Abstract
Medication non-adherence remains a critical healthcare challenge, contributing to approximately 125,000 preventable deaths each year and incurring over $100 billion in annual costs in the United States. Current adherence assessment strategies are constrained by limited scalability, suboptimal patient acceptability, and environmental sustainability concerns. To address these limitations, we developed SAFARI (Smart Adherence via FARaday cage And Resorbable Ingestible), a bioresorbable, passive RFID system that enables precise monitoring of medication ingestion events. The platform incorporates a novel cellulose–metal particle–based radiofrequency (RF) shielding layer, effectively creating a Faraday cage that facilitates reliable signal detection. By employing fully biodegradable materials, SAFARI obviates the need for device retrieval or battery replacement, thereby mitigating electronic waste. Moreover, the tags are compatible with standard gelatin or hydroxypropyl methylcellulose (HPMC) capsules, further enhancing clinical translational potential. In vivo evaluation in swine models corroborated SAFARI’s ability to accurately detect ingestion events while demonstrating complete biodegradation following administration. These findings highlight the system’s potential to improve patient adherence tracking without introducing significant logistical burdens or ecological impact. As such, SAFARI establishes a foundational framework for the development of next-generation, eco-conscious adherence monitoring solutions and associated interventions—ultimately aiming to bolster therapeutic outcomes and reduce healthcare expenditures.
Similar content being viewed by others
Introduction
Medication non-adherence represents a critical healthcare challenge, generating over $100 billion in additional costs annually in the United States alone and contributing to approximately 125,000 preventable deaths per year. Nearly 50% of patients with chronic conditions fail to take medications as prescribed1,2. The consequences are particularly severe for certain conditions - for example, poor adherence to HIV antiretroviral therapy can accelerate disease progression to AIDS, while high adherence has been shown to reduce HIV incidence by 54%3,4,5. While the World Health Organization (WHO) endorses directly observed therapy for tuberculosis treatment (where fixed doses are administered under healthcare provider supervision), this approach is both time-intensive and costly to implement at the scale needed to address widespread non-adherence6. Similarly, medication adherence among patients with type 2 diabetes mellitus remains consistently below average7, leading to increased complication risks and healthcare costs. Effective adherence monitoring strategies can enhance therapeutic outcomes while preventing misuse of high-risk medications, such as opioids8. Recent advances in non-invasive ingestible systems offer promising solutions for tracking capsule ingestion and/or device actuation which can improve medical adhesion without addition burden to the patient or clinician9.
Current approaches to medication adherence monitoring range from indirect methods, such as patient-reported adherence, pharmacy refill data and smart pill bottles, to direct methods like ingestible electronic devices that provide real-time data and biological measurement of drug in blood, urine or hair that can indicate adherence over time10,11,12. While electronic medication tracking has the potential to improve the accuracy of adherence measurement, many existing technologies rely on external devices that require patient engagement, leading to usability challenges13,14. Among these, ingestible sensors embedded in medication capsules offer a significant advancement by autonomously verifying ingestion events12. These capsules have been studied in humans with several studies demonstrating feasibility and acceptability of using these sensors as an adherence tool15,16. However, current designs predominantly use non-degradable polymers and rigid electronic components, which are excreted from the body intact. This design raises multiple concerns: frequent ingestion of non-degradable materials increases the potential risk of gastrointestinal (GI) injury, particularly for patients on multiple medications requiring long-term monitoring, and the accumulation of electronic waste poses significant environmental challenges17,18.
The transition toward biodegradable ingestible adherence monitoring systems presents a transformative opportunity in the field. By incorporating materials that naturally degrade within the GI tract or in the environment, new systems could maintain the benefits of real-time adherence tracking while mitigating safety and environmental concerns18. Biodegradable electronics, such as transient conductive polymers and dissolvable biosensors, have demonstrated feasibility for temporary physiological monitoring and could be adapted for medication adherence tracking13,17. These advancements would reduce the risk associated with prolonged exposure to rigid electronic components, eliminate concerns about device accumulation in the GI tract, and significantly lower electronic waste from medical devices. The development of such technologies represents a critical step toward sustainable, patient-friendly adherence monitoring solutions that seamlessly integrate into routine clinical practice11.
Bioresorbable devices for wearable sensors and electronics have been extensively studied, with various implants developed. Previous examples of transient electronics include devices for physiological monitoring, wound healing, electrical stimulation, and drug delivery19,20,21. Despite the existing need, development and use of biodegradable materials for ingestible electronics has been limited. To date, only a few technologies have utilized Generally Recognized as Safe (GRAS) or biodegradable materials, and even fewer have been tested in vivo on large animal models as using the concept of ingestible capsules22,23,24. For example, gelatin-based impedance sensors have been developed to detect healthy versus damaged esophageal tissue in ex vivo disease models24. A few edible conductive pastes, conceptual batteries for ingestible electronics have also been demonstrated as proof of concept devices25,26 but demonstrations in-vivo have remained limited. Recently, an implantable GI sensor, for early detection of gastric leakage was made from bioresorbable materials27, but embodiments that are fully ingestible remain rare. The development of bioresorbable, or transient, electronics holds multiple advantages for ingestible devices: it eliminates the need for secondary surgeries to remove the capsules, reducing the risks of infection and retention, and minimizes electronic waste in the GI tract28,29.
Commonly used biodegradable materials are often functional synthetic polymers, such as poly(lactic-co-glycolic acid) (PLGA), polyanhydride (PA), Polyethylene glycol (PEG) and poly(octanediol citrate) (POC), for substrates or as encapsulation30 Despite experience with these polymers, there are still unrealized risks surrounding manufacturability, scalability and exposure of microplastic components to the user and environment given their long degradation time31. Natural polymers, in contrast, are highly promising for ingestible and transient electronics due to their low cost, ease of processing32,33, and long established safety profile. Among these, cellulose stands out as a natural, biocompatible polymer with extensive use in healthcare, including the drug industry, diabetes management, electronics and pharmaceuticals34. This interest is reflected in the global market for cellulose production in pharmaceuticals and electroceuticals is projected to reach USD 1.52 billion, underscoring its importance in the field35.
