Skip to main content
Download PDF

A wearable enzyme sensor enabled by the floating-gate OECT with poly(benzimidazobenzophenanthroline) as the catalytic layer

Journal of Nanobiotechnology volume 23, Article number: 120 (2025) Cite this article

Abstract

With the advantages of miniaturization, simple device structure, and fast response, the organic electrochemical transistor (OECT) has become an emerging platform for developing wearable enzyme sensors for real-time health monitoring. The floating gate (FG) OECT employs a distinct signal acquisition and amplification structure, mitigating the effects of non-specific physical adsorption during the sensing process and preventing contamination of the electrolyte solution by side reaction products. The current work reports a feasible wearable enzyme sensor using a poly(benzimidazobenzophenanthroline) (BBL)-Nafion-enzyme-Nafion stacking structure as the sensing layer of the FG OECT. Based on the experimental results, the BBL film with an area of 3.14 mm2 and a thickness of 175 nm can generate an open circuit potential of 199.61 mV in 10− 1 M hydrogen peroxide compared with the blank control. Then, the FG OECT is integrated with the flexible microfluidic systems for on-skin detection of glucose, lactate, and uric acid with sensitivities of 92.47, 152.15, and 74.27 µA·dec− 1, respectively. This FG OECT-based wearable enzyme sensor will open new windows for multiplexed detection of sweat metabolites.

Introduction

Sweat contains many metabolites, such as glucose, uric acid, cholesterol, and lactate. Compared to blood, sweat can be obtained using non-invasive techniques and thus is of great significance for characterizing physiological states. The abnormal changes in sweat’s metabolites represent the possibility of diseases in the human body. For example, the glucose concentration in human sweat typically ranges from 10 µM to 1mM [1]. Deviation from this range indicates diabetes. Abnormal changes in lactate concentrations mean thrombosis [2, 3] and respiratory disease, which, in severe cases, can lead to shock. In addition, uric acid is a risk factor for cardiovascular disease and has been widely used in the clinical treatment of gout [4]. The uric acid level reported in the sweat of patients with gout is higher than that of healthy individuals. Therefore, great attention has been paid to developing non-invasive sensors that can monitor the changes in the concentration of sweat metabolites in real time [5, 6].

Organic electrochemical transistors (OECTs) are widely used as the core component of sweat sensors due to their intrinsic signal amplification ability, low operating voltage, and good biocompatibility [7,8,9]. By modifying specific enzymes on the OECTs, most sweat’s metabolic products can be detected. Macaya et al. first reported the application of OECTs in glucose sensors with the limit of detection (LOD) of 1 µM, where the glucose oxidase (GOx) was mixed with the electrolyte solution, and Pt served as the gate electrode material [10]. Yan et al. further improved the glucose detection sensitivity with a LOD down to 10 nM by modifying Pt gate electrodes of OECTs with graphene-based materials and GOx [11]. Welch et al. covalently bonded GOx onto the PEDOT: PSS semiconducting channel through a mixed polymer brush of poly(glycidyl methacrylate) and poly(2-hydroxyethylmethacrylate) [12]. The devices exhibited a solid response to glucose in the 10− 2 − 102 mM range with remarkable stability. In addition, OECTs do not require the introduction of typically large Ag/AgCl electrodes, enabling a high degree of array integration. Correspondingly, OECT-based biosensors with multi-analyte sensing capability were proposed. For instance, Pappa et al. designed a compact multi-analyte biosensing platform composed of an OECT microarray integrated with a “finger-powered” microfluidic for quantitative and simultaneous detection of common sweat metabolic, namely glucose, lactate, and cholesterol [13].

Although the enzyme sensors based on OECT have been widely studied, the remaining challenges urgently need to be addressed. The byproducts of enzyme-catalyzed reactions usually bring in contamination, which will deteriorate the OECT performance and limit the high-accuracy sensing. Among them, hydrogen ions and oxygen may alter the semiconducting layers’ electrochemical (de)doping state; reactants include d-glucose-1,5-lactone for glucose, pyruvate for lactate, and cholest-4-en-3-one for cholesterol may adsorb at the gate-electrode/electrolyte interface or the semiconducting-layer/electrolyte interface, changing the interface capacitance. Floating gate (FG) OECTs use an extended gate to conduct sensing, physically separated from the signal amplification process, enabling the optimization of both compartments independently regarding the materials and geometries and thus avoiding contamination [14]. Based on FG OECTs, yeast, and multiplexed ions have been detected successfully. Applying FG OECTs to construct enzyme sensors not only avoids the contamination of the transistor channel from the catalytic process when enzyme-catalyzed reactions are introduced but also fully utilizes the signal amplification ability of the semiconducting channel current on the gate voltage. Using FG OECT, Siew et al. achieved a more significant signal gain in hydrogen peroxide (H2O2) detections through FG OECT than the traditional OECT [15]. Although Siew et al. demonstrated glucose detection by mixing GOx in the PBS solution, none of these works reported embedding enzymes onto the FG OECT, limiting their future applications in wearable sweat sensors.

