wind energy to mechanical vibration, sonic/ ultrasonic waves, noise, fluidics, biomotions, muscle stretching, and more. Harvesting energy using piezoelectric materials has been demonstrated some time ago, but these structures are rather large and a tiny physical motion, such as the contraction of a blood vessel is not strong enough to drive the generator. More importantly, the demonstrated cantilever-based microelectromechanical systems (MEMS) work only under a specific driving resonance frequency that is determined by the cantilever. In a real biological system, the mechanical disturbance has a large frequency range and the mechanical vibration is time dependent. We have demonstrated a few approaches for harvesting mechanical energy from several different sources, one of which was based on an alternating-current (AC) nanogenerator using a two-ends-bonded piezoelectric nanowire (NW). The NW is laterally bonded on a flexible substrate, and the physical deformation of the NW is directly driven by the shape change of the substrate that is induced by external dynamic mechanical sources. When a ZnO NW is subject to a periodic mechanical stretching and releasing, the mechanical–electric coupling effect of the NW, combined with the gate effect of the Schottky contact at the interface, results in a alternating flow of the charge in the external circuit. The single-wire generator (SWG) acts as a ‘‘charging pump’’ that drives the electronmotion in accordance to the mechanical deformation of the NW. Recently, we have applied the AC generator to harvest mechanical energy from body movement under in vitro conditions. However, the applications of the nanogenerators under in vivo and in vitro environments are distinct. Some crucial problems need to be addressed before using these devices in the human body, such as biocompatibility and toxicity. To directly interface nanowires with cells, our studies have indicated that ZnO nanowires can be safely used for in vivo applications and they are biodegradable. In this Communication, in vivo biomechanical-energy harvesting using an AC nanogenerator has been achieved for the first time. We demonstrate the implanting of the nanogenerator in a live rat to harvest energy generated by its breath and heartbeat. This study shows the potential of applying nanogenerators for the scavenging of low-frequency dynamic muscle energy created by very small-scale physical motion for the possible driving of in vivo nanodevices. The fabrication process of a SWG was presented in detail in our previous publication. The piezoelectric ZnO NW was grown using a physical-vapor deposition process and had a diameter of 100–800 nm and a length of 100–500mm. The two ends of the NW were tightly fixed to the surface of a flexible polyimide substrate by applying silver paste and two lead wires, isolated from the environment, were connected to the ends. Because of the presence of bio-fluids under the in vivo working condition, the entire device was covered with a flexible polymer to isolate it from the surrounding medium and to improve its robustness. The short-circuit current (Isc) and open-circuit voltage (Voc) were measured to examine the performance of the SWG. All of the measurements were performed in a wellgrounded and screened environment and the noise level was carefully minimized. The output Voc and Isc of the SWGare typically less than 50mV and 500 pA, respectively, in most cases. Therefore, careful experiments have to be conducted to rule out possible artifacts introduced by factors such as themeasurement system, change in capacitance of the nanowire and the electric circuit, and/or the coupling of the SWGwith the measurement system. A series of testing criteria has been established before to identify the true signal generated by the SWG. An effective SWG must exhibit Schottky behavior at one end before and after measurement (Fig. 1c). The output voltage and current of a SWG should meet the switching polarity test. For easy notation and reference, we define the side of a SWGwith Schottky contact as the positive side.When the positive and negative probes of the measurement system are connected to the positive and negative sides of the SWG, respectively, the configuration is described as being a forward connection. The configuration with the two probes switched over is defined as a reverse connection. Both configurations were tested. The magnitude of the signal from different connecting configurations may differ because of the influence of a small bias current in themeasurement system. So the magnitude of the true signal is an average of those under forward and reverse connections. The first group of experiments was based on the conversion of the mechanical deformation related to the periodic expansion and contraction of the diaphragm of a rat into electricity (Fig. 1a). Adult rats (Hsd: Sprague Dawley SD, male, 200–224g) were used for the experiment. Our procedure in handling the animals followed National Institutes of Health (NIH) policies and guidelines, university policies, and Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)