In this work, we propose a bioresorbable RFID antenna combined with a cellulose-metal particle-based RF shielding coating that blocks RF signals until the device reaches the target site, enabling an advanced use of RFID technology to measure adherence in an environmentally conscious manner. This combination forms our platform, SAFARI: Smart Adherence via FARaday cage And Resorbable Ingestible, designed for assessing medication adherence. The bioresorbable RFID tags can be easily sized to integrated into standard-sized gelatin or HPMC capsules. The cellulose-based electromagnetic shielding system that functions as a Faraday cage, effectively blocking the RF radiation. Once the coating dissolves in target organ (e.g., the stomach), the tag radiation becomes active, the system enabling precise tracking of ingestible devices. This low-cost, proof-of-concept system has been successfully demonstrated in vivo using swine models, suggesting a straightforward pathway toward clinical translation. This innovative approach opens avenues for exploring applications, such as device localization, measurement of detection range for ingestible electronics, and RFID sensing for drug delivery. Our design leverages passive RFID functionality with a zinc-based tag, enabling RF communication while using bioresorbable materials, which eliminates the need for device removal or battery replacement. Large animal studies using swine were then conducted to establish clinical readiness of the SAFARI device. This study validates our device’s ability to monitor ingestion events through passive RF technology, presenting a viable platform for next-generation bioresorbable communication systems in the GI tract and enabling healthcare providers to more effectively monitor patient adherence and health status.
Results
In this study, we conceptualize a switching mechanism by introducing an electromagnetic interference (EMI) shielding coating that encapsulates the RFID tag and the gelatin carrier capsule. To facilitate utilization of wireless medication adherence technology, Fig. 1 shows an envisioned medical use of case of ingestible RFID technology. Our bioresorbable SAFARI platform has several components: zinc foil-based RFID tag, payload and 000 capsule and EMI shielding coating. The ingestible platform consists of i) Zinc based RFID tag, ii) RFID chip, iii) 000 capsule, and payload of interest (drug, sugar, dye etc.), iv) bioresorbable EMI shielding material. The tag can be incorporated into a gelatin or HPMC (hydroxypropyl methylcellulose) capsule along with the payload, in other words medication dose. Once it is encapsulated with ingestible EMI shielding coating, the tag signal is blocked, which is the OFF state. In the “ON State”, the bioresorbable RFID tag can be queried via an external reader and used to monitor events using RFID chip data. In the event of ingestion, the EMI shielding coating dissolves and the bioresorbable device can accept queries from the external reader, confirming the ingestion. During manufacturing, the RFID tag can be loaded with information regarding drug dosage, manufacture date, serial number and any other relevant information. Following payload delivery, the healthcare professional can track the medication dose that is delivered with the capsule using an external reader. Upon full dissolution of the capsule, the device is exposed to GI fluids. The EMI coating, capsule and metal antenna components dissolve in the stomach, while the RFID chip, a small biocompatible die, remains intact, making the overall system largely bioresorbable.
A Bio-RFID capsule administration. B Shielding coating dissolution and payload release C Monitoring of the Tag ID and frequency range, recording of the payload for tracking adherence. D Dissolution and biosorption of the coating, tag and the capsule.
Bioresorbable RFID tag
The Bio-RFID tag is a compact device that is composed of thin cellulose acetate substrate (35 μm), a bio-adhesive polyglycol sebacate (PGS) to bond the substrate with zinc traces, and Impinj Monza M700 RFID chip (≈ 400 x 400 μm, −24 dBm). Fig. 2a demonstrates the schematic illustration of the device layers and characterization of the tag. The tag includes Zinc antenna bonded to cellulose acetate substrate using bioadhesive PGS, an RFID chip and PLGA encapsulation. The electrical connection between the zinc antenna and RFID chip is maintained by wire bonding and supporting it with epoxy (Supplementary Fig. 1). This Zinc based RFID tag resonates at a frequency of 915 MHz. We have performed simulations (CST Microwave Studio) using different scenarios (Supplementary Note 1). Fig. 2b describes the general operation scheme of the RFID tags. In general, a signal is sent from an RFID reader and received by the RFID tag, where the signal backscattered to the reader. In order to verify the performance and specifically to read the tag name and the operational frequency range, first we set the transmitted power to 30 dBm and the tag to reader distance to ≈20 cm (Supplementary Fig. 2). The measurement setup is based on a programmable RFID reader (Rain R700) and a circularly polarized panel antenna with a gain of 9 (dBi). The setup can adjust the emitted power up to 33 dBm (constrained by the Federal Communications Commission at 30 dBm), within the frequency range of 860–928 MHz. We have monitored the Received Singal Strength Indicator (RSSI) for the RFID tag in air, rolled in a 000-gelatin capsule and inside of an ex-vivo swine stomach. For all cases RSSI values range in between 50–65 dBm for SGF and ex vivo cases, indicating that the RFID tag has sufficient performance for continuous reading (Fig. 2c). Fig. 2d shows the FEA simulation results when the tag is bent at an angle that enables it to fit in a 000 capsule. Bending the device to 168 degrees, to match that observed during capsule loading (Supplementary Note 2), shows that the electrode structures have the highest von Mises stress N/m2 at the zinc arm, indicating that the connectivity of the RFID chip to the substrate is unaffected during loading of the capsule. Fig. 2e, f demonstrate a series of images that show the process of dissolution of a tag during immersion in RGF and SGF at 37 °C. The zinc antenna and the encapsulation dissolves in less than a week in both Simulated Gastric Fluid (SGF) and Real Gastric Fluid (RGF), while the substrate degrades over several weeks. This agrees with the degradation mechanisms of the respective materials: Zinc degrades at a rate of 120 nm/day via the chemical reaction: \(\left[{Zn}+2{H}_{2}O\to {{Zn}({OH})}_{2}+{H}_{2}\right]\)36. Although flexible substrate degradation via hydrolysis may take months, the substrate’s dimensions (10 × 15 mm) and its softening upon wetting, facilitate its passage through the GI tract. The designed absence of thick encapsulation layers or humidity barriers enables the full device to be absorbed or passed with minimal complications.
a Antenna device architecture and simulated RF performance in equivalent tissue models (S11). b Schematic representation of RFID working principle. c RSSI measurement of RFID tags in air, in gelatin capsule and ex vivo in stomach. d FEM simulation of a device in 000 capsule. e Device dissolution in real gastric fluid (RGF) and f simulated gastric fluid (SGF) at pH=1.2 and 37 °C at 25 rpm. (Scale bars, 5 mm).
