Recent advances in materials and applications for bioelectronic and biorobotic systems

Abstract

The ultimate goal of the advancements in bioelectronics and robotics is the creation of seamless interfaces between artificial devices and biological structures. Current efforts in this area have been focused on designing biocompatible, mechanically compliant, and minimally invasive electronic and robotic systems for a range of applications, such as motor control and sweat sensing. The purposeful design of bioelectronic and robotic systems using the principles of biomimicry enables the creation of biocompatible and life‐like machines and electronics. The success of such approaches relies on the new development and applications of soft materials, as well as methods of actuation and sensing that are inspired, either by composition, function, or properties, of the naturally occurring organisms. A combination of rigid structural components, soft actuators, and flexible sensors can enable the integration of such devices with biological organisms and eventually human users. In this review, we highlight the recent advances in biomimetic soft robotics and bioelectronics. We describe the soft robotic fabrication toolbox and modern solution in bioelectronics that, in our opinion, will enable the fusion of these fields by creating robotic bioelectronic systems. Future development in this area will require substantial integration of adaptable and responsive components at the biointerfaces.

INTRODUCTION

Research in biomimetics aims to synthetically replicate the morphology and functionality of biological systems, from the nanoscale to the macroscale. The operational efficiency of biological organisms can be attributed to the millions of years of evolution that all biological systems underwent. However, through purposeful biomimicry, similar levels of structural performance and tissue-mimetic functionality can be achieved by design. Mimicry of biological tissue and organisms has resulted in the development of synthetic materials with improved biocompatibility, increased mechanical compliance, and greater energy efficiency. Furthermore, the development of biomimetic devices has resulted in a greater understanding of complex biophysical phenomena that are difficult to study in situ and allows sources of biological variability to be separated during the analysis.

Biomimetic systems often contain soft, nonrigid components, such as joints or volumetric substrates. Soft materials, such as polymers, possess higher mechanical compliance than rigid materials, such as steel. Emerging soft robotic and bioelectronic systems often utilize materials with self-healing properties and stimuli-responsive capability, increasing material functionality and lifespan. These systems can be systematically created using additive manufacturing, which allows precise integration of multiple materials and numerous sensors and actuation components into a single system.

Soft robotic and bioelectronic systems can combine compliant materials with soft actuation mechanisms to regulate atypical locomotion, such as jumping and crawling. Because of their greater compliance, soft materials can establish a seamless human–machine interface when interacting with biological tissue. Compliant materials also allow biomimetic systems to operate in dynamic, uncontrolled environments where traditional rigid systems may falter, such as rough terrain. Soft materials can also be engineered to exhibit specific tissue-like properties, such as strain-stiffening behavior. Additionally, biocompatible soft materials minimize immunogenic response and reduce tissue damage in vivo through matching mechanical properties and have the potential to expand the functions and lifetime of implantable prosthetic devices, e.g., neuromuscular sensors and stimulators.

This review focuses on the fundamentals of materials synthesis, fabrication, sensing, and actuation of biorobotic and bioelectronic systems, with a focus on novel devices (Figure 1). We conclude with an outlook on the next several years of biomimetics research, including future technologies and pertinent challenges.

FIGURE 1. The advances in the fabrication of robotic and bioelectronic components and their successful integration can enable biomedical devices that interact with patients seamlessly and enhance their living conditions

BIOROBOTICS

In this review, we define biorobotic devices as systems that interface with biological components, are conceived with bioinspired design, or are constructed in part with biological materials (biohybrid robots). Biorobotics are most commonly used for drug delivery, mimicry of biological phenomena, and operation in harsh environments. The following section will address common fabrication, sensing, and actuation methods for soft robotic systems. Recent breakthroughs and integration with bioelectronic systems are also discussed.

Materials

Biorobotic devices must possess sufficient amounts of compliance to achieve seamless biological integration and mimetic capability. Many robotic devices used to mimic biological functions, such as underwater adhesion and muscle-mimetics, utilize similar polymer substrates to those used in bioelectronic applications.^1,2 This section will focus on the most commonly used polymer substrates, including material characteristics and limitations (Figure 2A).

