Biofluid‐Activated Biofuel Cells, Batteries, and Supercapacitors: A Comprehensive Review

Recent developments in wearable and implanted devices have resulted in numerous, unprecedented capabilities that generate increasingly detailed information about a user's health or provide targeted therapy. However, options for powering such systems remain limited to conventional batteries which are large and have toxic components and as such are not suitable for close integration with the human body. This work provides an in‐depth overview of biofluid‐activated electrochemical energy devices, an emerging class of energy sources judiciously designed for biomedical applications. These unconventional energy devices are composed of biocompatible materials that harness the inherent chemistries of various biofluids to produce useable electrical energy. This work covers examples of such biofluid‐activated energy devices in the form of biofuel cells, batteries, and supercapacitors. Advances in materials, design engineering, and biotechnology that form the basis for high‐performance, biofluid‐activated energy devices are discussed. Innovations in hybrid manufacturing and heterogeneous integration of device components to maximize power output are also included. Finally, key challenges and future scopes of this nascent field are provided.


Introduction
24] DOI: 10.1002/adma.202303197Tissue-integrated sensors enable close monitoring of various metabolic, electrolytic, proteomic, electrical, kinematic, thermal, and vascular dynamics parameters which provide unprecedented insights into the biological processes occurring within the body. [25]Similarly, new classes of wearable and implantable drug delivery platforms offer precise administration of therapeutics for modulating physiological processes [26][27][28] and treating diseases while implantable devices such as pacemakers are becoming increasingly important due to growing numbers of patients with cardiac arrhythmias. [29]hile developments in materials science, biotechnology, and electronics are responsible for the realization of advanced tissue-integrated devices, [2,3,[30][31] powering these with suitable energy sources remains a major challenge.Conventional electrochemical energy devices such as coin cell batteries and supercapacitors are the first choices for such applications due to their ability to supply high and constant energy and their long operational lifespan.Unfortunately, these energy devices are bulky, rigid, and comprised of toxic materials.Their unyielding form factor and weight cause issues with their conformal attachment to soft, curvilinear biological tissues while their reliance on toxic materials poses serious safety concerns.
The plethora of research in the area of thin, flexible/stretchable batteries and supercapacitors fabricated on various platforms including textiles [32,33] and elastomers [34,35] represent excellent alternatives to conventional metal casing-based electrochemical energy devices.However, these conformal, lightweight devices still rely on the same toxic chemistries as their conventional counterparts and therefore inherit the same safety challenges.Moreover, their thin packaging makes them prone to leakage which reduces their operational lifespan and increases risk of health hazards.These limitations form the basis for developing biofluidactivated energy devices -a new, emerging class of electrochemical energy devices that are made of biocompatible materials and importantly utilize the surrounding biofluid as the electrolyte.The reliance on benign materials and biofluid as electrolytes renders these types of energy devices highly biocompatible and aptly suited for tissue-integrated applications.Abundant and continuously refreshed supplies of biofluids by the body ensure that these devices perform optimally and unlike other electrochemical A schematic representation of the biofluid-activated energy devices covered in this review article which include i) energy storage systems (batteries and supercapacitors); ii) energy harvesting systems (biofuel cells); and iii) integrated systems (combined energy storage and harvesting systems).Biofluids such as sweat, tears, gastric intestinal fluid, blood, urine, and saliva are composed of several ions and metabolites which render them as viable electrolytes and fuel sources for generating electricity.
energy devices, they do not face issues associated with electrolyte leakage.
In this review article, we discuss biofluid-activated electrochemical energy storage and harvesting devices that include biofuel cells, batteries, and supercapacitors (Figure 1).We cover the key innovations in materials engineering and biotechnology responsible for the realization of such devices.We also discuss systems engineering and hybrid manufacturing methodologies implemented to fabricate multi-modal, biofluid-activated energy devices that generate energy from multiple sources and thus offer a more reliable energy supply.Finally, we review key challenges associated with this field and suggest potential solutions to address them.We also discuss the various future research avenues that can help advance this nascent field.

Introduction
Biofuel cells (BFCs) are an exciting and unique class of biofluidactivated energy devices, since unlike other types of biofluidactivated energy devices, BFCs rely on ambient biofluids for both electrolyte and fuel.BFCs are electrochemical devices that utilize coupled half-reactions at an anode/cathode pair to generate power, typically driven by oxidation of a biological fuel at the anode and reduction of oxygen at the cathode. [36]44][45][46][47][48] Early examples of BFCs utilized microorganisms, relying on the aerobic metabolism of cells free floating in solution to consume oxidizable fuel and produce power, dubbed microbial fuel cells (M-BFCs).Whole cells can survive in a variety of media and conditions and contain multiple oxidoreductase enzymes which efficiently consume oxidizable biological molecules, but their cell membranes impede electron transfer from the enzyme active sites to the anode surface.The low power densities of M-BFCs drove research into functionalization of the relevant enzymes directly on the electrode surface for realizing enzymatic biofuel cells (E-BFCs).The most common configurations feature glucose oxidase (GOx) or lactate oxidase (LOx) on the anode to oxidize glucose and lactate respectively, and reduction of oxygen catalyzed by platinum (Pt), [49][50][51] laccase, or bilirubin oxidase (BOD) [36,[52][53][54][55][56][57][58][59] enzymes at the cathode.The cathodic reaction and factors affecting it are discussed in detail in a review by Mano et al. [60] In some cases, the electronic communication between enzymes and electrode surface is facilitated by redox mediating species present either in solution [61,62] or fixed to the electrode surface, [37,44,49,63] while others achieve current flow without such mediators via direct electron transfer (DET) between the enzyme and the underlying electrode. [64,65]Compared to mediated electron transfer (MET), systems using DET feature low overpotential for anodic oxidation reactions, simplified fabrication, nontoxicity, and improvements in stability over time by removing the possibility for mediator leaching.67] This requirement also restricts the amount of enzyme utilized in DET-based systems to a monolayer covering the electrode surface which in turn results in lower power output compared to MET systems. [68]Further discussion of DET is available in several focused reviews. [65,67,68]dvancements in the field include improved methods of enzyme immobilization on the anode to maximize enzyme loading; advanced redox mediation strategies to reduce the overpotential necessary for charge transfer between the enzyme and electrode; and methods to increase the cathodic oxygen reduction current. The following sub-sections describe major advancements in the field of E-BFCs and M-BFCs that harvest energy from various biofluids including blood, serum, extra cellular fluid, sweat, gastrointestinal (GI) fluid, saliva, tears, and urine.

