A Performance Comparison Between Organic Electrochemical Transistor and Electrode Configurations for Enzymatic Sensing

Organic electrochemical transistors (OECTs) excel at biosensing due to their high amplification factor, which allows for detecting low analyte concentrations and picking up weak physiological signals. One prominent use of OECT is in enzymatic metabolite sensing, with the OECT claimed to have a superior low limit of detection and enhanced sensitivity compared to conventional two or three electrode‐based setups. However, there has yet to be a direct comparative study on the performance metrics of these sensor configurations under unified conditions. Here, the glucose sensing performance of an enzyme‐immobilized electrode is systematically examined in two types of devices that have the same geometrical relations: the first one is a traditional 2‐ or 3‐electrode configuration where the sensing electrode is the working electrode, and in the second one, the enzymatic electrode serves as the gate electrode of an OECT. While benchmarking the performance of OECT technology for enzyme‐based metabolite sensing, this study provides insights into the operation mechanism of OECT‐based enzymatic sensors. These results can help to design more efficient OECT‐based circuits to transduce biological events that involve redox reactions.


Introduction
[3] The OECT is a three-terminal device with a source, DOI: 10.1002/adsr.202300188drain, and gate contact.The source and the drain are connected by a channel made of an organic (semi)conductor film that can transport ionic and electronic charges.The conductivity of the channel is controlled by the voltage applied at the gate electrode (V g ), which pushes or pulls ionic charges to or from the channel through an (often aqueous) electrolyte.OECT converts small voltage signals at the gate electrode (∂V g ) into a large change in the channel current (∂I d ).characterized by a high amplification factor quantified as transconductance, g m = V g I d .The high g m makes OECTs an active circuit element and attractive sensor component.
As one of the first sensor demonstrations that utilized the OECT circuity, [4] enzyme-based sensing has become popular due to the relatively simple device functionalization with a suitable catalytic enzyme.Enzymes have typically been immobilized on the gate electrode, and depending on the enzyme type integrated, OECT signals showed a correlation with the concentration of the metabolite such as glucose, [5] lactate, [6] and cholesterol. [7]Devices with electron-transporting polymerbased channels and gate electrode coatings detected very low concentrations of glucose [1] and lactate [6] in biological fluids through the integration of microfluidics. [8]n the current state of the art, most electrochemical enzymatic sensors are developed in a 3-electrode (3-e) configuration.This configuration consists of a working electrode (WE), which is functionalized with an enzyme, where the metabolite binding and redox reactions occur.The binding of the metabolite to its enzyme generates either free electrons that can be transferred to the electrode via a mediator or electrochemical species that react with the electrode.A counter electrode (CE), usually a large surface area metal such as platinum, completes the circuit, and the third electrode is the reference electrode (RE).The role of RE is important, as it maintains the electrochemical potential of the WE (E WE , which we define as the voltage measured in open circuit potentiometry versus RE) at a stable value so that the reactions at the WE are carried out always at the same electrochemical potential.This setup has been proven as the golden standard for electrochemical enzymatic sensing.On the other hand, enzymatic sensors based on OECTs have grown significantly in the last decade with a claim to exhibit superior performance compared to this conventional electrode configuration. [9]Yet, there are limited reports that evidence whether OECTs are inherently better than a 3-electrode setup for enzymatic sensing, [10] which require a systematic comparison under the same geometrical conditions.
Due to their amplification property, OECTs are expected to outperform the 3-e configuration by amplifying the signals arising from the analyte-binding events at the gate electrode, just as they improve the signal-to-noise ratio (SNR) of electrophysiological recordings. [11]However, this simplified comparison provides an incomplete picture since OECTs do not operate in the same way as a 3-e setup.The first difference is the absence of a RE; the OECT only contains the channel and the biofunctionalized gate electrode.OECTs, therefore, operate similarly to a 2electrode (2-e) setup, where there is no referenced electrochemical potential.Secondly, OECTs rely on a change in the electrochemical potential of the gate (E gate ), the source (E source ), and the drain (E drain ) to alter the conductivity of the channel and observe changes in the drain current (I d ).The 3-e setup, on the other hand, reports the faradaic current generated due to the enzymatic reaction, and the potentiostat compensates for the change in the E WE while operating in a chronoamperometric mode.While the two operation mechanisms differ significantly, both devices show a current change scaling, to some extent, with metabolite concentration.The shift in E gate triggered by the metabolite binding event has been argued to drive the OECT enzymatic sensing mechanism. [12]However, this may cause the electrochemical potential of the functionalized gate electrode to shift to potentials that are not favorable for the redox reaction, or to shift to values not favorable to operate the device at its maximum g m .Tan et al. suggested a method of separating the gate from the channel to create a superior sensor configuration that remains, at all times, electrochemically well-defined. [12]Ohayon et al. introduced a similar concept where sensing occurred at an enzymatic fuel cell, and the OECT only converted the potential change generated by the fuel cell into a large change in channel current. [1]owever, such device configurations might not always be possible, especially in applications requiring miniaturization (i.e., in vivo sensing devices) or when different form factors are desired.
To understand the enzymatic sensing mechanism of the OECT and how different its output is compared to the conventional setups, we systematically compared the figures of merit, namely, the limit of detection (LOD), SNR, stabilization time, and working range obtained using these three configurations.Our sensor electrode relies on the state-of-the-art material of organic bioelectronics, poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) integrated as the WE of a 2-or 3-e configuration or as the gate of a PEDOT:PSS OECT (Figure 1).We functionalized PEDOT:PSS with a mixture of glucose oxidase (GOx) as the biorecognition unit and ferrocene (Fc), i.e., the redox mediator.We detected each configuration's current output in response to different concentrations of glucose via chronoamperometry.Comparing the performance of each sensor, we find that OECTs outperform electrode configurations in LOD, SNR, and stabilization time but have a lower dynamic range compared to the 3-e design.