Cellulose based EMI shielding coating
Following the design of our RFID tag system, we then introduced a biodegradable EMI shielding coating to block the signal prior to capsule dissolution and to enable signal activation following coating dissolution. The ink uses two components: a polymer matrix of 2-Hydroxyethyl cellulose (HEC) and bioresorbable metal fillers. Cellulose and its derivatives are excellent candidates as a binder and have been used for pharmaceutical applications, as bioplastic, in printed electronics as well as electronic skin applications37,38. In the cellulose family, HEC was chosen due to is solubility in ethanol, and viscosity to facilitate gelatin capsule coating as well as its dissolution rate in acidic environment39,40. Fig. 3a shows the steps of preparing the shielded capsules, which can be easily scaled using a dip, brush or spray coating method as needed. It is also possible to print such inks using direct ink writing using a brush enabling mass manufacturing of the coated capsules. (Supplementary Fig. 3). We determined that a viscosity of 1–300 Pa.s (8% w/v HEC to solvent) was suitable for capsule coating (Supplementary Fig. 4). Blade coating the composites at a controlled thickness produced uniform films, with the composite containing 1–5 µm Mo particles exhibit the best performance, lowest sheet resistance (0.8 Ohms/sq) at the mass ratio of 1:11 (Fig. 3b, c), followed by W particles with size of 1–5 µm. Mo and W particles were selected as Fe, Mg and Zinc show less conductivity due to formation of surface oxides36,41. When the ratio passes beyond 1 to 11, the sheet resistance increases possibly due to non-uniform dispersion of particles42. The SEM image shows a uniform distribution of binder to filler ratio for the Mo based ink (Supplementary Fig. 5). The HEC/Mo composite with the Mo size of 3–7 µm showed the lowest sheet resistance of 6.63 Ohms/sq when the ratio 1:11. Similar trends are reported in the previous works and explained using electrical percolation conductivity formula19.
a Process flow for the functional coating on 000 gelatin capsules (Scale bars, 5 mm). SEM image shows homogenous distribution of bioresorbable metal particle distribution. b Sheet resistance vs. HEC to metal particle ratio with different metal particle sizes. Inset shows printed conductive path for powering an LED (Scale bars, 10 mm). c Summary of lowest sheet resistance values vs. the filler metal type and size. d Shielding data spectrum between the frequencies of 700 MHz and 1.2 GHz. e Total Shielding Effectiveness (SE) at 915 MHz for composite thin films (40 µm) using W and Mo particles. f The homogenous distribution of the Mo particles leads to high conductivity, which results in more absorption of the incident EM waves. g Coated capsule layers and representative thickness values.
To explore the EMI shielding properties of bioresorbable composites, we then fabricated thin films of HEC/Mo and W, each having thicknesses of 40–50 µm. We utilized a pair of wide-band near-field probes with a vector network analyzer to investigate the radio frequency (RF) transmission properties of the composites. To effectively reduce environmental RF noise interference, we constructed a Faraday cage using aluminum foil to encase the probing setup (Supplementary Fig. 6). We then scanned between 700 MHz and 1.2 GHz to analyze the S21 response in the mid-field range especially at the 915 MHz (Fig. 3d). The recorded EMI SE of HEC: W and HEC: Mo with metal particle size of 1–5 µm is given in Fig. 3e. HEC with Mo shows 25 dB at 915 MHz which is higher than the W composite that of 15 dB, which indicated the HEC: Mo is a good candidate to be used in shielding application. A comparison using different metals (including foils and polymer films) at UHF band is given in Supplementary Fig. 7. Although films of films of (3 N) W, Zn and Mo materials effectively attenuate RF signals, (measured S12 parameters) demonstrating their suitability for applications requiring high shielding effectiveness. However, printable inks are required to be applied on curved substrates. Several materials specifically conductive polymers, CNT and MXene family have been proposed as a thin film shielding material for X-Band and THz range43,44,45. The importance of the UHF band is that it is known as biomedical frequency and the commercial RFID tags are mainly operating at this range46,47. The measured EMI SE for Mo particles is comparable with PEDOT based polymers48 with the added benefit of full biodegradability.
Composites using microparticles of Mo represent a better choice compared to those of other bioresorbable metals and conductive polymer due to their high conductivity, biodegradability, their resistance to the formation of thick oxides, relatively slow absorption and low cost. Application of Mo based composites on the ingestible capsules creates a continues conductive layer, preventing RF transmission to and from the interior of the capsule (Fig. 3f) until layer dissolution. Fig. 3g shows the cross-section of the capsule architecture, which reserves ≈98 % volume for loading of drug payload.
Dissolution, degradation and in vivo Studies of SAFARI
To understand the dissolution mechanics of the full system, we tested the degradability of the capsule and its components in vitro at physiological temperature (37 °C) in SGF. The EMI encapsulation coating dissolved immediately after wetting in vivo tests (10–20 min). The tag’s zinc layer and Mo coating parts disintegrated into pieces in 24 h. We have demonstrated accelerated biosorption at an elevated temperature to convey the complete picture in case of dissolution (Fig. 4a). At 75 °C (corresponds to ≈16 times relative to 37 °C)49, the device and the components dissolved and disintegrated into small particles in a week. The proposed materials here are bioresorbable due to hydrolysis or enzymatic degradation in physiological and gastric conditions also considered as edible33,36. For instance, Zn and Mo undergo hydrolysis to form oxides36. Zinc and Molybdenum react with water (\({Zn}+ 2{H}_{2}O\to {{Zn}({OH})}_{2}+{H}_{2}\)), (\({Mo}+4{H}_{2}O\to {{MoO}}_{4}^{-2}+8{H}^{+}+6{e}^{-}\)). The substrate CA is a biodegradable commercial polymer that dissolves in water, upon swelling and is compatible with laser-processed electronics50. Bioadhesive PGS is a biodegradable synthetic polymer that undergoes surface erosion and effectively used in drug delivery51. The commercial cost-effective gelatin and HPMC (hydroxypropyl methylcellulose) capsules are considered as biodegradable and depending on the pH, their dissolution time varies52. HEC acts as a binder in the coating formulation and is a cellulosic material that degrades into glucose under the effect of enzymatic degradation40. HEC exhibits consistent swelling and dissolution behavior across the physiological gastric pH range (1–6), ensuring reliable degradation of the SAFARI capsule and timely activation of the RFID signal even under interpatient variability in stomach acidity and motility53,54. We have confirmed the degradation of the HEC/Mo film through a decrease in the characteristic glycosidic bond intensity, as shown by FTIR analysis, which is consistent with previously reported degradation behavior of cellulosic materials40 (Supplementary Fig. 8). In Supplementary Table 1, the biodegradation rates and daily intake recommendation of the device ingredients are given.