Poly(benzimidazobenzophenanthroline) (BBL) is an n-type, rigid ladder-type, and side-chain-free electron-transporting polymer, which is environmentally friendly and can be processed using spin-coating. BBL could also be a stable oxygen cathode material in the fuel cell, reducing the intermediate H2O2 into the water [16]. Considering the above advantages, we employ BBL to catalyze H2O2 molecules in our proposed wearable enzyme sensor, by which the metabolite could be detected through the cascade reactions. Previous studies have shown that stacking the Nafion layer between the electrode surface and enzyme layer can improve enzyme adhesion [17], and modifying the enzyme layer with Nafion can prevent potential enzyme leakage during long-term testing and improve the specificity of metabolite detection [11, 18,19,20]. Thus, in the current work, we use stacked layers of BBL-Nafion-enzyme-Nafion to modify the sensitive electrode of FG OECTs. As a proof-of-concept application, the OECT-based enzyme sensor is further integrated with a soft, flexible, sweat-delivery microfluidic chip, laying a solid foundation for its application in wearable devices. To our best knowledge, the current work first uses BBL to construct a stacked structure for enzyme immobilization, and for the first time, an FG OECT modified by the as-prepared stacked structure is used for building an enzyme sensor that could achieve multiplexed sensing of metabolites.

Scheme 1
scheme 1

The structure diagram of the wearable enzyme sensor based on the floating-gate (FG) organic electrochemical transistor (OECT)

Full size image

Results and discussion

The working principle of the enzyme sensor

As shown in Scheme 1, the FG OECT has two key elements: a signal amplification unit (AU) and a sensing unit (SU). The AU consists of a source (S), a drain (D), a semiconducting channel, a primary floating gate (FG1), and a primary electrolyte, while the SU consists of a secondary floating gate (FG2), a control gate (CG), and a secondary electrolyte. The secondary electrolyte provides a reaction pool for the SU, while the interface capacitance of the electric double layer (EDL) formed between the primary electrolyte and the semiconducting channel serves as the dielectric layer for the AU. The FG structure can separate the gate dielectric electrolyte (primary electrolyte) from the biochemical reaction electrolyte (secondary electrolyte), decouple the signal amplification and biochemical sensing processes, and achieve independent AU and SU structure design and performance optimization.

In particular, the Ag/AgCl electrodes are employed as CG and FG1 simultaneously, the poly (3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT: PSS) serves as the semiconducting material, the 100 mM NaCl serves as the primary electrolyte, and the secondary electrolyte is the aqueous solution containing metabolites to be tested. To convert analyte concentration into a voltage signal, we modify FG2 using the stacking structure of BBL-Nafion-enzyme-Nafion. If sweat metabolites, namely glucose, lactate, and uric acid, exist in the secondary electrolyte, embedded enzymes, such as the Gox, could catalyze glucose and produce H2O2. Then, the BBL film catalyzes the H2O2 and induces an electrochemical Nernst potential (Formula 1), which will further control FG1’s electrical potential:

$$\:{E}_{\text{Nernst}}={E}^{\text{0}}+\frac{{kT}}{{ne}}\text{ln(}\frac{\text{[OX]}}{\text{[Red]}}\text{)}$$
(1)

where [Ox] and [Red] are oxidized and reduced species concentrations, respectively, E0 is the formal potential, k is Boltzmann’s constant, T is the temperature, e is the fundamental charge, and n is the number of electrons transferred during the reaction.

FG OECT fabrication and sensitive layer optimization

Fig. 1
figure 1

The fabrication method (a) and optical image (b) of the as-prepared FG OECT

Full size image

Pt electrodes are widely used in the catalysis of H2O2. However, Pt electrodes are incompatible with conventional equipment, such as wire bonders, limiting their utilization in OECT. Thus, in most cases, the electrodes and leads of OECT are usually based on Au. Alternatively, Prussian blue (PB) can be easily electrodeposited on the surface of gold electrodes and work as a catalyst for H2O2. It is known to be ‘‘artificial peroxidase’’ due to its high catalytic activity and selectivity toward H2O2. However, preparing PB film requires highly toxic potassium ferrocyanide (K3Fe(CN)6). Herein, we used BBL, modifying the Au electrode, as the catalyst for H2O2 to overcome these above limitations.

A multi-step patterning process is used in the current work to realize the FG OECT’s fabrication (Fig. 1a). The first step is to use microfabrication technology to prepare Au microelectrodes on PI (Polyimide) film with a thickness of 7 μm supported by the SiO2/Si substrate (Note 1 and 2, Supporting Information). In the second step, a polyethylene terephthalate (PET) tape—called base-layer tape with holes aligned with all the electrode pads, the gap between the source and drain electrode, and all the microelectrodes except for the source and drain is patterned by a laser marking machine and attached to the microelectrodes chip, with no peeling in the whole fabrication process. In the third step, we prepare another PET tape—the first sacrificial tape—with holes aligned with the FG1 and CG electrodes and attach it to the base-layer tape. The Ag nanoparticles are sputtered onto FG1 and CG using a magnetron sputterer and soaked in 10− 1 M FeCl3 solution to deposit Ag/AgCl films [21, 22]. After removing the first sacrificial tape, non-polarized Ag/AgCl electrodes are formed on the FG1 and CG. In the fourth step, the third piece of perforated tape, named the second sacrificial tape (not shown), is used to achieve a patterned semiconducting layer. The second sacrificial tape presents an opening aligning with the source-drain electrode gap, in which the PEDOT: PSS mixture (Note 3, Supporting Information) will be spin-coated. A semiconducting channel could be obtained after peeling off the second sacrificial tape. In the final step, we use a third sacrificial tape (not shown) to pattern the stacked sensitive layer—BBL-Nafion-enzyme-Nafion, whose preparation procedure will be discussed next. Figure 1b shows the optical image of the prepared FG OECT.

