Miniaturized Iridium Oxide Microwire pH Sensor for Biofluid Sensing

Abstract

pH regulation in human biofluids is a crucial step for disease diagnosis and health monitoring. Traditional pH sensors are limited by their bulky size in wearable systems, and fragile glass tips require frequent calibration, thus limiting their use in continuous monitoring. Flexible sensors, particularly those utilizing microwires and thread-based substrates, present advantages for small sample analysis, including natural breathability and suitability for bandage or textile integration. This study examines iridium oxide and silver–silver chloride coated on thin gold wires, fabricated using sol–gel and dip-coating processes known for their simplicity. The flexible microwires demonstrated promising pH performance from a study of their pH characteristics, sensitivity, hysteresis, and potential drift. Electrodes tested in microwells allowed for small sample volumes and localized pH measurement in a controlled environment. Additional integration into fabrics for sweat sensing in wearables highlighted their potential for continuous, real-time health monitoring applications.

1. Introduction

Human biofluid exhibits numerous biomarkers for disease diagnosis and for tracking health status. Biofluids maintain strict pH regulation, and minor changes can indicate the early onset of diseases. Real-time data collection and analysis can significantly improve patient outcomes through early detection and intervention strategies. An instance is the early detection of sepsis infection, which can be alerted by monitoring pH changes in the blood or bodily fluids [1,2]. Sweat, tears, saliva, and urine are options for the detection of biomarkers for specific diseases [3,4,5,6]. Continuous sweat production can be achieved through iontophoresis and chemical methods [7,8], allowing for real-time skin monitoring within a tiny amount of sweat. Skin pH assessment can indicate the deterioration or healing of the tissues. Typically, healthy skin displays a mildly acidic pH of 5 [9], while wounded skin tends to be more alkaline, nearing a pH of 8 [10]. In clinical environments, portable handheld or table-top pH sensors featuring flat-tip designs are employed for monitoring skin health. However, the fragile glass tip necessitates skilled users to operate and regular calibration, and prevents the continuous monitoring of skin chemistry. The amount of fluid needed by the sensing electrodes is not achievable with the use of iontophoresis. This prohibits tracking the continuous time-variant pH changes that provide critical physiological information.

Flexible sensors offer an excellent solution for real-time health monitoring. Flexible polymeric substrates with electrodes, being created through laser engraving, lithography, and printing techniques, have been demonstrated [11,12,13]. Although planar polymeric substrates have laid the groundwork for flexible pH sensing, they face limitations when it comes to analyzing tiny samples, such as those in cell analysis, and in terms of substrate breathability. Microwire and thread-based pH sensors may offer a great alternative. Multiplexed pH measurements using microwire electrodes were demonstrated for cell growth study in cell culture [14]. Thread electrodes integrated into textile substrates provide natural breathability during skin sensing. The use of thread-based pH sensors allows for sensing on various diagnostic platforms, such as bandages [15], and knitted into textiles [16]. Thread-based pH sensors require an initial step of metallization to create a conductive surface, and then the pH-sensitive material is coated. Metal microwires do not require the metallization step during the fabrication process. The pH-sensitive material can be deposited by sputtering, electrodeposition, or sol–gel processes [17,18,19]. Among these, the sol–gel process stands out for its relative simplicity. Researchers can reproduce high-quality thin films without sophisticated deposition hardware. Additionally, the surface homogeneity and thickness can be controlled during the coating step [20]. This work utilizes the sol–gel process to deposit iridium oxide (IrO_x) on a gold bonding wire. The biocompatibility and high sensitivity features of IrO_x are crucial for biosensing, thus it has been a preferred coating material for implantable bioelectrodes [21]. Silver–silver chloride (Ag/AgCl) is widely used as a reference electrode material for stability, miniaturization, and fabrication ease [22].

Our group previously reported on IrO_x thin films coated on planar polyimide substrates [23]. This work explored IrO_x films conformally coated on 25 µm gold microwires. To complete the fabrication of the miniaturized sensing platform, Ag/AgCl films were deposited onto the microwires. The surface geometry of a microwire provided a three-dimensional sensing area beneficial for studying tiny samples. This was demonstrated by testing the electrodes immersed in microwells. Additionally, the microwires were weaved into fabric to illustrate wearable textile applications where sensing was effectively conducted using small sample volumes of solution. pH characteristics such as sensitivities, hysteresis, and drift were analyzed.

