Heart Energy Harvesting and Cardiac Bioelectronics: Technologies and Perspectives

: Nanogenerators are a recently emerging technology which is able to cost-effectively harvest energy from renewable and clean energy sources at the micro/nano-scale. Their applications in the ﬁeld of self-powered sensing systems and portable power supplying devices have been increasing in recent years. Wearable and implantable electromechanical/electrochemical transducers for energy harvesting represent a novel alternative to chemical batteries for low-power devices and to exploit the energy conveyed by human biomechanics. The human heart, in particular, is a compelling in vivo source of continuous biomechanical energy and is a natural battery which can power implantable or wearable medical devices. This review describes the recent advances in cardiac wearable/implantable soft and ﬂexible devices and nanogenerators for energy harvesting (piezoelectric nanogenerators, triboelectric nanogenerators, biofuel cells, solar cells, etc.), as well as cardiovascular implantable electronic devices in a more general sense, as components of more complex self-sustainable bioelec-tronic systems for controlling irregular heartbeats or for interventional therapy for cardiac diseases. The main types of soft heart energy harvesters (HEHs) and heart bioelectronic systems (HBSs) are covered and classiﬁed, with a detailed presentation of state-of-the-art devices, and the advances in terms of materials choice, chemical functionalization, and design engineering are highlighted. In vivo bioelectronic cardiac interfaces are outlined as well as soft devices for in vitro cardiac models (patch and organoids). Cutting-edge 3D/4D bioprinting techniques of cardiac tissue are also mentioned. The technical challenges for the practical application and commercialization of soft HBSs are discussed at the end of this paper.


Introduction
The human heart is a compelling in vivo source of biomechanical energy: its rhythmic cycle of contraction and expansion serves as a pump for blood circulation and it provides a periodic excitation for devices that can transduce this energy into electrical energy. This concept of heart energy harvesting has been recently considered as one of the possible alternatives which can power implantable biomedical devices (IBDs), such as pacemakers, with clean energy sources [1][2][3][4][5]. The most common IBD is the heart pacemaker, which is currently used to maintain hearts beating regularly and for the treatment and monitoring of bradyarrhythmia. Standard pacemakers are based on chemical batteries which need periodic (every 5-12 years) surgical replacements due to their restricted lifetime, besides not being environmentally friendly and unsuitable for positioning in hostile/unusual places, such as the human skin or the in-body environment [6][7][8][9][10]. Difficulty in charging, insufficient battery life, risk of restart, costs of the surgical interventions, and risks of complications and infections are the main problems related to the use of pacemakers. Flexible transducers and nanogenerators have been proposed in recent years as a novel useful technology for scavenging the heart's natural energy and providing power to selfsustained sensor systems or other miniaturized implantable devices, such as batteryless Nanoenergy Adv. 2022, 2, FOR PEER REVIEW 3 applicative perspectives and future challenges for the advancement and commercialization of the HEHs and HBSs are discussed in the conclusion.

Heart Anatomy and Heart Energy
Situated between the lungs, the heart is one of the most important organs in the body and pumps blood into and from the rest of the body through the network of arteries

Heart Anatomy and Heart Energy
Situated between the lungs, the heart is one of the most important organs in the body and pumps blood into and from the rest of the body through the network of arteries, arterioles, veins, venules, and capillaries, which form the cardiovascular system. Blood is carried from the heart to the periphery through the arteries to deliver oxygen and nutrients to cells and tissues, whereas it is returned to the heart through the veins to remove carbon dioxide and waste products. Figure 1 shows the anatomy of the heart: it is made of two upper chambers, i.e., left and right atriums, and two lower chambers, i.e., left and right ventricles (Figure 1b). These are connected through four valves. The right atrium receives non-oxygenated blood from the body (through the superior and inferior vena cava) and pumps it through the tricuspid valve into the right ventricle. From here, the blood is pumped through the pulmonary valve towards the lungs where it becomes oxygenated and enters the left atrium, which pumps it through the mitral valve into the left ventricle. Finally, the oxygen-rich blood is pumped by the left ventricle through the aortic valve to the aorta and is delivered to the rest of the body (Figure 1c). The surface of the heart, represented by a membrane called pericardium, is irrorated by oxygen-rich blood flowing through the coronary arteries, whereas the contraction/relaxation motions of the heart are governed by electrical signals delivered by nerves.
The heart is a periodic pump that acts as a cyclic source of biomechanical energy. As reported in [49], the heart energy system is characterized by two phases: the internal load, provided by the mechanical behaviour of the ventricles, and the external load, given by the hydraulic energy and vascular resistance of the blood flow through the arterial network [50]. The heart generates a pulse that establishes the necessary pressure for driving blood circulation, and this is accompanied by another complementary driving mechanism attributed to the vessels (see Figure 1d) [51][52][53][54][55]. On one hand, the diastole determines the degree of filling of the ventricles and the strength of the upcoming systole, and the latter generates a pulse which is energetic enough to overcome the friction forces opposing ejection. The external load is characterized by a steady component given by the energy lost due to vascular resistance, and a pulsatile component represented by the energy lost in arterial pulsations due to aortic hydraulic impedance. Moreover, the heart is an aerobic pump and needs myocardial oxygen to be activated; thus, it is important to consider the cardiac mechanical efficiency, i.e., the useful stroke work compared to the oxygen utilized, which is normally~25% and decreases in the case of diseases [49]. Cardiometabolic disorders are generally due to failure in energy transduction by the myocardium and the occurring of glycolysis and ketone oxidation. Besides the pharmacological administration of drugs, the cardiac metabolism can also be activated or enhanced by endocardial electrical stimulation [56].
The cardiac muscle has an average lifetime of 70 years, and it lasts for more than 1.8 billion cycles at a heart rate of 70 beats per minute [57]. The human heartbeat induces the myocardium deformation with a frequency of 1-3 Hz and a resulting strain of 15-23% in the radial, and 9-12% in the circumferential, direction [58]. The blood pressure gradient, in contrast, is in the range of 20-100 mmHg in the right and left ventricles and 40 mmHg in the arterial system. The blood vessels undergo remarkable diameter distensions (e.g., carotid artery 10%, and brachial artery 3.7%) which provide useable sources of biomechanical energy.

Standard Cardiac Technologies
In this section, the standard technologies (pacemakers, cardiac defibrillators) used in contact with the heart are described, with a focus on the materials and device fabrication.
Electrotherapy consists of using an external source of electricity to stimulate human tissues in order to achieve a beneficial therapeutic effect. The first attempt to treat bradycardia through electrotherapy consisted of external pulse generators connected to the heart with resulting mobility limitations and discomfort [59]. The devices were powered by series-wired mercury-zinc batteries which released hydrogen and, thus, were not compatible with sealant and encapsulations for implantable systems. The first fully-implantable pacemaker was introduced in 1958 in Sweden, and it was based on epicardial electrodes connected directly to the heart and guaranteed a functional operation for 3 h [60]. It was equipped with a rechargeable nickel-cadmium battery bearing a cell voltage of 1.25 VV and a capacity of 190 mAh. These first devices were bulky and had a limited lifetime; however, over the years, technological advancements improved the lead design, reduced the size, and increased the battery longevity, with a resulting increased functionality and reliability. To overcome the issue of restricted lifespan, nuclear-power (or isotope-based) pacemakers were introduced in 1970 [61], with a longer lifetime (>30 years) and synchronous pacing capabilities, although with safety and toxicity issues due to the use of isotopes (such as 238Pu). These pacemakers were then replaced in 1975 by lithium-iodine batteries which can provide power to the devices with a resulting stable pacing for 10 years or more, with the aid of software algorithms [62,63]. These batteries usually include a single central lithium anode surrounded by a cathode material made of 96% iodine and thermally fixed with a polymer [64]. The modern pacemakers have the capability of sensing the intrinsic rhythm in the atrium and ventricle and pace each chamber on demand Figure 2a. The essential key requirements for cardiac IBDs include: voltage, discharge current, intensity and duration of current pulses, power density, a long lifetime, and adaptability to different operating conditions, as well as a biocompatible, corrosion-resistant, hermetically encapsulated, and lightweight battery which can satisfy a power demand of 100-200 µW [58]. Lithiumsolid-cathode primary batteries are usually preferred over lithium-liquid-cathode systems for implantable devices [58], and the typical materials used in these batteries are copper oxide (CuO), manganese dioxide (MnO 2 ), vanadium oxide (V 2 O 5 ), and carbon monofluoride (CF)n.
Standard pacemakers and cardiac electronic interfaces require external power supplies: they need~15 µJ, resulting in an annual power consumption of 10-100 µW (0.5-2 Ah over 5-10 years); implantable cardiac defibrillators, and especially cardiac resynchronization therapy defibrillators, need even more power (~40 µJ) to generate the defibrillation shock [58]. Thus, they are associated with complications, such as defects in the batteries, housekeeping current drain, lead impedance, electromagnetic interferences with cellular devices, security systems, energy generators, and wireless power suppliers [65]. Electromagnetic fields can in fact hinder or inhibit the pacing provided by pacemakers. Another issue is related to the leads or generator insertion or malfunction, as well as risk of infections or thrombosis [66][67][68]. More recently, leadless pacemakers were developed (Figure 2b), in which the pulse generator, the battery, and the pacing electrodes are contained in a small capsule that is inserted into the ventricle via a steerable sheath through the femoral vein [69][70][71]. Thus, they do not have leads and their implantation does not require an open surgery. However, these devices require a large-bore venous delivery system and provide only single-chamber pacing, even though dual leadless pacemakers can be developed with wireless technologies for pacing in both chambers [72] (Figure 2c). Additionally, these pacemakers present the same drawbacks of normal pacemakers; in fact, their battery life lasts six-seven years and, after the battery is depleted, the pacemakers need to be extracted and replaced by a new device [73]. Other types of pacemakers, with a subcutaneous generator connected to endovascular pacing leads, can provide single-chamber, dual-chamber, biventricular, and other pacing modalities. Cingolani et al. [74] give a comprehensive overview of the standard electronic pacemakers and describe, also, the next-generation pacemakers which combine different sensing systems (e.g., accelerometer, impedance, blood pressure, etc.) and software algorithms. Evolution of electronic pacemakers: the first implantable devices were big with epicardial leads for asynchronous pacing, modern devices are smaller, endowed with catheter-like endocardial leads which can be implanted with minimally invasive techniques and capable of providing synchronous pacing. More recently, leadless pacemakers have been introduced which can wirelessly provide pacing and can be implanted percutaneously through the femoral vein. (c) Example of leadless pacemaker, with the indication of the electronic interface used to provide pacing. Adapted with permission from reference [72], 2018 Copyright Elsevier.