Glucose Sensing with the 3-Electrode Set-Up
Our sensing electrode is a square-shaped gold electrode coated with PEDOT:PSS film (500 × 500 μm).We used a chitosan matrix to entrap the GOx and Fc using 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide and N-hydroxysuccinimide (EDC and NHS) coupling reaction that covalently conjugated GOx into chitosan.[15] When glucose is added to the aqueous solution covering the electrode, it reacts with GOx and reduces it to GOx reduced (Figure 1).Fc shuttles electrons generated in this reaction from the enzyme active center to the PEDOT:PSS electrode, oxidizing GOx and reducing PE-DOT:PSS.A physical proximity between all redox centers and suitable electrochemical potentials are necessary for this redox chain reactions to take place.The 3-e system we designed has a GOx-immobilized PEDOT:PSS electrode as the WE, a Pt electrode as the CE, and an Ag/AgCl electrode as the RE (Figure 2a).
To evaluate whether the electrode is responsive to glucose and find optimal operating voltages, we recorded the cyclic voltammetry (CV) curves of the PEDOT:PSS film before any functionalization (pristine) and after the functionalization with Fc and GOx (biofunctionalized), as well as after the addition of glucose to the measurement solution covering the biofunctionalized film.Figure 2b shows that the pristine PEDOT:PSS CV curve has the typical capacitive shape at the selected voltage window (−0.2 V -0.5 V versus Ag/AgCl). [16,17]After the electrode has been biofunctionalized, redox peaks appear around 0.2 V versus Ag/AgCl, attributed to Fc activity. [18]The presence of glucose in the sys-tem increased the oxidation current of this electrode in the region between 0.1 V and 0.5 V, evidencing its sensitivity to glucose.To investigate whether the current signal is proportional to glucose concentration, we performed a chronoamperometry measurement as we added various glucose concentrations to the electrolyte.We monitored the current of the WE (I WE ) biased at 0.4 V versus Ag/AgCl, a voltage where we marked glucosetriggered changes in the CV curve and one that generated a high current read-out (Figure 2c).I WE scaled as a function of glucose content, and this is due to glucose oxidation with the enzyme, as the same conditions did not generate any response for the pristine PEDOT:PSS electrode (Figure S1, Supporting Information).We calculated the average change in I WE and plotted the calibration curve of the sensor, i.e., the change in I WE versus the concentration of glucose in the solution (Figure 2d).
The 3-e system is responsive to glucose only at 500 μM or concentrations above it.At higher concentrations, it shows a logarithmic relationship between its current and glucose concentrations.We estimated the sensitivity of the sensor by calculating the slope of the linear portion of the logarithmic calibration curve.The sensitivity of PEDOT:PSS electrode in the 3-e system is S = 104 ± 29.6 nA/decade.That is a change of 104 nA for each 10-fold increase in glucose concentration in the sensitive region (above 500 μM).We calculated the LOD of the sensor to be 334±88 μM, which is the x-intercept of the line formed on the linear portion of the logarithmic calibration curve.We also calculated the noise of the sensor to estimate the minimum current change needed to activate the sensor.The noise was calculated by averaging the variations in the blank signal across the samples and determined as  = 0.109 ± 0.062 nA.The response observed is much higher than the noise, giving the 3-e-based sensing an SNR of 59.5 dB, calculated according to Equation (1), where S is the sensitivity and  is the noise of blank measurement.
The time it takes for the sensor output to stabilize from the time of glucose addition is defined as the stabilization time.The stabilization point is defined arbitrarily as a low I WE drift of less than 0.1 nA minute −1 or less than 1% current drift per minute.When the stabilization point is reached, the next concentration is added.At a glucose concentration of 1 mM, the sensors take 1019 ± 310 s on average to stabilize, which is around 17 minutes (Figure S2, Supporting Information).This is a significantly long time for real-time sensing.This range falls outside of realtime glucose tracking, which usually requires stabilization times of less than a minute or even just a few seconds.This 3-e configuration for glucose sensing will be our reference when we evaluate the performance of the other sensor setups, namely, the 2-e configuration and the OECT.