FIGURE 2. A soft robotic design and fabrication toolbox. (A) Common materials for the fabrication of soft robotics. Polyurethane foams are general-purpose structural materials, while silicone elastomers can be used for flexible actuators. Hydrogels are promising for their uses in a wet environment and contact with biological structures. (B) Overview of additive manufacturing methods. Depending on the type of manufacturing method, different materials and resolutions will be achievable in the fabrication. (C) Examples of actuation mechanisms of soft robots. Pressure-based and combustion-based systems rely on inflation of the soft robotic parts. Electrotactic and thermotactic materials can intrinsically shrink upon receiving appropriate stimuli. (D) Approaches to soft sensing. Different types of sensors allow us to determine the configuration of a soft robot or sense its interactions with the environment

Polymers, substances composed of repeating macromolecule units, are favored in fabricating modern robotic systems for their elasticity, resilience, and conformability. Both the micro- and macrostructures of a given polymer chain dictate its resulting degree of viscoelasticity, a characteristic of materials that exhibits both viscous and elastic mechanical responses.³ Soft electronic devices must have similar Young's moduli to the biological tissue to achieve smooth interaction and seamless integration. Desirable substrates must exhibit Young's modulus between 1 kPa and 100 MPa. Matching mechanical stiffness, Young's moduli, and other relevant material properties (e.g., ultimate elongation, toughness, resilience modulus) will result in optimized tissue integration and prevent tissue damage.⁴ Polyurethanes, polydimethylsiloxane, and hydrogels represent three typical polymer substrates.

Polyurethanes, formed from urethane–urea linkages, are thermosetting polymers that are often solid at room temperature due to strong hydrogen bonding between chains.⁴ The main advantage of polyurethanes is their low cost because of their wide use in the polymer industry. While polyurethanes have been utilized for soft actuation and to develop foam substrates,⁵ the products of their decomposition and combustion are highly toxic,⁶ which prevents wider usage of the material. Another drawback of polyurethanes is their limited photostability, which is especially visible in the blends containing aromatic isocyanates.⁷ However, polyurethanes are still a promising material for fillings and structural support due to their high tensile strength and affordability.

Polydimethylsiloxane, or PDMS, is formed from siloxane linkages. Due to high elasticity, biocompatibility, gas permeability, inertness, and other characteristics, PDMS is used in countless biorobotic and bioelectronic applications.^4,8,9 While traditional PDMS is purely elastic, modified PDMS can exhibit life-like behaviors. For example, modification of PDMS side chains with cross-linking metal complexes has enabled self-healing properties,¹⁰ and co-polymerization with isophorone bisurea and bonding 4,4′-methylenebis(phenyl urea) has enabled viscoplastic properties, because of the balancing weak and strong chain interactions.¹¹ PDMS is a promising material for soft medical robotics. However, other softer silicone formulations, such as Ecoflex, can be considered due to their even higher compliance with soft biological tissues.¹²

Hydrogels are a large class of hydrophilic, cross-linked polymers with high water content. Hydrogels possess high biocompatibility and swell when placed in aqueous solutions, such as bodily fluids.⁴ They can act as structural materials and active elements, providing conductivity, sensing capabilities, and mechanical actuation.¹³ Due to their similarity to living tissues in terms of water content and mechanical compliance, hydrogels are preferable in biomedical and neuromodulation-related applications.^11,14 However, more research is required for these materials to enable stability, functionality, and ease of fabrication similar to classic polymeric materials.

Fabrication of biorobotic systems

Soft robots are generally assembled from stock engineering materials, and their components are molded to a specific shape. However, recent advances in additive manufacturing, a technology that fabricates objects by adding material layer-by-layer, have enabled rapid prototyping of molds and direct fabrication of soft robotic components (Figure 2B). Unlike traditional subtractive manufacturing processes, additive manufacturing allows for the construction of structures with shapes challenging to achieve through other means and significantly reduces material waste. Additive manufacturing is computer-controlled and allows for the fabrication of multimaterial structures with submillimeter precision.¹⁵ Fused deposition modeling (FDM), direct ink writing (DIW), direct inkjet printing, selective laser sintering (SLS), and stereolithography are the most commonly used additive manufacturing technologies. Significant effort is put into enabling three-dimensional (3D) printing with new materials, such as hydrogels or conductive polymers,¹⁶ and achieving ultra-high throughput fabrication.¹⁷

FDM is by far the most common 3D printing method. The process relies on the extrusion of a thermoplastic polymer filament through a heated nozzle, generally around 200 °C.¹⁸ Thermoplastics are the only biocompatible material that is printable via FDM because no other biocompatible material can withstand the severe temperature fluctuations required for extrusion. FDM resolution depends on the size of the printing nozzle and must be several times greater than its diameter to prevent structural imperfections, such as air pockets.⁴ Another common additive manufacturing method is DIW, which involves extruding a pressurized liquid polymeric precursor through a nozzle. After extrusion, a continuous or intermittent stimuli causes the material to solidify.¹⁸ As such, DIW can be used to fabricate hydrogels when photoexposure or heating is used to facilitate cross-linking. DIW allows incorporation of anisotropy into the resulting hydrogel structures and the production of shape-changing microstructures. Similar to FDM, the printing resolution is restricted by the nozzle size. DIW's disadvantage over FDM is that the low viscosity of DIW inks does not allow printing of hollow or overhanging structures, and sacrificial materials are required as support during printing.