Blood/Serum/Extra Cellular Fluid
E-BFCs are an appealing alternative to conventional batteries for powering pacemakers, drug delivery pumps, and other implanted medical devices, given their small size, biocompatibility, and ability to operate at physiological pH and temperature.While the first attempts at generating power from biofluids inside the body relied on inorganic materials for oxidation of glucose and reduction of oxygen, this approach necessitated an ionconducting membrane separating anodic and cathodic compartments due to the non-specific reactions.[77] An important advancement was the "wiring" of enzymes to both anode and cathode with carbon fiber anodes being modified with (polyvinylpyrrolidone-[osmium(N,N'-dialkylated-2,2'-biimidazole) 3 ] 2+/3+ ) and cathodes functionalized with (polyacrylamide-[osmium(4,4'-dichloro-2,2'bipyridine) 2 chloride] +/2+ ) redox polymers by Heller and Mano in 2003. [63]In this case, glucose was oxidized by GOx at the anode (Equation 1), and BOD reduced oxygen at the cathode (Equation 2): Transmission of the electron from enzyme active site to the electrode occurred via the redox polymers.Using the immobilized enzymes and mediators removed the need for a separating membrane, simplifying the device fabrication, reducing the cost, and allowing for miniaturization on thin, flexible carbon fiber electrodes which enabled facile implantation in biological mediums, as demonstrated by successful power generation inside of a grape.A later work using redox polymers produced 129 μW cm −2 at 0.38 V in human blood samples using anodes composed of polyvinylpyrrolidone-[Osmium(1,1′-dimethyl-2,2′bisimidazole) 2 −2-(6-methylpyridin-2yl)imidazole] 2+/3+ mediating glucose dehydrogenase (GDH) and cathodes featuring polyacrylamide-polyvinylimidazole-[Os(4,4′-dichloro-2,2′-bipyridine) 2 Chloride] +/2+ mediating BOD, for which the anodic reaction was represented by Equation (3) and the cathodic reaction by Equation (2). [53]−D − glucose The first in vivo E-BFC example was implanted in the retroperitoneal space of a rat, oxidizing glucose present in the extracellular fluid.The authors "mechanically confined" GOx and polyphe-nol oxidase (PPO) enzymes mixed with ubiquinone and quinhydrone redox mediators (anode and cathode, respectively) in a graphite/glycerol paste followed by mechanical pressing to form discs of 1.33 cm diameter and 0.1 cm thickness.Equation (4) and Equation ( 5) represent the redox reactions that generated power: Wrapping the discs in cellulose acetate dialysis membranes with molecular weight cut-off at 100 g mol −1 ensured no loss of water-soluble enzymes and mediators upon implantation.Doing so replaced the need to covalently link enzymes to the electrode surface as in previous in vitro designs.This allowed the substitution of PPO for BOD, overcoming the limitation imposed by BOD's inhibition by chloride ions present in extracellular fluid.This system achieved stable power output over 11 days but limited average power of 2 μW. [69]ow in vivo concentrations of glucose and oxygen are one of the major reasons for low energy production by E-BFCs.Miyake et al. addressed the low oxygen issue and greatly improved the power output to 131 μW cm −2 and open circuit voltage (OCV) of 0.56 V with a partially implanted E-BFC in which the glucose-oxidizing anode functionalized with GDH and polymeric vitamin K 3 redox mediator was inserted into a blood vessel in a rabbit's ear, while the oxygen reducing cathode remained outside the body where it could access the much higher atmospheric oxygen levels. [36]The hydrophobic carbon paper and ketjenblack surface of the cathode promoted high oxygen diffusion to the immobilized BOD and resulted in high cathodic current of ≈2 mA cm −2 .
Katz's group demonstrated that the E-BFCs could produce energy even when interfaced with tissues.The group placed 2.5 cm x 0.8 cm electrodes directly on rat cremaster tissue. [70]The buckypaper disc electrodes comprised compressed multiwalled carbon nanotubes (MWCNTs).The MWCNTs featured many randomly oriented active sites that could participate in DET with the active sites of pyrroloquinoline quinone-dependent glucose dehydrogenase (PQQ-GDH) at the anode and laccase at the cathode without the need for mediators or separators.This device produced 0.175 μW cm −2 at 0.140 V but the time scale of the experiment was limited to ≈1 h.Zebda et al. also used DET to remove the voltage losses associated with mediator-based electron transfer.Mixing GOx and laccase with MWCNTs in compressed paste electrodes without redox mediators and placing them in the abdominal cavity of a rat achieved 161 μW mL −1 peak power as compared to 24.4 μW mL −1 in their first work [69] and were able to power a digital thermometer and light emitting diode (LED). [78]aximizing the enzyme loading on electrode surface is a widely used strategy for obtaining high-power E-BFCs.Kwon et al. described a biscrolling method in which sheets of MWCNTs were twisted in a guest solution producing a porous layered 3D structure bringing a high amount of enzyme in contact with the conductive electrode material. [79]After coating a MWCNT sheet with poly(3,4-ethylenedioxythiophene) (PEDOT) to make the surface hydrophilic, guest solutions of GOx, poly(N-vinylimidazole)-[Os(4,4′-dimethoxy-2,2′-bipyridine) 2 Cl]) +/2+ and poly(ethylene glycol) diglycidyl ether (PEGDGE, for enzyme crosslinking) for anode oxidizing glucose and BOD, poly(acryl amide)-poly(Nvinylimidazole)-[Os(4,4′-dichloro-2,2′-bipyridine) 2 ]) +/2+ , and PEGDGE for cathode reducing oxygen resulted in a yarn with alternating layers of PEDOT-coated MWCNTs and enzyme with redox mediator.The innovative use of hydrophilic PEDOT coating on the hydrophobic MWCNTs was critical to enable immobilization of hydrophilic enzymes and redox polymers and bolstered the mechanical strength and flexibility of the MWCNT film which allowed the device to be incorporated into textiles or wrapped around rod-like features.In human serum, the E-BFC produced an impressive peak power of 2.18 mW cm −2 and maintained ≈77% of the initial power output after 28 days.
E-BFCs operating inside the body must operate under the physical and physiological conditions imposed by the surrounding cells and tissues.Guo et al. used a flexible carbon nanotube (CNT) fiber as the scaffold for a GOx/Pt-based E-BFC that oxidized glucose (mediated by tetrathiafulvalene (TTF)) and reduced oxygen inside a rat brain and demonstrated low internal stresses in the E-BFC upon compression by the surrounding tissue without impacting performance (91% of power output after 100 bending cycles) (Figure 2a). [50]This mechanical matching of fiber and tissue led to minimal tissue shearing during deformation and decreased the risk of inflammation.Coating the functionalized anode and cathode with polydopamine (PDA) and methacryloyloxyethyl phosphorylcholine (MPC) to render the surface of the device hydrophilic and zwitterionic prevented device passivation due to biofouling.After 1 month of implantation in a rat, PDA-MPC coated devices displayed a power output of 2.5 μW cm −2 as compared to device failure after one day for untreated devices.MPC coating also impeded blood clot formation and improved power output compared to untreated devices for an E-BFC inserted in a rabbit's vein. [36]iofouling is an important phenomenon that dictates the operational lifespan of an implanted E-BFC.Developing antibiofouling coatings is therefore an active area of research.El Ichi et al. fabricated cathodes that reduced oxygen featuring improved biocompatibility with impressive in vivo lifetimes. [80]By mixing laccase with chitosan, genipin, and MWCNTs, the group obtained a nanofibrous material with a nanofiber diameter of 30 nm and pore size of ≈6 nm enabled by genipin crosslinking of the chitosan.After 167 days of implantation in a rat, the cathode was removed and it retained 50% of the original catalytic current.Furthermore, new vascularized tissue without inflammatory cell growth covered the cathode surface, indicating excellent biocompatibility.Building on these findings, the group developed a complete E-BFC with the glucose oxidizing anode made by mixing GOx, catalase, naphthoquinone (NQ), and MWCNTs while the cathode was the same as their previous work (laccase mixed with chitosan, genipin, and MWCNTs).Both electrodes were then coated with a thin film of genipin-cross-linked chitosan and sealed within cellulose acetate dialysis bags.The researchers added catalase to the anode to enable rapid consumption of hydrogen peroxide (H 2 O 2 ) produced by GOx which can deteriorate GOx and aggravate the tissue inflammation response. [71,80]his system was implanted in a rabbit for 2 months with wireless electronic communication to monitor the E-BFC OCV and power output and perform daily discharging through a 100 kΩ resistor.After decreasing steadily for the first week, the OCV sharply increased to reach a plateau of 0.4 -0.5 V correlating to ≈16 μW mL −1 which persisted for two weeks, followed by a decrease in output to ≈2 μW mL −1 by day 60.Assessment of the biocompatibility upon device removal at the end of the experiment showed more inflammation response compared to an identical implanted device that was left as an open circuit (nonfunctioning).These results could be explained by the known inflammation stimulus of H 2 O 2 and gluconic acid, suggesting that the included catalase was insufficient to account for all of GOx's oxidation products. [81,82] new frontier of implantable E-BFCs is developing bioresorbable devices which dissolve harmlessly after use, thus obviating the need for an extraction surgery.Recently, laser-induced graphene (LIG) electrodes functionalized with gold nanoparticles and GOx/laccase enzymes oxidizing glucose and reducing oxygen at anode and cathode respectively supported by a poly(lacticco-glycolic acid) (PLGA) substrate and silver nanowire electrical connections produced a maximum power of 483.1 μW cm −2 and 0.77 V in PBS spiked with 5 mM glucose. [83]Porous, high surface area LIG functioned as an effective scaffold for decoration with AuNPs and high loading of enzymes.After 7 days in the same fluid, the device exhibited 33% of its maximum power, and the PLGA substrate fully dissolved after 44 days when implanted in a rat.Histological studies showed no inflammation response after implantation, though the LIG remained largely intact (with fibrous encapsulation by the surrounding tissue).The quick dissolution of the bioresorbable PLGA scaffold lessened the impact of the implanted devices on the body.Future devices will likely be comprised entirely of bioresorbable materials similar or identical to those used in recent reports of transient electronics. [84,85]

Sweat
[88][89] The majority of these devices require a battery which hinders the comfort and safety of wearable devices owing to the inherent bulky nature of batteries as well as their toxic electrolytes.Lactate in human sweat has been targeted as a promising alternative source of energy due to its high average concentrations of 14 mM increasing up to 50 mM at peak levels. [10]Sweat also contains glucose and ethanol, and these three fuel sources are readily oxidized by their respective oxidation enzymes (lactate oxidase, glucose oxidase, and alcohol oxidase).
Building off the development of E-BFCs for implantable applications, noninvasive wearable systems that tap into the easily accessible lactate fuel source in sweat became an area of much interest.The first example reported in 2013 involved a tattoobased E-BFC utilizing LOx-functionalized anode and Pt blackbased cathode to reduce oxygen. [51]The anode and cathode reactions are described by Equation (6) and Equation ( 7) respectively.lactate LOx Functionalizing tattoo paper with E-BFC components prepared a device with an extremely low footprint and high comfort while worn, critical parameters for systems on skin.Chitosan and  [50] Copyright 2022, Wiley-VCH GmbH.b) Schematic, photograph of flexible wearable sweat E-BFC when mounted onto a human arm and closeup photographs of device interfaced with skin (top); scale bar: 1 cm.Schematic showing the E-BFC array consisting of lactate oxidase-modified bioanodes and Pt alloy nanoparticle-modified cathodes (bottom).Reproduced with permission. [49]Copyright 2022, The American Association for the Advancement of Science (AAAS).c) Schematic illustrations of bacteria-powered biobattery activated by human sweat (top).Schematic illustration of sweat-activated microbial biofuel cell (M-BFC) layers above skin (bottom left) and photograph of experimental setup of M-BFC on a human hand powering a thermometer (bottom right).Reproduced with permission. [95]Copyright 2020, Elsevier.d) Schematic illustration of ingestible E-BFC powered capsule operating in a porcine model (top).Illustration showing the corresponding key components of the E-BFC and the enzymatic reactions responsible for energy generation (bottom left).Photograph of the ingestible E-BFC powered capsule for glucose sensing and wireless data transmission to an external receiver (bottom right).Reproduced with permission. [37]Copyright 2022, Springer Nature.
Nafion (anode, cathode, respectively) coatings reduced leaching of enzyme and redox mediators and prevented direct contact of the sensor reagents with the skin.On-body testing produced a power output of 5 to 70 μW cm −2 which were inversely correlated to participants' fitness levels.One critical aspect of this and later works is the ability of the device to function even after repeated mechanical deformations, as experienced while worn on the body and during physical activity.In this case, "chopped" car-bon fibers dispersed in a carbon ink provided sufficient tensile strength to endure stretching and bending cycles.A follow-up work by the same group used similar active materials oxidizing lactate and reducing oxygen functionalized on a textile base for incorporation into a headband or other articles of clothing. [40]The use of an improved redox mediator (tetrathiafulvalene-7,7,8,8tetracyanoquinodimethane (TTF•TCNQ)) increased power production to 100 μW cm −2 and was used to power a watch.The group further improved upon the concept by developing stretchable conductive inks printed directly on textiles which retained their conductivity even after 100 cycles of 100% strain. [44]The inks were obtained by mixing conductive elements (CNTs and silver/silver chloride (Ag/AgCl) ink) with mineral oil, polyurethane, and Ecoflex.
Glucose is another potential fuel in sweat, targeted by Shleev's group with an E-BFC design using cellobiose dehydrogenase (CDH)-modified anode and BOD cathode to provide power from the oxidation of glucose and reduction of oxygen in sweat. [56]hese enzymes were able to use DET with the 100 μm diameter gold (Au) microwire/gold nanoparticle (AuNP) modified electrode surface, producing an average power output of 0.26 ± 0.03 μW cm −2 at 0.5 V in sweat.The 17 nm diameter AuNPs formed a 3D nanostructure on the Au microwire surface and improved the electroactive surface area of the device by a factor of 100.This approach facilitated energy generation even with low volume of sweat produced.When the authors added 500 μM glucose to the sweat, the power increased to 0.47 ± 0.08 μW cm −2 at 0.5 V which likely meant that the low concentration of glucose in sweat hindered the power production.
Advancements in stretchable wearable systems set the stage for impressive work by the Wang lab producing a wearable E-BFC on a stretchable substrate that generated a record 1.2 mW cm −2 of power. [72]The high power density was possible due to a combination of thick film and thin film hybrid manufacturing techniques.The rigid, porous 3D CNT-based anode and silver oxide reducing cathode were assembled on lithographically patterned "island-bridge" architecture.The anode and cathode were densely packed with large quantities of LOx enzyme and NQ mediator to achieve high performance without being concerned about their brittleness since the islandbridge design accommodated strain while protecting the 3D anode and cathode.For the first time, the authors demonstrated a wearable E-BFC powering a Bluetooth Low Energy (BLE) radio, a promising step toward the practical use of sweat-powered BFCs.
The current state-of-the-art device was presented by Yu et al, in which hierarchical nickel microstructures were coated with layers of reduced graphene oxide (rGO) and Medola's blue-TTF modified CNTs (Figure 2b). [49]This use of nanomaterial structuring produced a flexible material with a high electrochemical surface area, and the pi-pi interaction of the rGO and redox mediator modified-CNTs enhanced the electron transfer rate from LOx enzyme to the electrode surface. [90]In particular, the modified CNTs enhanced the electrochemical surface area, reduced the overpotential of the lactate oxidation reaction, and maintained current stability during repeated cyclic voltammograms.The researchers used cobalt (Co) to stabilize Pt nanoparticles and form Pt-Co alloy nanoparticle functionalized surfaces which achieved minimal changes in onset potential of the cathodic oxygen reduction reaction in sweat samples over 30 h.As a result, this skin-conformal E-BFC achieved a power density of 3.5 mW cm −2 , the highest power output of a single-cell E-BFC in human sweat to date.The power was used to monitor sweat urea, ammonia, glucose, pH, and skin temperature, along with wireless transmission of the data via BLE.This sophisticated device achieves a central goal of the wearable E-BFC field by replacing the need for batteries in wearable sweat measurement systems.
Incorporating microfluidics within the BFCs that enable rapid transport of fuel to the electrode surface and removal of byproducts away from it is another strategy to enhance power output of BFCs in addition to improving electrode fabrication processes as described above.A review of E-BFC use in microfluidics provides further detail on the design and performance of such systems. [91]Minteer's lab demonstrated that the power output for a LOx/laccase E-BFC could be increased from 61.2 μW cm −2 to 305 μW cm −2 by integration of microfluidics. [92]In this study, the fluid used was 0.2 m phosphate buffer saline (PBS) at pH 5.6, but the concept of a microfluidic flow cell could be used to improve mass transfer and power in wearable systems with proper design considerations.Such a skin-interfaced microfluidic patch was designed by Zhou's group, comprised of flexible laser-cut polyimide and polyethylene terephthalate layers to define the microfluidic channels and E-BFC reservoir where alcohol oxidase (AOx) and BOD enzymes were immobilized on screen-printed carbon electrodes decorated with 3D coralloid nitrogen-doped hierarchicalmicro-mesoporous carbons aerogels (3D-NHCAs). [55]The system used Equations (8-10) to produce power.
The high surface area and blend of micro and meso pores of the 3D-NHCA surface significantly improved the electrocatalysis at both anode (>20 times higher current density compared to unmodified screen printed electrode (SPE)) and cathode (onset potential of 0.55 V vs. Ag/AgCl compared to no catalysis observed at SPE).The complete device featured conformal bending with movement and generated 1.01 μW cm −2 peak energy from ethanol in sweat during on-body exercise trials.
Low-cost paper substrates have also been utilized for wearable E-BFCs which offer good performance and could be used in cases where single use or disposability is required such as wound dressings or other medical applications requiring sterility.Shitanda et al. constructed lactate E-BFCs on paper substrates featuring screen-printed magnesium oxide (MgO) templated porous carbon electrodes with UV treatment to increase hydrophilicity. [54]Electrode pore size can be tuned in this fabrication process, and studies show that a mix of mesoscale and macroscale size pores favor efficient transport of substrate to the enzymes and provide a stable environment for enzyme immobilization, leading to improvements in electrode stability by reducing enzyme leaching. [93,94]NQ was solvent-cast and dried on the anode, followed by drop casting a solution of LOx.Cathodes required no redox mediation and were constructed by simply dropcasting BOD solution onto the untreated carbon.The facile fabrication methods enabled construction of an array of 6×6 cells which produced 4.3 mW at 2.44 V.
Combining paper with microbes, Choi's group recently illustrated a paper-based, microfluidic M-BFC utilizing bacteria native to human skin to generate power (Figure 2c). [95]Staphylococcus epidermidis organisms oxidize ammonia present in sweat, and despite possessing a thick cell membrane, can transfer electrons to the anode of the wearable microfluidic device.The M-BFC anode was coated with a PEDOT:PSS and 3-glucidyloxypropyldimethoxysilane (GLYMO) to achieve high conductivity and engineer a porous hydrophilic surface for S. epidermidis inoculation.Wax printing defined the outlines of anodic and cathodic compartments and separating membranes.Prussian blue, graphene, PEDOT:PSS, Nafion, and isopropanol were mixed and applied to the paper substrate to form the cathode.The cell delivered 41.74 ± 5.35 μW cm −2 demonstrating the promise of such systems which do not require purified enzymes or complicated immobilization schemes as is the case of E-BFCs.This group has also produced a heel-mounted patch version of their wearable M-BFC using Bacillus subtilis bacteria which could undergo sporulation, a genetically signaled process by which the bacteria form stable endospores that survive for long periods without nutrients-in this case during periods of low/no sweat flow. [90]Once sweating resumes, the spores germinate producing viable cells and continuing to produce power.The device produced 24.4 μW cm −2 peak power in sweat, and successfully stopped and started power production synchronously with sweat depletion and reintroduction into the device.

