The motivation behind the work is stark. The world generates roughly 50 million tons of electronic waste every year, and according to the United Nations Environment Programme that figure is climbing at about 5 percent annually. The 2024 Global E-Waste Monitor found that the volume of e-waste actually recycled remains far below the volume produced, even though the recoverable value locked inside discarded electronics was estimated at 57 billion US dollars in 2020 and is projected to reach 65.8 billion dollars by 2026. Beyond the lost resources, e-waste leaches heavy metals and persistent organic pollutants into soil and aquatic ecosystems, where they bioaccumulate and threaten human health. For the researchers behind the new material, the answer was not better recycling of toxic components but a redesign of the components themselves.

Piezoelectricity—the ability of certain materials to generate electrical charge when mechanically deformed—has long been attractive for self-powered sensing because it eliminates the need for batteries and external power sources. Piezoresistive and piezocapacitive sensors can be highly sensitive, but they demand continuous external power, complicating system design. Triboelectric generators also harvest mechanical energy but typically require intricate microstructural engineering to perform well. Natural, biodegradable piezoelectric materials such as collagen, chitosan, fish swim bladder, and the amino acid crystal γ-glycine have been explored, yet their piezoelectric coefficients fall substantially below those of ferroelectric ceramics, limiting their usefulness in real sensing applications. The new composite was designed to close exactly that gap.

The key insight lies in pairing two environmentally benign ingredients whose strengths compensate for each other’s weaknesses. The first is TEMPO-oxidized cellulose nanofiber, or TOCNF, extracted from bamboo biomass. Cellulose is renewable, biodegradable, and intrinsically piezoelectric thanks to its non-centrosymmetric crystal structure, but its piezoelectric coefficient of roughly 0.4 pC·N⁻¹ is far too weak for practical devices. The second ingredient is 2,2,3,3,4,4-hexafluoropentane-1,5-diol, abbreviated HFPD, a molecular ferroelectric crystal whose asymmetric distribution of carbon–fluorine bonds creates a strong molecular dipole and a much larger piezoelectric coefficient of about 138 pC·N⁻¹. Pure HFPD, however, is a brittle crystal that cannot be processed into flexible devices on its own. Combining the two in water and letting the solvent evaporate produces a flexible, robust film in which the crystal’s high piezoelectric activity is effectively transferred into the cellulose matrix.

The chemistry that makes this transfer possible is a dense, multilevel hydrogen-bonding network. Carboxylate groups introduced onto the cellulose nanofibers through TEMPO-mediated oxidation interact strongly with the hydroxyl groups of HFPD, creating a hierarchically ordered structure that channels mechanical stress efficiently and keeps molecular dipoles aligned. Density functional theory calculations quantified the strength of this partnership: the calculated binding energy for the HFPD/TOCNF pair was −25.89 kcal·mol⁻¹, considerably stronger than the interactions within HFPD alone (−6.28 kcal·mol⁻¹) or between cellulose fibers themselves (−19.28 kcal·mol⁻¹). Spectroscopic evidence reinforced the picture. Variable-temperature infrared measurements showed a band at about 3250 cm⁻¹ corresponding to strong, ordered hydrogen bonds within two-dimensional HFPD layers, which progressively vanished as the material was heated past its melting point—an order-to-disorder transition that mirrors the crystal’s role in the composite. Raman mapping revealed that the ratio of strong to weak hydrogen bonds jumped from 0.62 in pristine cellulose to 1.67 in the composite.

Composition proved critical. At an optimal HFPD-to-TOCNF mass ratio of 2:1, the film achieved its maximum piezoelectric coefficient of 14.0 ± 1.5 pC·N⁻¹, a more than fivefold improvement over pure cellulose nanofiber films, which measured just 2.55 pC·N⁻¹. Push the HFPD content higher, however, and performance drops. Differential scanning calorimetry revealed a telltale double peak at high loadings, indicating that some HFPD crystallizes outside the cellulose scaffold with random molecular orientations. Those misaligned dipoles cancel one another out, and the team observed a 32 percent reduction in open-circuit voltage at excessive HFPD levels, alongside increased brittleness. The result is a tunable system in which mechanical flexibility and electromechanical output can be balanced simply by adjusting the recipe.