Micro-Devices for Heart Energy Harvesting
One of the main issues of standard cardiac technologies and pacemakers is the need for frequent replacement of the chemical batteries. Usually, they are replaced every ~10 years; however, it also depends on the specific utilization of the device. The surgeries required for these interventions represent a cumulative risk for the patient's health, which could be avoided if the devices were self-powered without the need for recharging. This Evolution of electronic pacemakers: the first implantable devices were big with epicardial leads for asynchronous pacing, modern devices are smaller, endowed with catheter-like endocardial leads which can be implanted with minimally invasive techniques and capable of providing synchronous pacing. More recently, leadless pacemakers have been introduced which can wirelessly provide pacing and can be implanted percutaneously through the femoral vein. (c) Example of leadless pacemaker, with the indication of the electronic interface used to provide pacing. Adapted with permission from reference [72], 2018 Copyright Elsevier.

Micro-Devices for Heart Energy Harvesting
One of the main issues of standard cardiac technologies and pacemakers is the need for frequent replacement of the chemical batteries. Usually, they are replaced every~10 years; however, it also depends on the specific utilization of the device. The surgeries required for these interventions represent a cumulative risk for the patient's health, which could be avoided if the devices were self-powered without the need for recharging. This is the reason why, in recent years, many efforts have been made towards self-sustaining energy-scavenging technologies based on clean and affordable energy sources. Common oscillation generators (OGs) for harvesting the heart's kinetic energy are based on quartz clocks. Goto et al. [75] presented an OG device implanted on the right ventricular wall of a dog, which was capable of storing 13 pJ per heartbeat. Zurbuchen et al. [57], instead, described a mass-imbalance electromagnetic OG implanted on a sheep heart, consisting of clockwork from a commercially available automatic wrist watch and which was able to generate 11 µJ per heartbeat. The OGs, however, are bulky, heavy, and not reliable in the long-term. On the other hand, advanced heart energy harvesting systems (HEHs) are usually microfabricated devices which can be implemented on or inside the human body and can rely on different energy transducing mechanisms. They have been explored in plenty of preclinical studies in different animal models although additional tests would be required in the future to evaluate their reliability and robustness, in perspective, for clinical purposes. In this section, piezoelectric and triboelectric nanogenerators (PNGs and TNGs), solar-powered devices, and biofuel cells (BFCs) for cardiac applications are described, and the main examples are listed in Tables 1 and 2. Table 1. Summary of the main working principles and advantages/disadvantages of the microdevices for heart energy harvesting described in Section 4.

Heart Energy Harvesters (HEHs) Working Principle Advantages Disadvantages
Piezoelectric nanogenerators (PNGs) The energy generation is provided by the intrinsic properties of piezoelectric materials which convert even quasi-static mechanical deformation into electrical signals and power.

Piezoelectric Nanogenerators (PNGs)
Mechanical energy is ubiquitous and abundant, and the biomechanical energy conveyed by the heart is continuous, periodic, and endless. The rhythmic cycle of contraction/expansion provides the optimal conditions for exploiting the piezoelectric effect: when a PNG is subjected to a mechanical force, the piezoelectric active material intrinsically estab-lishes an internal electrical charge and potential [76][77][78]. In order to exhibit piezoelectricity, a material needs to have a non-centrosymmetric crystalline structure [79]: under a mechanical strain, a polarization occurs in the microstructure due to a displacement of the positive and negative charges (anions and cations), and the resulting potential distribution triggers the electrons flow through an external circuit, generating current and power [80]. Piezoelectric materials are commonly used and synthesized as thin films [81][82][83][84][85] or as nanostructured layers [86][87][88], and they belong, basically, to two big families, i.e., piezo-ceramics and piezopolymers [80]. Flexible PNGs are made either monolithically by piezoelectric polymers or by piezo-ceramics deposited onto flexible substrates [89]. Another possibility is to embed piezoelectric ceramic micro/nanoparticles into a polymeric soft matrix. Lead zirconate titanate (PbZr 1−x Ti x O 3 , PZT) and its derivatives are the most highly-performing piezoelectric materials; however, their lead content makes them not-environmentally-friendly and even toxic. Thus, many efforts have been made over the years to find lead-free alternatives, such as zinc oxide (ZnO) [81,90], barium titanate (BaTiO 3 ) [91], lithium niobite (LiNbO 3 ) [92,93], potassium sodium niobite (KNN) [94], and aluminium nitride (AlN) [95,96]. AlN [14,15,[97][98][99][100], for instance, can be deposited by room-temperature reactive sputtering in the form of transparent µm-thick films onto flexible substrates [43,95,[101][102][103], and it exhibits great properties such as biocompatibility [82,89], high-temperature and humidity resistance [104], and mechanical and chemical stability [105,106].
Li et al. [112] demonstrated the first PNG able to scavenge energy from respiratory and cardiac movements (Figure 3a): the authors performed the implanting of the devices in a live rat, showing the potential of applying nanogenerators for scavenging low-frequency dynamic muscle energy created by small-scale physical motion. The PNG was based on a ZnO nanowire with a diameter of 100-800 nm and length of 100-500 mm, with the two ends tightly fixed to the surface of a flexible polyimide substrate. The in vivo performances provided voltage and current outputs of around 3 mV and 30 pA: the presence of two kinds of peaks in the detected signal was explained by the nature of the heartbeat, which is not simply a stretching-release cycle.
Lu et al. [113] used an ultra-flexible PNG to scavenge energy from a pig's heart, placing the device from the left ventricular apex to the right ventricle ( Figure 3b). Their PNG were based on PZT films on soft substrates and was fabricated via transfer printing technology. In vivo experimental studies were performed in swine, when the animal's chest was being opened and closed and also when the swine resumed consciousness after anesthesia, revealing a peak-to-peak voltage of 3 V.
Kim et al. [114] demonstrated the implementation of a high-performance singlecrystalline flexible PNG for energy harvesting from heartbeats in a large animal model (pig) ( Figure 3c); in particular, the device based on PMN-PZT was sutured onto the epicardium after a median sternotomy and was able to generate an output voltage of 17.8 V and current of 1.75 µA. The detected signal of the output current was well-synchronized with the porcine ECG, which makes the device suitable as a heart monitoring sensor.
Zhang et al. [115] presented a first attempt to harvest the energy from the pulsation of the ascending aorta, which has the largest amplitude of deformation in all blood vessels. They used a PVDF-based PNG, wrapping it around the ascending aorta through in vivo study, without direct contact with the heart or with the blood, thus with no risks of blocking blood flow or thrombosis and stroke. The energy harvester generated an output voltage and current of 1.5 V and 300 nA, respectively, under a heart rate of 120 bpm and blood pressure of 160/105 mmHg, with a resulting instantaneous generated power of 30 nW for a long-lasting duration of 700 ms.
Li et al. [116] reported an integration strategy of a double-PMN-PZT-based piezoelectric nanogenerator to directly power a modern full-function cardiac pacemaker, also for myocardial stimulation and cardioversion therapy. The device implanted on a porcine heart generated a maximal output voltage of 20 V and current of 8 µA in series mode, and 12 V and 15 µA, respectively, in parallel mode.
Natta et al. [117] described an example of a PNG used in sensor mode and applied around a vessel. They proposed an ultrathin and flexible smart patch based on an AlN PNG, integrated on the extraluminal surface of a vascular graft: the system was tested in vitro using a pulsatile recirculating flow system mimicking the blood flow, and it was able to detect real-time variations in the hemodynamics parameters, exhibiting a sensitivity of 0.012 V Pa −1 m −2 .
Dong et al. [118] presented a unique design based on existing pacemaker leads tailored for compact energy harvesting through PNG. The device was made of flexible porous poly(vinylidene fluoride-trifluoroethylene) (PVDF-TrFE) placed within a dual-cantilever structure wrapping around the pacemaker lead. The maximum electrical voltage and current generated by the device were, respectively, 0.5 V and 43 nA at a frequency of 1 Hz.
Azimi et al. [119] proposed a biocompatible PNG based on the synergistic effects of ZnO, reduced graphene oxide (rGO), and PVDF nanofibers in enhancing the piezoelectric response, for harvesting energy from the left ventricle and powering, in vivo, a pacemaker ( Figure 3d). The composite with 0.1% wt of ZnO:rGO (90:10) exhibited a 10-fold increase in the output voltage compared to the pristine PVDF nanofiber mat, with a resulting power density of 138 µW/cm 3 . The PNG was implanted in a dog model, generating a maximum energy of 0.487 µJ from the heartbeat.
Xu et al. [120] reported a Kirigami-inspired energy harvester which was seamlessly integrated into a pacemaker lead and based on a piezoelectric composite film. This composite was made of a first pure PVDF-TrFE layer, and a second one with embedded ZnO nanoparticles and multi-wall carbon nanotubes (MWCNTs). A Kirigami pattern was handcrafted on the composite by blade shaping, conferring a great shape-adaptive property, and the final device was encapsulated in two PDMS layers. In vivo implantation on a porcine heart showed a generated voltage up to 0.7 V.