Gastrointestinal Fluid
The human GI tract is currently the focus of a great deal of research exploring the native microbiome's role in digestion, neurochemical production, and overall health. [96]The GI tract plays a critical role in human health, with 1 in 5 people experiencing GI disorders in their lifetime. [37]Monitoring digestive chemicals as they pass through the GI tract currently is limited to more invasive extraction methods but using the nutrient-rich GI fluid as fuel for E-BFCs could enable ingestible electronics that sample and continuously measure chemicals non-invasively.Preliminary work in 2018 described a fully edible E-BFC using plant sources for all active components. [97]The mushroom-derived alcohol oxidase/apple-derived PPO anode/cathode produced 282 μW cm −2 and power production remained above 80% of initial values for 4 h in a PBS buffer.The authors proposed this alcoholoxidizing E-BFC as a potential method of measuring the interactive effects of ethanol consumption and pharmaceuticals in the GI tract.
Glucose is an important GI parameter that De la Paz et al. monitored using a self-powered sensing approach based on GOx and BOD functionalized E-BFCs contained within a 2.6 cm x 0.9 cm ingestible pill (Figure 2d). [37]Despite relatively low power generation of 0.4 μW cm −2 , the inclusion of carefully designed electronics enabled wireless data transmission of glucose concentration without any other power source.As commonly observed, the use of a porous carbon electrode provided a high surface area with suitable pore size for maximum enzyme loading.Besides small form factors and sufficient power, passage through various stages of the GI tract necessitates system resiliency in both acidic (stomach) and neutral (small intestine) environments.In this case, an enteric pH-responsive coating protected the sensors during transit through the stomach and dissolved once the pill reached the small intestine.
Another work utilized a similar strategy of powering an ingestible pill using an enteric pH coating, but instead of glucose, relied on the metabolism of Bacillus subtilis bacteria to produce power. [74]The bacteria began in a dormant endospore form and germinated upon dissolution of the enteric coating when the pill entered the small intestine.Hydrophilic hydrogel absorbed nutrient-containing intestinal fluid and transported it to the spores, stimulating their metabolism and triggering operation of the M-BFC.To overcome the low oxygen concentrations of the GI tract, an oxygen-rich cathode was comprised of polychlorotrifluoroethylene (PCTFE), cerium dioxide (CeO 2 ), CNT, and Pt in the ionic liquid (IL) 1-butyl-3methylimidazolium hexafluorophosphate.This composite material catalyzed oxygen reduction, promoted oxygen diffusion to the electrode surface, and possessed its own native supply of oxygen to maximize the cathodic current in the E-BFC.As these groundbreaking studies illustrate, novel applications for E-BFCs and M-BFCs are emerging outside of the foundational powermaximizing paradigm responsible for developing the field to the present day.

Saliva, Tears, and Urine
Saliva is a readily accessible biofluid that contains glucose and has fast replenishment of oxygen and a relatively simple sample matrix as compared to blood.While the concentration of glucose in saliva can be as low as 50 μM, [98] studies have investigated its use as a fuel source for E-BFCs for powering wearable salivary devices. [99]Conghaile et al. functionalized anode graphite rods with acidtreated MWCNTs, redox polymer [Osmium(4,4′-dimethoxy-2,2′bipyridine) 2 (poly(vinylimidazole)) 10 Chloride] + , and glutaraldehyde vapor crosslinked deglycosylated recombinant pyranose dehydrogenase (PDH).This enzyme does not use oxygen in its catalysis of glucose, reducing competition at the cathode. [59]Cathodes were prepared on polished gold rods with several coats of gold nanoparticles followed by deposition of BOD.The E-BFC displayed 6.0 ± 0.5 μW cm −2 at 0.45 V, nearly identical levels as obtained from PBS spiked with 50 μM glucose, indicating a low effect from the saliva sample matrix on performance.Bollela et al. used a commercial screen-printed electrode base and constructed a miniaturized E-BFC by functionalizing cellobiose dehydrogenase from Corynascus Thermophilus (CtCDH) on anode and Trametes Hirsuta laccase (ThLac) on cathode, respectively. [100]oth enzymes were capable of DET and required only solution drop-casting of gold nanoparticles on the graphene electrodes prior to enzyme deposition for improving catalytic surface area.In human saliva samples, the power output was 1.10 ± 0.12 μW cm −2 at 0.41 V.
Glucose and lactate are found in tears and could be used as fuels for powering wearable ocular devices. [101,102]Glucose-powered a CtCDH/M.verrucaria bilirubin oxidase (MvBOx) E-BFC supported by gold microwire electrodes modified by gold nanoparticles and delivered 3.5 μW cm −2 at 0.2 V and produced 68% of the initial current when run at 0.51 V for 20 h. [73]An E-BFC with buckypaper anode and cathode shaped to fit around the outside of an elastomeric contact lens produced 8.01 μW cm −2 using lactate as fuel in an artificial tear solution. [57]Given the close proximity to the eye, the lactate dehydrogenase (LDH; on anode) and BOD (on cathode) were immobilized within octyl-modified linear poly-ethyleneimene and tetrabutylammonium bromide modified Nafion respectively to ensure biocompatibility and prevent leaching of chemicals into the eye.
Urine can also be used as fuel for E-BFCs as it features higher glucose concentration which is correlated with blood glucose levels and thus can be used to signal need for medical intervention in diabetic patients. [58]Shitanda et al. used MgO-templated mesoporous carbon-based inks to fabricate E-BFCs on paper substrates. [103]These E-BFCs provided 0.12 mW cm −2 and powered wireless transmission of glucose concentration in order to reduce the burden of care in nursing scenarios.In their device, graft polymerized poly(glycidyl methacrylate) (PGMA) covalently linked GDH and BOD enzymes to anode and cathode, respectively, and improved the power output stability as compared to devices without graft polymerization.A summary table comparing the performance of the discussed BFCs appears in Table 1.

Introduction
Biofluid-activated energy storage systems (B-ESSs), namely batteries and SCs, represent an exciting and innovative technology with the potential to revolutionize the field of energy storage.Unlike BFCs, B-ESSs already have electrical energy stored in them in the form of chemical energy or charge buildup and only require surrounding biofluids to serve as an electrolyte.High operating voltage and stable power generation are some of the key advantages of B-ESSs over BFCs.Performance of B-ESSs is governed by ionic conductivity of the biofluid, infiltrated volume, and evaporation rate.The use of biofluids as an electrolyte has the added advantage of electrolytes being readily available, as these fluids are constantly produced by living organisms.These attributes form the basis for recent interest in using E-BSSs for powering wearable, [104,105] ingestible, [106] and implantable medical devices. [107]electing the appropriate materials for B-ESSs is crucial to ensure their safe and efficient operation.In general, B-ESSs consist of three functional components: i) anode (oxidizing electrode), ii) cathode (reducing electrode), and iii) ion exchange membrane or separator that can readily absorb ambient biofluids as electrolytes. [108]The materials used must be biocompatible or biodegradable to prevent any harm to the body when implanted or in contact with bodily fluids.Furthermore, the selected materials should have suitable properties to ensure efficient ion transport and prevent fast degradation of electrodes. [109,110]Since the working voltage is determined by the potential at which the anode and cathode undergo oxidation and reduction, respectively, it is critical to carefully select materials for the anode and cathode that produce the desired potential difference.The biofluids can be acidic, alkaline, or neutral, and thus the material selection must consider the potential for corrosion and other forms of material degradation as well as the biocompatibility of degraded end products. [111]esearchers must also consider the mechanical properties of the materials used to ensure the durability and longevity of B-ESSs.