Fig. 2
figure 2

Optimization of the sensitive materials to catalyze hydrogen peroxide (H2O2). (a) Schematic diagram of the traditional OECT structure to calibrate the semiconducting channel’s performance and the obtained transfer and transconductance curves. (b) Calibrating the detection sensitivity of H2O2 when using Pt, poly(benzimidazobenzophenanthroline) (BBL), and Prussian blue (PB) as the sensitive materials, respectively. The open circuit potential (OCP) is obtained in PBS and 10− 1 M H2O2 by (c) BBL-modified Au electrode, (d) PB-modified Au electrode, and (e) Pt electrode. (f) The semiconducting channel current (IDS) versus time diagram is obtained under different concentrations of H2O2 at VCG = 0 V. (g) Electrochemical impedance spectroscopy (EIS) investigation results. (h) The equivalent capacitances. (i) OCPs for BBL-modified Au electrodes in different areas

Full size image

Based on the fabrication process, a semiconducting channel presenting a length of 20 μm, a width of 10 μm, and a height of 0.17 μm could be obtained. Using Ag/AgCl electrodes as the gate electrode, the transfer and transconductance curves of OECT using the traditional three-electrode structure (Fig. 2a) at drain bias (VD) of -0.60 V are investigated, presenting a peak transconductance up to 10.32 mS. When investigating FG OECT’s response, the control gate bias (VCG) is set at 0 V to fully utilize the electrochemical potential produced in the SU to regulate the AU, according to Siew’s recent work [23]. Since the non-polarizable Ag/AgCl electrodes serve as both FG1 and CG, the Nernst potential generated by the cascade reactions in the SU will be transmitted to the AU to regulate the IDS. Thus, the more significant the electrochemical potential that can be generated in H2O2 catalysis, the more pronounced the current response of FG OECT. We use FG2 of the same area for easy comparison, kept consistently at 12.56 mm2, which equals the 4 mm diameter commercial Pt electrode area. BBL and PB film need to be deposited on the substrate. Herein, gold sheets with a 12.56 mm2 exposed patterned area are employed as the substrate, and the manufacturing parameters are tuned to match the thickness of the BBL and PB films (Note 4, Supporting Information).

Figure 2b indicates that BBL has the highest sensitivity for H2O2 detection in the FG OECT configuration, consistent with the open circuit potential (OCP) investigation results (Fig. 2c and e). In detail, the OCPs are obtained in PBS and 10− 1 M H2O2 using working electrodes of different materials while keeping the Ag/AgCl electrode as the reference and counter electrodes. The experimental results show that BBL exhibits the maximum OCP difference (168.16 mV), and thus, the corresponding fluctuation amplitude of IDS is the largest when the H2O2 concentration increases from 10− 5 M to 10− 1 M (Fig. 2f). Note in Fig. 2f that the output current of the OECT obtained in H2O2 is different. We conducted electrochemical impedance spectra (EIS) investigations of various electrodes (Fig. 2g) in 10− 1 M H2O2 to elucidate the phenomenon. The experimental results show that, at low-frequency excitation (10− 1 Hz), the BBL-modified Au electrode has the most significant equivalent capacitance (3.21 × 10− 5 F), while the Pt electrode presents the most negligible equivalent capacitance (7.55 × 10− 7 F) (Fig. 2h), which was estimated from the Nyquist plot using the out-of-phase impedance (Z″=1/2πfCeff) [24]. According to the principle of series circuits (Note 5, Supporting Information), the semiconducting layer will undertake the maximum voltage division and produce the minimum output current when the BBL-modified Au electrode is used. In addition, we find that regardless of the electrode type, the absolute value of IDS decreases with the increase of H2O2 concentration in Fig. 2f. The reason is that when H2O2 is added dropwise in the SU, the Nernst potential generated by the electrochemical reaction increases with the increment of H2O2 concentration (Formula 1), resulting in an increasing effective FG voltage (VFGeff) and finally leading to a decrement of the output current of OECT [25].

Based on the electrode material optimization, we subsequently investigated the effect of the FG2 electrode geometry on the H2O2 catalysis. The BBL film was kept at a thickness of 175 nm with a spin-coating speed of 2000 rpm (Note 4 in the Supporting Information). The substrate areas vary from 3.14 mm2 to 50.24 mm2. As depicted in Fig. 2i, the OCP difference of BBL-covered Au electrodes in PBS and 10− 1 M H2O2 decreases with the increase of electrode area, consistent with our previous work regarding H2O2 sensors using Pt electrode as FG2 [14]. Moreover, the gate bias corresponding to the peak transconductance is 100 mV (Fig. 2a), falling well within the scope of electrochemical potentials (between − 5.00 and 194.61 mV) generated by the BBL-modified Au electrode of 3.14 mm2 area in PBS and 10− 1 M H2O2 (Fig. 2i). Therefore, in the following sensor investigations, we will use a gold sheet with an area of 3.14 mm2 and modify it with the BBL-Nafion-enzyme-Nafion stacked layer as the FG2 for constructing FG OECT-based enzyme sensors.