Triboelectric Nanogenerators (TNGs)
In a triboelectric nanogenerator, the periodic relative contact between two dissimilar materials induces a redistribution of charges at the surface of the materials due to the electrostatic induction: these charges can then be collected with an external circuit, resulting in an alternate current flow [121][122][123][124][125][126]. Each material has a different capability of attaining electrical charges, which depends on its surface triboelectric polarizability. A list of the main triboelectric materials ordered according to their electron affinity and work function can be found in the triboelectric series [121,127,128]. Different methods exist to enhance contact electrification and favour the charge transfer: in particular, the material surface can be nanostructured or patterned in order to increase the surface roughness or induce surface porosities and increase the friction area [129][130][131][132][133][134][135][136][137][138]. Traditional TNGs rely on the vertical contact-separation mode, where two triboelectric layers are brought into contact vertically and separated cyclically. However, other operation modes can be exploited, such as the sliding mode or the free-standing mode [14].
Zheng et al. [156] demonstrated in vivo biomechanical energy harvesting from respiratory movements using a TNG for the first time and implanted the device in a living rat under the left thoracic skin (Figure 4a). The device was based on 100 µm-thick PDMS films with patterned pyramid arrays, a Kapton substrate, a 400 µm-thick PET spacer, and an Al foil with nano-surface modification to serve as both the contact layer and electrode. The energy scavenged from breathing was directly used to power a prototype pacemaker, with an open circuit voltage of 3.75 V and a short circuit current of 0.14 µA.
Ma et al. [157] implanted a multifunctional TNG in the space between the pericardium and epicardium of a pig animal model to provide real-time and continuous monitoring functions, such as the rates of heartbeats and respiration ( Figure 4b). The device was based on a 50 µm-thick nanostructured PTFE film as one triboelectric layer, fixed on a Kapton substrate. Au was deposited on the back side of the Kapton film as one of the electrodes, whereas a 100 µm-thick Al foil was used both as a second triboelectric layer and electrode. A core-shell encapsulating multilayer strategy based on PDMS and parylene C was adopted to guarantee the in vivo durability, hermeticity, and bidirectional biocompatibility of the device. The TNG provided an open circuit voltage of 10 V and a short circuit current of 4 µA.
Ouyang et al. [158] reported on a fully-implanted symbiotic pacemaker based on energy-harvesting TNG which was able to successfully correct sinus arrhythmia and prevent deterioration (Figure 4c). The system was also based on a core-shell structure with two triboelectric layers: a support and the shell, made of two encapsulation layers. One tribo-layer was nanostructured PTFE, a 3D elastic EVA sponge was the spacer, and a memory alloy ribbon was the kneel; a Teflon film and PDMS were used to package the full device. The device generated 0.495 µJ for each cardiac motion cycle, which is enough to power the pacemaker since it is higher than the endocardial pacing threshold energy (0.377 µJ). In order to meet requirements for a minimally invasive implantation procedure and for better comfort in the long term, the implantable TNG would need a smaller size, higher energy density, efficient fixation with the biological tissues, and more effective power management units [159].
Ryu et al. [160] presented an inertia-driven in vivo TNG with the size of a commercial coin battery, made of amine-functionalised poly(vinil alcohol) (PVA) and perfluoroalkoxy (PFA) as triboelectric layers. Through a Bluetooth low-energy (BLE) informationtransmitting system, the implantable TNG could be used effectively to recharge a cardiac pacemaker. Five-stacked TNGs generated 136 V peak and 2 µApeak/cm 3 , with a resulting maximum power density of 4.9 µWRMS/cm 3 . The device charged energy storage components (capacitors) by scavenging energy from small movements while an adult mongrel was asleep.
Jiang et al. [161] reported various fully-bioabsorbable natural-materials-based TNGs made of cellulose, silk fibroin, chitin, rice paper, and egg white, and they used these devices to harvest energy and provide power for regulating the beating rates of dysfunctional cardiomyocytes in rats. The bioabsorbable capability makes these devices advantageous because, after biodegration, no additional surgery interventions are required and no side effects will occur after the first implantation. device generated 0.495 µJ for each cardiac motion cycle, which is enough to power the pacemaker since it is higher than the endocardial pacing threshold energy (0.377 µJ). In order to meet requirements for a minimally invasive implantation procedure and for better comfort in the long term, the implantable TNG would need a smaller size, higher energy density, efficient fixation with the biological tissues, and more effective power management units [159].  . Examples of TNGs for heart energy harvesting. (a) Implantable TNG based on patterned PDMS (i) for powering a pacemaker (ii) to regulate the heartbeat in a rat (iii). Adapted with permission from reference [156], 2014 Copyright Wiley. (b) Implantable TNG based on nanostructured PTFE triboelectric layer (i) for energy harvesting from the heartbeat in different implanting sites: lateral wall (LLW), right lateral wall (RLW), and posterior wall (PW) of the heart (ii). Adapted with permission from reference [157], 2016 Copyright American Chemical Society. (c) Symbiotic cardiac pacemaker system based on implantable TNG made of nanostructured PTFE and 3D elastic sponge structure (i); (ii) illustration of the pacemaker turned on by wireless passive trigger in vivo in a pig model, with the ECG of successful pacing. Adapted from reference [158]. Creative Commons Attribution 4.0 International License.
Besides being placed on the surface of the heart or close by, TNGs have also been employed transcutaneously in order to receive energy from ultrasounds and provide power to implantable devices, i.e., pacemakers. For instance, Hinchet et al. [162] investigated the use of a high-frequency vibrating and implantable TNG for harvesting ultrasounds in vivo underneath the skin. The TNG was based on a large membrane of PFA suspended on a thin Cu electrode and integrated with a rectifier, transformer, voltage regulator, and battery. The working principle was that ultrasound can induce micrometer-scale displacements of the polymer membrane and generate energy through contact electrification: the generated outputs were 2.4 V and 156 µA under porcine tissue, high enough for small medical implants.

Hybrid Piezo/Triboelectric Nanogenerators (HNGs)
The hybridization of piezoelectricity and triboelectricity has emerged recently as a promising strategy to increase the overall energy-generation performances of the single components. As demonstrated in [43], coupling a PNG and a single-electrode TNG results in a non-algebraic summative effect in the final hybrid signal generated by the full device. Additionally, the hybridization allows for a better adaptability of the overall device to different operating conditions. Mariello [42] provided a comprehensive review of the hybrid nanogenerators (HNGs) employed in several fields of applications. Examples of HNGs for implantable applications and, in particular, for cardiac bioelectronics or heart energy harvesting can be found in the literature [163,164], even though the deployment of these devices is still not fully explored because of the slightly more complex architectures compared to the single PNG or TNG components. Shi et al. [165], for example, presented a self-powered system made of a HNG based on BaTiO3 nanoparticle-embedded PDMS, and used it to power an electronic thermometer for detecting the subcutaneous temperature of a live rat.