a Images of dissolution of EMI shielded capsule and RFID tag in SGF (pH 1.2) solution at 37 °C (Scale bars, 10 mm). b Schematic representation of device measurement in vivo swine and endoscopic images of capsule dissolution and RFID tag under test (Scale bars, 5 mm). c X-Ray image of the ingested capsule. d RSSI recording of the devices after the EMI coating dissolution (Time of dissolution: 0.5–3 min). e Recorded frequency spectrum of the RFID devices communicated through the swine stomach. f Time dependent concentration of Zn and Mo ions in SGF immersion test by ICP-OES (n = 2). g Serum Zn and Mo concentration after the device administration in swine models (n = 3), ppm, parts per million.
To fully validate the technology, SAFARI capsules are tested in vivo. Swine models were selected due to the similarity of their GI tract size to that of humans. First, we administered the shielded 000 size capsules into the stomach. Fig. 4b shows the steps of dissolution of the device in swine stomach and data recording. Endoscopic images were taken to assess the dissolution trend of the coating and investigation of the RFID tag. After the administration of the capsule, the EMI shielding coating swells due to the cellulose content at the contact with gastric fluid (i). Then, the coating partially dissolves (ii), and the tag fully exposed for passive communication (iii). Once the tag is exposed, the panel antenna starts to record the tag data and following frequency range. The X-Ray image of the capsule in the swine stomach is shown in Fig. 4c. The image is taken immediately after the administration. In general, the tag and parts of the shielded capsule disintegrated in the stomach up to 24 h after administration (Supplementary Fig. 9), consistent with the in vitro degradation tests. To evaluate the visibility and physical footprint of the non-bioresorbable RFID chip in vivo, we performed X-ray imaging in swine, which confirmed the antenna’s visibility while the chip remained undetectable due to its sub-millimeter size (Supplementary Fig. 10). We anticipate that the RFID chip (0.16 mm²) will safely transit through the GI tract, as prior studies have demonstrated the safe passage of significantly larger ingestible capsules (diameter, 9 mm and length, 26 mm)22].
To acquire tag properties in live animals, the same measurement setup is maintained (Supplementary Fig. 11). The administrated RFID tags names are recorded prior to capsule ingestion, that simplify the data recording during the in vivo tests (Supplementary Fig. 12). Fig. 4d shows the RSSI data recorded from known tag ID, showing dB level of the different devices operating in the stomach (Supplementary Fig. 13). In addition, the frequency data of the tags during operation are recorded, while the capsule dissolves until the tag is immersed in gastric fluid (Fig. 4e). Although the devices fully immersed in gastric fluid or floating inside the stomach, the setup still is able to record the frequency range of 900–925 MHz. This shows the robust design of the proposed antenna configuration. Using the tag name, the payload along with the capsule can be recorded into a cloud system and noted. This passive way of recording the ingestion event enables electronic medication adherence and leads to a digital health application. Similar technologies, such as intrabody communication enabled-capsules for ingestible electronics, are compared in Supplementary Table 255,56.
To give more concrete insights into metal dissolution and degradation, inductive coupled plasma optical emission spectrometry (ICP-OES) analysis of the Zn antenna and Mo coating indicates material dissolution at 37 °C, 25 rpm over various time points. These findings align with previous reports conducted in PBS and are consistent with SGF immersion images57,58. The coated capsule with tag dissolves partially in 8 h, exhibiting corresponding 2 ppm Zinc and less than 1.5 ppm Mo (Fig. 4f). On Day I, peak Zn and Mo concentration reaches up to 5 ppm. The device dissolution saturates on Day III, showing 7 ppm Zinc and 3 ppm of Mo. To validate device safety, we have administrated the capsules in swine and measured the Zn and Mo levels in blood serum over time. We administered a special diet providing approximately 2000 ppm Zn and 1 ppm Mo daily. Baseline blood serum levels were measured at 0.4 ppm Zn (n = 4) and 0.05 ppm for Mo. We measured the metal levels in serum at 8 and 24 h after the capsule administration (Fig. 4g). No significant increase in the serum Zn and Mo levels was observed since the daily dietary intake of these metals exceeds the amounts present in the device (mZn = 20–25 mg, mMo = 20–30 mg). Both metals are essential trace elements with well-established metabolic handling and safety margins. The observed exposures correspond to levels previously demonstrated to be safe, tolerable and are readily cleared through normal physiological routes, consistent with prior reports of negligible tissue accumulation following molybdenum rich diet and implant degradation or high-zinc diets in animal models (Supplementary Table 3)59,60,61,62. For example, the measured serum metal concentrations align with previous swine study on high Zn load diet62. For Mo, the expectation of accumulation in organs is unlikely, as studies on Mo implants (diameter, 250 µm), in small animals have shown minimal accumulation and toxicity60. For a Mo foil-based implant (thickness, 25 µm), toxicity and Mo accumulation studies in a rodent model over 22 weeks post-implantation revealed no tissue damage, with only minimal accumulation observed in the organs19. Additionally, further research indicates that even with high Mo diet in humans, excess molybdenum is efficiently excreted by the body59.