Glucose Sensing with the 2-Electrode Set-Up
In the 3-e setup, the RE has the task of setting the reference point for the biasing of the WE.A saturated Ag/AgCl is often used due to high electrochemical stability, which is made by placing an Ag/AgCl wire inside a solution containing a stable concentration of chlorine (Cl − ) ions.This type of electrode can be difficult to implement for in vivo sensors as it raises biocompatibility concerns, but it is also hard to miniaturize. [19,20]A straightforward solution is to operate the sensor with just two electrodes (by connecting CE and RE) as in the 2-e configuration, as shown in Figure 3a.By doing that, the device is operated under the assumption that the electrochemical potential of the CE (E CE ) is stable enough.However, that is not the case for the PEDOT:PSS CE used in our system.We see that E WE and E CE shift towards more negative values when glucose is added to the system (Figure S3, Supporting Information).Note that we use a PEDOT:PSS CE to achieve our goal of comparing the electrode configuration directly with the OECT bearing a PEDOT:PSS gate and a PEDOT:PSS channel.We designed the electrode dimensions (CE:2000 × 2000 μm and WE:500 × 500 μm) such that the CE provides enough charges to compensate for the reaction at the WE (Figure 3a).
Figure 3b shows that the Fc redox peaks are shifted to 0 V versus CE (compared to 0.2 V versus Ag/AgCl) as PEDOT:PSS CE does not have the same electrochemical potential as Ag/AgCl.The I WE increases due to the enzymatic reaction with glucose (Figure 3b).However, the redox peak position shifts after each sensing cycle towards more positive potentials (Figure S4, Supporting Information).Since the CE potential is unstable (as measured versus Ag/AgCl in Figure S3, Supporting Information), the WE is biased at a different potential for every new measurement.We show the current-time profile of one 2-e sensor in Figure 3c.For the 2-e setup, the change in I WE for the same glucose concentrations is two orders of magnitude lower than that of the 3-e setup.The LOD is 429 ± 34 μM, which is higher than that measured using the 3-e system.While these sensors show increased I WE at higher levels of glucose, the variation among them is high, and the devices saturate at only 1 mM, presenting a narrow detection range (Figure 3d).The sensitivity of the sensor is calculated to be 2.33 ± 0.78 nA/decade.The noise level is 4.7 ± 1.1 pA, which is lower than in the 3-e setup.However, the SNR of this configuration is lower than that of the 3-e setup (53.9 dB).Although faster than the 3-e setup, the 2-e system also takes a long time to stabilize, especially for the high concentration range.

Glucose Sensing with the OECT
In Figure 4a, we show a schematic of an OECT with channel dimensions of L = 10 μm and W = 100 μm and the PEDOT:PSS coated gate (500 × 500 μm) that is functionalized with GOx and Fc.The OECT setup with a PEDOT:PSS coated gate is close to a 2-e setup because it does not have a RE.The gate electrode in this OECT has the same dimensions and is functionalized in the same manner as the WE in the 2/3-e setups.The distance be-tween the gate and the channel is the same as the distance between WE and CE in the 2-e setup to limit variations in sensing due to geometrical differences.The gate-to-channel area ratio is adjusted to ensure moderate g m at sensing conditions.Figure 4b shows the transfer curve and g m versus V g plot of this OECT.
For glucose sensing, the OECT is operated at a V d = −0.4V and V g = 0.4 V, similar to the voltage used for the 2/3-e configurations.However, E gate might be different due to differences in the reference potentials.Figure S5 (Supporting Information) shows that E gate is 0.11 V versus Ag/AgCl, compared to 0.37 V for E CE in the 2-e setup shown in Figure S3 (Supporting Information).We also observe that E gate shifts towards more negative values as the enzymatic reaction occurs, just as in the 2-e setup.Figure 4c shows that the I d responds to the addition of glucose with an LOD of 130 ± 43 μM.The device has a sensitivity of 1.31 ± 0.14 mA decade −1 with a noise level of 0.35 ± 0.26 μA.This means that OECTs have an SNR of 71.5 dB, much higher than the electrode setups.The stabilization time of the OECT is 125 ± 66s at 1 mM, significantly faster than 2-e and 3-e setups.However, the range of OECT sensors is limited to 1 mM.Beyond this concentration, the sensor cannot differentiate additional amounts of glucose.