Solar Cells (SCs)
Solar is the most abundant form of natural energy and, in recent years, remarkable efforts have been made for the research on solar/photovoltaic cells [166]. Differently from PNGs, TNGs, or their derivatives, which convert biomechanical energy and are activated by mechanical deformations or motions, solar-powered implantable devices utilize a subcutaneous solar module to convert transcutaneous light into electrical energy which can then be provided to cardiac pacemakers [74]. A typical solar cell consists of nonlinear semiconductor devices that generate electricity while exposed to light; in particular, it presents an anode and a cathode in a molecular dye. Solar energy harvesting for IBDs is still challenging due to problems related to the placement of the cell on the human body and its connection with the IBD, as well as their efficiency and durability; however, solar energy is an attractive source for wearable/implantable devices. Subdermal solar cells, in fact, can power bioelectronic implants as described in previous works related to in vitro tests under porcine skin [167,168], in pigs [44,169], and in mice [170,171]. Bereuter et al. [172] reported the first real-life validation data of energy harvesting by subcutaneous solar cells worn by volunteers in different seasons. They demonstrated that a 3.6-cm 2 subdermal solar cell would generate enough energy (19 µW cm −2 ) to power a cardiac pacemaker.
Tholl et al. [173] conducted Monte Carlo simulations of light distribution in human skin to estimate the power output of subdermal solar cells, which is affected by several factors, i.e., implantation depth, optical skin properties, geographical latitude, and solar panel size (Figure 5a). They concluded that, for the darkest skin type, a 2 cm 2 solar cell implanted subdermally at a depth of 3 mm would generate enough power under exposure to 11 min of midday for a modern standard cardiac pacemaker.
Haeberlin et al. [169] presented a device with a 3.24 cm 2 solar module for subcutaneous implantation into a pig at a depth of 2.8-3.84 mm. The output power was higher than 3500 µW/cm 2 , enough to successfully perform a battery-free VVI pacing. The same authors [44] proposed another 4.6-cm 2 solar module implanted in a pig, which was capable of empowering a pacemaker, even indoors. Song et al. [171] developed a subdermally-implanted flexible ultrathin photovoltaic device which delivered 647 µW to DC power a pacemaker in hairless mouse models (Figure 5b). The device was based on dual-junction solar microcells (0-76 mm × 0.76 mm × 5.7 µm) of GaInP/GaAs, transfer-printed on a flexible polyimide film, and interconnected through sputterd Ti/Au metals; finally, the device was encapsulated with multiple layers of biocompatible, transparent polymers (SU-8, NOA, PDMS, etc.).
Prominski et al. [174] reported on a fast method to obtain pure-silicon porositybased heterojunctions to transduce light pulses into electrical stimuli for optically-induced biomodulation and for overdriving heart pacing. The device was, in fact, interfaced conformally (thanks to capillary forces and without adhesives) with the left ventricle of an isolated cardiac tissue in a Langendorff apparatus. Under light pulses, the heart immediately synchronized to the light pulsing frequency, owing to the injection of photoelectrochemical currents and the depolarization of the myocardium. An amount of 4 mW mm −2 of a 532 nm laser light was enough to stimulate the heart. The proposed device could be used to scavenge energy from light and power a pacemaker or for direct heart stimulation.
Among the materials used for solar/photovoltaic cells, amorphous silicon is flexible, cost-effective, and highly sensitive to natural light, allowing for the fabrication of flexible photovoltaic devices [175], even though it has a relative low efficiency [176]. Dieffenderfer et al. [177] presented a solar-powered, wireless, wrist-worm device to produce photoplethysmogram signals from the radial artery: the system includes two monocrystalline solar cells to charge a 20 mAh lithium polymer battery and a supercapacitor. The solar modules were placed under pig skin flaps and generated output powers of 1963 µW/mm 2 , 206 µW/mm 2 , and 4 µW/mm 2 , respectively, for full sunlight outdoors, shade outdoors, and indoors.

Biofuel Cells (BFCs)
Implantable glucose biofuel cells are a subset of conventional biofuel cells (BFCs) that use hydrogen or alcohols as energy sources, and rely on glucose in the body to convert the biochemical energy of its bonds into electrical energy useable for supplying power to IBDs, e.g., cardiac pacemakers. These devices are generally based on two catalytic electrodes: the anode oxidizes glucose while the cathode reduces oxygen via electrocatalytic processes. Glucose BFCs, miniaturized and implanted inside the heart, can scavenge the energy of the glucose present in the blood stream to generate energy continuously, given the abundancy of glucose in the body. Glucose is, in fact, one of the most important sources of energy in living organisms; it is provided by the splitting of carbohydrates and it is oxidized to CO 2 and H 2 O through aerobic metabolic processes, releasing energy (up to 16 kWs/g). Cosnier et al. [178] gave a comprehensive review of implantable glucose biofuel cells, classifying them into abiotic and enzyme-catalyst BFCs. Abiotic glucose BFCs are based on noble metals, alloys, or activated carbon as a catalyst for oxygen reduction and glucose oxidation. Drake et al. [179] presented a tissue-implantable glucose BFC using a dialysis membrane for partial isolation of the cathode by phase separation of O 2 and CO 2 gases. The device was implanted in the flank of a dog and generated an open circuit voltage of 0.5 V and a stable power density of 2.2 µW cm −2 for 30 days of continuous operation, whereas short-term testing showed possible power densities in the range of 6.4 µW cm −2 . Simons et al. [180] presented a ceramic glucose fuel cell based on a crack-free, self-supported freestanding sub-400-nm-thick membrane consisting of a Pt anode, the proton-conducting electrolyte ceria (CeO 2 ), and a Pt cathode. The authors demonstrated the fabrication of 150 individual glucose BFCs on a Si chip, exhibiting a power density of 43 µW cm −2 : these devices represent the smallest potentially implantable energy harvesters to date.
Holade et al. [46] demonstrated an abiotic BFC composed of a catalytic electrode modified with inorganic nanoparticles of various compositions (Au x Pt y ) deposited on carbon black (CB) (Figure 6a). The electrode materials selected for the study were buckypaper composed of carbon nanotubes and carbon paper with carbon fibers: Au/CB on buckypaper was selected for catalyzing glucose oxidation and Au 60 Pt 40 /CB on carbon paper was chosen to catalyze oxygen reduction. The output open-circuit voltage, short-circuit current density, and power produced were, respectively, 0.35 V, 0.65 mAcm −2 , and 104 µW, in human serum at 5.4 mM glucose. An energy harvesting power-management circuit was used to amplify the generated voltage and to power a pacemaker.
However, metal-based catalysts generally exhibit low electrocatalytic activity at neutral pH; thus, abiotic glucose BFCs cannot produce enough power for the supply of electronic devices. This is the reason why research has moved towards biocatalysts and enzymecatalyst-based glucose BFCs [181,182], since enzymes are highly efficient at performing electrocatalysis when in contact with the corresponding electrode: the main enzymes for bioanodes are glucose oxidase (GOx) and glucose dehydrogenase (GDH), whereas those for biocathodes are copper oxidases (e.g., laccase or bilirubin oxidase (BOD)). The reactions in an enzyme-catalysts glucose BFC occur through two possible pathways: (i) mediated electron transfer (MET), where active molecules or polymers are used with their redox potential close to that of the enzyme [183,184], or direct electron transfer (DET), where electrons are directly exchanged by the catalytic site of the enzyme [181]. In order to achieve DET, different methods can be adopted, such as lowering the overpotentials of glucose oxidation and oxygen reduction or developing nanostructured 3D electrodes to increase the number of wired enzymes per surface or volume.
Cinquin et al. [185] presented the first functional glucose BFC implanted in the retroperitoneal space of freely moving Wistar rats (Figure 6b). The BFC was characterized by a mechanical confinement of various enzymes and redox mediators without bonding them covalently to the electrodes surface. In particular, a composite graphite pellet was used which contained glucose oxidase and ubiquinone (anode), and polyphenol oxidase and quinone (cathode). The generated power was 24.4 µWmL −1 , which was higher than the requirements for pacemakers. Carbon nanotubes (CNTs) have also been used to favour the DET due to their high specific surface and hydrophobicity, which allow for large enzyme densities, as well as their high porosity and electrical conductivity [186,187]. Halamkova et al. [187], for instance, used a PQQ-dependent glucose dehydrogenase and laccase immobilized on a CNTs-based buckypaper, and adopted a cross-linking agent to connect the enzymes to the CNT covalently. The device provided an open circuit voltage of 530 mV and a power density of 30 µWcm −2 . Zebda et al. [188] designed a BFC based on unconventional electrodes made of compressed, multi-walled CNTs and implanted it in a living rat, obtaining a voltage of 0.95 V and power density of 1.3 mWcm −2 ( Figure 6c). As an alternative approach for BFCs implanted in blood vessels, Castorena-Gonzalez et al. [189] prepared buckypaper electrodes with multi-walled CNTs coated with the enzyme and directly deposited the bioelectrodes of the BFC on the cremaster tissue of a rat during the surgical operation (Figure 6d).
Sales et al. [190] demonstrated an intravenous implantable glucose hybrid enzyme-Pt BFC in which the bioanode was made of a flexible carbon fiber microelectrode modified with neutral red redox mediator and glucose oxidase, whereas the biocathode was composed of another carbon fiber electrode modified with Pt nanoparticles stabilized on PAMAM-G4 dendrimer. Under physiological conditions, the implanted BFC captured glucose from rat blood and dissolved molecular oxygen acted as an oxidizing agent: in the jugular vein of the rat, the BFC supplied a power density of 95 µW cm −2 . Southcott et al. [47] assembled a flow BFC filled with serum solution, mimicking the human blood stream, based on biocatalytic electrodes composed of buckypaper modified with PQQ-dependent glucose dehydrogenase on the anode and with laccase on the cathode. The outputs of the BFC were 470 mV and 5 mA, which were enough to sustainably power a pacemaker through a charge pump and a DC/DC converter. MacVittie et al. [191] presented an implanted BFC operating in vivo in lobsters and in a model closely mimicking human physiology. A commercial pacemaker with an estimated power consumption of 90 mW and a required voltage of 2.8 V was cut open, and the internal battery was removed and replaced with a circuit of five biofuel cells connected in electrical series. The activation of the pacemaker was successfully demonstrated.
Therefore, implantable BFCs can provide the required energy for IBDs in many living organism; however, some challenges need to be addressed in the future, among which the selectivity of metal catalysts for abiotic BFCs, the biological stability of the employed enzymes (up to date it is approximately one month in vitro), the toxicity of some mediators used in BFCs, and mass transfer issues related to the available amount of "fuel" [192]. Table 1 summarizes the advantages and disadvantages of the aforementioned microdevices for heart energy harvesting. Figure 6. BFCs for heart energy harvesting. (a) Flow abiotic BFC coupled with an energy harvesting circuit (i) to power a pacemaker and make it apply electrical pulses (ii). Adapted with permission from reference [46], 2014 Copyright Wiley. (b) Glucose BFC based on graphite modified with ubiquinone, glucose oxidase, and catalase (anode), and with quinhydrone and polyphenol oxidase (cathode) (i), implanted in rats for energy harvesting (ii). Adapted from reference [185]. Creative Commons Attribution License. (c) BFC based on compressed carbon nanotube-enzyme electrodes (i). At the anode (ii), glucose is oxidized to gluconolactone, where the electrons are transferred from the GOX to CNT. Catalase decomposes hydrogen peroxide into oxygen and water. At the cathode, electrons are transferred Figure 6. BFCs for heart energy harvesting. (a) Flow abiotic BFC coupled with an energy harvesting circuit (i) to power a pacemaker and make it apply electrical pulses (ii). Adapted with permission from reference [46], 2014 Copyright Wiley. (b) Glucose BFC based on graphite modified with ubiquinone, glucose oxidase, and catalase (anode), and with quinhydrone and polyphenol oxidase (cathode) (i), implanted in rats for energy harvesting (ii). Adapted from reference [185]. Creative Commons Attribution License. (c) BFC based on compressed carbon nanotube-enzyme electrodes (i). At the anode (ii), glucose is oxidized to gluconolactone, where the electrons are transferred from the GOX to CNT. Catalase decomposes hydrogen peroxide into oxygen and water. At the cathode, electrons are transferred from CNT to laccase where dioxygen is reduced to water. Adapted from reference [188]. Creative Commons Attribution-NonCommercial-ShareAlike 3.0. (d) BFC based on buckypaper and MWCNTs (i), with the biocatalytic electrodes in contact with the cremaster tissue of a rat (ii). Reprinted with permission from reference [189], 2013 Copyright, Wiley.