Assembled into a sandwich-structured nanogenerator with copper electrodes and a polylactic acid encapsulation layer, the composite delivered an open-circuit voltage of 8.06 volts and a short-circuit current of 195 nanoamperes under 100 kilopascals of compressive stress—a 91.6 percent improvement in voltage and roughly a fourfold boost in current over pristine cellulose films. Maximum output power reached 1.52 microwatts, corresponding to a power density of 3.80 milliwatts per square meter. The device also proved remarkably durable, retaining 94 percent of its initial mechanical strength and 97 percent of its piezoelectric performance after 100,000 bending cycles. In capacitor-charging tests, the generator brought a 1.0-microfarad capacitor to 2.57 volts within about 400 seconds, demonstrating its potential as a sustainable power source for low-energy wearable electronics.

The sensing demonstrations are where the material becomes genuinely viral-worthy. Stuck to the throat, the sensor captured the subtle muscle contractions of speech, producing distinct voltage signatures for spoken words in different languages, including Mandarin, English, and Thai. Placed near joints, it tracked wrist, elbow, knuckle, and knee flexion with angle-dependent outputs—for example, roughly 0.17, 0.45, and 0.86 volts at wrist bending angles of 30, 60, and 90 degrees—and distinguished fast from slow motion with a temporal resolution of about 0.33 seconds. The team then wired the sensor into Internet of Things platforms and taught it Morse code: short and long finger presses generated dots and dashes, allowing users to transmit emergency messages such as SOS, COLD, PAIN, and HELP. A one-dimensional convolutional neural network trained on 5,000 signal sequences classified ten Morse code letters with 98.7 percent overall accuracy, reaching perfect recognition for several letters—a striking proof of concept for assistive communication among people with speech impairments or limited mobility.

Biocompatibility results support ambitions beyond the skin. In vitro tests with mouse embryonic fibroblast cells showed viability above 100 percent even at the highest tested concentration of 10 micrograms per milliliter, with live/dead staining confirming healthy cell morphology. Miniaturized devices measuring 5 by 10 millimeters were then implanted subcutaneously in rats, where a sensor placed on the quadriceps muscle generated stable outputs exceeding 100 millivolts during limb stretching, and one on the pectoralis major muscle recorded roughly 20-millivolt signals driven by breathing alone. Tail-vein injection of the composite into mice at 80 milligrams per kilogram produced no histological abnormalities in the heart, liver, spleen, lung, or kidney at 24 or 48 hours, with no signs of inflammation, hemorrhage, or edema.

Perhaps the most consequential result is what happens at the end of the device’s life. Because HFPD is water-dispersible and the cellulose network is hydrophilic, a used film completely disintegrates in water within six hours; vacuum filtration recovers crystalline HFPD as a white powder while the filtrate retains dispersible nanocellulose for centrifugation. Regenerated films retained mechanical strength and electrical output essentially unchanged through four recycling cycles, closing the loop without performance loss. In outdoor soil burial tests conducted over 25 days, the HFPD/TOCNF films began fracturing within five days and lost nearly 100 percent of their mass by day 25, while commercial PVDF control films showed no visible degradation or measurable weight change at all. By dissolving the long-standing trade-off between electromechanical performance and lifecycle sustainability, the bamboo-based composite offers a credible template for next-generation green electronics—wearable health monitors, implantable sensors, and human-machine interfaces that work hard while they are needed and quietly return to the earth when they are not.

Subject of Research: A recyclable, biodegradable cellulose-based piezoelectric nanogenerator for self-powered biosensing and human-machine interaction

Article Title: A Recyclable and Biodegradable High‐Performance Piezoelectric Nanogenerator for Self‐Powered Biosensing and Assistive Human‐Machine Interaction

Article References: Luo, T., Xiao, T., Zhang, L., Zhang, B., Lu, S., Xiao, H., Li, J., Liu, K., Chen, C., & Wu, H. (2026). A Recyclable and Biodegradable High‐Performance Piezoelectric Nanogenerator for Self‐Powered Biosensing and Assistive Human‐Machine Interaction. Advanced Science, Article e77780. https://doi.org/10.1002/advs.77780

Image Credits: AI Generated

DOI: 10.1002/advs.77780

Keywords: piezoelectric nanogenerator, cellulose nanofibers, biodegradable electronics, e-waste, self-powered sensors, wearable devices, molecular ferroelectric, hydrogen bonding, Morse code communication, implantable sensors, recyclable materials, human-machine interface