Discussion
The work presented here introduces a bioresorbable RF device technology for ingestible electronics. The materials selection, ease of fabrication and using printing methods provide foundations for the ingestible form factor devices for monitoring ingestion, device tracking. This technological platform serves as an important demonstration of digital medication adherence, utilizing cellulose-based electronics in the biomedical device field. SAFARI is not intended for mass-market deployment but rather for targeted, high-impact clinical contexts, where medication adherence is critical to patient outcomes and public health. These include infectious diseases, such as tuberculosis, hepatitis C, and HIV, where non-adherence can promote drug resistance; transplant and cardiovascular patients requiring strict immunosuppressive or antiplatelet regimens to prevent graft rejection or thrombosis; and complex patient populations with overlapping adherence and health risks, such as individuals with HIV and substance use disorder or poorly controlled hypertension. Demonstrations of the clinical use in large mammals indicate possible transition to clinical research to enabling transform postoperative monitoring to enable digitally optimized device tracking. Although SAFARI performs ingestion-verification function as prior systems, its advantage lies in the transient EMI-shielding architecture, which improves signal specificity and device scalability. From a safety and translational perspective, the SAFARI platform’s material composition aligns with established dietary and biomedical exposure limits. The zinc antenna and molybdenum shielding are used in microgram–milligram quantities per capsule recognized to be orders of magnitude below levels associated with subclinical toxicity. Both elements are essential micronutrients with tightly regulated absorption and renal elimination. Prior animal and human studies demonstrate minimal risk of long-term accumulation or organ retention, supporting the potential for chronic administration in adherence monitoring applications59,60,63. Ongoing work will expand these findings through chronic exposure modeling and pharmacokinetic analyses to confirm safety for lifelong use. Future work includes development of battery assisted RFID systems to extend the detection range, further improving the application of the electronic medication adherence technology. Placement of the external reader in a commercial product could be either integrated into the space surrounding a patient or implemented via a wearable system, as demonstrated in prior works. Additionally, wearable reader architectures, such as chest-worn patches or belt may further enable continuous signal capture during daily activities, and their integration into patients’ routines will require user-centered design and feasibility studies to ensure reliable and acceptable long-term use. Future clinical development will include user-centered design and feasibility studies to optimize RFID reader positioning for reliable signal capture.
Methods
Reagent and Materials
Mo particles (Thermo Scientific, 3–7 µm and 1–5 µm size) and W particles (Alfa Aesar, 1–5 µm) were used for bioresorbable conductive coatings. Hydroxyethyl cellulose (HEC, Mw = 90,000), glycerol, sebacic acid, DMSO were purchased from Sigma-Aldrich. Ethanol (%99, Decon labs) and DI water were used as solvent in all ink preparations. Cellulose acetate (35 μm) or polylactic acid (PLA) (50 μm substrate, and zinc foil (25 μm) were purchased from Goodfellow Corporation. PEDOT: PSS (PH1000, 1.3 wt%) was purchased from Heraeus Clevios GmbH.
Ink Preparation
PGS bio-adhesive synthesis: PGS adhesive was synthesized using equimolar mixtures of glycerol and sebacic acid under 120 °C and N2 for 24 h for forming prepolymer. Then the prepolymer is spread on the CA substrate to maintain adhesion to zinc film.
Gelatin based adhesive: 1 g gelatin, 1 g glycerol and 8 g of DI water first kept in oven at 70 °C for 1 h. Then, mixture is mixed on a hot plate at 70 °C with 250 rpm for 2 h. After it is fully dissolved the mixture mixed for 10 min at 40 °C to be used to bond the chip to the substrate.
Biodegradable, Conductive ink Preparation: The ink was prepared by mixing HEC (0.8 g) in ethanol (10 ml) at room temperature. After having a gel, Mo and W particles were subsequently added to the mixture at different volume ratios followed by mechanical mixing using Flacktek speed mixer (at 2500 rpm for 10 min). For conductive polymer EMI shielding comparison, 10 ml PH1000 (Sigma-Aldrich) 0.1 g glycerol and DMSO (5% v/v) are added to 10 ml PH1000. The ink was vigorously mixed overnight and drop-casted in petri dishes. The other metals, W, Mo films are purchased from Goodfellow Inc.
For conductivity measurements, samples were printed using a blade coater. First, a PET (125 µm) foil is placed to coating bed, then the ink spread with blade coater (distance kept as 0.2 mm between the glass bed and the PET foil). After the solvent evaporation in ambient air, the resulting film is 8 × 8 cm, and the film thickness range is between 40–80 µm achieved. The film was peeled off from the substrate, cut to appropriate dimensions and placed on four-point probe test bench for sheet resistance, conductivity and EMI shielding measurements.
Coating of 000 Capsule: the coating experiments were performed using a brush. For brush painting of gelatin capsules, a mold is fabricated to maintain better control on deposition. Ecoflex 00-30 is mixed 1:1 ratio and poured in plastic cups and degassed. 000 capsules are then pressed in the mixture and cured overnight. After curing, the sacrificial gelatin capsule is removed, and e-pill is replaced for coating. The ink was applied first to the upper face of the capsule, after evaporation of the solvent, the other half is coated with the ink. This process is repeated several times to maintain pinhole free coating of composite ink onto gelatin capsules to create continuous layer.
Immersion test, SGF preparation: SGF was prepared by dissolving 0.2% w/v NaCl in DI water and pH 1.2 was adjusted using HCl. For RGF tests, fresh gastric fluid is collected every other 2 days and replaced in every 24 h.
Device Fabrication
CA/PGS/Zn stack is used as a base for antenna manufacturing. U4 LPKF laser system is used for patterning and cutting of RFID antenna. After patterning the zinc, unwanted parts were peeled off from the substrate and the zinc antenna on CA substrate is achieved. Impinj Monza RFID chips were first gold wire bonded from the chip connections. At the other end the gold bonding left floating. The chip with gold connections is adhered to substrate using gelatin adhesive. The floating ends of the RFID chip with gold wires placed on the zinc and connected using conductive Mo based ink. For the in vivo test to maintain a good electrical connection, a biocompatible Ag/AgCl ink is used for the connection. As a last step, PLGA (1 mg in 1 ml acetone) is drop-casted on the RFID chip and electrical contacts. The resulting device is placed in a 000 gelatin or HPLC capsule, horizontally and placed for the molds for depositing shielding material. A painting brush (size #1) is used to print bioresorbable EMI shielding inks. After one pass of printing with the brush, the coating dried for 1 h. Then, the device removed from the Ecoflex mold, and the other face of the capsule is placed in the mold. Another coating is applied to coat the entire capsule with the shielding ink.