Bioelectronic Systems with Cardiac Interfaces
In the previous section, several principles for heart energy harvesting were presented, and the devices described all consisted of a generator connected to the cardiac pacemaker. These devices represent only a building block for more complex bioelectronic devices used for the interrogation and intervention of cardiac systems (i.e., sensing or stimulation). Rigid cardiac interfaces are constrained by structural and mechanical mismatches with the host soft tissues; thus, more recently, great interest has been raised in soft (flexible or stretchable) bioelectronics which can be symbiotically applied to treatment and rehabilitation. In this section, different classes of bioelectronic systems for cardiac interfaces are presented: in vivo epicardial systems, in vivo endocardial systems and minimally invasive cardiac catheters, in vitro systems for cardiac patch, and in vitro systems for cardiac organoids. A more comprehensive review of these bioelectronic devices is provided by Tang et al. [193], who classified them in terms of flexible or stretchable bioelectronics. They explain that, in order to accommodate the real mechanical motions of the heart and guarantee a longterm functionality, the modern bioelectronic interfaces must be designed with stretchable structures, where the sensing and computing components are patterned in rigid islands and the interconnections are generally made wavy, wrinkled, or in different strategic shapes (serpentines, star-shapes, etc.).

In Vivo Epicardial Systems
Flexible bioelectronics commonly rely on thin-film active inorganic layers deposited/ transferred onto or encapsulated into flexible polymeric substrates. Multilayer structures and design choices are generally adopted to minimize the stresses applied onto these active electronic components [194,195]. Epicardial cardiac bioelectronic interfaces which exploit the same concepts are based on soft polymers with functional thin films, attached conformally onto the surface of the heart and protected by ultra-high barrier encapsulations [196]. Viventi et al. [197] presented a high-performance, conformal bioelectronic system which was intimately integrated with dynamic, living biological tissues and used to measure cardiac electrophysiological activity in a porcine animal model with high temporal and spatial resolution (~500 µm), and with high durability (>10,000 cardiac contraction cycles) (Figure 7a). The device included an 18 × 16 array of amplified and multiplexed electrodes spaced 800 µm apart, and it was based on doped single crystal silicon nanoribbons deposited on a Si wafer and transfer-printed onto a thin polyimide sheet. Then, functional layers and multilayer encapsulations completed the structure, resulting in 2016 n-type metal-oxide-semiconductor field-effect transistors. Another example of an ultrathin cardiac bioelectronic interface being applied on the epicardium was given by Fang et al. [198]. They used a perfused rabbit heart and showed that an ultrathin, leakage-free, and biocompatible dielectric layer (SiO 2 ) can hermetically seal an underlying array of 396 multiplexed capacitive sensors based on silicon nanomembrane transistors. The electrophysiological measurements showed that it is possible to collect a signal even without direct metal contact with the tissue, thus extending the overall lifetime.
Choi et al. [199] reported a revolutionary, fully implantable and bioresorbable, leadless, battery-free, and externally programmable cardiac pacemaker based on a poly(lactide-goglycolide) (PLGA) thin-film dielectric interlayer, W/Mg biodegradable electrodes, and a Si nanomembrane diode, used for the treatment of atrioventricular (AV) nodal heart block. The authors showed that the devices could be tuned with tailored designs, be powered wirelessly via resonant inductive coupling, and provide effective pacing of hearts with stable signals throughout the desired period of use in several animal models (mice, rats, rabbits, canine, and human models). This type of soft bioresorbable bioelectronics paves the way for the next generation of cardiac technologies that will not require percutaneous hardware, thus reducing the risk of device-associated infections and dislodgement.
The same authors [200] presented a more complex, transient, and closed-loop system that embodied several time-synchronized components (Figure 7b), i.e., a temporary, bioresorbable, and stretchable epicardial pacemaker; a bioresorbable steroid-eluting interface to reduce inflammation; a subcutaneous bioresorbable energy harvesting module; an array of soft, skin-interfaced sensors to detect ECGs, heart rate, respiratory parameters, and cerebral hemodynamics; a wireless RF module for power transfer; a soft, skin-interfaced haptic actuator; a handheld device for real-time visualization of data. This system is an example of integration of different devices with cardiac interfaces and paves the way for closed-loop temporary electrotherapy through wirelessly connected, body-integrated bioelectronics.
Kim et al. [201] proposed a class of stretchable sensors and actuators based on epicardial webs that can conformally wrap around the heart surface and allow the interconnections to move jointly with the cardiac tissue, resulting in a stable ECG recording and pacing under normal rhythms, as well as for abnormal tachycardia rhythms. The epicardial web was fabricated through microelectronic processes and a silk sacrificial layer was used to transfer the pattern onto the surface of the heart: this layer dissolved after exposure to moisture and left the web as a laminate on the heart. In vivo studies on a porcine heart demonstrated that the adopted design provides ultralow effective stiffness and high degrees of deformability, a robust and reliable adhesion to the epicardium via capillary forces and without any additional adhesives, wide coverage of the tissue, and allowance for direct visual and optical interrogation.
Xu et al. [202] demonstrated a 1-mm 2 -area 3D multifunctional integumentary thin elastic membrane integrating 68 Au electrodes and custom-formed to completely and conformally envelope the heart, in the same way as the natural pericardium, guaranteeing robust and non-invasive contacts throughout dynamic cardiac cycles. The fabrication of this membrane included the 3D printing of a solid model of the heart to serve as a substrate for mounting ultrathin electronic or optoelectronic sensor systems (such as inorganic InGaNbased LEDs, Si nanomembranes for strain gauges, Au electrodes for electrical stimulation, IrOx pads for pH sensors), separately prefabricated through planar technologies and with the top face attached to the heart; then, a thin layer of silicone was cast on the heart model and, after curing, it was removed, resulting in a thin elastomeric envelope with integrated device components. An ex vivo test on perfused rabbit hearts showed the possibility to create a high-density epicardial platform for cardiac mapping/stimulation. Xu et al. [203] exploited fractal geometry to design and fabricate compliant, large-area, low-impedance electrodes for electrical stimulation (Figure 7c), based on platinum-iridium (Pt-Ir) alloys and poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS). The authors presented these very stretchable constructs integrated with high-performance sensors for cardiac electrotherapy ex vivo on Langendorff-perfused rabbit hearts, with high spatiotemporal resolution.
Lee et al. [204] proposed an ultrathin, stretchable grid-patterned active organic electrochemical transistor (OECT) matrix directly attached to the surface of a rat's beating heart. The 4 × 4 OECT array was used to map an ECG and exhibited a signal-to-noise ratio of 52 dB, a total thickness of 2.6 µm and durability and signal stability over long periods (1 h). In terms of microfabrication and materials structure, the system was based on a 1.2-µm-thick parylene honeycomb grid substrate which conferred mechanical stability and stretchability, and was dip-coated with 100 nm-tick poly(3-methoxypropyl acrylate) (PMC3A) in order to achieve antithrombotic properties.