The motivation is straightforward. Flexible and sustainable electronics has become one of the fastest-growing corners of the industry, powered by devices that are deliberately short-lived: smart packaging that reports the freshness of food, disposable environmental sensors, passive radio-frequency identification tags on retail shelves and ultra-low-cost tracking labels produced by the billion. Conventional substrate materials are a poor match for that throwaway future. FR4, the woven-glass-and-epoxy board that sits inside nearly every gadget ever made, is rigid and petrochemical-based, and PET foil is equally fossil-derived and stubbornly persistent in landfills. Paper, composed primarily of cellulose fibers, offers renewability, biodegradability, mechanical flexibility, low surface roughness and — crucially for mass manufacturing — direct compatibility with large-area printing technologies such as inkjet, screen and gravure printing. Paper electronics has already proven itself at low frequencies, notably in 13.56-megahertz HF RFID, but the application horizon now stretches toward millimeter-wave sensors, ultra-wideband antennas, wearable communications and Internet-of-Things nodes at 2.4 and 5.8 gigahertz. A circuit carrier you can print like a newspaper and compost like a newspaper is a genuinely disruptive proposition.

But at gigahertz frequencies, paper is an electrophysical minefield, and the new study confronts the problem head-on. The two parameters that decide everything in this regime are the relative permittivity, εr′, which sets how strongly a material stores electric-field energy, and the dielectric loss tangent, tanδ, which measures how much of that energy is dissipated as heat. Together they govern the impedance matching, bandwidth, quality factor and radiation efficiency of the antennas, resonators and transmission lines that would be printed on the substrate. Paper’s trouble begins with its microstructure: a porous, heterogeneous mat of cellulose fibers laced with bound and free water, trapped air and mineral fillers, all of which host frequency-dependent polarization mechanisms that push dielectric losses well above those of engineered polymers. Worse, cellulose is hygroscopic. Absorbed atmospheric moisture can drag the permittivity off target, inflate losses and even cause irreversible dimensional deformation, so that a circuit silently detunes over months of storage, handling or field deployment. Any variation in these properties — from moisture, temperature cycling or aging — shifts resonant frequencies, raises signal attenuation and erodes long-term reliability.

To interrogate the material without cutting, coating or otherwise contaminating it, the team used one of microwave metrology’s most trusted instruments: the split-cylinder resonator. The fixture is exactly what the name suggests — a cylindrical metal cavity sliced into two identical halves, with the flat specimen clamped in the axial gap between them. Inductive loop couplers weakly excite the cavity’s TE011 resonant mode, whose electric field lies in the plane of the sample, so the measurement directly probes the in-plane permittivity of the sheet. Measuring the empty cavity first allows the exact geometry and metal conductivity to be back-calculated from the resonance condition; inserting the dielectric then pulls the resonant frequency downward, and the size of that shift yields the real part of the permittivity. The loss tangent comes from the unloaded quality factor of the resonance, obtained by fitting the transmission response and correcting for conductor and coupling losses, scaled by the fraction of the cavity’s electric energy stored inside the specimen — a quantity computed by the same rigorous mode-matching solution of Maxwell’s equations that underpins modern permittivity extraction, eliminating any need for empirical calibration. The researchers used an Agilent 85072A 10-gigahertz split-cylinder resonator coupled to an Agilent PNA N5241A vector network analyzer, removed and reinserted each sample five times to average out positioning variability, and quoted measurement uncertainties of ±1.0 percent for permittivity and ±0.0001 for loss tangent. Because the fixture supports a whole ladder of TE0np modes, a single resonator can even characterize materials across a wide swath of the 1-to-30-gigahertz spectrum.

The material list reads like a guided tour of the global paper industry. The team selected seven commercially available substrate families spanning packaging, labeling and printed electronics, including two biodegradable candidates. There was Astralux Label WS, an 80-micrometer, one-side cast-coated, water-resistant, wood-free paper; Novatech Satin Matt, a 75-micrometer double-side coated, wood-free sheet; Terraprint Silk, a 65-micrometer coated, semi-matte grade made entirely from recycled fiber; and two aramid-based electrical insulation papers — the calendared classic Nomex 410 at 130 micrometers, and SynTherm YT510, a diamond-dotted meta-aramid evaluated in both 60- and 250-micrometer variants. The biodegradable contingent consisted of Cupforma Natura, a bleached virgin-fiber board with a three-layer construction tested at 184 and 214 grams per square meter, and EraCup Natural, a 370-micrometer uncoated paperboard with a sealable, water-based-dispersion inner side, designed for paper-cup manufacture and fully repulpable and recyclable. Because two families were tested in two thickness variants each, the experiment ultimately covered nine distinct substrate groups. Before any measurement, every specimen was conditioned at 23 degrees Celsius and 50 percent relative humidity for at least 48 hours, in line with the IEC 60212 standard, to establish a stable baseline.