Comparison of Configurations
In this study, we compare the sensors in an agnostic manner, focusing on figures of merit and explaining what context each sensor might be suitable for.While using such figures of merit is reductive, it gives us an idea of how sensors perform and highlights the strengths and weaknesses of a sensor.Table 1 summarizes the characteristics of all configurations.Note that by controlling fabrication conditions, we minimized variability in sensor functionalization between configurations.We did not optimize any of the sensors or use measurement techniques that may allow to reach higher performance.Therefore, observed differences result solely from the specific setups and their glucose responses.
When we need a sensor to detect low quantities, the configuration would be the one that leads to the lowest LOD.The 2-e has the highest LOD at 429 μM, and the 3-e setup has an LOD of 334 μM, while the OECT outperforms both with an LOD of 130 μM.OECTs perform exceptionally well in this category.Using the optimal OECT geometries and materials, it is possible to make a device with sensitivity to glucose at concentrations as low as 1 nM. [8]The amplification of OECTs is characterized by the maximum g m value.Generally, an OECT with higher g m is considered to lead to higher sensitivity, and the g m and the voltage that maximizes g m can be modulated with device dimensions (g m ∝ Wd/L). [21]We tested another OECT sensor, this time with channel dimensions of L = 50 μm and W = 300 μm and gate dimensions of 500 × 500 μm (≈1:16 area ratio channel:gate), which leads to a lower maximum g m than the device presented above (Figure S6a, Supporting Information).We found that the sensing performance of this device was not different; as such, a ca.22% increase in g m did not lead to a difference in sensitivity (Figure S6b, Supporting Information).Thus, the maximum g m of the OECT is not a predictor of sensitivity for enzymatic sensors used in this configuration.This discrepancy arises from the absence of a stable electrochemical reference, such as Ag/AgCl, within the system.With a PEDOT:PSS gate, i.e., a gate without a stable electrochemical potential, E gate will shift because of polarization, leading to a change in the position of maximum g m . [12]Consequently, the g m value of an OECT under a specific V g after adding glucose may not be the same as and even lower than the previously measured g m at the same V g (before adding glucose).
The 2-e and 3-e setups exhibit similar SNRs (53.9 dB and 59.5 dB, respectively), while the OECT configuration demonstrates a considerably higher SNR of 71.5 dB.PEDOT:PSS OECTs typically have very large I d on the order of 1 mA, whereas 2-and 3-e systems are often limited to I WE around 100 nA.All electrical systems pick up noise from power cables and electromagnetic sources between the sensor and the potentiostat, [22] at a level of generally 1-100 pA in our setup.With a maximum I WE of 100 nA, this noise significantly impacts 2-e and 3-e LODs.The I d in the mA range in OECTs renders this noise negligible.
An important metric for sensors is the working range, which is the concentration range in which the sensor responds predictably.The lower limit of the working range is the LOD, but the higher limit is set at the saturation point of the sensor.The saturation point is defined as the concentration where any additional glucose causes no change in current.For the 3-e setup, we do not reach a saturation point even after the addition of 10 mM of glucose, which is already much higher than the normal glucose range in the blood of 5-9 mM and going up to 17 mM for patients with diabetes. [23]For the 2-e system, however, the saturation point is quickly reached at 1 mM, without any enhancement of LOD over the 3-e.Thus, 2-e sensors make abysmal alternatives to 3-e in any application.For OECT sensors, we see a similar saturation at 1 mM.However, OECTs have a low LOD of 130 μM.Thus, OECTs can be complementary to the 3-e configuration in scenarios where lower concentrations need to be measured, like glucose in saliva, which varies between 100 and 500 μM. [23]he early saturation of the sensors at 1 mM for the 2-e and the OECT sensors is attributed to the absence of a RE.As shown in Figure S3 (Supporting Information), the electrochemical of the sensing electrode shifts with the oxidation of glucose.Without a RE, E gate and E WE are eventually pushed outside the glucose sensing window depicted in Figure 2b.We observe that above 1 mM of glucose, E gate no longer changes beyond 0.05 V versus Ag/AgCl, explaining why it is no longer possible to carry an additional glucose oxidation reaction (Figure S7, Supporting Information).For the 2-e sensor, we observe a reduction in E WE after adding 1 mM of glucose, stabilizing at around 0.15 V versus Ag/AgCl.The addition of 10 mM, however, slightly affects the E WE .Therefore, the change in E WE alone cannot explain why 2-e sensors saturate at 1 mM and not at higher concentrations, as the electrode is still biased within the glucose oxidation window.
Many sensors are often marketed as having a real-time response.The fast readout of analyte concentration fluctuations is an important sensor feature in order to generate immediate, realtime alerts for patients with diabetes, which may initiate insulin uptake.We characterized the stabilization time for the two concentrations detectable by all configurations, namely, 500 μM and 1 mM (Figure S2, Supporting Information).At 500 μM, the OECT has a lower average stabilization time among the three configurations, but it is not significantly different.At 1 mM, the stabilization time of the 3-e sensors is significantly longer than that of the 2-e and OECT sensors.The 3-e sensors take 8 times longer to stabilize than the OECT sensors.It is unclear why we observe this large difference in stabilization time between the configurations but it seems that using a RE increases stabilization time.We hypothesize that for the 3-e setup, the WE is pushed to a high electrochemical potential (thermodynamic state).The measured I WE is the regeneration of the redox-active species that reaches a stable dynamic equilibrium that defines the final stable I WE .However, in an OECT, the gate is pushed to a lower electrochemical potential, i.e., a different thermodynamic state than the 3-e setup.Because the two thermodynamic states are different, the stabilization time needed to reach these states could also be different.Furthermore, unlike with the 3-e setup, the current measured in an OECT is not the regeneration current because there is no RE keeping E gate constant.Further investigation of this phe-nomenon is required to understand exactly why RE influences the sensor speed.