SLS constructs objects by selectively melting grains of powder, which are then fused together. Substrate grains are sintered layer-by-layer until the structure is complete.¹⁸ In order to achieve the desired product, the thermoplastic granules must be uniformly sized. Unlike other additive manufacturing techniques, the resolution of SLS printing is determined by the size of the granules, not the print head/laser diameter.⁴ A related technique, also utilizing laser beams, is stereolithography (SLA), a family of polymerization technologies, which involves the polymerization of objects on the surface of liquid pre-polymer. SLA utilizes photopolymerization, which allows for the formation of thin features and geometries in the resin. While SLA is time-efficient and reduces material waste, few SLA-compatible materials are appropriate for soft robotics applications, and this method is the most useful for the fabrication of hard elements or molds. The surface roughness of SLA prints is a lot lower than for FDM and DIW, and such structures can, in general, be used directly without the need for post-print polishing. The last available method is direct inkjet printing, where ink droplets are jetted onto a substrate and experience vitrification, evaporation, or polymerization. Layers are solidified on top of one another until the resulting material has formed. Inkjet printing enables higher resolution than other 3D printing methods, and by printing with multiple nozzles simultaneously, multimaterial objects can be produced.

Actuation and sensing

In essence, advanced mechanical capabilities and mobility are what differentiate a biorobotic system from a static bioelectronic system. Such a biorobotic system requires efficient actuation processes and advanced sensing capabilities. This section will review relevant actuation and sensing techniques, with an emphasis on fundamental components and pressing challenges.

Actuation

Effective soft actuation is critical to the creation of functioning of biorobotic systems. Devices ranging from drug delivery microrobots to untethered octopi robots utilize soft actuation mechanisms to achieve efficient mechanical performance while retaining compliance with soft systems.¹⁹ This section will review the fundamentals of commonly used soft actuation mechanisms (Figure 2C).

The most commonly used pneumatic actuators utilize pressurized air to inflate and deflate embedded air cavities, resulting in linear or rotational deformation.^20,21 Due to their high responsiveness, high work density, and large strain capability, pneumatic actuators are the favorite approach to achieve biorobotic locomotion and for fabrication of gripping devices (Figure 3A).²² Generally, pneumatically actuated systems must be tethered to an external pump or another pressure source to facilitate actuation. Built-in pumps require high power input, and lightweight batteries with a high power density that could enable long-term independent operation are not yet available. These drawbacks greatly limit the applicability of pneumatic actuators in untethered systems.²³

FIGURE 3. Selected examples of soft robotic actuators. (A) Pneumatic soft actuator networks are a popular solution in robotics research. Reproduced with permission from Mosadegh et al.22 Copyright 2013, Wiley. (B) Three-dimensional (3D) printed combustion-actuated soft robot. On the right, an image montage of a robot making a jump. Reproduced with permission from Bartlett et al.24 Copyright 2015, AAAS. (C) Input shape memory polymer–metal composite. The frames show switching between four distinct configurations of the actuator. Reproduced with permission from Shen et al.28 Copyright 2016, Springer Nature. (D) Muscle mimetic, hydraulically amplified dielectric elastomer actuators enable movement with a high strain rate. Reproduced with permission from Kellaris et al.2 Copyright 2017, AAAS

An alternative to a pressure source is applying combustion, an exothermic reaction between a fuel and oxygen, resulting in a release of large amounts of heat and pressure. Combustion has been used as an actuation mechanism in jumping robots (Figure 3B),²⁴ and to generate pressure for pneumatic actuation in an untethered robot.²⁵ The aforementioned soft robot utilized a logic board to distribute the pneumatic pressure from the combustion to different pneumatic chambers, representing a combination of biorobotic and bioelectronic design. Further development must be undertaken to lengthen the operational period and locomotion capabilities of combustion-based systems.²³ However, the direct transformation from fuel to movement is a more energy-efficient approach than the application of electric pumps.

A relevant new approach is the creation of biohybrid actuators, which utilize biological tissue to optoelectronically/electrically actuate soft robotic systems.²⁶ Due to their high biocompatibility and responsiveness, biohybrid actuators can be used in biomedical applications. Devices can be designed to exhibit tissue-specific mechanical behavior, such as the periodic contraction, which mimics the behavior of cardiomyocytes. Biohybrid substrates may also be incorporated into bioelectronic systems,²⁷ such as neural implants, to minimize immunogenicity and achieve seamless integration between the electronic device and the organism. However, current biohybrid systems possess lower levels of controllability and mechanical performance than soft actuation methods.