Fig. 3
figure 3

Characterizations of the stacked layer. (a) The scanning electron microscope (SEM) image of (i) BBL layer deposited on the Au substrate, (ii) bottom Nafion layer, (iii) glucose oxidase (GOx) layer, (iv) upper Nafion layer, (b) The cross-sectional view SEM. (c) Fourier transforms infrared (FT-IR) spectra of each layer. (d) The wettability characterizations of (i) BBL film, (ii) BBL-Nafion stacked layer, (iii) BBL-Nafion-Gox stacked layer and (iv) BBL-Nafion-GOx-Nafion stacked layer, (e) The EIS characterizations of the BBL-Nafion-GOx and BBL-Nafion-GOx-Nafion stacked layers, (i) the Nyquist plot, (ii) the Bode plot

Full size image

We then construct different stacked FG2 structures based on the optimized parameters for detecting glucose, lactate, and uric acid. First, the BBL is prepared by electrodeposition following the methodology described in Note 4 in the Supporting Information. Afterward, the preparation of the stacked Nafion-enzyme-Nafion layer is based on the methods described in the literature [26, 27]. In brief, a first layer of Nafion is fabricated by drop casting 10 µL of the Nafion solution onto the as-prepared BBL film and air-dried overnight at room temperature. The enzymatic layer is prepared by drop casting 10 µL of a solution containing 20 mg·mL− 1 of enzyme in distilled water onto the first Nafion layer. Afterward, the stacked layer is left to dry for 6 h at 4 ℃. Finally, another 10 µL of the Nafion solution is applied to make the upper layer that entraps the enzymatic layer and let dry overnight at room temperature. The stacked FG2 electrodes are kept at 4 ℃ when not in use.

The scanning electron microscope (SEM) characterizes the electrode morphology during the layer-by-layer modification. Herein, GOx is still employed as an example for the demonstration. As shown in Fig. 3a-i and a-ii, the Nafion layer is successfully deposited on the BBL film. Afterward, new dendritic patterns are generated by the physical adsorption of GOx (Fig. 3a-iii). Finally, the top Nafion layer is deposited (Fig. 3a-iv), forming a similar pattern to Fig. 3a-ii. Focused-ion beam scanning electron microscopy (FIB-SEM) is used to verify stacked FG2 structures at the nanoscale (Fig. 3b). Stacking different layers (BBL, bottom Nafion, GOx, and top Nafion layers) is distinguishable from the image. From FIB-SEM analyses, the thickness of the BBL layer, bottom Nafion layer, GOx layer, and top Nafion layer are found to be about 175 nm, 225 nm, 275 nm, and 437 nm, separately.

We have employed the Fourier transforms infrared (FT-IR) spectrometry to verify the composition of each layer. As shown in Fig. 3c, the fingerprint band at 1700 cm− 1 is attributed to the C = O vibration, the band at 1645 cm− 1 is attributed to negative polarons in BBL, while the 1545 cm− 1 band is attributed to the C = C vibration [28]. Then, the Nafion layer is spin coated, where the band at 629 cm− 1 is assigned to the stretching C-S groups; the band at 975 cm− 1 is assigned to C-O-C stretching [29]; the band at 1203 cm− 1 is assigned to the asymmetric C − F stretching vibrational modes [30]. Subsequently, the GOx covering the bottom Nafion layer is spin-coated, and the FTIR results present apparent contributions of amide I (at 1647 and 1693 cm− 1) and amide II (at 1540 cm− 1) [31].

Figure 3d shows the changes in wettability during the preparation process of the stacked layer. The BBL is hydrophilic with a contact angle of 64.71° when using deionized ultrafiltered water as the testing liquid. However, the contact angle slightly increases when the BBL film is covered with Nafion membranes due to the inherent hydrophobic properties of Nafion [32, 33]. However, after dropping cast GOx onto the bottom Nafion layer, the contact angle of the stacked structure is significantly reduced to 19.96° due to the hydrophilicity of GOx. Stacking the upper Nafion layer restores the contact angle to 95.71°, comparable to Fig. 3d-ii. Figure 3e shows the corresponding EIS investigation results in the last two steps, conducted in 1 mM glucose solution. When the upper Nafion layer covers the GOx layer, the slope of the Nyquist plot decreases from 1.18 to 0.90, indicating that covering the upper Nafion layer is less conducive to transport the analyte solution [34], consistent with the wettability investigation result. It could also be verified by the Bode plot, where the electrode impedance slightly increases at low frequencies when the upper Nafion layer covers the GOx layer.