Operation of OECT Enzymatic Sensors
While these configurations contain the same functional electrode as the active electrode and generate a current signal correlated with glucose concentrations, their individual sensing mechanisms are not identical.In a 3-e setup, when glucose is added to the system, oxidation of glucose occurs, leading to the transfer of electrons to the WE.To maintain the electrochemical potential of the WE, an opposing current flows to the CE, regenerating all redox species on the WE and maintaining constant E WE .This is only possible in a 3-e with a RE, as the RE provides a stable electrochemical reference potential.The resulting change in current observed in I WE is generated to maintain the WE at the set electrochemical potential as described by the Cottrell equation.In an OECT sensor, there is no RE, and thus, the current change measured in an OECT does not originate from the same mechanism as in a 3-e sensor.
In an OECT, I d depends on the electrochemical potential of the channel, namely, E source and E drain .Since E drain is defined in relation to E source (E drain = E source + V d ), we refer only to changes in E source to represent changes in the electrochemical potential of the channel.In Figure S8a (Supporting Information), we measure the current-voltage characteristics of an OECT at two different E source values.In case #1, E source is more reduced (at more negative potentials versus Ag/AgCl reference) compared to case #2.PEDOT:PSS, when biased at negative potentials, gets dedoped, which results in a smaller I d (Figure S8b, Supporting Information).During glucose sensing, as shown in Figure S6 (Supporting Information), E source shifts towards even more negative potentials; therefore, the channel is further de-doped upon the addition of glucose.Thus, a reduction in I d is expected and observed in Figure 4c.This sensing mechanism is explained in Figure S9 (Supporting Information), where in steady state, the OECT has a particular E gate and E source , where E gate = E source + V g , and both can be measured against Ag/AgCl.When glucose is added, E gate drops to to E gate *, which in turn brings E source to E source *, to a more negative sate, which results in a smaller channel current.
Understanding the operation mechanism of OECTs, what advantages they offer over traditional systems, and how these advantages can enhance sensing capabilities is crucial.OECT sensors excel at detecting small changes induced by low analyte concentrations.While a system with low noise can achieve sensitivity similar to that of OECTs, [8] OECTs possess the unique ability to amplify signals even in non-ideal conditions and thus have a large SNR.The large SNR makes OECTs ideal for low-concentration detection.However, OECTs, as currently used, do not contain a RE, which limits the working range and reliability of enzymebased sensors, as we observed in our study.Thus, new configurations are necessary, for example, integrating a RE electrode into an OECT while allowing for shifts in E gate , E drain , and E source , which lead to changes in I d .Tan et al. [12] and Ji et al. [24] showed successful examples of RE electrode-integrated OECT configurations for Faradaic sensing.The challenge is to generate stable and miniaturized reference electrodes and design simpler configurations compatible with different sensing applications, including in vivo conditions.