Another approach to creating robotic movement is electrothermal actuation, which can be achieved with heat-responsive smart materials, such as shape memory polymers (SMPs) and shape memory alloys (SMAs). Once heated, these materials return from a deformed state to their original state.²⁵ Use of resistive heating allows for the operation of SMAs and SMPs in untethered robots (Figure 3C).²⁸ Unfortunately, thermal energy storage in smart materials generally prevents rapid deactivation, which increases the time required to return to the original state resulting in slow movement.²³ SMAs and SMPs may be used broadly in MEMS (microelectromechanical systems) and NEMS (nanoelectromechanical systems) to fabricate devices, such as microactuators and micropumps.^4,29

Electrical actuation technologies, such as ionic polymer–metal composites (IPMCs) and dielectric elastomers (DEAs), utilize electrical stimulation to deform soft materials and are promising for the next-generation biorobotic systems. IPMCs consist of an ionic polymer encased by two electrodes. When a voltage is applied, water migrates toward the cathode, which causes the IPMC to bend. Even though IPMCs exert low actuation force, they are suitable for applications requiring underwater movement.²³ By contrast, DEAs harness coulombic attraction to deform elastomers layered between two compliant electrodes.²⁰ Due to their high power density and responsiveness, DEAs are favored in muscle-mimetic (Figure 3D)² and crawling applications.³⁰

Sensing

Sensing capabilities are essential in functional soft robotic devices. In order to perform complex actuation, soft robotics must utilize soft sensors to establish position and deformation state.⁴ Sensing components must possess similar compliance to the elastomeric substrate to avoid compromising system performance. We will review the essential characteristics of commonly used soft sensors, as well as notable challenges (Figure 2D).

Resistive and capacitive sensors are commonly used to measure strain in soft robots. Hydrogels loaded with electrolytic solvent can serve as ionic resistive sensors. However, hydrogel-based resistive sensors suffer from readout instabilities. Colloidal inks, which benefit from high electrical conductivity, may be used as resistive sensors (Figure 4A),³¹ but require precise control over their morphology. Incorporating conductive precursors into similar inks results in nanoparticle networks that can operate across larger strain differentials. Alternatively, soft capacitors can be created by adding a dielectric layer between elastomeric electrodes (Figure 4B).³² Unlike soft resistive sensors, soft capacitive sensors can detect minute changes in strain.² Soft capacitors are inexpensively fabricated, but as with resistive sensors, care must be taken to avoid manufacturing faults. Many of these electronic sensing materials and mechanisms have been adopted in bioelectronics device designs. Magnetic sensors, although not yet demonstrated in soft robotic systems, could allow for remote measurement of deformation state, similarly how existing nonsoft applications utilize Hall effect sensors to measure the positioning of embedded magnets. Position-sensing capabilities may be achieved in soft robotics likely with magnetic nanoparticles instead of embedded magnets.

FIGURE 4. Selected examples of electrical sensors that may be used broadly for robotics. (A) Inkjet-printed soft resistive sensors show high sensitivity and small hysteresis. Reproduced with permission from Lo et al.31 Copyright 2019, Wiley. (B) Silicone-fabric-based capacitive strain sensors enable the detection of hand gestures. Reproduced with permission from Atalay et al.32 Copyright 2017, Wiley. (C) Compliant pulse oximetry sensor delivers high-fidelity physiology recordings. Reproduced with permission from Lochner et al.35 Copyright 2014, Springer Nature

Due to the prevalence of pressurized chambers in the design of soft robotic actuators, air pressure sensors, which are created by plugging air-filled chambers with volumetric sensors, can detect simple movements of robots and contacts with the environment.³³ Compliant and movement-specific 3D-printed accordion structures can be used in conjunction with an air pressure sensor to quantify deformation. Movements such as pulling, pressing, and bending are easily detected with this method. However, such sensors have low specificity, and multidirectional or concurrent movements (e.g., simultaneous compression and twisting) are difficult to distinguish.³¹ It is possible that such motion could be deconvoluted from multiple sensors using machine learning approaches.

Lastly, deformable optoelectronic sensors based on stretchable waveguides have been developed. Signal attenuation (penetration distance) is based on the refractive properties of the core and cladding materials and waveguide dimensions. In order to achieve effective data collection, the refractive indices of core material must be greater than the refractive indices of the cladding of the optical fiber. Distributed optic fiber networks made of parallel waveguides allow for precise readout of the position and configuration of soft electronic systems.³⁴ Also, optoelectronic circuits based on soft organic diodes can be utilized to measure biomedical data (Figure 4C).³⁵ Development of increasingly compliant distributed optic sensors makes them an ideal candidate for sensing mechanical strain and temperature in biorobotic systems in real time. Due to their high accuracy across large strain differentials, these sensors may also be incorporated into larger bioelectronic devices, such as gloves and socks.