2. Materials and Methods

Figure 1 illustrates the device configuration. The preparation of the sol–gel solution, the dip-coating process, and the heat treatment were adapted from previous work [24]. After coating IrCl₃ on a gold (Au) microwire, heat treatment at 325 °C facilitated oxide growth. The IrO_x layer formed on the wire surface as illustrated in Figure 1a. The scanning electron microscope (SEM) image in Figure 1b shows an electrode diameter of 25 µm with a layer of IrO_x. A SEM image in Figure 2a shows the surface detail and successful IrO_x growth. Figure 2b shows the elemental peaks for Ir, O, and Au by energy dispersive spectroscopy (EDS) (JEOL IT500HR Field Emission).

A potentiostat (CH Instruments, Bee Cave, TX, USA) was used for open-circuit potential (OCP) and cyclic voltammetry (CV) measurements. Phosphate-buffered saline (PBS) 1X (Fisher, Waltham, MA, USA), with a pH of 7.4 and a salt concentration (0.137 M) close to that of human body fluids, was chosen for its biological relevance [25]. Artificial sweat (Reagents, Charlotte, NC, USA) with an initial pH of 4.7 was adjusted from pH 4 to 9 using hydrochloric acid (HCl) and sodium hydroxide (NaOH). A commercial pH meter (Apera, Columbus, OH, USA) was used to verify the pH levels of the adjusted sweat solutions. A commercial glass rod (Basi, West Lafayette, IN, USA) Ag/AgCl reference electrode was used for tests in buffer pH 4, 7, and 10 solutions (Fisher, Manchester, NH, USA). Due to the large size of the commercial electrode, 50 mL solutions were prepared for verifying their pH values. The Ag/AgCl microwire was prepared by coating gold (Au) bonding wires with commercial ink (CHI., CA, USA). The 25 µm diameter microwires with IrO_x and Ag/AgCl enabled sensing in volumes as small as 0.2 mL. To test the microwire in these small volumes, 3-D printed microwells with a diameter of 10 mm and a 0.2 mL capacity were created. To further demonstrate wearable applications, the microwires were assembled in fabrics and tested with small solution volumes of 0.2 mL. Three IrO_x-Ag/AgCl microwire pairs were weaved into a fabric pad with a size of $5\times 5$ cm². Each microwire pair was connected to a single input channel of a data acquisition card (NI, Austin, TX, USA). Each channel consists of two ports for the IrO_x working electrode and the Ag/AgCl reference electrode. This electrical interface connecting the electrode pairs embedded in the pad allows for three different pH-level solutions of 4, 6, and 9 to be dripped onto the pad and for their local potentials to be simultaneously recorded.

3. Results and Discussion

3.1. pH-Sensing Performance by Open-Circuit Potential Measurement

Figure 3a shows the OCP responses and hysteresis study across two test cycles using the IrO_x microwire vs. a commercial Ag/AgCl electrode. This was to verify solely the IrO_x performance using the commercial Ag/AgCl electrode as a standard reference. Each cycle involved testing in the sequence of pH 4-7-10-7-4. A 50 mL beaker filled with solutions was used for all of the tests. The electrodes underwent cleaning in deionized (DI) water between tests. Previously, hysteresis (dV) was defined and reported for IrO_x films on planar substrates [24]. The dV was characterized as the standard deviation between stabilized potentials at identical pH values in the test solutions, as indicated in Figure 3a. The correlation between pH and dV at pH 4, 7, and 10 was utilized to compute the pH fluctuation (dpH). The dV values of ±1.9, ±11, and ±0.3 mV corresponded to pH variations (dpH) of ±0.03, ±0.2, and ±0.06 pH, respectively, as depicted in Figure 3b. The highest dpH of ±0.2 pH occurred at pH 7, in contrast to both acidic and alkaline environments. The “±” sign here indicates the error range from its average output value. The difference in ionic concentrations of pH 10 and pH 7 likely caused high dV while returning from pH 10 to pH 7 during the cyclic testing. Alkaline pH 10 predominantly contains OH⁻ ions compared to neutral pH 7, with equal amounts of OH⁻ and H⁺ ions. OH⁻ ions attached to the oxide states and tended to be more adherent on the surface even with deionized water cleaning to the electrode. Another study also reported that higher hysteresis in a neutral pH 7 solution in cyclic tests, compared to acidic or alkaline levels, existed in field effect transistor sensors [26]. The correlation between pH and potential was utilized for determining sensitivity. A sensitivity of −54 mV/pH for the IrO_x microwire electrode was comparable to the previously demonstrated IrO_x planar films at −51 mV/pH [24].