Tag data measurement
A 902–928 MHz panel antenna (TE Connectivity, 9dBic) connected with Impinj RAIN R700 is used to communicate with the fabricated RFID tags. Impinj Itemtest software is used for capturing tag name, RSSI data and frequency. We have also validated the tag name using a handheld tag reader (TSL 1128 UHF RFID Reader) to verify the electrical connection and tag name for ease of detection during in vivo demonstrators (Supplementary Fig. 12).
EMI shielding measurement
The EMI shielding measurements were measured over the UHF band (700 MHz– 1.2 GHz) frequency by using probes that are connected to vector network analyzer (Keysight E5080B ENA). We utilized a pair of wide-band near-field probes (NFP-3 Near Field Probe Kit 30 MHz–3 GHz, RIGOL) with the VNA to investigate the RF transmission properties of various metallic shielding materials. To effectively reduce environmental RF noise interference, we constructed a Faraday cage using aluminum foil to encase the probing setups. This setup ensured that our measurements were precise and minimally impacted by external noise. Prior to dissolution of the shielding layer, no RFID signal was detected in either in vitro or in vivo settings, consistent with the effectiveness of the complete EMI shielding design.
In vivo testing
All swine studies were approved by and performed in accordance with the Committee on Animal Care at the Massachusetts Institute of Technology. In vivo studies were performed in female Yorkshire swine, n= five (4–8 months, weighing ≈50–70 kg) due to the anatomical similarity of their GI tract to that of humans. Before any procedure, the animals were given a liquid diet (Ensure, Abbott Laboratories) for 24 h, followed by an overnight fast. For capsule placement, animals were sedated using an intramuscular injection of midazolam 0.25 mg kg−1 and dexmedetomidine 0.03 mg kg−1. Following sedation, the animal was intubated, placed on isoflurane (2%) in oxygen and connected to a vital sign monitoring system and provided thermal support. The device was endoscopically delivered through an overtube to the stomach and visualized using a PENTAX EC-3870TLK (160 cm) while the animal was positioned in the left lateral. After delivering the capsule and performing endoscopic imaging, the panel antenna was positioned 10 cm from the animal to capture tag events.
Finite element analysis
3D FEA was used to simulate the zinc tag bending using commercial software COMSOL Multiphysics. In the model, Solid Mechanics Interface was used to calculate stress when the bending conditions match with the bending of the tag to place in a 000 capsule.
ICP-OES
Agilent 5100 DVD Inductively Coupled Plasma-Optical Emission Spectrometer is used to quantify the metals. For quantifying the amount of Zn and Mo in blood, samples are digested in immersion trace–grade nitric acid (HNO3; >69%; Sigma Aldrich) and then heating 37 °C overnight at 25 rpm. The addition of ultrapure H2O (18.2 megohm∙cm) yielded a final solution of 2.0% nitric acid. Zn and Mo solutions (1000 μg/ml; Inorganic Ventures, Christiansburg, VA, USA) were used to create calibration curves.
Characterization Methods
SEM was conducted by Hitachi FlexSEM TM-1000 II (Tokyo, Japan) using low voltage imaging (3–5 kV). The dissolution of the HEC polymer matrix in SGF was assessed via Fourier Transform Infrared spectroscopy (FTIR) using the Agilent Cary 630 FTIR with the MicroLab Expert Software (version: 1.1.0.0). The rheological property of the electrode ink was measured by DHR-3 Rheometer (TA Instruments. New Castle, Delaware USA). Flow sweeps were implemented to test the viscosity from 1 1/s to 100 1/s. Optical microscope images were taken by (Keyence VHX-X1 Digital Microscope). The surface roughness and step height profiles were quantitatively analyzed by laser confocal scanning using a 3D surface profilometer (Keyence VK-X3000. Osaka, Japan). Conductivity measurements were taken by Ossila Four Point Probe System. Thickness values were measured with micrometer (Mitutoyo).
Ethics Declarations
Every experiment involving animals has been carried out under a protocol approved by the Committee on Animal Care at the Massachusetts Institute of Technology.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data supporting the findings of this study are available within the article and its supplementary files. Any additional requests for information can be directed to, and will be fulfilled by, the corresponding author. Source data are provided with this paper.
References
Brown, M. T. & Bussell, J. K. Medication adherence: WHO Cares? Mayo Clin. Proc. 86, 304–314 (2011).
Lars, O. & Terrence, B. Adherence to medication. New Engl. J. Med. 126, 1196–1206 (2005).
Abdool Karim, Q. et al. Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science 329, 1168–1174 (2010).
Kirtane, A. R. et al. Development of an oral once-weekly drug delivery system for HIV antiretroviral therapy. Nat. Commun. 9, 2 (2018).
Bangsberg, D. R. et al. Adherence–resistance relationships for protease and non-nucleoside reverse transcriptase inhibitors explained by virological fitness. AIDS 20, 223 (2006).
Verma, M. et al. A gastric resident drug delivery system for prolonged gram-level dosing of tuberculosis treatment. Sci. Transl. Med. 11, eaau6267 (2019).
Al-Qerem, W., Jarab, A. S., Badinjki, M., Hyassat, D. & Qarqaz, R. Exploring variables associated with medication non-adherence in patients with type 2 diabetes mellitus. PLoS ONE 16, e0256666 (2021).
Chai, P. R. et al. Digital pills to measure opioid ingestion patterns in emergency department patients with acute fracture pain: a pilot study. J. Med. Internet Res. 19, e19 (2017).
Abramson, A. et al. An ingestible self-orienting system for oral delivery of macromolecules. Science 363, 611–615 (2019).