Biofluid-Activated Batteries
Batteries provide high energy density, low self-discharge, and constant voltage output which are crucial factors for providing sustained power to electronics.Most conventional batteries are composed of toxic components such as heavy metals (lead, cobalt, cadmium, and nickel) and intrinsically flammable electrolytes, feature unyielding form factors, pose serious safety concerns, and require thick packaging.These factors make them undesirable for tissue-integrated applications. [112,113]Furthermore, mechanical failures could cause leakage of encapsulated toxic electrolytes, which is extremely concerning, particularly in implantable use cases. [114]Therefore, environmentally benign materials are imperative for the development of safer and highly biocompatible battery technology.

Sweat
Sweat has attracted major attention for biofluid-activated batteries compared to other biofluids, due to its safe, simple, and noninvasive mode of collection as well as its ionic conductivity (≈17 mS cm −1 ). [115,116]One example of skin-integrated B-ESSs includes our previous work, where we exploited sweat as an electrolyte for a thin, flexible, biocompatible battery integrated with microelectronic and microfluidic platforms (Figure 3a). [117]The sweat-activated cell (SAC) consisted of magnesium (Mg) as the anode, Ag/AgCl as the cathode, and sodium chloride (NaCl) impregnated dry cellulose membrane as the separator.Mg was chosen as anode material due to its excellent biocompatibility, its nutritional value (minimum daily intake: 100 mg day −1 ), nontoxic products from corrosion, low thrombogenicity, high theoretical capacity of 2205 mAh g −1 , and low standard electrode potential of -2.38 V vs. Standard Hydrogen Electrode (SHE).Furthermore, Mg is an abundant element and eco-friendly.Equation (11) and Equation ( 12) provide the half-cell reactions for the Mg-Ag/AgCl cell.
The absorbent nature of SACs overcomes major issues commonly faced with conventional energy storage devices such as electrolyte leakage, self-discharge, and poor shelf-life.In the presence of sweat, the SAC provided an output voltage of 1.6 V, a specific energy of ≈580 Wh kg −1 , and a specific capacity of ≈67 Ah kg −1 which is comparable to the specific capacity of commercial cells (CR2032; ≈73Ah kg −1 ).The high capacity and operating voltage enabled powering of a wireless, wearable heart rate sensor.Similarly, a paper-based SAC using Mg as the anode and AgCl as the cathode was developed. [118]The electrodes were connected by a hydrophilic paper that is capable of holding biofluid.The battery operated in the sweat conductivity range of 5-160 mM equivalent NaCl, generated 1 mW power and displayed dual use as a sensor and energy storage device.Others have also reported similar SACs which are integrated with cotton yarn and demonstrated a stretchable battery in skin-integrated electronics. [119,120]Integrating SACs into textiles and garments offers several benefits  including intrinsic collection of sweat via wicking properties of textiles and ample surface area to integrate arrays of SACs in various series/parallel combinations for producing high power output.To achieve stretchability, biocompatible materials zinc (Zn; anode) and copper (Cu; cathode) were embedded in low-cost water-absorbable nylon fabrics as a functional layer, and a stretchable silicone shell-based soft hydrophilic cotton containing potassium chloride (KCl) powder for capturing sweat was developed.These SACs produced a high capacity of 42.5 mAh and power density of 7.46 mW cm −2 and thus enabled sufficient power for skin electronics.Huang et al. recently demonstrated a garmentintegrated B-ESS comprised of Mg anode and graphene cathode with KCl-impregnated cotton membrane as a separator. [121]Ultrathin architecture (1.25 mm), low weight (0.65 g), and high flexibility of batteries allowed good interface with skin for long-term use.Mg oxidation reaction occurred at the anode while oxygen was reduced at the graphene-based cathode.Anode half-cell re-action was described by Equation ( 11) and the several cathodic pathways were described by Equations ( 13)- (15).
The uniform coating of graphene on paper enhanced the performance of the SAC by catalyzing oxygen reduction and adsorbing electrolytes.Upon activation by sweat, the cell exhibited a high capacity of 14.33 mAh, a maximum power density of 3.17 mW cm −2 , and continuously powered 120 LEDs over 3 h.Reproduced with permission. [117]Copyright 2020, Springer Nature.b) Schematic illustration of the subcutaneous fluid-activated biodegradable cell showing key components (top), photographs of the device showing front and side view (bottom left), and photographs of the implanted cell at the sciatic nerve of the rat (bottom right); scale bar: 5 mm.Reproduced with permission. [124]Copyright 2020, AAAS.c) Exploded view illustration of dual electrolyte magnesium-iodine (Mg-I 2 ) cell with constituent materials (top left) and schematic presenting the chemical reactions at the Mg anode and I 2 cathode when immersed in biofluid (top right).Schematic diagram of the in vivo experimental setup for the evaluation of implanted dual-electrolyte cells in mice (bottom left) and optical image of the stacked cell implanted inside a mouse (bottom right).Reproduced with permission. [125]Copyright 2022, The Royal Society of Chemistry.d) Illustration of a tear-activated cell embedded in a smart contact lens showing the corresponding porous electrodes (top).Photograph of the cell on an artificial eye and an optical microscope image showing the cross-sectional view of the cell (bottom).Reproduced with permission. [127]Copyright 2021, American Chemical Society.

Gastrointestinal Fluid
GI fluid is comprised of metabolites, electrolytes, enzymes, proteins, ions, and gases which are replenished by human consumption of food and the various digestive processes, making it an excellent electrolyte for B-ESSs. [122]Nadeau et al. demonstrated a GI fluid-activated cell to power an ingestible pill for real-time, wireless monitoring of GI tract temperature in pigs. [123]Due to the nutritional value, ease of manufacturing, affordability, and relatively low position in the electrochemical series, Mg was chosen as anode, and Cu as a cathode material.The anodic reaction occured with metal dissolution as given by Equation (11), while the cathodic reaction involved either reduction of dissolved oxygen (Equation 13) or hydrogen evolution, given by Equation (16).
By utilizing GI fluid as the electrolyte, the fabricated cell delivered a peak power of 13 μW mm −2 and an average cell voltage of 0.23 V in the stomach of a pig.However, increasing the voltage and power of the cell by careful selection of materials and integrating membranes to improve proton exchange while controlling corrosion of electrodes could improve the cell performance for ingestible electronic applications.

Subcutaneous Fluid
A fully biodegradable cell activated by subcutaneous fluid as an electrolyte for continuous electrical stimulations for peripheral nerve regeneration was presented by Wang et al. as shown in Figure 3b. [124]Mg and iron-manganese (FeMn) alloys were selected for fabricating the cell, due to their favorable electrochemical potential, excellent biocompatibility, and desirable degradation rates in biological surroundings.As compared to iron, FeMn thin films exhibited a lower charge transfer resistance, a more negative corrosion potential, and a higher corrosion current (≈2.5 times faster corrosion current than pure iron films).The relevant anodic and cathodic reactions are given in Equation ( 13) and Equation ( 16).The biofluid-activated cell consisted of biodegradable Mg as anode and FeMn alloy as a cathode on a porous polycaprolactone (PCL) scaffold.The cell provided an average OCV of 0.984 V and generated an electric field of ≈20 to 250 mV mm −1 which was distributed along the conduit to stimulate surrounding nerve tissue.In in vivo sciatic nerve injury (10-mm nerve gap) assessments in rats, the cell demonstrated successfully accelerated neuroregeneration and enhanced function recovery, and hence these cells could be used as a reliable power source for promoting healing in damaged tissues and organs.
A novel biocompatible material and dual-electrolyte cell architecture with high-performance, resorbable cells capable of operating in intrascapular regions using subcutaneous fluids were developed by Huang et al. (Figure 3c). [125]Mg was chosen as the anode material and iodine (I 2 ), dispersed carbon black, chitosan, and poly(lactide-co-glycolide) (PLGA) were chosen for the cathode.I 2 has a high standard electrode potential of 0.536 V vs SHE and carbon black imparted conductivity and minimized leaching of I 2 due to strong adsorptive properties.The water-permeable PLGA enhanced the effective surface area of I 2 in the composite and formed a 3D porous network, and the strong affinity of chitosan to I 2, as well as the amine and water barrier formed by hydroxyl groups, prevented I 2 dissolution.The half-cell reactions of the Mg-I 2 cell are given by Equations ( 11) and (17).
The fabricated cell consisted of a partitioned dual-electrolyte system in which the catholyte consisted of chloride/urea-based ionic liquid (IL) and subcutaneous fluids media served as the anolyte.This innovative design strategy greatly enhanced the durability and improved the capacity compared to traditional systems that use a single electrolyte.The cell produced a stable operating voltage of ≈1.8 V, areal capacity of ≈9.8 mAh cm −2 , areal energy density of ≈17.7 mWh cm −2 , areal power density ≈0.7 mW cm −2 , volumetric energy density ≈93.0 mWh cm −3 , and volumetric power density ≈3.8 mW cm −3 , all of which are significantly higher than those of previously reported examples.Implantation in the intrascapular region of a rat and its ability to power a bioresorbable pacemaker demonstrated the ability of the cell to power implantable devices.
Li et.al. developed an implantable zinc-oxygen (Zn-O 2 ) battery for neural regeneration, which served as a power source for electrical stimulation. [126]This battery, with its tubular shape and compact size of 0.86 mm 3 , could be directly wrapped around a nerve to stimulate nerve fibers in situ.It consisted of a flexible Zn wire anode, a carbon nanotube/platinum (CNT/Pt) film cathode, and subcutaneous body fluid utilized as an electrolyte.Zn was selected as the anode material due to its biocompatibility, low corrosion rate in the body, and higher density compared to the biocompatible Mg. [123] The CNT film was chosen as the substrate for its excellent electrical conductivity, lightweight nature, large specific surface area, and good biocompatibility.Additionally, the porous structure of the CNT film facilitates the exchange of oxygen, nutrients, and waste.During discharge, the Zn anode undergoes oxidation to Zn 2+ , while the surrounding body fluid continuously supplies O 2 .The CNT/Pt cathode worked as a catalyst for the reduction of O 2 to hydroxide, enabling the flow of electrons through an external load.The battery provided a discharge voltage of 1.06 V at a current density of 28 mA g −1 , and a high energy density of 231.4 mWh cm −3 .In simulated body fluid the battery exhibited a discharge-specific capacity of 683 mAh and an energy density of 511 mWh at a discharge current density of 112 mA.In vivo evaluations in rats for 12 weeks demonstrated that the Zn-O 2 battery-based nerve conduit promoted the synthesis of myelin basic protein and facilitated the maturation of myelinated nerve fibers in injured nerves and ultimately accelerated the regeneration of sciatic nerves.