Then came the abuse. To compress years of field deployment into weeks of laboratory time, substrates were cycled through a dual-chamber thermal shock system, a CTS TSS-70/32 in which an elevator-type carrier automatically transfers samples between separate hot and cold zones, subjecting each sheet to violent, repeated thermal contraction and expansion. A second reliability regimen exposed samples in a climatic chamber to 85 degrees Celsius and 85 percent relative humidity — the notorious “85/85” damp-heat test long used to accelerate the aging of electronic components. Crucially, the researchers resisted the temptation to measure the sheets while still damp. After exposure, samples were removed and allowed to equilibrate for 48 hours under laboratory conditions before any dielectric measurement, with no controlled re-humidification applied. That deliberate choice means the reported values capture the irreversible, retained changes in the material — the electrical scars left by heat and moisture — rather than the fleeting, reversible effect of water currently sitting in the fibers. Moisture-uptake mass was not recorded, keeping the dataset cleanly focused on post-exposure dielectric behavior rather than in-situ humidity response.

What elevates the study beyond a routine materials table is its insistence on spatial resolution. Average dielectric drift, the authors argue, conceals the very thing that kills gigahertz circuits: non-uniformity. A substrate can shift by a modest amount on average while developing local patches where the drift is far larger, and an antenna or resonator laid over such a patch will detune no matter what the datasheet average says. The team therefore combined the comparative aging program with post-stress homogeneity mapping of the dielectric properties across each sheet, supported by microstructural and mechanical observations, revealing not only how far each material’s electrical properties moved but where the degradation lodged and which patterns it retained. The authors present this pairing of accelerated thermal-shock and 85/85 aging with mapped, high-frequency dielectric data — across coated and uncoated, bleached and recycled, cellulose and aramid papers alike — as the work’s central novelty, aimed squarely at a literature gap that has left engineers designing paper circuits for gigahertz applications largely on faith, with little understanding of how the substrates endure long-term thermal and humidity stress.

The engineering stakes scale directly with frequency. At 2.4 and 5.8 gigahertz, where IoT devices operate, and in the millimeter-wave and ultra-wideband bands now being explored, even minor inconsistencies in substrate composition, surface roughness or fiber orientation translate into significant variations in dielectric response and overall circuit performance. Because the resonant frequency of a printed antenna or filter depends directly on the permittivity of the substrate beneath it, even a modest drift in εr′ can detune a narrowband link, while a rise in tanδ silently bleeds away signal power and drags down the quality factor. The measurement campaign was correspondingly scrupulous about the dominant error source: thickness, which the IPC-TM-650 standard identifies as the leading contributor to uncertainty. A Mitutoyo digital micrometer with one-micrometer resolution applied a controlled 5-to-10-newton ratchet force to every sheet, guaranteeing repeatable contact pressure without crushing the compliant paper; the thickness was recorded in the clamped state and fed directly into the permittivity extraction. Samples were cut to match the lateral dimensions of the resonator aperture, ensuring full coverage and suppressing the edge-field perturbations that would otherwise corrupt the resonance.

Beyond the immediate numbers, the study sketches a design philosophy for the coming generation of green microwave hardware. By quantifying how each paper family responds to thermal shock and tropical humidity — and by showing which regions of a sheet hold their electrical ground and which surrender it — the work gives circuit designers, packaging engineers and life-cycle analysts a common evidence base for selecting substrates and derating their designs accordingly. It also hands the emerging field of printed, biodegradable radio-frequency electronics something it has rarely possessed: a reliability methodology borrowed from the unforgiving world of component qualification and adapted to materials that began life as packaging. The authors frame the results as practical guidance for material selection, gigahertz-range circuit design and the life-cycle engineering of sustainable electronic systems — a step toward eco-efficient components that ask no compromise in electromagnetic performance for the sake of the planet. If disposable sensors and compostable antennas are to blanket the infrastructure of the coming decade, they will first have to survive heat waves, monsoons and warehouse summers. Now, at last, there is a rigorous map of how the humblest candidate material holds up — measured at 10 gigahertz, one scar at a time.

For over 20 years, Professor Choi’s dedicated research into "papertronics"—electronic circuits fabricated on biodegradable paper substrates—has sought to marry functionality with environmental responsibility. Unlike conventional electronics that persist as e-waste, transient electronics are meant to dissolve safely inside the human body or in natural settings after fulfilling their function, a property central to biomedical devices and ecological sensors. The challenge has always been the battery, the lifeblood of any electronic circuit. Traditional power sources, particularly lithium-ion batteries, pose significant biocompatibility and toxicity issues, undermining the devices’ transient promises.