Responses of FG OECTs to typical analytes

Fig. 4
figure 4

Enzyme sensors characterizations. (a) Sensitivity calibrations. The error bar indicates the standard error of three independent devices. (b) Selectivity investigations. Stability investigations are conducted when (c) cycling the gate biases and (d) cycling the analyte concentration

Full size image

Finally, we assemble the as-prepared sensing element of stacked BBL-Nafion-enzyme-Nafion layers, Ag/AgCl electrode, and semiconducting channel to construct FG OECT (Fig. 1) and prepare glucose, lactate, and uric acid sensors using the same processing approach by changing the enzymes types in the stacked layer. As shown in Fig. 4a (i-iii), the obtained glucose, lactate, and uric acid detection sensitivities are 92.47, 152.15, and 74.27 µA·dec− 1, respectively. The selectivity of the glucose/lactate/uric acid sensor is verified by the negligible response of non-target metabolites and ionic compositions, which have the same concentration as the substance to be detected (1 mM for i, 100 mM for ii, 100 µM for iii, respectively) (Fig. 4b). In addition, we conducted stability testing on the sensor over five days, with 5000 cycles of pulse testing conducted daily. In detail, the VG step from − 0.60 to 0.60 V, with a pulse width and duty cycle of 0.25 s and 50%, respectively (Fig. 4c). The outputs of glucose, lactate, and uric acid sensors decay by 8.05%, 2.83%, and 2.09%, respectively. The stability testing under cycling the analyte concentrations has also been investigated (Fig. 4d). Herein, the hysteresis width was defined as the semiconducting channel current (IDS) difference between the initial and final electrolyte solution at the same analyte concentration. For glucose detections, the hysteresis width in the 0–1000 µM loop obtained in the third cycle is 18.30 µA, about 0.71% of the initial IDS (2.58 mA). Based on the glucose detection sensitivity (92.47 µA·dec− 1), this hysteresis width only introduces a deviation of about 0.20 dec glucose. For the lactate acid, the hysteresis width obtained in the three cycles is 1.21%, 2.58%, and 1.02%, respectively. Regarding the uric acid, the hysteresis width obtained in the three cycles is 1.28%, 1.39.%, and 2.34%, respectively. It is worth noting that the electrode modified with the BBL layer can quickly reach an equilibrium state in hydrogen peroxide (Fig. 2f). However, if the BBL-Nafion-enzyme-Nafion stacked structure is placed in the analyte aqueous solution, the open circuit potential to reach equilibrium will significantly increase(Fig. 4a). The slow catalytic rate of enzymes towards analytes and the hindering effect of the Nafion layer on the analyte transport, as indicated by the wettability and EIS investigations results in Fig. 3, may cause the systematic drift of IDS independently from the detected metabolites (Note 6 and Note 7, Supporting Information).

Fig. 5
figure 5

The wearable enzyme sensors. (a) The experimental setup for investigating the wearable enzyme sensor. (b) The retention of flexible sensor current under different bending conditions. (c) The structural schematic (i), the optical photograph of wearable enzyme sensors (i), and the investigation results of sweat metabolites on volunteers during exercises (ii and iii). In Fig. 5c-ii, BE stands for “before exercise”. The error bar indicates the standard error of three independent tests

Full size image

The current output of the flexible enzyme sensor is electrically characterized under different bending states with a probe station (Fig. 5b) to evaluate the durability. Herein, we still take glucose sensing as an example. The maximum attenuation occurs at a bending curvature of 10 mm. For 10, 50, 200, 400, 800, and 1000 μm glucose aqueous solution, the sensor outputs are − 1.85 ± 0.01, -1.82 ± 0.02, -1.78 ± 0.01, -1.75 ± 0.02, -1.71 ± 0.02, and − 1.68 ± 0.01 mA. Compared to the flat state, the attenuation is 18.10%, 18.25%, 18.41%, 17.91%, 18.33%, and 18.58%, respectively. In addition, after 500 bending cycles, at R = 30 mm, the attenuation is 14.31%, 14.12%, 15.33%, 16.24%, 14.82%, and 15.31% for 10, 50, 200, 400, 800, and 1000 μm glucose detections, respectively. Since different enzymes could be fixed at the SUs of FG OECT, each specific enzyme will catalyze the sweat metabolites, generating the Nernst potential that drives the FG OECTs. Therefore, quantitative assessment of sweat metabolites, including glucose, lactate, and uric acid, could be achieved in this FG OECT-based enzyme sensor.

The wearable enzyme sensor consists of four layers: (i) a skin-compatible adhesive layer with an inlet opening for sweat collection, (ii) a microfluidic channel for sweat delivery, (iii) an enzyme-based FG OECT for sweat metabolite sensing (Fig. 5c), which will be attached on the volunteer’s forehead, and (iv) A medical-grade adhesive layer is used to enable the robust and seamless adhesion of the wearable enzyme sensor on the forehead [35]. The inlet opening (2 mm in diameter) improves access to the sweat glands and maximizes the sweat collection [36]. The microfluidic device is comprised of a double-sided adhesive layer (thickness, 200 μm) with four channels (width, 1 mm) for sweat delivery. The laser engraving generates the patterns on the adhesive layers.