Tears
Tears contain various ions such as sodium and potassium, which make them a suitable candidate to serve as an electrolyte for biofluid-activated batteries.Given the close proximity to a sensitive and important organ such as the eyes, the need to remove the risk of toxic electrolyte exposure associated with conventional batteries is critical.To this end, a flexible aqueous cell activated by tears was integrated into smart contact lenses by Yun et al. [127] The cell consisted of copper hexacyanoferrate (CuHCFe) as a cathode, and iron hexacyanoferrate (FeHCFe) as an anode and was embedded in a UV-polymerized hydrogel (2-hydroxyethyl methacrylate) that served as both a soft contact lens and an ionpermeable separator (Figure 3d).These electrode materials were chosen to obtain high voltage in an oxygen-permeable aqueous battery.CNTs were dispersed within polyvinylidene fluoride (PVDF) and N-Methylpyrrolidone (NMP) to produce a slurry and polymer matrix that was printed onto the lens to form current collectors.The oxygen reduction reaction was given by Equation ( 18): The cell oxygen evolution reaction was given by Equation ( 19): The CuHCFe cathode exhibited an oxidation capacity of 176.6 μAh between 0.6 and 1.1 V, and the FeHCFe anode exhibited a reduction capacity of 181.8 μAh between 0.1 and 0.6 V.The fabricated cell offered a discharge capacity of 155.4 μAh at 100 μA with cutoff voltages of 0.2 and 0.9 V and successfully powered a low-power static random-access memory designed to operate in the voltage range of 0.26 to 1.2 V. Another approach that could be incorporated into a smart contact lens was described by Pourshaban et al. in which an Mg-Air cell was activated with a sliding electrolyte (here tears) periodically dragged on top of the electrodes during the natural eye-blinking motion. [128]Pt was selected as the cathode material due to its high power density of 70 μW cm -2 and Mg was selected as the anode material.When the eye was closed, tears contacted with anode and cathode and provided a closed circuit to activate the cell.The anodic and cathodic reactions are given in Equation (11) and Equation ( 15) respectively.The tear-activated cell showed an open-circuit voltage of 2.2 V and a specific capacity of 3561 mAh g -1 at a discharge current density of 5 mA cm -2 .The cell exhibited a maximum power density of 1.3 mW cm -2 at a load of 740 Ω and delivered 8 times higher energy output and >3 times longer lifetime than the static Mg-Air cell designs due to the abundant supply of dissolved oxygen to the cell during every eye blinking cycle.A summary table comparing the performance of the discussed batteries is shown in Table 2.

Biofluid-Activated Supercapacitors
Supercapacitors (SCs) are a type of electrochemical energy storage device that have several advantages over batteries including faster charge-discharge rates, lower internal resistance, higher power density, and better cycling stability.Unlike batteries, which store electrical energy in the form of chemical energy, SCs store energy through electrostatic charge buildup without the need for chemical phase or composition changes.This feature results in a high degree of cyclability for SCs.However, some SCs based on pseudocapacitance can exhibit Faradic behavior like that of a battery. [129]Furthermore, nanostructured electrode materials can increase the electrochemical double-layer capacitance, and introducing redox molecules, polymers, and conductive materials can increase the pseudocapacitive behavior. [130]Conventional SCs have commonly employed toxic electrolytes, which as discussed can pose safety risks when integrated into wearable or implantable systems and necessitate special packaging to prevent electrolyte leakage.In contrast, biofluid-activated SCs utilize biological fluids such as sweat, subcutaneous fluid, and blood as biocompatible electrolytes.In this section, we discuss emerging biofluidactivated SCs and their chemistries, materials engineering, and applications.

Sweat
A sweat-activated, wearable SC was demonstrated using two polyanilines (PANI)/CNT electrodes by pasting one of the electrodes on Zn foil to provide chemical energy and another electrode on filter paper (Figure 4a). [131]Polyvinyl alcohol (PVA) gel electrolyte was used to prepare the quasi-solid-state SC by partially immersing the electrodes.The galvanostatic chargedischarge (GCD) curves for 3.0 m and 1.0 m NaCl showed capacitance of 112.1 mF cm −2 and 110.5 mF cm −2 , respectively and the cyclic test showed a decrease in discharge time from 120 s (cycle 1) to 50 s (cycle 2000) while the GCD curve remained symmetric.The high capacitance of the PANI/CNT SC was mainly attributed to the pseudocapacitance of PANI and the increased surface area of CNTs formed during PANI in situ polymerization which resulted in CNTs with rougher surface and increased adsorption sites for the electrolyte ions.The self-charging performance of the sweat-activated SC was demonstrated by integration onto a watchband where it charged the watch to 3.2 V.In addition to excellent electrical performance, the SC showed robust mechanical behavior in static and cyclic bending tests.Integrating SCs into textiles offers numerous benefits, including the ability for clothes to produce energy without the need for separate wearable energy devices.Additionally, textiles have a much larger surface area of contact with the skin, allowing for larger SC that can capture sweat from various regions of the skin.Manjakkal et al. presented a flexible SC that used a polyester cellulose cloth as substrate and dimethyl sulfoxide (DMSO) doped and key materials necessary for developing it (bottom).The separator-free supercapacitor consists of polyaniline/carbon nanotube (PANI/CNT) electrodes and electrolyte gel.Reproduced with permission. [131]Copyright 2021, Elsevier.b) Photograph of a wearable, flexible supercapacitor for self-powered smart textiles (top left) and mechanism of the PEDOT:PSS-coated cellulose cloth-based separator that absorbs sweat as an electrolyte (top right).Photograph (bottom left) and SEM images of the PEDOT:PSS-coated cloth and a magnified view showing a single thread in the inset (bottom right).Adapted with permission. [132]Copyright 2022, Wiley-VCH GmbH.c) Schematic illustration of an implantable body fluid-activated supercapacitor cell with oxidized single-walled carbon nanotubes (SWCNTs) as the electrode material (top).Photograph of the fabricated biocompatible supercapacitor device comparing its size with a dandelion head and SEM images of oxidized-SWCNTs revealing the increased amorphous carbon and sidewall defects of the nanotubes after electrochemical oxidation (bottom).Adapted with permission. [133]Copyright 2022, The Royal Society of Chemistry.d) Schematic illustration of the fabrication method for a biscrolled PEDOT:PSS/ferritin/MWCNT (PFM) yarn supercapacitor (left) and SEM image of a biscrolled PFM fiber electrode (top right); Scale bar: 50 μm.Photograph showing the implantation of PFM fiber supercapacitor into the mouse abdominal cavity (bottom right).Reproduced with permission. [135]opyright 2018, Elsevier.
PEDOT:PSS as both an active electrode and current collector and sweat as electrolyte (Figure 4b). [132]Drop-coated PEDOT:PSS has been shown to have stable adhesion to cellulose/polyester cloth due to its strong electrostatic and hydrogen bonding interactions between charged PEDOT:PSS and the hydroxyl groups of cellulose.Furthermore, the high sweat absorption rate of cloth cellulose fibers (55% cellulose and 45% polyester) in the SC successfully charged the device even at low sweat volumes (20 μL).The redox reaction in the conjugated polymers and high con-ductivity of the electrode allowed for both pseudocapacitance and electrochemical double-layer (EDL) formation to occur on the surface of PEDOT:PSS electrode.As shown in Figure 4b, as cations from the sweat electrolyte (Na + or K + ) enter the PE-DOT:PSS channel, the oxidized PEDOT + reduces to its natural state by ion exchange with the sweat electrolyte, as given by Equation (20).
The PEDOT:PSS-coated cloth-based SC with sweat electrolyte showed energy and power densities of 1.36 Wh kg −1 and 329.70 W kg −1 , respectively, for 1.31 V, and its specific capacitance was 5.65 F g −1 .Furthermore, the SC was washable and offered a charging-discharging stability of 5000 cycles, with capacitive retention of 70% and 45% after 2000 cycles and 5000 cycles, respectively.

Subcutaneous Fluids
A subcutaneous fluid-activated flexible SC was developed as an implantable energy storage system as shown in Figure 4c. [133]uckypaper comprised of oxidized CNTs was used as implantable electrodes as they can efficiently store energy, provide high electrical conductivity and chemical stability, and possess flexural strength.Furthermore, the mesoporous structure can provide a large surface area, fast mass/charge transfer kinetics, and a high abundance of active sites.Oxidized CNTs were used to make the buckypaper since electrochemical oxidation of CNTs resulted in opening nanotube caps, cleaning their surfaces, etching sidewalls, and decreasing carbonaceous impurities.Furthermore, electrochemical oxidation reduced the amount of iron impurities and produced carboxylic groups on the outer walls of the CNTs thereby decreasing cytotoxicity which is critical for implantable electronic medical devices.To evaluate their long-term safety and effectiveness, electrodes made of oxidized CNTs were implanted into mice and monitored for 6 months.The cytotoxicity of the electrodes based on oxidized CNTs was found to be lower than pristine CNTs with no signs of inflammation or fibrosis.The SC when implanted in the subcutaneous region of the dorsal part of the thoracic vertebrae of healthy mice exhibited an areal capacitance of 51.3 mF cm −2 , an energy density of 7.12 mW h cm −2 , a power density of 500 mW cm −2 , and cycling stability for over 50 000 cycles.

Blood/Serum
Blood is another biofluid that has been utilized as an electrolyte in biofluid-activated SCs.Pushparaj et al. presented a nanocomposite paper-based SC that used blood as the electrolyte. [134]A nanoporous CNT-embedded cellulose paper was used as electrode material due to its inherent flexibility and high porosity.The blood-activated SC showed a specific capacitance of 18 F g −1 .Another blood-activated SC described by Sim et al. was developed as an in vivo energy storage device that worked in blood vessels as shown in Figure 4d. [135]A flexible, fiber-type SC was comprised of PEDOT:PSS/ferritin nanoclusters trapped within MWCNT sheets.Here the ferritin served as the energy material while PEDOT:PSS and MWCNTs imparted mechanical strength and electrical conductivity.Ferritin is composed of ferrihydrite and FeO(OH) nanoparticles and has good biocompatibility and physicochemical properties for in vivo applications.The PEDOT:PSS/ferritin/MWCNT (PFM) fiber-based SC exhibited an areal capacitance and areal energy density of 32.9 mF cm −2 and 0.82 μWh cm −2 respectively, in PBS solution.In human serum, an areal capacitance of 31 mF cm −2 was observed at 10 mV s −1 .When implanted into a mouse abdominal muscle the PFM fiber-based SC retained ≈90% of its initial capacitance eight days after implantation.
Another example of biofluid-activated SC involved an implantable, ultrathin protein-based bioelectrochemical capacitor (bEC) that utilized blood serum or urine as an electrolyte. [136]t comprised of rGO sheets interlayered with chemically modified mammalian proteins.rGO sheets are a promising candidate for bEC as they possess a high theoretical capacitance of ≈550 F g −1 and proteins provide pseudocapacitive behavior for charge-discharge through their heteroatom-rich nanostructures and abundance of the protonatable charged amino acid residues.Cation-derivatized bovine serum albumin was attached to the surface of rGO as a nano-spacer and a heme protein myoglobin was deposited in a layer-by-layer process to fabricate the ultrathin graphene-based bECs.The device showed a high volumetric capacitance of 655 F cm −3 at a scan rate of 100 mV s −1 in biofluids, 534 F cm −3 at a current density of 2.5 A g −1 in human urine, and 563 F cm −3 at 100 mV s −1 in calf serum.The bEC provided an energy density of 1.8 mWh which is 3-11 times higher than commercially available thin-film electrochemical capacitors.A summary table comparing the performance of the discussed SCs is included in Table 3.