Park et al. [205] introduced an epicardial mesh for electromechanical cardioplasty and electrotherapy based on a blend of sliver metal nanowire network (AgNWs) and styrene-butadiene-styrene (SBS) rubber as stretchable interconnects and electrodes that resembled the elastic and electrical properties of cardiac tissue. The blend was prepared using the ligand exchange of AgNWs with polyvinylpyrrolidone (PVP) partially exchanged to hexylamine (Ham) through NOBF4, which allowed for a phase transfer of the water-dispersed AgNW to the organic phase (SBS). The epicardial mesh was integrated in postinfarct rats and was able to deliver synchronized electrical impulses to restore cardiac contractile functionality, and to terminate ventricular tachyarrhythmia serving as an epicardial defibrillator.
Choi et al. [206] reported on Ag-Au nanocomposites made of ultralong Au-coated Ag nanowires embedded in an elastomeric block-copolymer matrix which exhibited a high percolation connectivity and an optimized stretchability. The Au coating prevented the Ag surface oxidation and ion leakage. The authors demonstrated the fabrication of 2D fan-shaped implantable devices with 42 multichannel electrodes which was based on these nanocomposites and was applied on swine heart for continuous electrophysiological recording (surface ECG and intracardiac ECG).
Gutruf et al. [207] provided an example of integration of miniaturized wireless energyharvesting and digital communication electronics to fabricate pacing devices which were implanted subdermally in rats without performance losses and with optimal biocompatibility (Figure 7d). The stimulation electrodes were made of Pt-coated copper, while parylene C was used as an encapsulation layer. The stretchable devices were capable of delivering pacing stimuli in rodents electrically, and they also included microscale inorganic LEDs for optogenetic applications, i.e., to stimulate ChR2-expressing hearts.
Sim et al. [208] described an epicardial bioelectronic patch entirely made of stretchable rubbery materials with cardiac tissue-like mechanical softness, used for spatiotemporal mapping of electrophysiological activity as well as strain/temperature sensing and electrical pacing on a beating porcine heart. The materials employed were: PDMS, conductive robbery paste, AgNWs/PDMS rubbery conducting layer, PDMS mixed with poly(3-hexylthiophene-2,5-diyl) nanofibrils (P3HT-NFs) acting as a stretchable semiconductor, and an ionic gel as rubbery dielectric.
Liu et al. [209] developed a fully elastic electrode with tissue-like softness (~10 kPa Young's modulus) made of intrinsically stretchable materials to improve biocompatibility, reduce immune responses, and avoid tissue damage. In particular, they fabricated a device (called elastrode) with 64 electrodes packed in a 0.25 cm 2 area and made of PEDOT:PSS hydrogel and fluorinated elastomers: it allowed for robust and intimate contact with the cardiac tissue, and, at the same time, demonstrated optimal electrical properties during cardiac cycles on porcine hearts. The same authors, Liu et al. [210], showed, for the first time, a fully stretchable organic light-emitting electrochemical cell array driven by a solutionprocessed organic thin-film transistor active-matrix; even though it was developed for wearable applications, the processes are compatible for implantable devices and, especially, for cardiac bioelectronic interfaces.  [197], 2010 Copyright, The American Association for the Advancement of Science.
(b) Transient closed-loop system for temporary cardiac pacing based on a wireless bioresorbable pacemaker (i), for treatment of temporary bradycardia on a Langendorff-perfused human wholeheart model (ii). Reprinted with permission from reference [200], 2022 Copyright, The American Association for the Advancement of Science. (c) 3D multifunctional integumentary membrane on a Langendorff-perfused rabbit heart (i), with indication of the maximum strain during deformation (ii). Reprinted with permission from reference [203], 2015 Copyright Wiley. (d) Wireless, battery-free multimodal and multisite electrical and optical pacemaker for providing successful pacing in rats and mice. Adapted from reference [207], 2019 Creative Commons CC-BY License.

In Vivo Endocardial Systems
Endocardial bioelectronic interfaces consist of devices integrated into basket and balloon catheters, and they are generally delivered non-invasively through blood vessels [211][212][213]. Traditional catheters are rigid, unsuitable for being coupled with soft tissues without any damage, require precise and guided placement for effective diagnosis and therapy, and have limited space for the integration of sensing or stimulation elements. Standard balloons, in particular, are attractive because they enable minimally invasive insertion into lumens and other organs, and their size and shape can be tailored to have controlled inflation and specific non-destructive interaction with the tissue. However, they are limited in their therapeutic potential because they rely only on mechanical forces applied to the walls of blood vessels, and they lack electronically active materials. Soft bioelectronics can replace or be integrated into these systems and implanted for high-resolution endocardial recording/mapping and stimulation [214].
Klinker et al. [215] developed a novel smart balloon catheter system containing stretchable electrodes and thermal-based blood flow sensors for monitoring hemodynamic parameters, electrical stimulation, and ablation therapy (Figure 8a).
Kim et al. [201,216] demonstrated an array of integrated stretchable radio-frequency electrodes and multimodal (temperature, flow, and tactile) sensors on a balloon catheter for controlled local cardiac ablation therapy. The tactile sensors were used to detect circumferential contact between the balloon surface and the endocardium, avoiding the use of X-ray and radioactive contrast agents during the ablation procedure. Thin-film Au serpentine tracks were designed to achieve optimal stretchability (up to 200% tensile strains) and were used for recordings of the anterior right ventricle and the left atrium surfaces.
Han et al. [217] reported on endocardial balloon catheter-integrated soft multilayer electronic arrays for multiplexed sensing and actuation during minimally-invasive forms of cardiac surgery. They used plastic heart models and Langendorff animal and human hearts to validate the devices and demonstrate its high-density spatiotemporal mapping of temperature, pressure and electrophysiological parameters, programmable electrical stimulation, and other multifunctional diagnostic and therapeutic functions.
Liu et al. [218] presented a flexible, self-powered endocardial pressure sensor for real-time electrophysiology monitoring and the detection of ventricular fibrillation or ventricular premature contraction (Figure 8b).

In Vitro Systems for Cardiac Patch
Cardiac patch is a functional engineered tissue grown by the primary culture of cardiomyocytes in 3D tissue scaffolds, and it represents a realistic in vitro cardiac model which mimics the structure and function of the heart [193,[219][220][221][222][223]. The utility of a soft bioelectronics-integrated cardiac patch is to overcome the issues related to the conventional scaffold materials and conventional cardiac electronics, for which their mechanical and structural disparities, in fact, prevent the long-term 3D interrogation of cardiac tissue with high spatiotemporal resolution. Liu et al. [224] proposed a new strategy to prepare ordered 3D interconnected microporous nanoelectronics networks from ordered 2D nanowire nanoelectronics precursors in order to achieve the seamless and minimally invasive integration of 3D bioelectronics within host tissues. The microporous network was mixed with organic gels and polymers to form hybrid materials containing nanowire devices (nanoelectronics scaffold) with ~14-nm resolution, without damaging or altering the tissue.