De Bleser, L., De Geest, S., Vandenbroeck, S., Vanhaecke, J. & Dobbels, F. How accurate are electronic monitoring devices? A laboratory study testing two devices to measure medication adherence. Sensors 10, 1652–1660 (2010).
Mason, M. et al. Technologies for medication adherence monitoring and technology assessment criteria: narrative review. JMIR mHealth uHealth 10, e35157 (2022).
Chai, P. R. et al. Utilizing an ingestible biosensor to assess real-time medication adherence. J. Med. Toxicol. 11, 439–444 (2015).
Aldeer, M., Javanmard, M. & Martin, R. P. A review of medication adherence monitoring technologies. Appl. Syst. Innov. 1, 14 (2018).
Conway, C. M. & Kelechi, T. J. Digital health for medication adherence in adult diabetes or hypertension: an integrative review. JMIR Diab. 2, e8030 (2017).
Chai, P. R. et al. DigiPrEP: a pilot trial to evaluate the feasibility, acceptability, and accuracy of a digital pill system to measure PrEP adherence in men who have sex with men who use substances. JAIDS J. Acquired Immune Defic. Syndromes 89, e5 (2022).
Browne, S. H., Behzadi, Y. & Littlewort, G. Let visuals tell the story: medication adherence in patients with type II diabetes captured by a novel ingestion sensor platform. JMIR mHealth uHealth 3, e4292 (2015).
Sharova, A. S. & Caironi, M. Sweet electronics: honey-gated complementary organic transistors and circuits operating in air. Adv. Mater. 33, 2103183 (2021).
Bettinger, C. J. Materials advances for next-generation ingestible electronic medical devices. Trends Biotechnol. 33, 575–585 (2015).
Song, J. W. et al. Bioresorbable, wireless, and battery-free system for electrotherapy and impedance sensing at wound sites. Sci. Adv. 9, eade4687 (2023).
Zhang, Y. et al. Self-powered, light-controlled, bioresorbable platforms for programmed drug delivery. Proc. Natl. Acad. Sci. USA 120, e2217734120 (2023).
Boutry, C. M. et al. Biodegradable and flexible arterial-pulse sensor for the wireless monitoring of blood flow. Nat. Biomed. Eng. 3, 47–57 (2019).
You, S. S. et al. An ingestible device for gastric electrophysiology. Nat. Electron 7, 497–508 (2024).
Nadeau, P. et al. Prolonged energy harvesting for ingestible devices. Nat. Biomed. Eng. 1, 1–8 (2017).
Balakrishnan, G. et al. Gelatin-based ingestible impedance sensor to evaluate gastrointestinal epithelial barriers. Adv. Mater. 35, 2211581 (2023).
Rezaie, M., Rafiee, Z. & Choi, S. A biobattery capsule for ingestible electronics in the small intestine: biopower production from intestinal fluids activated germination of exoelectrogenic bacterial endospores. Adv. Energy Mater. 13, 2202581 (2023).
Cataldi, P. et al. An electrically conductive oleogel paste for edible electronics. Adv. Funct. Mater. 32, 2113417 (2022).
Li, S. et al. Bioresorbable, wireless, passive sensors for continuous pH measurements and early detection of gastric leakage. Sci. Adv. 10, eadj0268 (2024).
Ho, K. K. & Joyce, A. M. Complications of capsule endoscopy. Gastrointest. Endosc. Clin. North Am. 17, 169–178 (2007).
Abdigazy, A. et al. End-to-end design of ingestible electronics. Nat. Electron 7, 102–118 (2024).
Zhang, Y. et al. Advances in bioresorbable materials and electronics. Chem. Rev. 123, 11722–11773 (2023).
Cox, K. D. et al. Human consumption of microplastics. Environ. Sci. Technol. 53, 7068–7074 (2019).
Gao, D., Lv, J. & Lee, P. S. Natural polymer in soft electronics: opportunities, challenges, and future prospects. Adv. Mater. 34, 2105020 (2022).
Sharova, A. S., Melloni, F., Lanzani, G., Bettinger, C. J. & Caironi, M. Edible electronics: the vision and the challenge. Adv. Mater. Technol. 6, 2000757 (2021).
Greenway, F. L. et al. A randomized, double-blind, placebo-controlled study of gelesis100: a novel nonsystemic oral hydrogel for weight loss. Obesity 27, 205–216 (2019).
Andrew, L. J., Lizundia, E. & MacLachlan, M. J. Designing for degradation: transient devices enabled by (nano)cellulose. Adv. Mater. 37, 2401560 (2025).
Yin, L. et al. Dissolvable metals for transient electronics. Adv. Funct. Mater. 24, 645–658 (2014).
Say, M. G. et al. Ultrathin paper microsupercapacitors for electronic skin applications. Adv. Mater. Technol. 7, 2101420 (2022).
Li, T. et al. Developing fibrillated cellulose as a sustainable technological material. Nature 590, 47–56 (2021).
Arca, H. C. et al. Pharmaceutical applications of cellulose ethers and cellulose ether esters. Biomacromolecules 19, 2351–2376 (2018).
Diao, Y. et al. Enzymic degradation of hydroxyethyl cellulose and analysis of the substitution pattern along the polysaccharide chain. Carbohydr. Polym. 169, 92–100 (2017).
Lee, Y. K. et al. Room temperature electrochemical sintering of ZN microparticles and its use in printable conducting inks for bioresorbable electronics. Adv. Mater. 29, 1702665 (2017).
Kim, J. et al. Electroceuticals for regeneration of long nerve gap using biodegradable conductive conduits and implantable wireless stimulator. Adv. Sci. 10, 2302632 (2023).
Hong, X. et al. High-permittivity solvents increase mxene stability and stacking order enabling ultraefficient terahertz shielding. Adv. Sci. 11, 2305099 (2024).
Hosseini, E., Arjmand, M., Sundararaj, U. & Karan, K. Filler-free conducting polymers as a new class of transparent electromagnetic interference shields. ACS Appl. Mater. Interfaces 12, 28596–28606 (2020).
Yun, T. et al. Electromagnetic shielding of monolayer MXene assemblies. Adv. Mater. 32, 1906769 (2020).