Conclusion
OECTs have been claimed to be superior in biosensing performance over the traditional 2-e and 3-e configurations due to their high amplification.However, there have been no direct comparisons of both configurations under similar geometrical constraints.In this study, we compared an OECT to a 3-e and 2-e configuration where we used an enzyme-immobilized PEDOT:PSS coated gold electrode as the gate electrode of the OECT as well as the working electrode of the conventional electrochemical sensor setups.In cases where an RE cannot be used, we propose that the OECT can make a better alternative to 3-e setups than the 2-e setup.The OECT's active amplification yields superior detection limits compared to the 2-and 3-e systems by mitigating electrical noise.The OECT also reaches steady-state faster, enabling realtime sensing if further optimized.However, its working range falls short of the 3-e setup, making the OECT less ideal for high analyte concentrations.Moreover, to ensure repeatability, each gate electrode of the OECT has to exhibit the same electrochemical potential before sensing.The 2-e setup offers no advantages over either the OECT or 3-e platforms for enzymatic detection.Consequently, the OECT and 3-e systems represent a trade-off between lower and upper detection limits based on the application.These features make OECTs ideal for real-time sensing of glucose in environments where the concentrations are low, like saliva, while the 3-e configuration is more suited when a wider operational range is needed, e.g., when measuring blood glucose.Future designs should incorporate reference electrodes to OECTs detecting faradic events at the gate electrode for reliable analyte concentration estimations.

Figure 1 .
Figure 1.Left: Glucose sensing mechanism of a PEDOT:PSS electrode functionalized with GOx and Fc.Right: The functionalized PEDOT:PSS electrode as the WE of the 3-e and 2-e system (top and middle, respectively) and as the gate electrode of the OECT (bottom).In our designs, the sensor electrode dimensions (WE in 2-e and 3-e systems and the gate in the OECT) and the distance between the channel and the gate and that between the CE and the WE are the same.The areal ratio of gate:channel and the lateral CE:WE are 250 and 16, respectively.

Figure 2 .
Figure 2. a) The schematic of the 3-e configuration.b) CV curves of a PEDOT:PSS electrode before (pristine) and after the immobilization of Fc/GOx/chitosan (biofunctionalized) and after adding 10 mM of glucose to the electrolyte (+ glucose).The scan rate was 100 mV s −1 .c) The realtime current change of a biofunctionalized PEDOT:PSS electrode in the 3-e configuration as different glucose concentrations were added successively to the measurement solution.The WE was biased at 0.4 V versus Ag/AgCl.The arrows mark the addition of glucose.d) The calibration curve of the 3-e glucose sensor.The current output was measured from chronoamperometry curves, such as the one shown in c (n = 3).

Figure 3 .
Figure 3. a) Schematic of the 2-e configuration.b) CV curves of a PEDOT:PSS electrode before (pristine) and after the immobilization of Fc/GOx/chitosan (biofunctionalized) and after adding 10 mM of glucose to the electrolyte (+ glucose).The scan rate was 100 mV s −1 .c) The real-time current change of a biofunctionalized PEDOT:PSS electrode in a 2-e configuration as different glucose concentrations were added successively to the measurement solution.The WE was biased at 0.4 V versus CE.The arrows mark the addition of glucose.d) Calibration curve of the 2-e glucose sensor.The current output was measured from chronoamperometry curves, such as the one shown in c (n = 3).

Figure 4 .
Figure 4. a) Schematic of the PEDOT:PSS OECT channel and the PEDOT:PSS gate electrode functionalized with chitosan/GOx/Fc.b) V g -dependent I d and g m characteristics of the OECT monitored at V d = −0.4V. c) The real-time change in the channel current when glucose is added to the measurement electrolyte.The arrows mark the addition of each concentration.The operating conditions were V d = −0.4V and V g = 0.4 V d).Calibration curve of OECT glucose sensor with the output measured from chronoamperometry curves, such as the one shown in c (n = 3).

Table 1 .
Comparison of the metabolite sensor figures of merit obtained with 3 different configurations.