Integrated Systems
Presented with the advances in materials engineering there clearly exist many options to generate energy from biofluids using BFCs, batteries, and SCs.However, each approach faces some limitations: B-ESSs offer a limited and diminishing supply of power while BFCs cannot store energy and generate low power.Therefore, to produce solutions that offer robust functionality, integrated systems must be used that leverage the benefits of biofluid-activated energy devices while circumventing their weaknesses.In this section, we discuss recent progress and prospects of approaches that range from simple electronic circuits to combinations of electrochemical energy devices with mechanical, solar, and thermoelectric energy harvesting.Table 4 describes the power requirements of several representative wearable, implantable, and ingestible electronic devices reported in recent publications in order to provide context for potential applications of these integrated systems.

Electronics Integration and Cell Arrays for Improving Performance of BFCs
Most reported BFCs struggle to produce sufficient voltages to power the intended devices, and as such, many groups have increased the voltage output of their BFCs using custom [37,137,138] or commercial DC-DC voltage boosting circuits which solve this issue in a simple manner.Boosting circuits can bring low voltage sources into the range of conventional batteries, enabling powering of common electronics.They require limited space and have been integrated into very small spaces like the inside of a slug [38] and in stretchable, flexible wearable platforms. [39,49,72]A key limitation is the power loss (≥ 20% depending on initial and desired final voltage of the BFC) associated with such circuits.
Another option to increase BFC voltage is to connect multiple cells in series.Two yarn-based GOx E-BFCs in series increased power and voltage output from 128 μW and 0.70 V to 229 μW and 1.32 V in serum. [79]Stretchable LOx BFCs printed on socks displayed increasing power output up to ≈240 μW cm −2 from sweat with 6 cells in series. [44]However, MacVittie et al. showed that for multiple PQQ-GDH E-BFCs connected in series inside one fluidic compartment (the body of a lobster in this case), fluidic short-circuiting via the surrounding tissue resulted in minimal voltage increases, which was solved by moving the E-BFCs to two separate lobsters (i.e., two separate fluidic compartments) while keeping the electrical series connection. [139]This arrangement produced sufficient power to drive an electronic watch for several hours.In the same work, 5 E-BFCs located in flow cells containing human serum (with separate fluidic flow to each) powered a commercial pacemaker drawing 90 μW and 1.4 V for 5 h.Choi's group presented a method to circumvent fluidic shorting in a single compartment by using 3D printed ingestible pill with screw threads and UV glue sealing to connect two M-BFCs in series.The researchers powered a calculator using the pill-based M-BFC. [74]arallel wiring of BFCs increases the system current instead of voltage and may offer improvements over series configurations where total current output is important.Szczupak et al. connected three E-BFCs in clams to power an electric motor, Spinal cord stimulator 1-10 mW [207]   Pacemaker 0.01-0.03mW [207]   Implantable drug delivery system 0.1 μW-1 mW [208]   Artificial urinary sphincter 0.2 mW [207]   Wearable blood pressure sensor 24 mW [209]   Wearable pulse oximeter 4.8 mW [209]   Wearable heart rate monitor 50 mW [209]   producing 800 mV, 25 μA, and 5.2 μW when the E-BFCs were arranged in series. [140]A parallel arrangement of the same E-BFCs improved the power output due to the increase in current (360 mV, 300 μA, 37 μW) but this arrangement was more sensitive to the load resistance, with output decreasing to similar power as the series wiring outside of a 900-2000 Ω range.In another study, eight implantable GOx E-BFCs in parallel produced an impressive 3.5 mW cm −2 , but the output voltage remained low (≈0.3V). [83] Jia et al. solved this issue of low voltage (0.376 V) despite boosted power (from 3 to 6 μW) in a parallel configuration of wearable textile-based E-BFCs operating in sweat with inclusion of a BQ25504 boost converter to increase the voltage to 3.2 V. [40] Shitanda et al. utilized both series and parallel wiring in a 6 × 6 array of LOx E-BFCs (5.4 cm × 11.5 cm) on paper which produced a total output of 2.4 V and 4.3 mW. [54]Their approach connected 6 E-BFCs in series and each E-BFC in parallel with 5 others.Each 6-cell series strip was fed by a single strip of paper which ensured sweat flow to the entire stack via capillary action.
The array powered a commercial activity meter drawing 1.44 mW using artificial sweat for 1.5 h.

Integrated BFC -SC Systems
Although combinations of multiple BFCs can help increase the power generated from biofluids, intermittent power generation stemming from transient fuel flow rates and concentration remains a critical problem.Therefore, B-ESSs are important additions to any resilient system which seeks to provide usable power from biofluids.
Coupling the power generation of BFCs with SCs is a promising option to manage variable power generation of the former by accumulating energy in the latter.In addition to a series configuration of M-BFCs, Gao et al. also integrated a paper-based SC. [141] A printable CNT, rGO, PEDOT:PSS hydrogel, and nickel/tin nanoparticle ink enabled a flexible screen printed design for the SC with capacitance of 9.81 mF at 10 mV s −1 scan rate.This combination charged the SC to 1.2 V in 51 s and produced 5.53 μAh at a discharge rate of 0.1 mA.Over 20 cycles the system retained 96% of its initial capacitance, demonstrating the concept for applications that require a biocompatible energy supply for short durations.
Longer-term use of such BFC-SC configurations would also be desirable, as envisioned by Lv et al. in a stretchable armband functionalized with stretchable conductive inks forming sweatlactate-powered E-BFCs and SCs. [142]In this work, silver flakes were mixed with styrene-isoprene rubber (SIS) to form the conductive ink which was resilient to repeated stretching and deformation on the textile surface and screen printable.The stretchable SC featured areal energy and power density of 17.5 μW h cm −2 and 0.4 mW cm −2 , respectively.In solutions of 1.25 mM lactate, the E-BFC charged the SC to 0.28 V in 10 min, and a maximum of 0.46 V held by the SC when the BFC was exposed to 10 mM lactate.Higher concentrations of lactate produced no further increase in SC charging due to the saturated enzyme kinetics of the E-BFC.The SC could be discharged for ≈1000 s at 50 μA.Other teams have opted for commercially available capacitors to integrate with BFCs in wearable and implantable systems. [54,72]iosupercapacitors (BSCs) are an emerging approach that integrate E-BFCs and SCs into a single device by immobilizing zymes on highly capacitive anodes and cathodes.While the anode and cathode are disconnected from an external load the underlying electrodes build up positive and negative charge from the E-BFC-catalyzed redox reactions, then once sufficient charge is accumulated, connecting the external load to the cell discharges the energy stored in the BSC.One such device was demonstrated featuring a PQQ-GDH anode and BOD cathode immobilized on flexible indium tin oxide (ITO) electrodes spray-coated with ITO nanoparticles. [143]In PBS with 50 μM glucose, the BSC was charged and discharged through an external 1 kΩ resistor for 100 cycles (17 h) over which an initial stable OCV of 0.150-0.170V (correlating to 30 μW cm −2 ) decreased by 39% by cycle 100 mainly due to consumption of the glucose.The device could also turn a low-voltage liquid crystal display ON and OFF when discharging and charging.A stretchable, wearable BSC was developed by Lv et al. harvesting energy from lactate in sweat, generating 343 μW cm −2 by individuals performing exercise trials (Figure 5a). [144]Electrodeposited polypyrrole (anode) and high surface area Pt (cathode) on printed CNT electrodes provided high capacitance to the system while LOx mediated by NQ catalyzed the oxidation of lactate.The device's power output saturated at 10 mM lactate, generating 80 μW cm −2 at 0.38 V.During human trials, sweat lactate charged the BSC to a maximum of 0.48 V after 1 hour of exercise and continued to produce ≈150 μW cm −2 of power for 4 h after removal from the body.Combining two BSCs in parallel with a DC-DC voltage boosting circuit turned on an LED (≈1.6 V turn-on voltage) in 1 s discharge/10 s charge cycles, exhibiting the fast and repeatable power generation of the system.
Storing the energy generated from a sweat BFC in an SC is useful for applications that require brief bursts of energy.To improve upon the concept, Yin et.al. instead used a stretchable, rechargeable Zn-AgCl battery connected to 6 LOx E-BFCs encapsulated in elastomeric layers to generate and store biofluid energy for hours (Figure 5b). [41]The Zn-AgCl redox chemistry is well-suited to wearable applications due to the neutral pH electrolyte (polyvinyl alcohol-polyacrylamide hydrogel was prepared and soaked with 1 m potassium chloride and 0.5 m zinc sulfate) required, and the battery displayed stretchability up to 20% and no loss of coulombic efficiency after 120 charge-discharge cycles.Compared to previously reported SC-integrated approaches, this design accumulated over 150 times the charge and could provide tens of hours of energy from 20 min of exercise.These material achievements produced wearable devices which could be worn with minimal user discomfort and could offer a seamless integration of biofluid-activated power generators into everyday life.