Transient electronics must be bioresorbable; they require components that can dissolve or degrade without releasing harmful substances. While prior work by Choi’s team demonstrated the feasibility of biobatteries powered by electricity-generating bacteria, uncertainties remained surrounding the safety of these bacteria in natural ecosystems. The pivotal question has been whether these microorganisms could safely return to the environment without disrupting existing microbial milieus or causing health concerns. Inspired by this, Choi’s latest research took a bold new direction by employing probiotics—live beneficial microbes already proven safe for human consumption and environmental release.

Building on her groundbreaking doctoral research, Maedeh Mohammadifar developed early prototypes of dissolvable microbial fuel cells that utilized electricity-producing bacteria classified under Biosafety Level 1. This ensured the bacteria’s safe handling but did not fully assess safety post-deployment in an open environment. Addressing lingering concerns, current PhD candidate Maryam Rezaie embarked on extensive experimentation with a commercial blend of fifteen probiotic strains, seeking to verify their electrogenic capabilities. Probiotics present a unique advantage: their established safety profile avoids biosafety risks while potentially enabling electrical energy generation.

Initial results were underwhelming, reflecting the inherent challenges in coaxing probiotics to produce substantial electricity. However, the research team innovated by engineering a novel electrode surface, integrating polymers and nanoparticles to create a porous, roughened material favorable for bacterial adhesion and growth. This tailored environment enhanced electron transfer efficiency by increasing the electrode’s surface area and fostering a synergistic interface between electrogenic probiotics and the conductive substrate, a crucial step for augmenting bioelectricity production.

A further innovation involved coating the dissolvable paper’s electrode with a low pH-sensitive polymer designed to activate electricity generation only under acidic conditions, such as those found in polluted waters or the human gastrointestinal tract. This selective functionality not only optimizes power output but also ensures the battery’s operation aligns with specific biomedical or environmental scenarios, minimizing energy waste and enhancing safety. Such precision-controlled biodegradability marks a significant advancement in transient electronic design.

Although the resultant biobattery produces modest power levels relative to commercial batteries, its proven operation verifies the concept’s viability, setting a platform for future enhancements. Choi emphasizes the importance of continued research focusing on isolating probiotic strains with enhanced electric-generating genes and examining synergistic bacterial interactions that could amplify electricity output. Additionally, assembling multiple biobattery units in series or parallel configurations might exponentially improve the power supply, unlocking broader application potentials.

This marriage of microbiology and materials science promises transformational impacts on biomedical devices such as implantable sensors, environmental monitoring tools, and disposable electronics where sustainability and biocompatibility are paramount. Unlike conventional batteries, dissolvable probiotic biobatteries can safely disintegrate within living tissues or natural ecosystems, thereby eliminating hazardous waste accumulation. This breakthrough aligns with global pushes toward greener technologies and responsible innovation.

Professor Choi’s work illustrates how interdisciplinary approaches can solve longstanding technological challenges. By shifting focus from traditional toxic components to biologically benign probiotics, his team creates a paradigm shift in transient electronics—transforming power sources from environmental liabilities to eco-friendly, biocompatible assets. Though hurdles remain in scaling and optimizing these devices, the foundational research lays critical groundwork for the next generation of sustainable electronics.

Beyond fundamental scientific advances, this research holds promise for real-world applications that previously seemed unattainable. For example, ingestible medical devices powered by probiotic batteries could monitor internal health parameters in real-time, then harmlessly dissolve, averting surgical removal procedures. Similarly, environmental sensors deployed in sensitive habitats could deliver data and then biodegrade, reducing human impact on fragile ecosystems.

In summary, Professor Seokheun “Sean” Choi and his team have demonstrated a pioneering approach to powering transient electronics with dissolvable, probiotic-fueled biobatteries. Their efforts highlight the feasibility of biocompatible, environmentally benign power sources capable of safely disintegrating post-use. As transient electronics move from theory to practical deployment, innovations like these will be vital in ensuring that the electronics of tomorrow are as sustainable as they are sophisticated.

Subject of Research: Cells

Article Title: Dissolvable Probiotic-Powered Biobatteries: A Safe and Biocompatible Energy Solution for Transient Applications

News Publication Date: 26-Mar-2025

Web References:
DOI: 10.1002/smll.202502633

Image Credits: Seokheun “Sean” Choi

Keywords:
Batteries, Energy storage, Electrical engineering, Engineering, Bacteria, Microbiology, Bacteriology, Probiotics, Microorganisms