Recent robotic research

Biorobotic devices bring the promise in many industrially valuable applications, including fabrication of biomedical robotics. Additionally, replication of biological phenomena in artificial systems is an emerging direction of research. Advances in biomedical robotic devices, which are commonly used for drug/cargo delivery, have resulted in greater biocompatibility and energy efficiency in such solutions. One example of biomimetic actuation is a work by Yang et al. The authors developed an agglutinate magnetic spray coating that allows inanimate objects, such as origami and kirigami structures, to be magnetically actuated (Figure 5A).³⁶ In response to an external magnetic field, these spray-coated millirobots successfully mimicked walking and crawling motions and were utilized for drug delivery and catheter navigation. As mentioned in Section 2.3.2, purposefully distributed magnetic nanoparticles may also serve as strain sensors in future magnetically actuated robots.

FIGURE 5. A few novel robotic devices and interfaces. (A) Magnetic spray coating for actuation of biomedical millirobots. Reproduced with permission from Yang et al.36 Copyright 2020, AAAS. (B) Janus platelet chemophoretic micromotors for targeted drug delivery. Reproduced with permission from Tang et al.37 Copyright 2020, AAAS. (C) Biomimetic remora disc for underwater adhesion. Reproduced with permission from Wang et al.1 Scale bars: 10 mm. Copyright 2017, AAAS

Moreover, a number of Janus particle-based drug/cargo delivery robots that are both magnetically and chemically actuated have been reported in recent literature. Recently, Alapan et al. developed a Janus microroller-based cargo delivery platform with functionalized anticancer drug and antibody surfaces for targeted therapeutics.¹⁹ The Janus microrollers successfully navigated through physiologically relevant blood flow via magnetic actuation, at approximately 600 μm/s. Tang et al. achieved similar functionalities by utilizing enzyme-powered chemophoretic motion, which allows for self-propelled drug delivery (Figure 5B).³⁷ Unequal deposition of urease on the surface of the Janus platelets resulted in uneven decomposition of urea, resulting in an enzyme-powered drug-delivery mechanism. Notably, chemophoretic propulsion allows for drug delivery without the usage of external magnetic fields. In the future, drug/cargo delivery robots are expected to translate into clinical applications, such as chemotherapy, targeted immunotherapeutics, and catheter navigation.

Biorobotic systems have recently been used to replicate biological phenomena, such as flight and underwater adhesion. The biomimicry process allows for the development of new technologies and facilitates the study of biological processes. To better understand the kinematics of the avian wing, Chang et al. developed a biohybrid morphing wing with avian feathers. The biohybrid wings were used to construct a functional flying robot, which mimicked the structure of the common pigeon.³⁸ The 'PigeonBot' utilized a servo motor-based propulsion system, alongside GPS and radio transceiver electronics. By designing a biorobotic system with multiple noncontrolled degrees of freedom, researchers discovered that asynchronous wing joint movement might result in a greater turning efficiency. Similarly, Wang et al. synthetically replicated the adhesive disc from the remora suckerfish in order to optimize its morphology for underwater adhesion (Figure 5C).¹ The adhesive disc, which utilized pneumatically actuated lamellae to increase adhesion performance, exhibited high pull-off forces on both smooth and textured surfaces. Undoubtedly, future mimicry of biological organisms will continue to inform the development of novel locomotion devices, jumping robotics, gripping robotics, adhesion mechanisms, and more.

BIOELECTRONICS

Bioelectronics allow the exchange of information between digital and biological systems. Bioelectronic probes can read out the physiology of organisms, and stimulation devices can actively modulate their response creating two-way communication between machines and the biological system. This section discusses recent progress in bioelectronic probes and their outlook for integration with robotics systems.

Materials

Conductive yet soft materials are essential to the functioning of bioelectronic systems. Therefore, the development and optimization of novel conductive soft materials will greatly advance the field of bioelectronics as a whole. Despite the biocompatibility and compliance requirements, there are many materials whose properties overlap between biorobotic and bioelectronic substrates. Therefore, this section will focus only on recent developments and the characterization of soft conductive materials.

Traditional silicon and inorganic electronic devices are far more rigid than biological tissue, resulting in poor interfaces between hard electronics and soft tissue. For reference, silicon semiconductors have Young's modulus of approximately 130 GPa, whereas biological tissues have Young's modulus from between 1 kPa and 20 GPa.³⁹ In the past, bioelectronic devices often achieved desired levels of compliance at the expense of conductivity as highly conductive materials, such as metals, have a much higher Young's modulus than compliant materials. However, traditional soft bioelectronic substrates, such as PDMS, are mostly insulating.²³ For this reason, developing conductive yet compliant materials has been a focus of recent bioelectronics research.