Integrated BFC -Biomechanical Energy Harvesting Systems
Physical motion is also a commonality in most human experiences, and a host of devices are being explored which scavenge energy from basic human movements. [145,146]Wang's group has recently developed a long sleeve shirt-based system that incorporates triboelectric generators (TEG) with E-BFCs, SCs, and custom electronic circuits to quickly generate and sustain power from the human body without external inputs (Figure 5c). [39]EGs generate energy from electrostatic charge buildup induced by mechanical friction between two surfaces of differing roughness. [147,148]Each component (TEG, E-BFC, SC) was fabricated using distinct flexible conductive inks in a wearable form factor that could withstand the rigors of physical activity while meeting the goal of multi-modal, biofluid-activated energy production.Initial electronics start-up energy demands were met by the high voltage output (160 V) of the TEG that depended on the swinging motion of the negative mover patches on the forearm section of the sleeves of the shirt moving against the positive stator sections on the waist of the shirt.Once sweating began, a flexible "island-bridge" style LOx E-BFC patch adapted from earlier work placed on the inside of the shirt in the chest region provided power from sweat lactate even after movement stopped. [72]Utilizing these material innovations to produce flexible electrochemical devices enabled energy production from daily human activities and offers a promising route to producing wearable energy harvesting systems.These energy harvesters working in concert charged a SC and the whole system was able to power a wristwatch with a liquid crystal display (power in the μW range) or electrochromic display (power in the mW range) for over 30 minutes after a single 10 min activity session.
Despite this impressive performance, a system based on physical activity to generate power offers little use in the more sedentary portions of an average person's day such as office work and sleep.Therefore, Yin et al. proposed a biofluid-biomechanical system that produced power with essentially no physical input from the wearer by placing small, flexible E-BFCs on a wearer's fingertips to utilize the relatively high local sweat rates. [149]A PVA hydrogel layer reduced the Laplace pressure of sweat droplets on the fingertip, facilitating continuous transport of sweat to flexible, porous CNT foam E-BFCs.Integrating a lead zirconate titanate piezoelectric generator enabled further energy generation from the mechanical energy of a finger press.The PVA layer also protected the E-BFC from mechanical damage during repeated presses.This bifunctional device was able to produce 28.4 mJ from 1 hour of typing, 389 mJ during 10 h of sleep, and could power potentiometric and amperometric sensing systems with included electrochromic displays.Photograph of the on-body BSC-powered LED during and after the pulsed discharge (bottom right).Reproduced with permission. [144]Copyright 2021, Wiley-VCH GmbH.b) Photographs of a bioenergy-based E-skin energy patch on the arm (top left) and illustration of the integrated system composed of two bioenergy harvesting-storage modules and sweat sensors interconnected via microfluidics (top right).Photograph showing the integrated E-skin wearable patch (bottom left); Scale bars, 5 mm.On-body testing of the bioenergy module harvesting energy from 20 min exercise followed by discharging the battery (bottom right).Reproduced with permission. [41]Copyright 2022, Wiley-VCH GmbH.c) Photographs illustrating a multi-modal integrated microgrid system on a shirt consisting of triboelectric generators (TEGs) on the side of torso, a SC module at the chest area, and E-BFC modules inside the shirt for direct contact with sweat (top).Schematic and photograph, showing key components of SC and its working principle (bottom left).On-body charging of supercapacitors with 75 μF from 0 V in a 30-min exercise session (bottom right).Reproduced with permission. [39]Copyright 2021, Springer Nature.d) Schematic illustration of an implantable electronic medical device with biocompatible electrodes wrapped in a cellulose separation film and integrated with a solar cell as an energy harvesting device (top).Photograph of the implanted biocompatible SC in a rat model and solar cell (bottom left).Solar charging and discharging properties of the supercapacitor at 2 mA (bottom right).Adapted with permission. [154]Copyright 2017, Elsevier.

Integrated SC -Battery Systems
ESSs are often charged using an electrical grid, which may not be readily available in outdoor or remote settings.To address this challenge, researchers are exploring the integration of SCs with other generators such as piezo-generators, [150] tribogenerators, [151] and thermo-generators. [152,153]However, many of these hybrid devices require complex equipment and configurations, making their self-charging operation impractical for wearable devices.Consequently, there is a high demand for energy harvesting-storage hybrid devices that can be effectively selfcharged in a wearable mode with little dependence on the usage scenario.In this regard, a flexible, on-skin, sweat-chargeable, hybrid device was developed by integrating a separator-free SC (energy storage unit) and a Zn-air cell (energy generator). [131]The integrated system utilized a biocompatible NaCl electrolyte and freestanding in situ polymerized, high-conductivity PANI/CNT electrodes.In contrast to prior examples that relied on hard wiring different energy source components, in this work, the researchers demonstrated a seamlessly integrated system.The right part of the laser-cut PANI/CNT electrodes were coated with electrolyte gel to form a separator-free SC for energy storage, while the middle part was covered with scotch tape.On the left part of the first electrode, a piece of filter paper was attached, while on the left part of second electrode, a Zn foil was added using silver paste to form a Zn-Air cell.Finally, the two electrodes were coated with electrolyte gel and pressed together to complete the fabrication.The Zn-Air cell showed a maximum power density of 0.696 mW cm −2 at 0.676 V in 1.0 m NaCl.Despite having a lower capacitance of 110.5 mF cm −2 compared to a conventional SC, the biocompatible SC offered a wide potential window ranging from 0.8 V to 1.4 V.

Integrated of SC -Photovoltaics Systems
Integration of biofluid-activated SCs with photovoltaics, which feature high power-to-weight ratios and can operate using only sunlight as input, would also be advantageous in scenarios without electrical grid access.In this context, a novel energy device for implanted medical devices comprised of a subcutaneous fluidactivated SC and a solar cell was developed (Figure 5d). [154]The system consisted of two biocompatible electrodes, an MWCNT sheet decorated with manganese dioxide nanoparticles (MnO 2 NP) was used as positive electrode to overcome the drawbacks of toxicity and limited capacitance, while phosphidated activated carbon was used as the negative electrode due to its high biocompatibility.The composite positive electrode positioned the MnO 2 NPs in a 3D conductive network formed by the MWCNTs, which helped to create a high electroactive surface area and avoid leaching of the MnO 2 NPs.A mini solar panel that offered a voltage range of up to 1.0 V and a current supply of ≈2 mA was used to charge the SC implanted in the body.The in vivo implanted (rat model) SC maintained a stable potential window ranging from 0.2 to 1.0 V at a current of 2 mA and showed 99% cycling stability over 1000 cycles at a constant current of 10 mA.

Grand Challenges and Potential Solutions
Despite impressive innovations in biofluid-activated energy harvesting and storage systems, providing sufficient energy to stably power electronics inside and on the body remains a challenge.For implanted scenarios, BFCs and B-ESSs require further work to improve their areal and volumetric energy density to enable powering a wide range of energy-intensive electronics; improve the stability of electrochemical reactions in vivo; and address the deleterious effects of biofouling.A recent review of the challenges faced by E-BFCs compiles a host of specific reviews which discuss in detail the various challenges faced by these devices and specific strategies to immobilize enzymes, design electrode materials, and tailor redox mediation. [155]he interaction between electrode surface and enzyme molecules which drives E-BFC operation is complex and requires careful design to maximize performance.Primary challenges facing E-BFCs are providing a suitable environment for enzymatic catalysis, protection from interfering species present in biofluids, optimization of functional enzyme amounts on device, and stability of power output.The characteristics of the host electrode material will control catalysis in several ways.First, the surface area of the material will define the amount of enzyme which can be physically loaded on the surface of the electrode and participate in electrochemical charge transfer. [156,157]][159] The effect of pore size on performance of bioelectrodes has been the subject of several studies, and the consensus is that electrodes with pores larger than the enzyme produce higher catalytic currents by maximizing enzyme loading and allowing high diffusion of fuel. [93,160,161]lectrostatic and hydrophobic/hydrophilic interactions between enzyme and electrode surface will dictate the enzyme's orientation relative to the electrode.In systems using DET, this orientation is critical to bring the active site within 1.5 nm of the conductive electrode material to enable electron transfer. [66]-ray crystallographic structures provide the 3D structure of a given enzyme, which in turn determines the surface charge and dipole moment. [162]By tuning the surface charge of the electrode via functional groups, the ensuing electrostatic interactions between enzyme and electrode can be tailored to bring the active site close to the electrode surface and optimize charge transfer.For example, BOD from Myrothecium verrucaria exhibited improved rates of catalysis on electrodes functionalized with positively charged pyrenemethylamine-modified multiwalled carbon nanotubes (MWCNTs). [163]Hydrophobic interactions between enzyme surface and electrode functional groups can similarly improve catalysis through optimal active site orientation, as seen with aromatic docking sites coupling with the hydrophobic pocket surrounding the active site of laccase. [164]espite attempts to optimize the electrode for electronic communication with the enzymes on the surface, the real amount of enzyme which can participate in electrochemical reactions with the electrode will likely be less than the amount applied.A detailed procedure for this evaluation was presented by Mazurenko et al. in which adsorption isotherms and noncatalytic currents were used to evaluate the real number of electroactive enzymes in a DET system. [165]As previously discussed, high surface area electrodes and particularly porous materials will result in more surface area contact between enzyme and electrode, thereby increasing the likelihood of active site proximity to the electrode surface.Further optimization of the electroactive fraction of enzymes participating in DET can be obtained by accounting for the dipole moment of the enzyme of interest, and use of oppositely charged surface groups to favor electrostatic attraction between the enzyme active site and electrode surface. [163]nterference from species present in biofluids can impact the enzymatic catalysis of E-BFCs.Chloride ions (Cl − ) bind with the T1 active site of laccase and BOD enzymes, in one example 100 mM Cl − caused a 30% decrease in catalytic current. [166]This inhibition can be avoided by protecting enzymes in nanostructured electrode features [167] or through electrostatic repulsion of Cl − by redox polymers. [168,169]Laccase is also inhibited at pH ≥ 7, the mechanism is the subject of ongoing research. [60,170,171]The issue of laccase pH inhibition can be overcome using engineered enzymes [172][173][174] or by creating localized acidic pH regions around the enzyme. [175]On the other hand, BOD shares a similar structure to laccase but its activity is less sensitive to pH. [60] Electrooxidizable chemicals in biofluids such as ascorbate, uric acid, dopamine, cysteine, and acetaminophen can interfere with E-BFC operation due to their ease of oxidation at conductive surfaces. [73,176]Covering the exposed conductive surfaces with negatively charged polymers or a dense coating of enzymes are possible strategies to avoid these nonspecific oxidation reactions. [73]The small microenvironment of implanted single-compartment E-BFCs can cause self-poisoning interference from H 2 O 2 and other reactive oxygen species.Commonly used in E-BFCs, GOx and LOx both produce H 2 O 2 as a byproduct during oxidation of glucose and lactate respectively, which has a detrimental effect on surrounding tissue [177] and the catalysis of BOD and laccase enzymes. [178]Strategies to mitigate H 2 O 2 inhibition include its decomposition by catalase or peroxidase enzymes co-immobilized in GOx and LOx E-BFCs [69,78,179] or utilization of alternative anode enzymes which do not use O 2 for catalysis such as glucose dehydrogenase [180] or cellobiose dehydrogenase. [181]ther more general challenges are shared by BFCs and B-ESSs alike.One widespread strategy to improve power output of BFCs involves maximizing the electrode surface area to accommodate high reagent loading while providing a conducive environment for the enzymes that are responsible for energy generation.In this context, new nanomaterials and functional polymers are currently being explored.SCs and batteries may benefit from similar approaches given their need for high electroactive surface area to store large amounts of charge and chemical energy, respectively.
Degradation and leaching of the energy materials resulting in low operational lifespan are major challenges that affect energy devices.These issues are especially pronounced in biofluid-activated energy devices due to their direct contact with the surrounding biofluids which can cause the energy materials to rapidly leach.Moreover, ions, electrooxidizable compounds, and pH shifts in the biofluids can lead to expedited degradation of the energy materials.Advanced immobilization strategies that effectively mitigate leaching and engineered surface coatings that reduce exposure of the energy materials to harmful chemicals present in the biofluids are some methods that could help resolve these issues.E-BFCs in particular suffer from instability (decreasing power output over time) caused by leaching of enzyme and redox mediators. [182]mprovements in enzyme stability are obtained by decreasing their conformational flexibility [183] through covalent bonding, surface modification, crosslinking, and entrapment in polymers or sol-gel matrixes [184,185] or by nanostructuring electrode surfaces. [186,187]Covalent bonding is often achieved via 1-ethyl-3(3-(dimethylamino)propyl)carbodiimide (EDC) linker and creates a covalent bond between amino and carboxylic groups, strongly linking enzyme and electrode to reduce device instability to a certain extent. [188]Modifying the electrode surface with charged or hydrophilic/hydrophobic functional groups increases the electrostatic or hydrophobic attraction between electrode and enzyme and controls the orientation of the enzyme, thereby improving catalysis. [189]Cross-linking using small bidirectional molecules such as glutaraldehyde is a simple, commonly used approach to form rigid enzyme aggregates which exhibit high stability. [190]Physical entrapment in polymers or solgel matrixes can effectively mimic the natural environments of enzymes, [191,192] but typically requires addition of some conductive filler to transfer charge through the otherwise insulating material. [193]Hierarchically porous electrode surfaces featuring pores of similar size as the immobilized enzyme create protective microenvironments that improve stability by rigidification of the enzyme structure and shielding from bulk solution pH changes. [93,194]iofouling is another important challenge, especially in the case of implanted systems, that significantly reduces the performance of biofluid-activated energy devices.Similar to the degradation and leaching challenge discussed above, biofluid-activated energy devices are more prone to biofouling-induced degradation of performance than their conventional counterparts.This is again due to their open design in which the biofluids come in direct contact with the energy materials.Over time, implanted devices are subjected to an inflammatory response from the host's body-first protein absorption on the time scale of seconds followed by cellular infiltration over several hours, then tissue repair cells after 1-2 weeks, and finally collagen-dominated fibrous encapsulation and granulation tissue formation after 3-4 weeks. [195]hysical impediment of biofluid flow resulting from thickening and densification of the scar tissue on the device deteriorates performance of biofluid-activated energy devices.Additional decrement in performance of implanted BFCs occurs due to consumption of glucose and oxygen by the white blood cells associated with the inflammation response. [196]Since no long-term in vivo studies have been performed with B-ESSs, little is known about how this inflammation process may impact them.Selection of biocompatible materials in all aspects of the design and specially engineered coatings may limit initial protein adhesion which has been shown to reduce the long-term buildup of scar tissue. [197,198]ore work should investigate zwitterionic polymers which have been shown to offer antifouling properties, [199] and judiciously designed nanostructured surfaces that can deter cell adhesion without any chemical modifications. [200]xternally located BFCs and B-ESSs operating in sweat, saliva, tears, and urine face additional challenges arising from variations in ambient conditions, and temporal changes in supply and composition of biofluid.For example, temperature, humidity, and solar radiation will impact device operation through changes in energy materials activity (especially in the case of enzymes) and evaporation rates of fuel and electrolytes.Researchers must therefore design strategies to mitigate the effects of these factors.[203][204][205] Moreover, integrated systems that combine BFCs with B-ESSs could further assist in offering a continuous supply of stable power even under a variable supply of biofluids.Fluctuating pH of unbuffered biofluids such as sweat, tears, saliva, and urine is another major challenge that researchers must take into account while developing biofluid-activated energy devices since enzymatic activity as well as battery chemistry are dependent on pH.Incorporating microfluidics or hydrogels with buffering capacity may correct this issue. [205]