In Vitro Systems for Cardiac Patch
Cardiac patch is a functional engineered tissue grown by the primary culture of cardiomyocytes in 3D tissue scaffolds, and it represents a realistic in vitro cardiac model which mimics the structure and function of the heart [193,[219][220][221][222][223]. The utility of a soft bioelectronics-integrated cardiac patch is to overcome the issues related to the conventional scaffold materials and conventional cardiac electronics, for which their mechanical and structural disparities, in fact, prevent the long-term 3D interrogation of cardiac tissue with high spatiotemporal resolution. Liu et al. [224] proposed a new strategy to prepare ordered 3D interconnected microporous nanoelectronics networks from ordered 2D nanowire nanoelectronics precursors in order to achieve the seamless and minimally invasive integration of 3D bioelectronics within host tissues. The microporous network was mixed with organic gels and polymers to form hybrid materials containing nanowire devices (nanoelectronics scaffold) with~14-nm resolution, without damaging or altering the tissue.
Tian et al. [225] developed a microporous, flexible and free-standing Si nanowire nanoelectronics scaffold (nanoES) to mimic the structure of natural tissue scaffolds. The device was based on field-effect transistors on SU8 polymer backbones, and then the nanoES was integrated with natural or synthetic biomaterials, i.e., electrospun nanofibers or alginate hydrogels. A cardiac patch was grown on the nanoES from primary rat cardiomyocytes and, inside this hybrid scaffold, the nanoES could monitor the extracellular electrical activity (local field potentials). Dai et al. [226] presented a tissue-scaffold-mimicking 3D nanoelectronics array with 64 devices to enable real-time sub-millisecond recording of extracellular action potential in developing rat cardiac tissues and in a transient arrhythmia disease model.
Wang et al. [227] reported another example of a 3D porous, flexible electronic scaffold formed through the compressive buckling of 2D precursor structures (Figure 9a), with microelectrodes integrated within thermoresponsive extracellular matrix-based hydrogels acting as structural support for engineered cardiac tissues, with the final purpose of sensing, stimulation, drug release, and regulation of tissue function. Feiner et al. [228] integrated a cardiac patch, with flexible, free-standing electronics (32 spatially distributed Au electrodes) and a 3D nanocomposite scaffold, in order to record cellular electrical activities, stimulate cardiac cells to obtain synchronized contraction, regulate cardiac functions, and provide programmable drug release through the use of electroactive polymers deposited onto the metal electrodes. The same authors [229] introduced elastic, biodegradable, and electronic scaffolds composed of electrospun albumin fibers as substrate and as encapsulation for Au electrodes (Figure 9b). The scaffolds were integrated with cardiac patches grown by cardiomyocytes and, after one-three weeks of implantation into adult rats, they biodegraded, leading to the dissociation of the composing inorganic materials.
Lind et al. [230] proposed a method based on multimaterial 3D printing for fabricating novel instrumented cardiac microphysiological devices (heart-on-chip electronics): they prepared six functional inks, made of biocompatible, piezo-resistive soft materials, such as thermoplastic polyurethane, PDMS, and carbon black nanoparticles, and they embedded, by 3D printing, mechanical sensors within the self-assembled laminar cardiac tissues to provide non-invasive, electronic readouts of tissue contractile stresses in vitro.

In Vitro Systems for Cardiac Organoids
Cardiac organoids are biological constructs derived from the self-assembly of embryonic or pluripotent stem cells and represent in vitro cardiac models that mimic the structure and functions of the heart, and which are useful for studies in regenerative medicine, drug delivery, and cell therapy [231][232][233][234][235][236]. The development and maturation of cardiac organoids can be characterized and monitored through different multimodal methods, such as fluorescence imaging, single-cell RNA sequencing, and bioelectronic organs-on-chip, to assess the environment pH, O 2 , temperature, protein biomarker, or morphology [237]. Soft bioelectronics are preferable for integration with cardiac (as well as brain) organoids because they can adapt to their rapid variation in volume, and they need to detect 3D electrophysiology with cellular-scale spatial resolution and millisecond temporal resolution in order to capture the change in the cell's structure and connectivity [238].
tions, and provide programmable drug release through the use of electroactive polymers deposited onto the metal electrodes. The same authors [229] introduced elastic, biodegradable, and electronic scaffolds composed of electrospun albumin fibers as substrate and as encapsulation for Au electrodes (Figure 9b). The scaffolds were integrated with cardiac patches grown by cardiomyocytes and, after one-three weeks of implantation into adult rats, they bio-degraded, leading to the dissociation of the composing inorganic materials.  Kalmykov et al. [239] introduced a flexible organ-on-electronic-chip device (Figure 10a): it consisted of 3D self-rolled biosensor arrays composed of field-effect transistors or passive graphene-based microelectrodes applied onto human cardiac spheroids in order to enable electrophysiological investigation of the dynamic 3D cellular assemblies and provide continuous and stable multiplexed (12-channel) recordings of field potentials.
Li et al. [240] reported on the creation of cyborg organoids, i.e., the 3D assembly of stretchable sub-µm-thick mesh nanoelectronics (SMN) across the organoids through organogenesis and cell-cell attractive forces (Figure 10b). They fabricated the SMN by using a releasable layer, and demonstrated that the SMN can grow and transform from an initial 2D cell-seeded layer into a 3D organoid-like structure, maintaining an intimate and seamless contact with the cardiac tissue and providing a stable, high-resolution electrical recording.
Kim et al. [241] proposed an unconventional method for multimodal mechanophysiological characterization of 3D human-induced pluripotent stem-cell-derived cardiac organoids using a single device composed of an active-matrix array of pressure-sensitive transistors and 3D liquid metal-based electrodes (Figure 10c). This intra-organoid electronic interface was used for electrophysiological recording and stimulation. Figure 10. Bioelectronics integrated with cardiac organoids as in vitro cardiac models. (a) 3D self-rolled biosensor array (organ-on-e-chip) (i) for electrical interrogation of human embryonic stem cell-derived cardiomyocytes spheroids (ii). Adapted with permission from reference [239], 2019 Copyright Creative Commons Attribution-Noncomercial License. (b) Cyborg organoids based on stretchable mesh nanoelectronics (i), consisting of SU8, Pt, PEDOT, Au, integrated into organoids through organogenesis, in particular into human cardiac organoids (ii). Stage I: transferring and laminating the mesh-like plane of nanoelectronics onto a 3D sheet of stem cells. Stage II: cell aggregation, proliferation and migration shrink the cell sheet into a cell-dense plate and compress the nanoelectronics into a closely packed architecture. Stage III: Contraction and curling of the interwoven cell-nanoelectronics structure. Stage IV: 3D spherical morphology. **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. Reprinted with permission from reference [240], 2019 Copyright, American Chemical Society. (c) Multimodal sensing device based on directly printed 3D electrodes and pressure-sensitive transistor arrays (i), for simultaneous electrical stimulation of cardiac organoids with multimodal sensing (ii). Reprinted with permission from reference [241], 2022 Copyright, American Chemical Society, Creative Commons CC-BY-NC-ND License.

Bioprinting of Cardiac Tissue
Bioprinting of cardiac tissue is an emerging field that attempts to address a crucial requirement of studies on the functions of the heart. Cardiac tissue is in fact dynamic and made of complex cellular and extracellular structures; thus, there is an increasing interest in the creation of biomimetic and dynamic scaffolds for tissue organization and regeneration for applications in drug screening, transplants, disease models, electrophysiological investigations, etc. 3D bioprinting is generally based on a layer-by-layer additive approach and is advantageous over conventional scaffold fabrication technologies thanks to better control of compositions and structural complexities. 3D bioprinting supported by imaging technologies (3D scanners, computer tomography, magnetic resonance imaging, and computer-aided design) can be useful for the creation of in vitro constructs that mimic the complex architectures of cardiac tissues and for a better understanding of cardiac abnormalities. Jafari et al. [242] summarized the main 3D bioprinting techniques and the biomaterials available for the realization of cardiac constructs. Kato et al. [243] reviewed recent approaches of 3D cardiac tissue construction based on biomaterials, e.g., alginate, gelatin, collagen, and decellularized extracellular matrix, coupled with bioprinting methods (i.e., extrusion, inkjet, laser-assisted, etc.).
Jia et al. [244] explored the use of an extrusion-based bioprinting technique to deposit perfusable vascular structures with highly ordered arrangements (Figure 11a). They developed a cell-responsive bioink consisting of gelatin methacryloyl (GelMA), sodium alginate, and 4-arm poly(-ethylene glycol)-tetra-acrylate (PEGTA), and this was used in combination with a multilayered coaxial extrusion system. The stable bioprinted constructs were created through the ionic-and photo-crosslinking of GelMA and PEGTA, and they demonstrated the capability of supporting the spreading and proliferation of encapsulated endothelial and stem cells.
Sanjuan-Alberte et al. [245] developed a 3D printed bioink composed of MWCNTs and decellularized extracellular matrix. Electrical stimulation of 3D printed structures with this bioink, containing human pluripotent stem cell-derived cardiomyocytes, was performed: in the absence of stimulation, the conductive properties of MWCNTs improved the contractile behavior of the cardiac cells, and this was enhanced in the case of stimulation.
Current scaffolds are mainly static and, although they can be integrated with nanobioelectronic meshes or stretchable ultrasensitive devices, they are unable to adapt to changes in the in vitro/in vivo cardiac environment [229,[246][247][248][249]. Bioprinting of 4D smart biomaterials for dynamic scaffolds represents an affordable strategy in this context. These adaptive materials can change their form, structure, shape, or properties in response to biological, magnetic, or electrical stimuli, as well as light, temperature, ultrasound, pH, UV light, or osmotic pressure. Miao et al. [250] provided an extensive summary of the state-of-the-art technological advances in the field of transformation-preprogrammed 4D printing and 4D printing of shape memory polymers for tissue and organ regeneration. In particular, bisphenol A diglycidyl ether, poly(propylene glycol) bis(2-aminopropyl) ether, and decylamine can be used to create dynamic myocardial tissue-like scaffolds.
Wang et al. [251] developed a digital light processing-based printing technique to fabricate 4D near-infrared (NIR) light-sensitive cardiac constructs which can change their curvature and shape on-demand to mimic the curved topology of the myocardial tissue (Figure 11b). The NIR light-sensitive 4D material was based on a shape memory polymer and graphene and was tested with human-induced pluripotent stem cell-derived cardiomyocytes, and mesenchymal stem cells and endothelial cells.
Ma et al. [252] presented 4D printed shape memory composites based on a polylactic acid (PLA) matrix and semi-crystalline linear polymer polycaprolactone (PCL): the material was temperature-sensitive and was used to create a controllable drug release system. Pedron et al. [253] used a diacrylated triblock copolymer composed of poly(ethylene glycol) and poly(lactic acid) (PLA-PEG-PLA), and poly(N-isopropyl acrylamide) (PIPAAm), for the creation of biocompatible constructs for cardiac microtissue transplantation.
The combination of bioprinted scaffolds of 4D printed materials with bioelectronics and biosensors is also a promising method to create dynamic constructs for the study of the cardiovascular system. The bioelectronic devices can integrate additional functions to favour or accelerate tissue regeneration and biosensing of the complex cell signalling networks which contribute to the dynamic cardiac environment and evolution [254][255][256].
Nanoenergy Adv. 2022, 2, FOR PEER REVIEW 29 Wang et al. [251] developed a digital light processing-based printing technique to fabricate 4D near-infrared (NIR) light-sensitive cardiac constructs which can change their curvature and shape on-demand to mimic the curved topology of the myocardial tissue (Figure 11b). The NIR light-sensitive 4D material was based on a shape memory polymer and graphene and was tested with human-induced pluripotent stem cell-derived cardiomyocytes, and mesenchymal stem cells and endothelial cells.
Ma et al. [252] presented 4D printed shape memory composites based on a polylactic acid (PLA) matrix and semi-crystalline linear polymer polycaprolactone (PCL): the material was temperature-sensitive and was used to create a controllable drug release system. Pedron et al. [253] used a diacrylated triblock copolymer composed of poly(ethylene glycol) and poly(lactic acid) (PLA-PEG-PLA), and poly(N-isopropyl acrylamide) (PIPAAm), for the creation of biocompatible constructs for cardiac microtissue transplantation.
The combination of bioprinted scaffolds of 4D printed materials with bioelectronics and biosensors is also a promising method to create dynamic constructs for the study of the cardiovascular system. The bioelectronic devices can integrate additional functions to favour or accelerate tissue regeneration and biosensing of the complex cell signalling networks which contribute to the dynamic cardiac environment and evolution [254][255][256].