Thin films, with thickness under 5 μm, can achieve compliance with soft systems due to extremely low bending stiffness.^14,40 Thin films are often constructed from metals used in traditional circuitry such as gold and deposited on compliant substrates.¹⁴ Thin films are most commonly incorporated into biomedical devices for applications such as skin-mimetic and electrophysiological sensing/modulation.^14,39 Using conformal additive stamp printing, researchers have successfully deposited thin films onto curved substrates, namely, for the development of smart contact lenses.⁹ Thin-film robots, which incorporate somatosensory capabilities and light-responsive actuation, have recently been used to mimic human gripping and animal locomotion.⁴¹

A widespread method of achieving compliant, conductive bioelectronic materials is through the use of nanomaterial-embedded elastomer composites (nanocomposites). Nanofilms and nanoparticles are often embedded into soft substrates to develop conformable pressure/strain sensing electronics, which can measure local and global resistivity.⁴² Recently, Wang et al. employed an aramid-based nanocomposite to develop a conformal structural battery.⁴³ The aramid ion conductors exhibit high strength, connectivity, and percolating behavior, resulting in a battery capacity 72 times greater than a Li-ion battery with identical volume. Such compliant materials are promising energy sources for biorobotics and bioelectronic devices alike.

Conductive gels that contain ionic liquids are yet another viable soft conductor.²³ Unlike thin film and nanocomposite materials, these gels are purely aqueous. Recently, Lee et al. developed a bioelectronic spiderweb that utilizes an ionic organogel to regulate electrostatic sensing, self-cleaning, and capturing of prey.⁴⁴ By harnessing the properties of electrostatic adhesion, vibration, and induction, researchers enhanced the ionic spiderweb's capabilities. Similar soft conductors may be used as multipurpose substrates for biorobotic applications, which combine self-regulation and actuation functionalities using one material. Conductive polymers can also be tuned to match the mechanical properties of biological tissues⁴⁵ and are being successfully applied to fabricate fully compliant elastomeric electrode arrays for tissue recording with cellular resolution.⁴⁶ Additionally, it is important to improve the conductivity between the bioelectronic device and the tissue. Graphene-based conductive adhesives have been recently developed to improve such contact strength and electrical conductivity at the biointerface.⁴⁷

Fabrication of bioelectronic systems

Similar to biorobotic systems, the manufacture of bioelectronic systems can be classified into additive and subtractive processes, termed top-down and bottom-up fabrications. This section will focus on less explored and potentially useful fabrication methods for current or future bioelectronics.

Top-down fabrication

Top-down fabrication involves industry-standard lithography techniques and many steps of deposition–patterning–removal to achieve the desirable structures. Critical to the process are etching methods, all of which remove material either chemically or physically.⁴⁸ A notable form of dry etching, ion milling uses a focused ion beam to machine nanomaterials and was used in bioelectronic applications, such as antireflection nanocoatings (Figure 6A).^49,50 Ion milling, however, suffers from slow milling speeds, among other drawbacks, which reduce its utility.⁴⁵ Howver, when vertical and angled dry etching techniques are combined, a 3D structure can be produced, as opposed to a 2D structure (Figure 6B).^48,51 As an alternative to dry etching, anisotropic wet etching can also be used to generate a wide range of 3D nanostructures (Figure 6C).⁵² These advanced top-down fabrication methods can be applied to improve the performance of bioelectronic devices for recording and stimulation applications.

FIGURE 6. Less-explored fabrication methods for bioelectronics. (A) 'Sketch and peel'-based focused ion beam milling of plasmonic nanoparticles. Reproduced with permission from Chen et al.50 Copyright 2016, American Chemical Society. (B) Fabrication of freestanding nanostructures via angled ion beam etching. Reproduced with permission from Atikian et al.51 Copyright 2017, American Institute of Physics. (C) Anisotropic wet etching for fabrication of a variety of silicon nanostructures. Reproduced with permission from Lin et al.52 Copyright 2013, Royal Society of Chemistry. (D) Synthesis of polyaniline nanostructures using self-assembled catalytic DNA nanotemplates. Reproduced with permission from Wang et al.54 Copyright 2013, American Chemical Society

Bottom-up fabrication

While top-down fabrication methods allow for precise control over material etching and deposition, the processes are complex and yield poor results with biomaterials, which are highly chemically and heat sensitive.⁴⁸ Unlike top-down processes, bottom-up processes utilize self-assembly, vapor deposition and other material-based approaches to achieve device assembly.⁵³ DNA origami, for example, exploits properties of DNA base pairing to form 3D nanostructures and is a common method of self-assembly. Similar self-assembled structures can function as templates for 3D nanomaterial growth.⁵⁴ Furthermore, chemical and physical vapor deposition (CVD/PVD, respectively) are commonly used for bottom-up fabrication processes. PVD occurs when the coating material is vaporized, then solidified on the target material. In CVD, gaseous substances must react above the material's surface to produce as-deposited or self-assembled structures. Lastly, CVD- and PVD-coated microstructures can be shrunken into the nanoscale through dehydration of gel target material by 10–20×.⁴⁸ CVD-grown nanowires and nanocrystalline membranes have been successfully used for photovoltaic and photothermal-based bioelectronic stimulations.⁵⁵

While top-down fabrication methods, such as photolithography, have been used in the semiconductor industry for several decades, bottom-up processes are far newer. The key advantage of bottom-up processes is that the structure arises from the design of nanosized building blocks, such as DNA, and not from the direct patterning of the material (Figure 6D).⁵⁴ However, such assembly processes are still poorly controlled. Future advances in self-assembly techniques will certainly allow for the construction of more complex bioelectronic devices.