We have conducted three times of on-skin validation experiments on the same volunteer over three days. Due to the wired connections required between our chip and the source-measure unit, the volunteer needs to pause the exercise after sweating. Thus, on each day, the volunteer needs to engage in a five-cycle jogging exercise and each cycle ends with visual sweating on the forehead. After each exercise cycle, the metabolites are detected sequentially after connecting the sensor with the source-measure unit. During the measurements, a 30-second test is performed for each analyte to obtain a stable current response, and each measurement cycle takes approximately 2 min. The first current value in Fig. 5(c) is recorded before the exercise, indicating the semiconducting channel current obtained under only the action of the drain bias, while without the impact of sweat. As shown in Fig. 5(c), the signals of glucose, lactate, and uric acid reach their peak during the first exercise cycle, indicating that the volunteer undertakes the highest training intensity during the first exercise cycle. As the exercise progresses, the volunteer subconsciously reduces the intensity of the training. Based on the changes in current values recorded, the glucose, lactate, and uric acid concentration changes during consecutive measurements are approximately 77.41 µM, 6.41 mM, and 36.63 µM, respectively. Both the trend and the variation in metabolite concentration are consistent with literature reports [37,38,39,40,41]. The result suggests that the wearable enzyme sensor could collect sweat from the skin and independently detect the metabolites secreted by the sweat glands in SUs of the FG OECT. The entire experimental process is recorded in the attached movie (Suppmentary movie 1). The limitation of the current devices is the wired connection with the external readout devices. Wireless signal transfer represents a possibility for remote data collection.

Conclusions

This work developed a wearable enzyme sensor by modifying OECT’s floating gate with the BBL-Nafion-enzyme-Nafion stacking structure. BBL provided better sensing sensitivity and processing compatibility than Pt electrodes and PB. In addition, the stacking structure rendered the as-prepared enzyme sensor to be a wearable device. Through multi-step microfabrication technology, we have achieved the patterning of semiconductor channels, Ag/AgCl unpolarized electrodes, and BBL-Nafion-enzyme-Nafion stacking structures on flexible substrates. The floating gate OECTs-based enzyme sensor is further integrated with the soft, flexible, and wearable sweat-delivery microfluidic system for on-skin, multiplexed detection of sweat metabolites. The detection sensitivities of this wearable enzyme sensor for glucose, lactate, and uric acid are 92.47, 152.15, and 74.27 µA·dec− 1, respectively., with good selectivity, stability, and recovery.

Data availability

Data is provided within the manuscript or supplementary information files.

References

  1. Lee H, Song C, Hong YS, Kim M, Cho HR, Kang T, et al. Wearable/disposable sweat-based glucose monitoring device with multistage transdermal drug delivery module. Sci Adv. 2017;3:e1601314.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Certo M, Tsai C-H, Pucino V, Ho P-C, Mauro C. Lactate modulation of immune responses in inflammatory versus tumour microenvironments. Nat Rev Immunol. 2021;21:151–61.

    Article  CAS  PubMed  Google Scholar 

  3. Gupta GS. The Lactate and the Lactate dehydrogenase in inflammatory diseases and Major Risk factors in COVID-19 patients. Inflammation. 2022;45:2091–123.

    Article  CAS  PubMed  Google Scholar 

  4. Yang Y, Song Y, Bo X, Min J, Pak OS, Zhu L, et al. A laser-engraved wearable sensor for sensitive detection of uric acid and tyrosine in sweat. Nat Biotechnol. 2020;38:217–24.

    Article  CAS  PubMed  Google Scholar 

  5. Ji W, Zhu J, Wu W, Wang N, Wang J, Wu J, et al. Wearable Sweat Biosensors Refresh Personalized Health/Medical Diagnostics. Research. 2021;2021:20219757126.

    Article  Google Scholar 

  6. Gao F, Liu C, Zhang L, Liu T, Wang Z, Song Z, et al. Wearable and flexible electrochemical sensors for sweat analysis: a review. Microsyst Nanoeng. 2023;9:1.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Sophocleous M, Contat-Rodrigo L, Garcia-Breijo E, Georgiou J. Organic Electrochemical transistors as an emerging platform for Bio-sensing Applications: a review. IEEE Sens J. 2021;21:3977–4006.

    Article  CAS  Google Scholar 

  8. Chen X, Ji J, Peng Y, Gao Z, Zhao M, Tang B, et al. Flexible pH sensors based on OECTs with a BTB dye-embedded ion-gel gate dielectric. J Mater Chem C. 2023;11:7722–31.

    Article  CAS  Google Scholar 

  9. Ji J, Wang Z, Zhang F, Wang B, Niu Y, Jiang X, et al. Pulse electrochemical synaptic transistor for supersensitive and ultrafast biosensors. InfoMat. 2023;5:e12478.

    Article  CAS  Google Scholar 

  10. Macaya DJ, Nikolou M, Takamatsu S, Mabeck JT, Owens RM, Malliaras GG. Simple glucose sensors with micromolar sensitivity based on organic electrochemical transistors. Sens Actuators B. 2007;123:374–8.

    Article  CAS  Google Scholar 

  11. Liao C, Zhang M, Niu L, Zheng Z, Yan F. Highly selective and sensitive glucose sensors based on organic electrochemical transistors with graphene-modified gate electrodes. J Mater Chem B. 2013;1:3820.

    Article  CAS  PubMed  Google Scholar 

  12. Welch ME, Doublet T, Bernard C, Malliaras GG, Ober CK. A glucose sensor via stable immobilization of the GOx enzyme on an organic transistor using a polymer brush. J Polym Sci Part A: Polym Chem. 2015;53:372–7.