Conclusions and Future Challenges
This paper reviews different types of bioelectronic interfaces used for monitoring, pacing, or studying the heart and its in vitro models. The most recent technologies for heart energy harvesting are described in comparison with the standard ones, and the main types of micro-devices used for this purpose are presented. The necessity of novel cardiac energy generators is motivated by several factors, especially by the need for reliability and the powering of implantable biomedical electronic devices, in order to avoid the reliance on chemical batteries, characterized by restricted lifetime and requiring frequent replacements with risky surgical interventions. Battery-less pacemakers represent the first important advancement in this field, and they are based on micro-devices for harvesting the energy conveyed by the cardiac contraction/expansion cycle; in particular, the cardiac systole is a source of periodic pressure and endless clean energy. Implantable/wearable nanogenerators and heart energy harvesters are, thus, described with a focus on materials and state-of-the-art examples from previous works. Against other electromagnetic generators, piezoelectric and triboelectric nanogenerators, solar cells, and biofuel cells represent the most promising technologies for the realization of biocompatible, self-powered biomedical devices. In particular, PNGs and TNGs are the most widely used alternatives to chemical batteries or wireless inductive charging, because they offer attractive properties, e.g., light weight, simple architectures, biocompatibility, optimal performances, compatibility with most microfabrication processes, a wide variety of employable materials, easy processability, suitability to be conformally applied onto the heart or in endocardial systems, and sensitivity to quasi-static mechanical deformations. Thus, they could easily be applicatied as heart energy harvesters. However, they also present some disadvantages: in particular, their scarce resistance to humidity makes their deployment for implantable applications quite challenging, revealing the need for hermetic encapsulations for protecting them from the surrounding tissues. The research of new biocompatible high-performance materials also represents another future challenge for PNGs and TNGs as HEHs.
Moreover, more general soft bioelectronic cardiac interfaces are presented in the second part of the review, classified as epicardial and endocardial bioelectronics, and devices for applications with in vitro cardiac models (cardiac patches and cardiac organoids). Thanks to their flexibility, stretchability, and reliability, these systems can accommodate the mechanical deformations and movements of the heart, of blood vessels, and of the growing cells in cardiac models, and they can provide a continuous, robust, reliable, and high-density electrophysiological mapping with high spatiotemporal resolution. The most recent advances in the 3D and 4D bioprinting of cardiac tissues are, finally, highlighted with emphasis on the possible combination and hybrid integration of bioprinted scaffolds and dynamic biomaterials with morphing bioelectronics. The main applications of bioprinting of cardiac tissues reside in regenerative medicine and tissue engineering; however, the seamless integration of biomimetic bioelectronics allows for multiple additional functions, e.g., biosensing, electrical stimulation, and drug delivery.
Heart energy harvesters (HEHs) and heart bioelectronic systems (HBSs) are promising technologies for the next-generation implantable biomedical devices, and future research work is needed to assess their efficiency, adaptivity to the environment, biocompatibility, flexibility and conformality, and complexity of their electronic interfaces and circuits. Future developments in materials science and nanotechnology will improve the realization of more biocompatible and more efficient cardiac bioelectronic devices.
The main challenges related to these technologies are summarized and illustrated in Figure 12. Concerning HEHs and cardiac nanogenerators, the most crucial points to take into account for their future development and commercialization regard the output generation performances, their reliability, their electronic interfaces and power management circuitry, their biocompatibility, and their connection into more complex closed-loop systems. Increasing their performances is possible through the development of novel energy transducing materials, through nanostructuring techniques that enhance the charge generation [261,262], or even by acting on the design and architectures of the devices [262,263]. The use of hermetic packages and encapsulations is a key aspect to guarantee long-term operation and stable reliability of cardiac implantable devices [264]. The electronic circuitry serves to extract the signal from the HEH and collect the charges to generate power; therefore, it requires minimized losses and maximized signal-to-noise ratios and efficiencies. Additionally, the advances of soft bioelectronics and microfabrication techniques allow for the integration of the circuitry and power managements modules directly in the space of the real device, minimizing the hindrance and making the interconnections flexible or stretchable according to the needs. Connecting HEHs into more complex HBSs requires the processing of the signal of multiple devices working simultaneously, i.e., the energy harvester should be synchronized with the IBD under consideration, and, in the case of multiple IBDs, multiplexing instrumentation is needed on a miniaturized scale. The advent of artificial intelligence and deep learning approaches will allow for the creation of intelligent systems of sensors/transducers/devices [265].
Regarding, more generally, the cardiac bioelectronic interfaces, the main challenges for the future can be ascribed to the following points: (i) increasing the number of electrodes per unit area to achieve high-density electrical recording/stimulation; (ii) enhancing multifunctionality and multimodality in order to detect simultaneously different cardiac physiological activities (electrophysiology, temperature, metabolism, mechanical forces, etc.); (iii) developing wireless, biodegradable devices with online data processing algorithms; (iv) miniaturizing technologies up to the cellular level to perform in situ molecular phenotype characterization for a better cardiac diagnosis and therapy; (v) enabling large-scale production for a better cost-effectiveness, in order to enable commercialization and clinical applications; (vi) conducting more studies on animal-derived cells and animal models, clinical trials and biocompatibility analyses to assess the long-term safety, the immune reaction, and the devices' stability over time, and establishing validation protocols. Figure 12. Illustrations of challenges and issues to be solved for the future employment of heart energy harvesters (HEHs) and heart bioelectronic systems (HBSs). OGs: oscillator generators; PNGs: piezoelectric nanogenerators; TNGs: triboelectric nanogenerators; SCs: solar cells; BFCs: biofuel cells; Epi: epicardial bioelectronics; Endo: endocardial bioelectronics; CP: cardiac patch; CO: cardiac organoids.
Funding: This research received no external funding.

References
Regarding, more generally, the cardiac bioelectronic interfaces, the main challenges for the future can be ascribed to the following points: (i) increasing the number of electrodes per unit area to achieve high-density electrical recording/stimulation; (ii) enhancing multifunctionality and multimodality in order to detect simultaneously different cardiac physiological activities (electrophysiology, temperature, metabolism, mechanical forces, etc.); (iii) developing wireless, biodegradable devices with online data processing algorithms; (iv) miniaturizing technologies up to the cellular level to perform in situ molecular phenotype characterization for a better cardiac diagnosis and therapy; (v) enabling largescale production for a better cost-effectiveness, in order to enable commercialization and clinical applications; (vi) conducting more studies on animal-derived cells and animal models, clinical trials and biocompatibility analyses to assess the long-term safety, the immune reaction, and the devices' stability over time, and establishing validation protocols.