Recent progress

The last several years have seen a surge in bioelectronic advances. In this section, we will review notable advancements from the past 2 years in the subfields of biomedical devices, neuroelectronics, and e-skin devices.

Developments in biomedical devices, both wearable and implantable, allow monitoring of vital health information and disease biomarkers in real time. By constructing carbon nanotube-based helical fiber implants that mimic the soft tissue structure, Wang et al. successfully monitored the presence of disease biomarkers in vivo (Figure 7A).⁵⁶ These novel carbon nanotube (CNT) bundles contained specific fibers for real-time monitoring of glucose, hydrogen peroxide, and calcium content of the bloodstream. Similarly, several groups have successfully demonstrated the efficacy of smart contact lenses for monitoring of temperature, glucose content of tears, and other physiological information (Figure 7B).^9,57,58 Usage of smart contact lenses eliminates the need for expensive diagnostic devices, such as tonometers, and allows for noninvasive blood sugar monitoring (for diabetic patients). In the future, biomedical electronics will likely increase our ability to identify early signs of disease/illness and result in more comprehensive health data collection.

FIGURE 7. Representative examples for bioelectronic applications. (A) Carbon nanotube-based single-ply sensing fibers (SSFs) designed with specificity for different biomarkers (e.g., H2O2, glucose, Na+, K+, pH). (i and ii) Schematic of the multiprobe assembly. (iii) SEM picture of the assembly. Scale bar: 50 μm. Reproduced with permission from Wang et al.56 Copyright 2020, Springer Nature. (B) Electronic circuit insert and assembly of smart contact lenses for near-field wireless biomedical sensing. Reproduced with permission from Vásquez Quintero et al.58 Copyright 2020, Wiley. (C) Morphing bioelectronics for neuromodulation of growing nerves. Reproduced with permission from Lin et al.11 (D) Multimodal ion-electronic skin facilitates simultaneous temperature and capacitance measurement with high spatial resolution even under a considerable shear. Reproduced with permission from You et al.60 Copyright 2020, AAAS

Similarly, advances in neuroelectronics have led to precise, biocompatible neuromodulation and sensing devices. Recently, Abbott et al. developed a CMOS neuroelectronic interface capable of recording the intracellular activity of over 1700 neurons simultaneously.⁵⁹ The array, which consisted of platinum-black nanoelectrodes on top of a CMOS chip, was utilized to measure the ion channel current changes and record action/postsynaptic potentials. Future neuroelectronic interfaces may improve this design by replacing traditional silicon architectures with biohybrid substrates to improve biocompatibility and signal transduction. Historically, neuromodulation devices have often had to be replaced to accommodate tissue growth. However, Liu et al. have utilized morphing bioelectronic electrodes, which continued to stimulate the sciatic nerve during a 2.4× increase in diameter (Figure 7C).¹¹ Morphing electrodes are a marked advance from traditional neuromodulation devices, which must be replaced after significant tissue growth. As novel neuroelectronic devices are becoming more stable, we expect them to translate into the clinical setting in the near future.

Lastly, skin-mimetic electronics are an emerging technology that brings the possibility of monitoring skin temperature, mechanical stress/strain, and touch, among other metrics. Notably, Lee et al. constructed an ultrathin, compliant nanomesh sensor to monitor finger pressure without sensory interference.¹⁴ The sensor was composed of nanomesh passivation, top electrode, intermediate, and bottom electrode layers. In particular, accurately measuring finger pressure exertion is extremely important for the development of prosthetic devices. Further, You et al. recently measured and differentiated between temperature and mechanical strain in a soft ionic somatosensory system designed to record inputs such as pinching and spreading (Figure 7D).⁶⁰ The skin-mimetic receptors were made from two opposing layers of an electrode film coated with the ion conductor EMIM TFSI (1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide). Skin-mimetic electronics have many prospective applications in the health care and technological sectors, such as touch-sensitive devices and health care data collection.