    Article  CAS  Google Scholar 

  13. Pappa A, Curto VF, Braendlein M, Strakosas X, Donahue MJ, Fiocchi M, et al. Organic transistor arrays integrated with finger-powered Microfluidics for Multianalyte Saliva Testing. Adv Healthc Mater. 2016;5:2295–302.

    Article  CAS  PubMed  Google Scholar 

  14. Li H, Jin Z, Jiang X, An M, Ji J, Huang D. Influence of reaction cell electrodes on organic electrochemical transistors. Appl Phys Lett. 2024;124:093509.

    Article  CAS  Google Scholar 

  15. Tan STM, Giovannitti A, Melianas A, Moser M, Cotts BL, Singh D, et al. High-gain chemically gated Organic Electrochemical Transistor. Adv Funct Mater. 2021;31:2010868.

    Article  CAS  Google Scholar 

  16. Vagin M, Gueskine V, Mitraka E, Wang S, Singh A, Zozoulenko I, et al. Negatively-doped conducting polymers for Oxygen reduction reaction. Adv Energy Mater. 2021;11:2002664.

    Article  CAS  Google Scholar 

  17. Liu C-T, Liu C-H, Lai Y-T, Lee C-Y, Gupta S, Tai N-H. A salivary glucose biosensor based on immobilization of glucose oxidase in Nafion-carbon nanotubes nanocomposites modified on screen printed electrode. Microchem J. 2023;191:108872.

    Article  CAS  Google Scholar 

  18. Xu Z, Zhang M, Zou H, Liu J, Wang D, Wang J, et al. Non-enzymatic electrochemical detection of uric acid with electrodeposited Nafion film. J Electroanal Chem. 2019;841:129–34.

    Article  CAS  Google Scholar 

  19. Tsou K-L, Chen K-Y, Chou Y-D, Cheng Y-T, Tsai H-E, Lee C-K. Inkjet-printed flexible non-enzymatic lactate sensor with high sensitivity and low interference using a stacked NiOx/NiOx-Nafion nanocomposite electrode with clinical blood test verification. Talanta. 2022;249:123598.

    Article  CAS  PubMed  Google Scholar 

  20. Chang T, Li H, Zhang N, Jiang X, Yu X, Yang Q, et al. Highly integrated watch for noninvasive continual glucose monitoring. Microsyst Nanoeng. 2022;8:25.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Tirado P, Chavez-Urbiola IR, Alcantar-Peña JJ. Ionic gel effect on a reference electrode in a flexible solid-state pH sensor. MRS Commun. 2022;13:41–6.

    Article  Google Scholar 

  22. Ji X, Lin X, Rivnay J. Organic electrochemical transistors as on-site signal amplifiers for electrochemical aptamer-based sensing. Nat Commun. 2023;14:1665.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Tan STM, Keene S, Giovannitti A, Melianas A, Moser M, McCulloch I, et al. Operation mechanism of organic electrochemical transistors as redox chemical transducers. J Mater Chem C. 2021;9:12148–58.

    Article  CAS  Google Scholar 

  24. Gao G, Yu J, Yang X, Pang Y, Zhao J, Pan C, et al. Triboiontronic Transistor MoS2 Adv Mater. 2019;31(7):1806905.

  25. Bernards DA, Macaya DJ, Nikolou M, DeFranco JA, Takamatsu S, Malliaras GG. Enzymatic sensing with organic electrochemical transistors. J Mater Chem. 2008;18:116–20.

    Article  CAS  Google Scholar 

  26. Parrilla M, Cánovas R, Andrade FJ. Paper-based enzymatic electrode with enhanced potentiometric response for monitoring glucose in biological fluids. Biosens Bioelectron. 2017;90:110–6.

    Article  CAS  PubMed  Google Scholar 

  27. Ait Yazza A, Blondeau P, Andrade FJ. Simple Approach for Building High Transconductance Paper-based Organic Electrochemical Transistor (OECT) for Chemical sensing. ACS Appl Electron Mater. 2021;3:1886–95.

    Article  CAS  Google Scholar 

  28. Yang C-Y, Stoeckel M-A, Ruoko T-P, Wu H-Y, Liu X, Kolhe NB, et al. A high-conductivity n-type polymeric ink for printed electronics. Nat Commun. 2021;12:2354.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Laporta M. PerÑuorosulfonated membrane (NaÐon): FT-IR study of the state of water with increasing humidity.

  30. Singh RK, Kunimatsu K, Miyatake K, Tsuneda T. Experimental and theoretical Infrared Spectroscopic Study on Hydrated Nafion membrane. Macromolecules. 2016;49:6621–9.

    Article  CAS  Google Scholar 

  31. Delfino I, Portaccio M, Ventura BD, Mita DG, Lepore M. Enzyme distribution and secondary structure of sol–gel immobilized glucose oxidase by micro-attenuated total reflection FT-IR spectroscopy. Mater Sci Engineering: C. 2013;33:304–10.

    Article  CAS  Google Scholar 

  32. Paul DK, Karan K, Docoslis A, Giorgi JB, Pearce J. Characteristics of self-assembled ultrathin Nafion films. Macromolecules. 2013;46:3461–75.