Concluding Remarks
This review has surveyed advancements in the field of biofluidactivated energy systems including BFCs, batteries, SCs, and multi-modal integrated systems featuring combinations of energy generating and energy storing systems operating in blood, serum, extracellular fluid, subcutaneous fluid, GI fluid, sweat, saliva, tears, and urine.The technologies detailed within use materials innovations to produce systems that link electrical to biological systems in exciting new ways and promise to facilitate further integration of electronic devices with the human body by meeting fundamental needs for power.As further work explores the capabilities of biofluids to power the devices which operate in their proximity, remaining challenges of increased and stable power output as well as biofluid-specific tailoring to account for the complex biological environments are expected to be met.Progressing the field of biofluid-activated energy devices will have a broad positive impact on personalized healthcare.

Figure 1 .
Figure 1.A schematic representation of the biofluid-activated energy devices covered in this review article which include i) energy storage systems (batteries and supercapacitors); ii) energy harvesting systems (biofuel cells); and iii) integrated systems (combined energy storage and harvesting systems).Biofluids such as sweat, tears, gastric intestinal fluid, blood, urine, and saliva are composed of several ions and metabolites which render them as viable electrolytes and fuel sources for generating electricity.

Figure 2 .
Figure 2. Biofluid-activated biofuel cells.a) Illustration of flexible, fiber-based enzymatic biofuel cell (E-BFC) and corresponding reaction mechanism (top).SEM images of CNT fiber-based cathode (top) and anode (bottom) with scale bars equal to 30 μm (bottom left) and a photograph showing the implantation of the fiber-based E-BFC into a mouse brain (bottom right); scale bar: 1 cm.Reproduced with permission.[50]Copyright 2022, Wiley-VCH GmbH.b) Schematic, photograph of flexible wearable sweat E-BFC when mounted onto a human arm and closeup photographs of device interfaced with skin (top); scale bar: 1 cm.Schematic showing the E-BFC array consisting of lactate oxidase-modified bioanodes and Pt alloy nanoparticle-modified cathodes (bottom).Reproduced with permission.[49]Copyright 2022, The American Association for the Advancement of Science (AAAS).c) Schematic illustrations of bacteria-powered biobattery activated by human sweat (top).Schematic illustration of sweat-activated microbial biofuel cell (M-BFC) layers above skin (bottom left) and photograph of experimental setup of M-BFC on a human hand powering a thermometer (bottom right).Reproduced with permission.[95]Copyright 2020, Elsevier.d) Schematic illustration of ingestible E-BFC powered capsule operating in a porcine model (top).Illustration showing the corresponding key components of the E-BFC and the enzymatic reactions responsible for energy generation (bottom left).Photograph of the ingestible E-BFC powered capsule for glucose sensing and wireless data transmission to an external receiver (bottom right).Reproduced with permission.[37]Copyright 2022, Springer Nature.

Figure 3 .
Figure 3. Biofluid-activated batteries.a) Exploded view illustration of sweat-activated cell (SAC) showing the essential components (left), photograph of the device (top right), and the device mounted on a human chest to wirelessly capture real-time heart rate signals (bottom right); scale bar: 1 cm.Reproduced with permission.[117]Copyright 2020, Springer Nature.b) Schematic illustration of the subcutaneous fluid-activated biodegradable cell showing key components (top), photographs of the device showing front and side view (bottom left), and photographs of the implanted cell at the sciatic nerve of the rat (bottom right); scale bar: 5 mm.Reproduced with permission.[124]Copyright 2020, AAAS.c) Exploded view illustration of dual electrolyte magnesium-iodine (Mg-I 2 ) cell with constituent materials (top left) and schematic presenting the chemical reactions at the Mg anode and I 2 cathode when immersed in biofluid (top right).Schematic diagram of the in vivo experimental setup for the evaluation of implanted dual-electrolyte cells in mice (bottom left) and optical image of the stacked cell implanted inside a mouse (bottom right).Reproduced with permission.[125]Copyright 2022, The Royal Society of Chemistry.d) Illustration of a tear-activated cell embedded in a smart contact lens showing the corresponding porous electrodes (top).Photograph of the cell on an artificial eye and an optical microscope image showing the cross-sectional view of the cell (bottom).Reproduced with permission.[127]Copyright 2021, American Chemical Society.

Figure 4 .
Figure 4. Biofluid-activated supercapacitors.a) Illustrations of a sweat-activated supercapacitor(top) and key materials necessary for developing it (bottom).The separator-free supercapacitor consists of polyaniline/carbon nanotube (PANI/CNT) electrodes and electrolyte gel.Reproduced with permission.[131]Copyright 2021, Elsevier.b) Photograph of a wearable, flexible supercapacitor for self-powered smart textiles (top left) and mechanism of the PEDOT:PSS-coated cellulose cloth-based separator that absorbs sweat as an electrolyte (top right).Photograph (bottom left) and SEM images of the PEDOT:PSS-coated cloth and a magnified view showing a single thread in the inset (bottom right).Adapted with permission.[132]Copyright 2022, Wiley-VCH GmbH.c) Schematic illustration of an implantable body fluid-activated supercapacitor cell with oxidized single-walled carbon nanotubes (SWCNTs) as the electrode material (top).Photograph of the fabricated biocompatible supercapacitor device comparing its size with a dandelion head and SEM images of oxidized-SWCNTs revealing the increased amorphous carbon and sidewall defects of the nanotubes after electrochemical oxidation (bottom).Adapted with permission.[133]Copyright 2022, The Royal Society of Chemistry.d) Schematic illustration of the fabrication method for a biscrolled PEDOT:PSS/ferritin/MWCNT (PFM) yarn supercapacitor (left) and SEM image of a biscrolled PFM fiber electrode (top right); Scale bar: 50 μm.Photograph showing the implantation of PFM fiber supercapacitor into the mouse abdominal cavity (bottom right).Reproduced with permission.[135]Copyright 2018, Elsevier.

Figure 5 .
Figure 5. Integrated systems.a) Schematic representation and photograph of skin-mounted, sweat-powered, stretchable biosupercapacitor (BSC) patch (top); scale bar: 2 cm.Schematic presenting the key materials of the device and mechanism of energy generation from sweat lactate and capacitive energy storage (bottom left).Photograph of the on-body BSC-powered LED during and after the pulsed discharge (bottom right).Reproduced with permission.[144]Copyright 2021, Wiley-VCH GmbH.b) Photographs of a bioenergy-based E-skin energy patch on the arm (top left) and illustration of the integrated system composed of two bioenergy harvesting-storage modules and sweat sensors interconnected via microfluidics (top right).Photograph showing the integrated E-skin wearable patch (bottom left); Scale bars, 5 mm.On-body testing of the bioenergy module harvesting energy from 20 min exercise followed by discharging the battery (bottom right).Reproduced with permission.[41]Copyright 2022, Wiley-VCH GmbH.c) Photographs illustrating a multi-modal integrated microgrid system on a shirt consisting of triboelectric generators (TEGs) on the side of torso, a SC module at the chest area, and E-BFC modules inside the shirt for direct contact with sweat (top).Schematic and photograph, showing key components of SC and its working principle (bottom left).On-body charging of supercapacitors with 75 μF from 0 V in a 30-min exercise session (bottom right).Reproduced with permission.[39]Copyright 2021, Springer Nature.d) Schematic illustration of an implantable electronic medical device with biocompatible electrodes wrapped in a cellulose separation film and integrated with a solar cell as an energy harvesting device (top).Photograph of the implanted biocompatible SC in a rat model and solar cell (bottom left).Solar charging and discharging properties of the supercapacitor at 2 mA (bottom right).Adapted with permission.[154]Copyright 2017, Elsevier.

Table 2 .
Performance comparisons of various biofluid-activated batteries.

Table 4 .
Power requirements for representative wearable, implantable, and ingestible electronic devices.