Ag/AgCl microwire was then used to replace the commercial Ag/AgCl electrode. Figure 4a shows the OCP responses from the IrO_x vs. Ag/AgCl microwire pair and (b) illustrates a sensitivity of −55 mV/pH. Figure 4b indicates the highest hysteresis (dV) of ±4.3 mV at pH 7, resulting in a pH variation (dpH) of ±0.07. The higher hysteresis for pH 7 was again a similar phenomenon, as mentioned before. From the OCP measurements, the IrO_x vs. Ag/AgCl microwires produced a lower pH variation of ±0.07 compared to ±0.2 by the commercial Ag/AgCl electrode. This is probably due to more residues on a larger surface area of the commercial Ag/AgCl electrode than the limited surface of the microwire.

3.2. pH-Sensing Performance by Cyclic Voltammetry

Cyclic voltammetry is a versatile technique widely used to study the oxidative/reductive processes indicated by the peak potentials. To validate the functionality of the small microwires, a sensitivity study from CV was performed. The sensitivities were obtained from the CV profiles using the peak potential and pH relationship. CV was performed in a three-electrode system (CHI., USA) with an IrO_x microwire as the working electrode, commercial Ag/AgCl or microwire Ag/AgCl as the reference electrode, and platinum as the counter electrode. The tests were performed in the same buffer solutions in the potential range of −0.4 to +0.8 V at a scan rate of 30 mV/s.

The pH performance was compared between the IrO_x microwire vs. the commercial Ag/AgCl and IrO_x vs. Ag/AgCl microwires electrode pairs. For the IrO_x microwire vs. the commercial Ag/AgCl electrode, Figure 5a shows a sensitivity of −66.7 mV/pH, calculated from the peak potentials in the CV curves and pH values. A similar sensitivity of −63.7 mV/pH for the IrO_x vs. Ag/AgCl microwire pair is shown in Figure 5b. The increased sensitivity observed in the CV tests in contrast to the OCP measurements is ascribed to the differences in the oxide states. In CV scans, the oxide composition undergoes alterations owing to redox reactions. Conversely, in OCP measurements, the high input impedance of the instrument prevents substantial changes in the oxide states [27]. Nonetheless, the OCP and CV measurements demonstrated that IrO_x microwire performance was comparable between against a commercial reference electrode and the Ag/AgCl microwire made in our lab.

3.3. Potential Drift Study

To study the potential drift behavior, Figure 6a shows an IrO_x microwire vs. a commercial Ag/AgCl electrode in 50 mL solution. Figure 6b shows the IrO_x vs. Ag/AgCl microwires dipped in microwells filled with 0.2 mL solution in each well. The larger volume of 50 mL was used to accommodate the size of the commercial Ag/AgCl electrode. PBS 1× solution of pH 7.4 and a salt concentration of 0.137 M, close to human body fluid, was used in all stages of the drift studies.