    Article  CAS  Google Scholar 

  33. Sigwadi R, Dhlamini MS, Mokrani T, Nemavhola F. Enhancing the mechanical properties of zirconia/Nafion® nanocomposite membrane through carbon nanotubes for fuel cell application. Heliyon. 2019;5:e02112.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Cai X, Xiao Y, Sun W. Pore-size effect of activated carbons on the electrochemical performances of symmetric supercapacitors under compression. J Energy Storage. 2022;55:105438.

    Article  Google Scholar 

  35. Koh A, Kang D, Xue Y, Lee S, Pielak RM, Kim J et al. A soft, wearable microfluidic device for the capture, storage, and colorimetric sensing of sweat. Sci Transl Med [Internet]. 2016 [cited 2024 Aug 5];8. Available from: https://www.science.org/doi/https://doiorg.publicaciones.saludcastillayleon.es/10.1126/scitranslmed.aaf2593

  36. Yuan Q, Fang H, Wu X, Wu J, Luo X, Peng R et al. Self-Adhesive, Biocompatible, Wearable Microfluidics with Erasable Liquid Metal Plasmonic Hotspots for Glucose Detection in Sweat. ACS Appl Mater Interfaces. 2023;acsami.3c11746.

  37. Min J, Tu J, Xu C, Lukas H, Shin S, Yang Y, et al. Skin-interfaced wearable sweat sensors for Precision Medicine. Chem Rev. 2023;123:5049–138.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Chen Y, Xiao H, Fan Q, Tu W, Zhang S, Li X et al. Fully Integrated Biosensing System for Dynamic Monitoring of Sweat Glucose and Real-Time pH Adjustment Based on 3D Graphene MXene Aerogel. ACS Appl Mater Interfaces. 2024;acsami.4c13013.

  39. Komkova MA, Eliseev AA, Poyarkov AA, Daboss EV, Evdokimov PV, Eliseev AA, et al. Simultaneous monitoring of sweat lactate content and sweat secretion rate by wearable remote biosensors. Biosens Bioelectron. 2022;202:113970.

    Article  CAS  PubMed  Google Scholar 

  40. Xuan X, Pérez-Ràfols C, Chen C, Cuartero M, Crespo GA. Lactate Biosensing for Reliable On-Body sweat analysis. ACS Sens. 2021;6:2763–71.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Bi Y, Sun M, Wang J, Zhu Z, Bai J, Emran MY, et al. Universal fully Integrated Wearable Sensor arrays for the multiple electrolyte and metabolite monitoring in raw sweat, saliva, or urine. Anal Chem. 2023;95:6690–9.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

This work is financially supported by the National Natural Science Foundation of China (5217554,52203316, 52305609), Science and Technology Cooperation and Exchange Special Project of Shanxi Province (No.202304041101032), Fund Program for the Scientific Activities of Selected Returned Overseas Professionals in Shanxi Province(20240007), Research Project Supported by Shanxi Scholarship Council of China (No. 2024-062), Patent Transformation Special Program of Shanxi Province (No.202304012), Natural Science Foundation of Shanxi Province (No. 20210302123136,202103021223068), China Postdoctoral Science Foundation (No. 2024M762331) and the Fund for “Shanxi Provincial Doctoral Innovation Station”.

Author information

Authors and Affiliations

  1. College of Integrated Circuits, Taiyuan University of Technology, Taiyuan, China

    Jianlong Ji, Jing Xiao, Fan Zhang, Zhaoqun Wang, Tianyuan Zhou, Xiaorong Niu, Wendong Zhang, Shengbo Sang & Xiaojie Chai

  2. 6D Artificial Intelligence Biomedical Research Institute, Taiyuan, China

    Tianyuan Zhou & Xiaorong Niu

  3. Institute for Advanced Study Shenzhen University, Shenzhen, China

    Sheng Yan

Authors
  1. Jianlong Ji
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Jing Xiao
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Fan Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Zhaoqun Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Tianyuan Zhou
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Xiaorong Niu
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Wendong Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Shengbo Sang
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Xiaojie Chai
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Sheng Yan
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

J. J. and S. Y. wrote the manuscript. J. X., F. Z., Z. W., T. Z., N. X, and W. Z. contributed to all experiments, data processing, and Figures. X. C., S. Y., and S. S. supervised the whole project and revised the manuscript. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Shengbo Sang, Xiaojie Chai or Sheng Yan.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors agreed to submit this manuscript.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

12951_2025_3189_MOESM2_ESM.docx

Supplementary Material 2: Reagents and instruments (Note 1); Detailed enzymatic sensors preparation process (Note 2); The preparation procedure of the PEDOT: PSS mixture (Note 3); The fabrication method of electrodes used for comparing the catalytic performance (Note 4); The equivalent circuit of the floating-gate OECTs (Note 5); The relationship between Nernst potential and hydrogen peroxide concentration (Note 6); The illustration of the reason for the systematic drift of IDS (Note 7). On-skin signal acquisition of wearable OCET enzyme devices (Movie 1).

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/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ji, J., Xiao, J., Zhang, F. et al. A wearable enzyme sensor enabled by the floating-gate OECT with poly(benzimidazobenzophenanthroline) as the catalytic layer. J Nanobiotechnol 23, 120 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12951-025-03189-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12951-025-03189-1

Keywords

Journal of Nanobiotechnology

ISSN: 1477-3155

Contact us