PERSPECTIVE

Prototypical examples of future-bioinspired robots combine effective movement, renewable energy collection learned abilities/morphologies, and safe disposal (e.g., biodegradability, biointegration, or recyclability). In order to successfully develop a “synthetic life cycle” for future robots, we must intelligently draw inspiration from the ways in which organisms develop and behave. To improve future biomimetic systems, we must extract and apply complex biological principles in their designs instead of focusing purely on the replication of biological features/organisms. Further, robotic intelligence must not be perceived as an assembly of electronic components but rather as an integrated system. By further integrating sensing, actuation, and computational capabilities, greater biological and environmental integration levels can be achieved. Importantly, integrating scientific and technological fields beyond current multidisciplinary research approaches will push the envelope of bioinspired and biohybrid robotics. As the field of biomimetics progresses, novel devices may be conceived through the translational application of biorobotic design methodology, materials, and fabrication techniques in bioelectronic devices and vice versa.

On a larger scale, biomimetic devices have the potential to influence the manner in which medicine is practiced. The current US healthcare system is designed to treat patients in a preventive or purely reactionary manner, based on the symptoms they present. Conversely, the revolutionary field of precision healthcare aims to prevent disease progression through active monitoring and individual risk analysis. In the ideal precision healthcare system, family medical history, genetic predispositions, and other factors determine a patient's risk for various diseases. These risk data are supplemented by biomedical monitoring equipment, such as implants, wearable devices, biomarker detection devices, and in-home sensors (e.g., diet, sleep, urine). The resulting data are analyzed to determine appropriate action items to prevent disease progression. If a medical intervention is deemed necessary, doctors may provide additional treatment on case-by-case basis. Future development of precision health monitoring devices must focus on relevant data collection and, importantly, user engagement. Further integration of biorobotic and bioelectronic medical devices in the healthcare sector serves to increase the patient's quality of care. The combination of drug delivery therapeutics, passive sensing devices, and biocompatible implants will provide medical professionals with more bioinformatic data, enabling new treatment methods to combat challenging diseases.

Advances in biocompatible and biohybrid materials, sensing and actuation components, and control systems/algorithms will revolutionize future bioinspired systems. Biomimetics is a growing field, and the proliferation of bioinspired technologies in the consumer, healthcare, and industrial sectors is sure to continue. Without a doubt, robotic systems with incorporated bioelectronic devices will improve the living conditions of many patients in the future. A synergy between biorobotic and bioelectronic systems and design methodologies will spearhead the creation of highly biocompatible and precise devices with extended lifespans.

Lastly, due to the demand for greater cellular specificity, biocompatibility, and stability of bioelectronic devices, living (cell-based) bioelectronics research is set to proliferate. The development of living bioelectronics is a critical step toward the seamless integration. As an example of this seamless integration, living electrodes were grown from cortical neurons inside biocompatible hydrogel cylinders. The electrodes accurately performed optical sensing and neuromodulation in the rat brain for extended periods.⁶¹ In the future, living electrodes may be combined with traditional electronic circuits to create a closed-loop neural interfaces. Additionally, biological vasculature, which may increase biointegration and cell viability near implants, should be considered in future bioelectronic and biorobotic systems. Enhanced vascular regeneration was recently achieved by Lee et al., who combined mechanobiological conditioning and pharmacological treatment to increase vascular regeneration of mesenchymal stem cells. The enhanced growth was achieved with increased angiogenic paracrine signaling and a larger endothelial-pericytes population.⁶²

SUMMARY

In this review, we have discussed the materials and fabrication processes, along with sensing and actuation components required for the construction of prototypical biorobotic and bioelectronic systems. Much of this paper details the basic properties of compliant and bioinspired materials, common fabrication methods of bioelectronic and biorobotic devices from the nanoscale to the macroscale, as well as the mechanics and use cases of key sensing and actuation components. Additionally, we have introduced the principles of bioinspired design and examined a variety of groundbreaking biomimetic devices for mimicry of biological phenomena, passive sensing, drug delivery, and devices for an active interface.

Beyond reviewing recent developments in bioelectronics and biorobotics, this article details how biorobotic design methodologies may be utilized to create novel bioelectronic systems and how biorobotic systems may be supplemented with bioelectronic devices. We postulate that future biomimetic devices, such as morphing bioelectronics and biomedical devices, may combine the actuation capabilities of biorobotic systems with the electronic sensing and logic of bioelectronics to achieve higher level functionalities. These systems may also incorporate biohybrid substrates for increased biocompatibility and decreased immunogenicity. It is clear, then, that the future of biomimetic devices lies not in the individual subfields of biorobotics and bioelectronics but the combination of each field's strengths in design, fabrication, and control methodologies.

ACKNOWLEDGMENTS

This work was supported by the US Office of Naval Research (N000141612958) and the National Science Foundation (NSF CMMI-1848613, NSF DMR-2011854). Aleksander Prominski acknowledges support from the NSF MRSEC Graduate Fellowship (NSF DMR-2011854).

CONFLICT OF INTEREST

The authors declare no conflict of interest.