Potential drift (V′) was previously defined as the slow non-random change in the output potential and was measured as the change between the initial and settled potentials [24]. Our group previously demonstrated a reset technique on planar IrO_x films that could improve the drifting issue [28]. Figure 6c,d show how the reset method applied to the IrO_x microwire vs. a commercial Ag/AgCl electrode, and the IrO_x vs. Ag/AgCl microwires, respectively, can improve the drift phenomena. The experimental protocol for both scenarios was the same: after the recording, they continued for 1.5 h with the pairs in solution. The blue-shaded areas show the potential drift ranges. Then, an electrical reset was applied to the electrode pairs, indicated by the yellow lines. The reset was performed by applying an external voltage of 0.2 V across the working and reference electrodes. The 0.2 V reset voltage was chosen because it has been shown to produce a stable OCP with low drifts compared to other reset values, demonstrated in previous work for IrO_x [28]. After 2 min, the voltage was turned off and the recording wires were reconnected to the electrode pairs to record the potentials for another 1.5 h. The red-shaded areas show the improved performance of output potentials with less drifting. The drift (V′) improvements after the reset for an IrO_x microwire vs. a commercial Ag/AgCl electrode in Figure 6c and the IrO_x vs. Ag/AgCl microwire pair in Figure 6d were changed from 11.8 mV to 4.2 mV and from 16.5 mV to 2.2 mV, reducing drifting by 64% and 87%, respectively. The potential drift improvement was accompanied by a faster settling time towards their stable potential values. The change in the output potential after reset can be attributed to the charge exchange processes at the OH⁻-rich (IrO_x) surface previously observed for the planar electrode [28]. This study confirmed that the voltage-reset technique applied to planar electrodes can also improve the performance of the microwire electrodes [28].

3.4. Microwires in Small-Volume Sensing

Artificial sweat with an initial pH of 4.7 was used to create new solutions, with their pH levels adjusted to range from 4 to 9 using HCl and NaOH. A pH meter verified the adjusted pH levels. A pipette was used to drop in or draw out solutions or water from the microwells. Deionized water was used to clean the microwells and electrodes between tests to remove residues. A single IrO_x vs. Ag/AgCl microwire pair was immersed in the microwell and continuously tested from pH 4 to 9. Figure 7a shows the OCP step responses using the same microwire pair. The sensitivity of −57.4 mV/pH, shown in Figure 7b for the microwire, was comparable with the previous planar film design, which produced a sensitivity of −59 mV/pH, tested in artificial sweat [29].

To demonstrate potential wearable applications, three pairs of IrO_x vs. Ag/AgCl microwire pairs were weaved into a $5\times 5$ cm² meshed fabric pad, as illustrated in Figure 8a. The meshed fabric was a stretchable cloth material for breathability that was important for long-term wound dressing. Each pair was connected to the data acquisition (DaQ) card. Each pair utilized a single channel. Solutions with pH 4, 6, and 9 were dripped on the fabric pad simultaneously with pipettes. The zoomed-in drawing in Figure 8a illustrates the pipette dripping a small volume, 0.2 mL, of artificial sweat solution onto the fabric pad containing the microwires. Figure 8b shows that the electrodes produce a sensitivity of −56 mV/pH across the three pH levels tested. These experiments demonstrated small-volume in situ sensing in two different scenarios.

4. Conclusions

This study investigated the development and pH-sensing characteristics of microwire-based pH sensors. Our group previously demonstrated IrO_x based pH sensors built on planar surfaces. This manuscript makes a significant advancement by creating a much smaller, wire-shaped 25 µm IrO_x pH sensor. To complete the sensor system, a Ag/AgCl reference electrode was built utilizing the microwire. Sol–gel and dip coating processes were utilized to produce IrO_x and Ag/AgCl films on gold microwires.

Electroanalytical tests, such as open-circuit potential and cyclic voltammetry, indicated that IrO_x and Ag/AgCl microwire pairs exhibited performance comparable to their planar and commercial counterparts. When tested in control solutions, such as commercial buffers and adjusted pH solutions in PBS and artificial sweat, the microwire pairs showed near-Nernstian sensitivity and low hysteresis. Improved drifting was achieved after the voltage reset was applied and was also similar to the planar films. The reset technique enables long-term stability without frequently removing the sensor from the solution for calibration. The small-volume sensing ability was demonstrated using wells and embedding the microwires into $5\times 5$ cm² fabrics.

The flexible microwire with a diameter of 25 µm enables localized pH measurements. The miniaturized configuration offers distinct advantages for both wearable and micro-volume sensing applications. Integrating these sensors into fabrics embedded in a bandage presents an approach for wearable monitoring systems. Their compact size also supports minimally invasive applications for real-time sensing in wound dressing. Moreover, the capacity to operate in low sample volumes is also beneficial for microfluidic systems where sample availability is limited. These attributes collectively position the microwire pH sensor as a promising tool for advancing biomedical research and clinical diagnostics. Future applications involve inserting these microwires into biological cells to interface directly with cellular environments without disrupting normal cellular functions.