Biswas, B., Karmakar, A. & Chandra, V. Miniaturised wideband ingestible antenna for wireless capsule endoscopy. IET Microw. Antenna Prop. 14, 293–301 (2020).
Hadi, A., Cil, E., Ozdil, Z. C. C., Nikolayev, D. & Dumanli, S. Gastrointestinal Segment Tracking of Ingestible Capsules Using Biodegradable Superstrates. in Proc. 19th European Conference on Antennas and Propagation (EuCAP) 1–5 (IEEE, Stockholm, Sweden). https://doi.org/10.23919/EuCAP63536.2025.10999460. (IEEE, 2025).
Bora, P. J., Anil, A. G., Vinoy, K. J. & Ramamurthy, P. C. Outstanding absolute electromagnetic interference shielding effectiveness of cross-linked PEDOT: PSS film. Adv. Mater. Inter 6, 1901353 (2019).
Hukins, D. W. L., Mahomed, A. & Kukureka, S. N. Accelerated aging for testing polymeric biomaterials and medical devices. Med. Eng. Phys. 30, 1270–1274 (2008).
Yang, Q. et al. High-speed, scanned laser structuring of multi-layered eco/bioresorbable materials for advanced electronic systems. Nat. Commun. 13, 6518 (2022).
Pomerantseva, I. et al. Degradation behavior of poly(glycerol sebacate. J. Biomed. Mater. Res. 91A, 1038–1047 (2009).
Franc, A., Vetchý, D. & Fülöpová, N. Commercially available enteric empty hard capsules, production technology and application. Pharmaceuticals 15, 1398 (2022).
Naeem, A. et al. Hydroxyethyl cellulose-based hydrogels as controlled release carriers for amorphous solid dispersion of bioactive components of radix paeonia alba. Molecules 28, 7320 (2023).
Chen, Y.-C., Ho, H.-O., Liu, D.-Z., Siow, W.-S. & Sheu, M.-T. Swelling/floating capability and drug release characterizations of gastroretentive drug delivery system based on a combination of hydroxyethyl cellulose and sodium carboxymethyl cellulose. PLoS ONE 10, e0116914 (2015).
Hafezi, H. et al. An ingestible sensor for measuring medication adherence. IEEE Trans. Biomed. Eng. 62, 99–109 (2015).
Lamanna, L., Cataldi, P., Friuli, M., Demitri, C. & Caironi, M. Monitoring of drug release via intra body communication with an edible pill. Adv. Mater. Technol. 8, 2200731 (2023).
Lee, G. et al. A bioresorbable peripheral nerve stimulator for electronic pain block. Sci. Adv. 8, eabp9169 (2022).
Xu, K. et al. Bioresorbable electrode array for electrophysiological and pressure signal recording in the brain. Adv. Health. Mater. 8, e1801649 (2019).
Turnlund, J., Keyes, W., Peiffer, G. & Chiang, G. Molybdenum absorption, excretion, and retention studied with stable isotopes in young men during depletion and repletion. Am. J. Clin. Nutr. 61, 1102–1109 (1995).
Schauer, A. et al. Biocompatibility and degradation behavior of molybdenum in an in vivo rat model. Materials 14, 7776 (2021).
Jay Murray, F., Tyl, R. W., Sullivan, F. M., Tiwary, A. K. & Carey, S. Developmental toxicity study of sodium molybdate dihydrate administered in the diet to Sprague Dawley rats. Reprod. Toxicol. 49, 202–208 (2014).
Burrough, E. R., De Mille, C. & Gabler, N. K. Zinc overload in weaned pigs: tissue accumulation, pathology, and growth impacts. J. VET Diagn. Invest. 31, 537–545 (2019).
Nordberg, M. & Nordberg, G. F. Toxicology and Biological Monitoring of Metals. in General, Applied and Systems Toxicology (John Wiley & Sons, Ltd). https://doi.org/10.1002/9780470744307.gat145.(Wiley, 2011).
Acknowledgements
This work was supported in part by the following grants including: a grant from Novo Nordisk; the Karl van Tassel (1925) Career Development Professorship, the Department of Mechanical Engineering, Massachusetts Institute of Technology (MIT); the Division of Gastroenterology, Brigham and Women’s Hospital; the Advanced Research Projects Agency for Health (ARPA-H) under Award Number D24AC00040-00. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Advanced Research Projects Agency for Health. This work was carried out in part through the use of MIT.nano’s facilities. We thank V. E. Fulford, Alar Illustration, for their work on the figures. P.C. is funded by NIH DP2DA056107. Y.C acknowledges K. Lisa Yang Brain-Body Center Fellowship. M.G.S acknowledges support from the Knut and Alice Wallenberg Foundation (KAW 2021.0317) for postdoctoral research at MIT. The authors would like to thank Dr. Deniz Hur for his valuable guidance on the synthesis of PGS.
Author information
Authors and Affiliations
Contributions
M.G.S., S.S.Y., and G.T. conceived the concept and designed the overall research; M.G.S. and Y.C designed, developed, and characterized the device and process; M.G.S., S.S.Y., Y.C., A.E., I.M., B.M., F.V.Z., A.Z., A.P., K.S., N.F., A.H., P.C., and G.T. planned and performed the experiments; M.G.S. and S.S.Y. conducted the data analysis; Y.C. and A.E. performed the modeling and simulations; M.G.S. and G.T. wrote the manuscript.
Corresponding author
Ethics declarations
Competing interests
M.G.S., S.S.Y. and G.T. are co-inventors on a patent application describing the reported system: US Provisional Application No: PCT/US2025/049378. Complete details of all relationships for profit and not for profit for G.T. can be found at https://www.dropbox.com/sh/szi7vnr4a2ajb56/AABs5N5i0q9AfT1IqIJAE-T5a?dl=0. The remaining authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Bappaditya Mandal, Sema Dumanli, Kristin Riekert and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Say, M.G., You, S.S., Cai, Y. et al. Bioresorbable RFID capsule for assessing medication adherence. Nat Commun 17, 52 (2026). https://doi.org/10.1038/s41467-025-67551-5
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41467-025-67551-5






